diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,129533 +1,3 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.2090713025292465,
-  "eval_steps": 500,
-  "global_step": 18500,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 6.535520554212144e-05,
-      "grad_norm": 4.307169437408447,
-      "learning_rate": 1.088139281828074e-08,
-      "loss": 0.7613,
-      "step": 1
-    },
-    {
-      "epoch": 0.00013071041108424287,
-      "grad_norm": 4.215761184692383,
-      "learning_rate": 2.176278563656148e-08,
-      "loss": 0.668,
-      "step": 2
-    },
-    {
-      "epoch": 0.0001960656166263643,
-      "grad_norm": 3.614027500152588,
-      "learning_rate": 3.264417845484222e-08,
-      "loss": 0.6272,
-      "step": 3
-    },
-    {
-      "epoch": 0.00026142082216848575,
-      "grad_norm": 3.7208406925201416,
-      "learning_rate": 4.352557127312296e-08,
-      "loss": 0.6818,
-      "step": 4
-    },
-    {
-      "epoch": 0.00032677602771060717,
-      "grad_norm": 3.8955652713775635,
-      "learning_rate": 5.44069640914037e-08,
-      "loss": 0.6793,
-      "step": 5
-    },
-    {
-      "epoch": 0.0003921312332527286,
-      "grad_norm": 3.875957727432251,
-      "learning_rate": 6.528835690968444e-08,
-      "loss": 0.6872,
-      "step": 6
-    },
-    {
-      "epoch": 0.00045748643879485,
-      "grad_norm": 3.9089701175689697,
-      "learning_rate": 7.616974972796518e-08,
-      "loss": 0.6867,
-      "step": 7
-    },
-    {
-      "epoch": 0.0005228416443369715,
-      "grad_norm": 3.8417911529541016,
-      "learning_rate": 8.705114254624592e-08,
-      "loss": 0.6714,
-      "step": 8
-    },
-    {
-      "epoch": 0.0005881968498790929,
-      "grad_norm": 4.269287109375,
-      "learning_rate": 9.793253536452667e-08,
-      "loss": 0.7382,
-      "step": 9
-    },
-    {
-      "epoch": 0.0006535520554212143,
-      "grad_norm": 4.398168087005615,
-      "learning_rate": 1.088139281828074e-07,
-      "loss": 0.7119,
-      "step": 10
-    },
-    {
-      "epoch": 0.0007189072609633358,
-      "grad_norm": 3.9808433055877686,
-      "learning_rate": 1.1969532100108813e-07,
-      "loss": 0.6287,
-      "step": 11
-    },
-    {
-      "epoch": 0.0007842624665054572,
-      "grad_norm": 3.8891844749450684,
-      "learning_rate": 1.305767138193689e-07,
-      "loss": 0.7767,
-      "step": 12
-    },
-    {
-      "epoch": 0.0008496176720475786,
-      "grad_norm": 3.9056179523468018,
-      "learning_rate": 1.4145810663764961e-07,
-      "loss": 0.7177,
-      "step": 13
-    },
-    {
-      "epoch": 0.0009149728775897,
-      "grad_norm": 3.930377960205078,
-      "learning_rate": 1.5233949945593037e-07,
-      "loss": 0.6845,
-      "step": 14
-    },
-    {
-      "epoch": 0.0009803280831318215,
-      "grad_norm": 3.7490644454956055,
-      "learning_rate": 1.6322089227421112e-07,
-      "loss": 0.6027,
-      "step": 15
-    },
-    {
-      "epoch": 0.001045683288673943,
-      "grad_norm": 4.5679473876953125,
-      "learning_rate": 1.7410228509249185e-07,
-      "loss": 0.7149,
-      "step": 16
-    },
-    {
-      "epoch": 0.0011110384942160643,
-      "grad_norm": 4.164201736450195,
-      "learning_rate": 1.8498367791077258e-07,
-      "loss": 0.7281,
-      "step": 17
-    },
-    {
-      "epoch": 0.0011763936997581858,
-      "grad_norm": 4.343791961669922,
-      "learning_rate": 1.9586507072905333e-07,
-      "loss": 0.7351,
-      "step": 18
-    },
-    {
-      "epoch": 0.0012417489053003071,
-      "grad_norm": 4.092403411865234,
-      "learning_rate": 2.0674646354733408e-07,
-      "loss": 0.6856,
-      "step": 19
-    },
-    {
-      "epoch": 0.0013071041108424287,
-      "grad_norm": 4.36989688873291,
-      "learning_rate": 2.176278563656148e-07,
-      "loss": 0.7318,
-      "step": 20
-    },
-    {
-      "epoch": 0.00137245931638455,
-      "grad_norm": 3.744257688522339,
-      "learning_rate": 2.2850924918389557e-07,
-      "loss": 0.6401,
-      "step": 21
-    },
-    {
-      "epoch": 0.0014378145219266715,
-      "grad_norm": 3.8534998893737793,
-      "learning_rate": 2.3939064200217627e-07,
-      "loss": 0.6965,
-      "step": 22
-    },
-    {
-      "epoch": 0.0015031697274687928,
-      "grad_norm": 3.3089194297790527,
-      "learning_rate": 2.502720348204571e-07,
-      "loss": 0.6115,
-      "step": 23
-    },
-    {
-      "epoch": 0.0015685249330109144,
-      "grad_norm": 3.912757158279419,
-      "learning_rate": 2.611534276387378e-07,
-      "loss": 0.7333,
-      "step": 24
-    },
-    {
-      "epoch": 0.0016338801385530357,
-      "grad_norm": 3.9144887924194336,
-      "learning_rate": 2.7203482045701853e-07,
-      "loss": 0.6414,
-      "step": 25
-    },
-    {
-      "epoch": 0.0016992353440951572,
-      "grad_norm": 3.714334011077881,
-      "learning_rate": 2.8291621327529923e-07,
-      "loss": 0.6527,
-      "step": 26
-    },
-    {
-      "epoch": 0.0017645905496372785,
-      "grad_norm": 4.35928201675415,
-      "learning_rate": 2.9379760609358004e-07,
-      "loss": 0.6958,
-      "step": 27
-    },
-    {
-      "epoch": 0.0018299457551794,
-      "grad_norm": 3.618541717529297,
-      "learning_rate": 3.0467899891186074e-07,
-      "loss": 0.6722,
-      "step": 28
-    },
-    {
-      "epoch": 0.0018953009607215214,
-      "grad_norm": 3.630235195159912,
-      "learning_rate": 3.155603917301415e-07,
-      "loss": 0.7036,
-      "step": 29
-    },
-    {
-      "epoch": 0.001960656166263643,
-      "grad_norm": 3.5564510822296143,
-      "learning_rate": 3.2644178454842224e-07,
-      "loss": 0.6338,
-      "step": 30
-    },
-    {
-      "epoch": 0.0020260113718057644,
-      "grad_norm": 3.5693376064300537,
-      "learning_rate": 3.3732317736670295e-07,
-      "loss": 0.7431,
-      "step": 31
-    },
-    {
-      "epoch": 0.002091366577347886,
-      "grad_norm": 3.2249109745025635,
-      "learning_rate": 3.482045701849837e-07,
-      "loss": 0.5889,
-      "step": 32
-    },
-    {
-      "epoch": 0.002156721782890007,
-      "grad_norm": 3.57615065574646,
-      "learning_rate": 3.590859630032645e-07,
-      "loss": 0.7309,
-      "step": 33
-    },
-    {
-      "epoch": 0.0022220769884321286,
-      "grad_norm": 3.233099937438965,
-      "learning_rate": 3.6996735582154515e-07,
-      "loss": 0.692,
-      "step": 34
-    },
-    {
-      "epoch": 0.00228743219397425,
-      "grad_norm": 3.3310513496398926,
-      "learning_rate": 3.8084874863982596e-07,
-      "loss": 0.738,
-      "step": 35
-    },
-    {
-      "epoch": 0.0023527873995163717,
-      "grad_norm": 3.067564010620117,
-      "learning_rate": 3.9173014145810666e-07,
-      "loss": 0.6536,
-      "step": 36
-    },
-    {
-      "epoch": 0.0024181426050584928,
-      "grad_norm": 3.765641689300537,
-      "learning_rate": 4.026115342763874e-07,
-      "loss": 0.6706,
-      "step": 37
-    },
-    {
-      "epoch": 0.0024834978106006143,
-      "grad_norm": 3.6483607292175293,
-      "learning_rate": 4.1349292709466817e-07,
-      "loss": 0.7605,
-      "step": 38
-    },
-    {
-      "epoch": 0.002548853016142736,
-      "grad_norm": 3.1373181343078613,
-      "learning_rate": 4.2437431991294887e-07,
-      "loss": 0.6633,
-      "step": 39
-    },
-    {
-      "epoch": 0.0026142082216848573,
-      "grad_norm": 2.975259304046631,
-      "learning_rate": 4.352557127312296e-07,
-      "loss": 0.7719,
-      "step": 40
-    },
-    {
-      "epoch": 0.0026795634272269784,
-      "grad_norm": 2.4532039165496826,
-      "learning_rate": 4.461371055495103e-07,
-      "loss": 0.6455,
-      "step": 41
-    },
-    {
-      "epoch": 0.0027449186327691,
-      "grad_norm": 2.310720682144165,
-      "learning_rate": 4.5701849836779113e-07,
-      "loss": 0.6816,
-      "step": 42
-    },
-    {
-      "epoch": 0.0028102738383112215,
-      "grad_norm": 2.1589982509613037,
-      "learning_rate": 4.678998911860719e-07,
-      "loss": 0.6649,
-      "step": 43
-    },
-    {
-      "epoch": 0.002875629043853343,
-      "grad_norm": 2.4565489292144775,
-      "learning_rate": 4.787812840043525e-07,
-      "loss": 0.7088,
-      "step": 44
-    },
-    {
-      "epoch": 0.002940984249395464,
-      "grad_norm": 2.1987106800079346,
-      "learning_rate": 4.896626768226333e-07,
-      "loss": 0.6978,
-      "step": 45
-    },
-    {
-      "epoch": 0.0030063394549375857,
-      "grad_norm": 1.9119884967803955,
-      "learning_rate": 5.005440696409141e-07,
-      "loss": 0.6952,
-      "step": 46
-    },
-    {
-      "epoch": 0.003071694660479707,
-      "grad_norm": 2.0680160522460938,
-      "learning_rate": 5.114254624591948e-07,
-      "loss": 0.6818,
-      "step": 47
-    },
-    {
-      "epoch": 0.0031370498660218287,
-      "grad_norm": 1.882829189300537,
-      "learning_rate": 5.223068552774755e-07,
-      "loss": 0.6843,
-      "step": 48
-    },
-    {
-      "epoch": 0.0032024050715639503,
-      "grad_norm": 1.6339185237884521,
-      "learning_rate": 5.331882480957563e-07,
-      "loss": 0.5162,
-      "step": 49
-    },
-    {
-      "epoch": 0.0032677602771060714,
-      "grad_norm": 1.7776230573654175,
-      "learning_rate": 5.440696409140371e-07,
-      "loss": 0.6381,
-      "step": 50
-    },
-    {
-      "epoch": 0.003333115482648193,
-      "grad_norm": 1.873404622077942,
-      "learning_rate": 5.549510337323178e-07,
-      "loss": 0.6518,
-      "step": 51
-    },
-    {
-      "epoch": 0.0033984706881903144,
-      "grad_norm": 1.8278203010559082,
-      "learning_rate": 5.658324265505985e-07,
-      "loss": 0.6048,
-      "step": 52
-    },
-    {
-      "epoch": 0.003463825893732436,
-      "grad_norm": 1.652664303779602,
-      "learning_rate": 5.767138193688793e-07,
-      "loss": 0.5727,
-      "step": 53
-    },
-    {
-      "epoch": 0.003529181099274557,
-      "grad_norm": 1.7577887773513794,
-      "learning_rate": 5.875952121871601e-07,
-      "loss": 0.6717,
-      "step": 54
-    },
-    {
-      "epoch": 0.0035945363048166786,
-      "grad_norm": 1.3795794248580933,
-      "learning_rate": 5.984766050054407e-07,
-      "loss": 0.6046,
-      "step": 55
-    },
-    {
-      "epoch": 0.0036598915103588,
-      "grad_norm": 1.725850224494934,
-      "learning_rate": 6.093579978237215e-07,
-      "loss": 0.6673,
-      "step": 56
-    },
-    {
-      "epoch": 0.0037252467159009216,
-      "grad_norm": 1.3847562074661255,
-      "learning_rate": 6.202393906420022e-07,
-      "loss": 0.6046,
-      "step": 57
-    },
-    {
-      "epoch": 0.0037906019214430427,
-      "grad_norm": 1.551095962524414,
-      "learning_rate": 6.31120783460283e-07,
-      "loss": 0.6481,
-      "step": 58
-    },
-    {
-      "epoch": 0.0038559571269851643,
-      "grad_norm": 1.5927350521087646,
-      "learning_rate": 6.420021762785637e-07,
-      "loss": 0.6342,
-      "step": 59
-    },
-    {
-      "epoch": 0.003921312332527286,
-      "grad_norm": 1.6861894130706787,
-      "learning_rate": 6.528835690968445e-07,
-      "loss": 0.7077,
-      "step": 60
-    },
-    {
-      "epoch": 0.003986667538069407,
-      "grad_norm": 1.5050122737884521,
-      "learning_rate": 6.637649619151251e-07,
-      "loss": 0.5497,
-      "step": 61
-    },
-    {
-      "epoch": 0.004052022743611529,
-      "grad_norm": 1.658370852470398,
-      "learning_rate": 6.746463547334059e-07,
-      "loss": 0.7689,
-      "step": 62
-    },
-    {
-      "epoch": 0.00411737794915365,
-      "grad_norm": 1.7540634870529175,
-      "learning_rate": 6.855277475516866e-07,
-      "loss": 0.6374,
-      "step": 63
-    },
-    {
-      "epoch": 0.004182733154695772,
-      "grad_norm": 1.380056381225586,
-      "learning_rate": 6.964091403699674e-07,
-      "loss": 0.5967,
-      "step": 64
-    },
-    {
-      "epoch": 0.004248088360237893,
-      "grad_norm": 1.303650975227356,
-      "learning_rate": 7.072905331882482e-07,
-      "loss": 0.6135,
-      "step": 65
-    },
-    {
-      "epoch": 0.004313443565780014,
-      "grad_norm": 1.3145781755447388,
-      "learning_rate": 7.18171926006529e-07,
-      "loss": 0.5947,
-      "step": 66
-    },
-    {
-      "epoch": 0.004378798771322136,
-      "grad_norm": 1.2271623611450195,
-      "learning_rate": 7.290533188248096e-07,
-      "loss": 0.562,
-      "step": 67
-    },
-    {
-      "epoch": 0.004444153976864257,
-      "grad_norm": 1.2204347848892212,
-      "learning_rate": 7.399347116430903e-07,
-      "loss": 0.6246,
-      "step": 68
-    },
-    {
-      "epoch": 0.004509509182406378,
-      "grad_norm": 1.298128604888916,
-      "learning_rate": 7.508161044613712e-07,
-      "loss": 0.6229,
-      "step": 69
-    },
-    {
-      "epoch": 0.0045748643879485,
-      "grad_norm": 1.0342358350753784,
-      "learning_rate": 7.616974972796519e-07,
-      "loss": 0.5551,
-      "step": 70
-    },
-    {
-      "epoch": 0.004640219593490621,
-      "grad_norm": 1.2292605638504028,
-      "learning_rate": 7.725788900979327e-07,
-      "loss": 0.5936,
-      "step": 71
-    },
-    {
-      "epoch": 0.004705574799032743,
-      "grad_norm": 1.0556893348693848,
-      "learning_rate": 7.834602829162133e-07,
-      "loss": 0.6333,
-      "step": 72
-    },
-    {
-      "epoch": 0.004770930004574864,
-      "grad_norm": 0.9866349697113037,
-      "learning_rate": 7.943416757344941e-07,
-      "loss": 0.6039,
-      "step": 73
-    },
-    {
-      "epoch": 0.0048362852101169855,
-      "grad_norm": 1.0882861614227295,
-      "learning_rate": 8.052230685527748e-07,
-      "loss": 0.5994,
-      "step": 74
-    },
-    {
-      "epoch": 0.0049016404156591075,
-      "grad_norm": 0.8924552202224731,
-      "learning_rate": 8.161044613710556e-07,
-      "loss": 0.5833,
-      "step": 75
-    },
-    {
-      "epoch": 0.004966995621201229,
-      "grad_norm": 1.0262629985809326,
-      "learning_rate": 8.269858541893363e-07,
-      "loss": 0.6139,
-      "step": 76
-    },
-    {
-      "epoch": 0.0050323508267433505,
-      "grad_norm": 0.7652501463890076,
-      "learning_rate": 8.37867247007617e-07,
-      "loss": 0.5224,
-      "step": 77
-    },
-    {
-      "epoch": 0.005097706032285472,
-      "grad_norm": 0.8548263311386108,
-      "learning_rate": 8.487486398258977e-07,
-      "loss": 0.5988,
-      "step": 78
-    },
-    {
-      "epoch": 0.005163061237827593,
-      "grad_norm": 0.8868620991706848,
-      "learning_rate": 8.596300326441785e-07,
-      "loss": 0.6644,
-      "step": 79
-    },
-    {
-      "epoch": 0.005228416443369715,
-      "grad_norm": 0.7825864553451538,
-      "learning_rate": 8.705114254624592e-07,
-      "loss": 0.5369,
-      "step": 80
-    },
-    {
-      "epoch": 0.005293771648911836,
-      "grad_norm": 0.8478065729141235,
-      "learning_rate": 8.8139281828074e-07,
-      "loss": 0.6551,
-      "step": 81
-    },
-    {
-      "epoch": 0.005359126854453957,
-      "grad_norm": 0.8154911994934082,
-      "learning_rate": 8.922742110990207e-07,
-      "loss": 0.6163,
-      "step": 82
-    },
-    {
-      "epoch": 0.005424482059996079,
-      "grad_norm": 0.8136587142944336,
-      "learning_rate": 9.031556039173014e-07,
-      "loss": 0.5972,
-      "step": 83
-    },
-    {
-      "epoch": 0.0054898372655382,
-      "grad_norm": 0.8032965064048767,
-      "learning_rate": 9.140369967355823e-07,
-      "loss": 0.6475,
-      "step": 84
-    },
-    {
-      "epoch": 0.005555192471080322,
-      "grad_norm": 0.7939515113830566,
-      "learning_rate": 9.24918389553863e-07,
-      "loss": 0.6002,
-      "step": 85
-    },
-    {
-      "epoch": 0.005620547676622443,
-      "grad_norm": 0.8008838295936584,
-      "learning_rate": 9.357997823721438e-07,
-      "loss": 0.5833,
-      "step": 86
-    },
-    {
-      "epoch": 0.005685902882164564,
-      "grad_norm": 0.6839145421981812,
-      "learning_rate": 9.466811751904245e-07,
-      "loss": 0.4914,
-      "step": 87
-    },
-    {
-      "epoch": 0.005751258087706686,
-      "grad_norm": 0.7718969583511353,
-      "learning_rate": 9.57562568008705e-07,
-      "loss": 0.5712,
-      "step": 88
-    },
-    {
-      "epoch": 0.005816613293248807,
-      "grad_norm": 0.6582275032997131,
-      "learning_rate": 9.68443960826986e-07,
-      "loss": 0.5437,
-      "step": 89
-    },
-    {
-      "epoch": 0.005881968498790928,
-      "grad_norm": 0.7192770838737488,
-      "learning_rate": 9.793253536452666e-07,
-      "loss": 0.5673,
-      "step": 90
-    },
-    {
-      "epoch": 0.00594732370433305,
-      "grad_norm": 0.707466185092926,
-      "learning_rate": 9.902067464635474e-07,
-      "loss": 0.6058,
-      "step": 91
-    },
-    {
-      "epoch": 0.006012678909875171,
-      "grad_norm": 0.6474989652633667,
-      "learning_rate": 1.0010881392818283e-06,
-      "loss": 0.5328,
-      "step": 92
-    },
-    {
-      "epoch": 0.006078034115417293,
-      "grad_norm": 0.6916255950927734,
-      "learning_rate": 1.011969532100109e-06,
-      "loss": 0.5966,
-      "step": 93
-    },
-    {
-      "epoch": 0.006143389320959414,
-      "grad_norm": 0.6603280305862427,
-      "learning_rate": 1.0228509249183896e-06,
-      "loss": 0.5382,
-      "step": 94
-    },
-    {
-      "epoch": 0.0062087445265015355,
-      "grad_norm": 0.6775929927825928,
-      "learning_rate": 1.0337323177366705e-06,
-      "loss": 0.513,
-      "step": 95
-    },
-    {
-      "epoch": 0.0062740997320436575,
-      "grad_norm": 0.6800976991653442,
-      "learning_rate": 1.044613710554951e-06,
-      "loss": 0.4817,
-      "step": 96
-    },
-    {
-      "epoch": 0.0063394549375857786,
-      "grad_norm": 0.7037233710289001,
-      "learning_rate": 1.055495103373232e-06,
-      "loss": 0.5381,
-      "step": 97
-    },
-    {
-      "epoch": 0.0064048101431279005,
-      "grad_norm": 0.6074303984642029,
-      "learning_rate": 1.0663764961915126e-06,
-      "loss": 0.5317,
-      "step": 98
-    },
-    {
-      "epoch": 0.006470165348670022,
-      "grad_norm": 0.6007310152053833,
-      "learning_rate": 1.0772578890097933e-06,
-      "loss": 0.5476,
-      "step": 99
-    },
-    {
-      "epoch": 0.006535520554212143,
-      "grad_norm": 0.6698499917984009,
-      "learning_rate": 1.0881392818280741e-06,
-      "loss": 0.5226,
-      "step": 100
-    },
-    {
-      "epoch": 0.006600875759754265,
-      "grad_norm": 0.5805163383483887,
-      "learning_rate": 1.0990206746463548e-06,
-      "loss": 0.4965,
-      "step": 101
-    },
-    {
-      "epoch": 0.006666230965296386,
-      "grad_norm": 0.6301014423370361,
-      "learning_rate": 1.1099020674646356e-06,
-      "loss": 0.5327,
-      "step": 102
-    },
-    {
-      "epoch": 0.006731586170838507,
-      "grad_norm": 0.6079239845275879,
-      "learning_rate": 1.1207834602829163e-06,
-      "loss": 0.4963,
-      "step": 103
-    },
-    {
-      "epoch": 0.006796941376380629,
-      "grad_norm": 0.5865227580070496,
-      "learning_rate": 1.131664853101197e-06,
-      "loss": 0.4875,
-      "step": 104
-    },
-    {
-      "epoch": 0.00686229658192275,
-      "grad_norm": 0.6349737644195557,
-      "learning_rate": 1.1425462459194778e-06,
-      "loss": 0.5909,
-      "step": 105
-    },
-    {
-      "epoch": 0.006927651787464872,
-      "grad_norm": 0.6744226813316345,
-      "learning_rate": 1.1534276387377586e-06,
-      "loss": 0.5864,
-      "step": 106
-    },
-    {
-      "epoch": 0.006993006993006993,
-      "grad_norm": 0.6469442844390869,
-      "learning_rate": 1.1643090315560393e-06,
-      "loss": 0.6008,
-      "step": 107
-    },
-    {
-      "epoch": 0.007058362198549114,
-      "grad_norm": 0.7091807126998901,
-      "learning_rate": 1.1751904243743201e-06,
-      "loss": 0.5771,
-      "step": 108
-    },
-    {
-      "epoch": 0.007123717404091236,
-      "grad_norm": 0.5648328065872192,
-      "learning_rate": 1.1860718171926008e-06,
-      "loss": 0.5274,
-      "step": 109
-    },
-    {
-      "epoch": 0.007189072609633357,
-      "grad_norm": 0.5575926899909973,
-      "learning_rate": 1.1969532100108814e-06,
-      "loss": 0.5327,
-      "step": 110
-    },
-    {
-      "epoch": 0.007254427815175479,
-      "grad_norm": 0.6258563995361328,
-      "learning_rate": 1.2078346028291623e-06,
-      "loss": 0.5631,
-      "step": 111
-    },
-    {
-      "epoch": 0.0073197830207176,
-      "grad_norm": 0.6843926310539246,
-      "learning_rate": 1.218715995647443e-06,
-      "loss": 0.5581,
-      "step": 112
-    },
-    {
-      "epoch": 0.007385138226259721,
-      "grad_norm": 0.5899907946586609,
-      "learning_rate": 1.2295973884657238e-06,
-      "loss": 0.5119,
-      "step": 113
-    },
-    {
-      "epoch": 0.007450493431801843,
-      "grad_norm": 0.5911238789558411,
-      "learning_rate": 1.2404787812840045e-06,
-      "loss": 0.5373,
-      "step": 114
-    },
-    {
-      "epoch": 0.007515848637343964,
-      "grad_norm": 0.5474612712860107,
-      "learning_rate": 1.251360174102285e-06,
-      "loss": 0.5233,
-      "step": 115
-    },
-    {
-      "epoch": 0.0075812038428860855,
-      "grad_norm": 0.551700234413147,
-      "learning_rate": 1.262241566920566e-06,
-      "loss": 0.5032,
-      "step": 116
-    },
-    {
-      "epoch": 0.0076465590484282074,
-      "grad_norm": 0.5410349369049072,
-      "learning_rate": 1.2731229597388466e-06,
-      "loss": 0.4826,
-      "step": 117
-    },
-    {
-      "epoch": 0.0077119142539703285,
-      "grad_norm": 0.5772070288658142,
-      "learning_rate": 1.2840043525571275e-06,
-      "loss": 0.5321,
-      "step": 118
-    },
-    {
-      "epoch": 0.0077772694595124505,
-      "grad_norm": 0.6064501404762268,
-      "learning_rate": 1.2948857453754083e-06,
-      "loss": 0.5671,
-      "step": 119
-    },
-    {
-      "epoch": 0.007842624665054572,
-      "grad_norm": 0.5898416042327881,
-      "learning_rate": 1.305767138193689e-06,
-      "loss": 0.5738,
-      "step": 120
-    },
-    {
-      "epoch": 0.007907979870596693,
-      "grad_norm": 0.5656580924987793,
-      "learning_rate": 1.3166485310119698e-06,
-      "loss": 0.479,
-      "step": 121
-    },
-    {
-      "epoch": 0.007973335076138814,
-      "grad_norm": 0.5723082423210144,
-      "learning_rate": 1.3275299238302503e-06,
-      "loss": 0.537,
-      "step": 122
-    },
-    {
-      "epoch": 0.008038690281680937,
-      "grad_norm": 0.5626292824745178,
-      "learning_rate": 1.338411316648531e-06,
-      "loss": 0.4921,
-      "step": 123
-    },
-    {
-      "epoch": 0.008104045487223058,
-      "grad_norm": 0.5675073862075806,
-      "learning_rate": 1.3492927094668118e-06,
-      "loss": 0.4907,
-      "step": 124
-    },
-    {
-      "epoch": 0.008169400692765179,
-      "grad_norm": 0.6061102151870728,
-      "learning_rate": 1.3601741022850926e-06,
-      "loss": 0.5864,
-      "step": 125
-    },
-    {
-      "epoch": 0.0082347558983073,
-      "grad_norm": 0.5727167725563049,
-      "learning_rate": 1.3710554951033733e-06,
-      "loss": 0.547,
-      "step": 126
-    },
-    {
-      "epoch": 0.008300111103849421,
-      "grad_norm": 0.6190813183784485,
-      "learning_rate": 1.3819368879216541e-06,
-      "loss": 0.5945,
-      "step": 127
-    },
-    {
-      "epoch": 0.008365466309391544,
-      "grad_norm": 0.6756262183189392,
-      "learning_rate": 1.3928182807399348e-06,
-      "loss": 0.6157,
-      "step": 128
-    },
-    {
-      "epoch": 0.008430821514933665,
-      "grad_norm": 0.5898597240447998,
-      "learning_rate": 1.4036996735582157e-06,
-      "loss": 0.5841,
-      "step": 129
-    },
-    {
-      "epoch": 0.008496176720475786,
-      "grad_norm": 0.5485091209411621,
-      "learning_rate": 1.4145810663764963e-06,
-      "loss": 0.4964,
-      "step": 130
-    },
-    {
-      "epoch": 0.008561531926017907,
-      "grad_norm": 0.5897760987281799,
-      "learning_rate": 1.4254624591947772e-06,
-      "loss": 0.472,
-      "step": 131
-    },
-    {
-      "epoch": 0.008626887131560028,
-      "grad_norm": 0.5610101222991943,
-      "learning_rate": 1.436343852013058e-06,
-      "loss": 0.5105,
-      "step": 132
-    },
-    {
-      "epoch": 0.00869224233710215,
-      "grad_norm": 0.6131514310836792,
-      "learning_rate": 1.4472252448313385e-06,
-      "loss": 0.5259,
-      "step": 133
-    },
-    {
-      "epoch": 0.008757597542644272,
-      "grad_norm": 0.520524799823761,
-      "learning_rate": 1.4581066376496191e-06,
-      "loss": 0.522,
-      "step": 134
-    },
-    {
-      "epoch": 0.008822952748186393,
-      "grad_norm": 0.5163785815238953,
-      "learning_rate": 1.4689880304679e-06,
-      "loss": 0.4493,
-      "step": 135
-    },
-    {
-      "epoch": 0.008888307953728514,
-      "grad_norm": 0.5651346445083618,
-      "learning_rate": 1.4798694232861806e-06,
-      "loss": 0.544,
-      "step": 136
-    },
-    {
-      "epoch": 0.008953663159270635,
-      "grad_norm": 0.5506546497344971,
-      "learning_rate": 1.4907508161044615e-06,
-      "loss": 0.5048,
-      "step": 137
-    },
-    {
-      "epoch": 0.009019018364812757,
-      "grad_norm": 0.6239061951637268,
-      "learning_rate": 1.5016322089227423e-06,
-      "loss": 0.5648,
-      "step": 138
-    },
-    {
-      "epoch": 0.00908437357035488,
-      "grad_norm": 0.5096587538719177,
-      "learning_rate": 1.512513601741023e-06,
-      "loss": 0.4776,
-      "step": 139
-    },
-    {
-      "epoch": 0.009149728775897,
-      "grad_norm": 0.646660327911377,
-      "learning_rate": 1.5233949945593038e-06,
-      "loss": 0.5995,
-      "step": 140
-    },
-    {
-      "epoch": 0.009215083981439122,
-      "grad_norm": 0.5526559948921204,
-      "learning_rate": 1.5342763873775845e-06,
-      "loss": 0.5316,
-      "step": 141
-    },
-    {
-      "epoch": 0.009280439186981243,
-      "grad_norm": 0.5419862270355225,
-      "learning_rate": 1.5451577801958654e-06,
-      "loss": 0.5316,
-      "step": 142
-    },
-    {
-      "epoch": 0.009345794392523364,
-      "grad_norm": 0.512829065322876,
-      "learning_rate": 1.5560391730141458e-06,
-      "loss": 0.4442,
-      "step": 143
-    },
-    {
-      "epoch": 0.009411149598065487,
-      "grad_norm": 0.48366427421569824,
-      "learning_rate": 1.5669205658324266e-06,
-      "loss": 0.4226,
-      "step": 144
-    },
-    {
-      "epoch": 0.009476504803607608,
-      "grad_norm": 0.5660455226898193,
-      "learning_rate": 1.5778019586507073e-06,
-      "loss": 0.4859,
-      "step": 145
-    },
-    {
-      "epoch": 0.009541860009149729,
-      "grad_norm": 0.5128727555274963,
-      "learning_rate": 1.5886833514689882e-06,
-      "loss": 0.4619,
-      "step": 146
-    },
-    {
-      "epoch": 0.00960721521469185,
-      "grad_norm": 0.5074844360351562,
-      "learning_rate": 1.5995647442872688e-06,
-      "loss": 0.4757,
-      "step": 147
-    },
-    {
-      "epoch": 0.009672570420233971,
-      "grad_norm": 0.4872952699661255,
-      "learning_rate": 1.6104461371055497e-06,
-      "loss": 0.4593,
-      "step": 148
-    },
-    {
-      "epoch": 0.009737925625776094,
-      "grad_norm": 0.5077556371688843,
-      "learning_rate": 1.6213275299238303e-06,
-      "loss": 0.445,
-      "step": 149
-    },
-    {
-      "epoch": 0.009803280831318215,
-      "grad_norm": 0.5790615081787109,
-      "learning_rate": 1.6322089227421112e-06,
-      "loss": 0.5307,
-      "step": 150
-    },
-    {
-      "epoch": 0.009868636036860336,
-      "grad_norm": 0.5457535982131958,
-      "learning_rate": 1.643090315560392e-06,
-      "loss": 0.5595,
-      "step": 151
-    },
-    {
-      "epoch": 0.009933991242402457,
-      "grad_norm": 0.5287249088287354,
-      "learning_rate": 1.6539717083786727e-06,
-      "loss": 0.4813,
-      "step": 152
-    },
-    {
-      "epoch": 0.009999346447944578,
-      "grad_norm": 0.6061781048774719,
-      "learning_rate": 1.6648531011969535e-06,
-      "loss": 0.5805,
-      "step": 153
-    },
-    {
-      "epoch": 0.010064701653486701,
-      "grad_norm": 0.572921633720398,
-      "learning_rate": 1.675734494015234e-06,
-      "loss": 0.5791,
-      "step": 154
-    },
-    {
-      "epoch": 0.010130056859028822,
-      "grad_norm": 0.5639303922653198,
-      "learning_rate": 1.6866158868335148e-06,
-      "loss": 0.5556,
-      "step": 155
-    },
-    {
-      "epoch": 0.010195412064570943,
-      "grad_norm": 0.5342715382575989,
-      "learning_rate": 1.6974972796517955e-06,
-      "loss": 0.471,
-      "step": 156
-    },
-    {
-      "epoch": 0.010260767270113064,
-      "grad_norm": 0.5192899107933044,
-      "learning_rate": 1.7083786724700763e-06,
-      "loss": 0.4722,
-      "step": 157
-    },
-    {
-      "epoch": 0.010326122475655185,
-      "grad_norm": 0.511927604675293,
-      "learning_rate": 1.719260065288357e-06,
-      "loss": 0.4848,
-      "step": 158
-    },
-    {
-      "epoch": 0.010391477681197307,
-      "grad_norm": 0.48781517148017883,
-      "learning_rate": 1.7301414581066378e-06,
-      "loss": 0.4455,
-      "step": 159
-    },
-    {
-      "epoch": 0.01045683288673943,
-      "grad_norm": 0.5133116245269775,
-      "learning_rate": 1.7410228509249185e-06,
-      "loss": 0.4575,
-      "step": 160
-    },
-    {
-      "epoch": 0.01052218809228155,
-      "grad_norm": 0.5270615220069885,
-      "learning_rate": 1.7519042437431994e-06,
-      "loss": 0.4897,
-      "step": 161
-    },
-    {
-      "epoch": 0.010587543297823672,
-      "grad_norm": 0.5281715393066406,
-      "learning_rate": 1.76278563656148e-06,
-      "loss": 0.4943,
-      "step": 162
-    },
-    {
-      "epoch": 0.010652898503365793,
-      "grad_norm": 0.550919771194458,
-      "learning_rate": 1.7736670293797609e-06,
-      "loss": 0.5126,
-      "step": 163
-    },
-    {
-      "epoch": 0.010718253708907914,
-      "grad_norm": 0.5056918859481812,
-      "learning_rate": 1.7845484221980413e-06,
-      "loss": 0.4466,
-      "step": 164
-    },
-    {
-      "epoch": 0.010783608914450037,
-      "grad_norm": 0.5478942394256592,
-      "learning_rate": 1.7954298150163222e-06,
-      "loss": 0.4994,
-      "step": 165
-    },
-    {
-      "epoch": 0.010848964119992158,
-      "grad_norm": 0.6065598130226135,
-      "learning_rate": 1.8063112078346028e-06,
-      "loss": 0.535,
-      "step": 166
-    },
-    {
-      "epoch": 0.010914319325534279,
-      "grad_norm": 0.5542285442352295,
-      "learning_rate": 1.8171926006528837e-06,
-      "loss": 0.5126,
-      "step": 167
-    },
-    {
-      "epoch": 0.0109796745310764,
-      "grad_norm": 0.5846586227416992,
-      "learning_rate": 1.8280739934711645e-06,
-      "loss": 0.5165,
-      "step": 168
-    },
-    {
-      "epoch": 0.011045029736618521,
-      "grad_norm": 0.5893979072570801,
-      "learning_rate": 1.8389553862894452e-06,
-      "loss": 0.5743,
-      "step": 169
-    },
-    {
-      "epoch": 0.011110384942160644,
-      "grad_norm": 0.5135499238967896,
-      "learning_rate": 1.849836779107726e-06,
-      "loss": 0.432,
-      "step": 170
-    },
-    {
-      "epoch": 0.011175740147702765,
-      "grad_norm": 0.5777080059051514,
-      "learning_rate": 1.8607181719260067e-06,
-      "loss": 0.5251,
-      "step": 171
-    },
-    {
-      "epoch": 0.011241095353244886,
-      "grad_norm": 0.5141600370407104,
-      "learning_rate": 1.8715995647442875e-06,
-      "loss": 0.4484,
-      "step": 172
-    },
-    {
-      "epoch": 0.011306450558787007,
-      "grad_norm": 0.5639011263847351,
-      "learning_rate": 1.8824809575625682e-06,
-      "loss": 0.519,
-      "step": 173
-    },
-    {
-      "epoch": 0.011371805764329128,
-      "grad_norm": 0.5748486518859863,
-      "learning_rate": 1.893362350380849e-06,
-      "loss": 0.5692,
-      "step": 174
-    },
-    {
-      "epoch": 0.011437160969871251,
-      "grad_norm": 0.5394601821899414,
-      "learning_rate": 1.9042437431991295e-06,
-      "loss": 0.5143,
-      "step": 175
-    },
-    {
-      "epoch": 0.011502516175413372,
-      "grad_norm": 0.5378672480583191,
-      "learning_rate": 1.91512513601741e-06,
-      "loss": 0.4956,
-      "step": 176
-    },
-    {
-      "epoch": 0.011567871380955493,
-      "grad_norm": 0.5230949521064758,
-      "learning_rate": 1.926006528835691e-06,
-      "loss": 0.4408,
-      "step": 177
-    },
-    {
-      "epoch": 0.011633226586497614,
-      "grad_norm": 0.5612627863883972,
-      "learning_rate": 1.936887921653972e-06,
-      "loss": 0.4842,
-      "step": 178
-    },
-    {
-      "epoch": 0.011698581792039735,
-      "grad_norm": 0.5498820543289185,
-      "learning_rate": 1.9477693144722527e-06,
-      "loss": 0.5395,
-      "step": 179
-    },
-    {
-      "epoch": 0.011763936997581857,
-      "grad_norm": 0.5984400510787964,
-      "learning_rate": 1.958650707290533e-06,
-      "loss": 0.4864,
-      "step": 180
-    },
-    {
-      "epoch": 0.01182929220312398,
-      "grad_norm": 0.5528789162635803,
-      "learning_rate": 1.969532100108814e-06,
-      "loss": 0.5236,
-      "step": 181
-    },
-    {
-      "epoch": 0.0118946474086661,
-      "grad_norm": 0.5016252994537354,
-      "learning_rate": 1.980413492927095e-06,
-      "loss": 0.4496,
-      "step": 182
-    },
-    {
-      "epoch": 0.011960002614208222,
-      "grad_norm": 0.5519850850105286,
-      "learning_rate": 1.9912948857453757e-06,
-      "loss": 0.5349,
-      "step": 183
-    },
-    {
-      "epoch": 0.012025357819750343,
-      "grad_norm": 0.5407207608222961,
-      "learning_rate": 2.0021762785636566e-06,
-      "loss": 0.5174,
-      "step": 184
-    },
-    {
-      "epoch": 0.012090713025292464,
-      "grad_norm": 0.4994608759880066,
-      "learning_rate": 2.013057671381937e-06,
-      "loss": 0.4631,
-      "step": 185
-    },
-    {
-      "epoch": 0.012156068230834587,
-      "grad_norm": 0.548252284526825,
-      "learning_rate": 2.023939064200218e-06,
-      "loss": 0.5058,
-      "step": 186
-    },
-    {
-      "epoch": 0.012221423436376708,
-      "grad_norm": 0.5238258838653564,
-      "learning_rate": 2.0348204570184983e-06,
-      "loss": 0.4947,
-      "step": 187
-    },
-    {
-      "epoch": 0.012286778641918829,
-      "grad_norm": 0.5684877634048462,
-      "learning_rate": 2.045701849836779e-06,
-      "loss": 0.5531,
-      "step": 188
-    },
-    {
-      "epoch": 0.01235213384746095,
-      "grad_norm": 0.5372764468193054,
-      "learning_rate": 2.05658324265506e-06,
-      "loss": 0.4918,
-      "step": 189
-    },
-    {
-      "epoch": 0.012417489053003071,
-      "grad_norm": 0.5252590775489807,
-      "learning_rate": 2.067464635473341e-06,
-      "loss": 0.4838,
-      "step": 190
-    },
-    {
-      "epoch": 0.012482844258545194,
-      "grad_norm": 0.5476294159889221,
-      "learning_rate": 2.0783460282916213e-06,
-      "loss": 0.5394,
-      "step": 191
-    },
-    {
-      "epoch": 0.012548199464087315,
-      "grad_norm": 0.5454583168029785,
-      "learning_rate": 2.089227421109902e-06,
-      "loss": 0.5217,
-      "step": 192
-    },
-    {
-      "epoch": 0.012613554669629436,
-      "grad_norm": 0.5394318103790283,
-      "learning_rate": 2.100108813928183e-06,
-      "loss": 0.4354,
-      "step": 193
-    },
-    {
-      "epoch": 0.012678909875171557,
-      "grad_norm": 0.5456311702728271,
-      "learning_rate": 2.110990206746464e-06,
-      "loss": 0.5058,
-      "step": 194
-    },
-    {
-      "epoch": 0.012744265080713678,
-      "grad_norm": 0.528677225112915,
-      "learning_rate": 2.1218715995647448e-06,
-      "loss": 0.4879,
-      "step": 195
-    },
-    {
-      "epoch": 0.012809620286255801,
-      "grad_norm": 0.5452241897583008,
-      "learning_rate": 2.1327529923830252e-06,
-      "loss": 0.522,
-      "step": 196
-    },
-    {
-      "epoch": 0.012874975491797922,
-      "grad_norm": 0.5905510187149048,
-      "learning_rate": 2.1436343852013056e-06,
-      "loss": 0.5218,
-      "step": 197
-    },
-    {
-      "epoch": 0.012940330697340043,
-      "grad_norm": 0.48563042283058167,
-      "learning_rate": 2.1545157780195865e-06,
-      "loss": 0.4231,
-      "step": 198
-    },
-    {
-      "epoch": 0.013005685902882164,
-      "grad_norm": 0.49927181005477905,
-      "learning_rate": 2.1653971708378674e-06,
-      "loss": 0.4182,
-      "step": 199
-    },
-    {
-      "epoch": 0.013071041108424285,
-      "grad_norm": 0.4863174259662628,
-      "learning_rate": 2.1762785636561482e-06,
-      "loss": 0.4262,
-      "step": 200
-    },
-    {
-      "epoch": 0.013136396313966408,
-      "grad_norm": 0.5405679941177368,
-      "learning_rate": 2.187159956474429e-06,
-      "loss": 0.5124,
-      "step": 201
-    },
-    {
-      "epoch": 0.01320175151950853,
-      "grad_norm": 0.5243266224861145,
-      "learning_rate": 2.1980413492927095e-06,
-      "loss": 0.4921,
-      "step": 202
-    },
-    {
-      "epoch": 0.01326710672505065,
-      "grad_norm": 0.5484087467193604,
-      "learning_rate": 2.2089227421109904e-06,
-      "loss": 0.5368,
-      "step": 203
-    },
-    {
-      "epoch": 0.013332461930592772,
-      "grad_norm": 0.5455852150917053,
-      "learning_rate": 2.2198041349292712e-06,
-      "loss": 0.4925,
-      "step": 204
-    },
-    {
-      "epoch": 0.013397817136134893,
-      "grad_norm": 0.5337474942207336,
-      "learning_rate": 2.230685527747552e-06,
-      "loss": 0.4396,
-      "step": 205
-    },
-    {
-      "epoch": 0.013463172341677014,
-      "grad_norm": 0.5302766561508179,
-      "learning_rate": 2.2415669205658325e-06,
-      "loss": 0.4705,
-      "step": 206
-    },
-    {
-      "epoch": 0.013528527547219137,
-      "grad_norm": 0.5174290537834167,
-      "learning_rate": 2.2524483133841134e-06,
-      "loss": 0.4584,
-      "step": 207
-    },
-    {
-      "epoch": 0.013593882752761258,
-      "grad_norm": 0.545461893081665,
-      "learning_rate": 2.263329706202394e-06,
-      "loss": 0.5213,
-      "step": 208
-    },
-    {
-      "epoch": 0.013659237958303379,
-      "grad_norm": 0.5689192414283752,
-      "learning_rate": 2.2742110990206747e-06,
-      "loss": 0.5755,
-      "step": 209
-    },
-    {
-      "epoch": 0.0137245931638455,
-      "grad_norm": 0.558050274848938,
-      "learning_rate": 2.2850924918389556e-06,
-      "loss": 0.4973,
-      "step": 210
-    },
-    {
-      "epoch": 0.013789948369387621,
-      "grad_norm": 0.5459199547767639,
-      "learning_rate": 2.2959738846572364e-06,
-      "loss": 0.4719,
-      "step": 211
-    },
-    {
-      "epoch": 0.013855303574929744,
-      "grad_norm": 0.5807853937149048,
-      "learning_rate": 2.3068552774755173e-06,
-      "loss": 0.5219,
-      "step": 212
-    },
-    {
-      "epoch": 0.013920658780471865,
-      "grad_norm": 0.5296688675880432,
-      "learning_rate": 2.3177366702937977e-06,
-      "loss": 0.4791,
-      "step": 213
-    },
-    {
-      "epoch": 0.013986013986013986,
-      "grad_norm": 0.5390053391456604,
-      "learning_rate": 2.3286180631120786e-06,
-      "loss": 0.4881,
-      "step": 214
-    },
-    {
-      "epoch": 0.014051369191556107,
-      "grad_norm": 0.5217953324317932,
-      "learning_rate": 2.3394994559303594e-06,
-      "loss": 0.4584,
-      "step": 215
-    },
-    {
-      "epoch": 0.014116724397098228,
-      "grad_norm": 0.5571444034576416,
-      "learning_rate": 2.3503808487486403e-06,
-      "loss": 0.5203,
-      "step": 216
-    },
-    {
-      "epoch": 0.014182079602640351,
-      "grad_norm": 0.49948057532310486,
-      "learning_rate": 2.3612622415669207e-06,
-      "loss": 0.4676,
-      "step": 217
-    },
-    {
-      "epoch": 0.014247434808182472,
-      "grad_norm": 0.5300005674362183,
-      "learning_rate": 2.3721436343852016e-06,
-      "loss": 0.4667,
-      "step": 218
-    },
-    {
-      "epoch": 0.014312790013724593,
-      "grad_norm": 0.5889579653739929,
-      "learning_rate": 2.383025027203482e-06,
-      "loss": 0.4937,
-      "step": 219
-    },
-    {
-      "epoch": 0.014378145219266714,
-      "grad_norm": 0.5454444885253906,
-      "learning_rate": 2.393906420021763e-06,
-      "loss": 0.4882,
-      "step": 220
-    },
-    {
-      "epoch": 0.014443500424808835,
-      "grad_norm": 0.540638267993927,
-      "learning_rate": 2.4047878128400437e-06,
-      "loss": 0.4539,
-      "step": 221
-    },
-    {
-      "epoch": 0.014508855630350958,
-      "grad_norm": 0.5287627577781677,
-      "learning_rate": 2.4156692056583246e-06,
-      "loss": 0.5135,
-      "step": 222
-    },
-    {
-      "epoch": 0.01457421083589308,
-      "grad_norm": 0.4872298836708069,
-      "learning_rate": 2.426550598476605e-06,
-      "loss": 0.4097,
-      "step": 223
-    },
-    {
-      "epoch": 0.0146395660414352,
-      "grad_norm": 0.5071539878845215,
-      "learning_rate": 2.437431991294886e-06,
-      "loss": 0.4748,
-      "step": 224
-    },
-    {
-      "epoch": 0.014704921246977322,
-      "grad_norm": 0.5522286891937256,
-      "learning_rate": 2.4483133841131668e-06,
-      "loss": 0.51,
-      "step": 225
-    },
-    {
-      "epoch": 0.014770276452519443,
-      "grad_norm": 0.50446617603302,
-      "learning_rate": 2.4591947769314476e-06,
-      "loss": 0.4605,
-      "step": 226
-    },
-    {
-      "epoch": 0.014835631658061564,
-      "grad_norm": 0.55495685338974,
-      "learning_rate": 2.4700761697497285e-06,
-      "loss": 0.5133,
-      "step": 227
-    },
-    {
-      "epoch": 0.014900986863603687,
-      "grad_norm": 0.621311366558075,
-      "learning_rate": 2.480957562568009e-06,
-      "loss": 0.5095,
-      "step": 228
-    },
-    {
-      "epoch": 0.014966342069145808,
-      "grad_norm": 0.4835759401321411,
-      "learning_rate": 2.4918389553862898e-06,
-      "loss": 0.3879,
-      "step": 229
-    },
-    {
-      "epoch": 0.015031697274687929,
-      "grad_norm": 0.5162561535835266,
-      "learning_rate": 2.50272034820457e-06,
-      "loss": 0.4907,
-      "step": 230
-    },
-    {
-      "epoch": 0.01509705248023005,
-      "grad_norm": 0.5197626948356628,
-      "learning_rate": 2.5136017410228515e-06,
-      "loss": 0.4758,
-      "step": 231
-    },
-    {
-      "epoch": 0.015162407685772171,
-      "grad_norm": 0.5136451125144958,
-      "learning_rate": 2.524483133841132e-06,
-      "loss": 0.4617,
-      "step": 232
-    },
-    {
-      "epoch": 0.015227762891314294,
-      "grad_norm": 0.5694302320480347,
-      "learning_rate": 2.5353645266594124e-06,
-      "loss": 0.5283,
-      "step": 233
-    },
-    {
-      "epoch": 0.015293118096856415,
-      "grad_norm": 0.588736355304718,
-      "learning_rate": 2.5462459194776932e-06,
-      "loss": 0.5085,
-      "step": 234
-    },
-    {
-      "epoch": 0.015358473302398536,
-      "grad_norm": 0.5073676109313965,
-      "learning_rate": 2.557127312295974e-06,
-      "loss": 0.4521,
-      "step": 235
-    },
-    {
-      "epoch": 0.015423828507940657,
-      "grad_norm": 0.5521355867385864,
-      "learning_rate": 2.568008705114255e-06,
-      "loss": 0.4846,
-      "step": 236
-    },
-    {
-      "epoch": 0.015489183713482778,
-      "grad_norm": 0.5055086016654968,
-      "learning_rate": 2.5788900979325354e-06,
-      "loss": 0.4133,
-      "step": 237
-    },
-    {
-      "epoch": 0.015554538919024901,
-      "grad_norm": 0.5076500773429871,
-      "learning_rate": 2.5897714907508167e-06,
-      "loss": 0.4314,
-      "step": 238
-    },
-    {
-      "epoch": 0.015619894124567022,
-      "grad_norm": 0.512229859828949,
-      "learning_rate": 2.600652883569097e-06,
-      "loss": 0.4631,
-      "step": 239
-    },
-    {
-      "epoch": 0.015685249330109143,
-      "grad_norm": 0.5498212575912476,
-      "learning_rate": 2.611534276387378e-06,
-      "loss": 0.4705,
-      "step": 240
-    },
-    {
-      "epoch": 0.015750604535651264,
-      "grad_norm": 0.5373813509941101,
-      "learning_rate": 2.6224156692056584e-06,
-      "loss": 0.508,
-      "step": 241
-    },
-    {
-      "epoch": 0.015815959741193385,
-      "grad_norm": 0.5663868188858032,
-      "learning_rate": 2.6332970620239397e-06,
-      "loss": 0.5417,
-      "step": 242
-    },
-    {
-      "epoch": 0.015881314946735507,
-      "grad_norm": 0.4976261556148529,
-      "learning_rate": 2.64417845484222e-06,
-      "loss": 0.3833,
-      "step": 243
-    },
-    {
-      "epoch": 0.015946670152277628,
-      "grad_norm": 0.516413152217865,
-      "learning_rate": 2.6550598476605005e-06,
-      "loss": 0.4281,
-      "step": 244
-    },
-    {
-      "epoch": 0.01601202535781975,
-      "grad_norm": 0.5108800530433655,
-      "learning_rate": 2.6659412404787814e-06,
-      "loss": 0.4669,
-      "step": 245
-    },
-    {
-      "epoch": 0.016077380563361873,
-      "grad_norm": 0.517380952835083,
-      "learning_rate": 2.676822633297062e-06,
-      "loss": 0.4966,
-      "step": 246
-    },
-    {
-      "epoch": 0.016142735768903994,
-      "grad_norm": 0.5903849601745605,
-      "learning_rate": 2.687704026115343e-06,
-      "loss": 0.5843,
-      "step": 247
-    },
-    {
-      "epoch": 0.016208090974446115,
-      "grad_norm": 0.5836236476898193,
-      "learning_rate": 2.6985854189336236e-06,
-      "loss": 0.5589,
-      "step": 248
-    },
-    {
-      "epoch": 0.016273446179988237,
-      "grad_norm": 0.5302822589874268,
-      "learning_rate": 2.7094668117519044e-06,
-      "loss": 0.4774,
-      "step": 249
-    },
-    {
-      "epoch": 0.016338801385530358,
-      "grad_norm": 0.48977068066596985,
-      "learning_rate": 2.7203482045701853e-06,
-      "loss": 0.4425,
-      "step": 250
-    },
-    {
-      "epoch": 0.01640415659107248,
-      "grad_norm": 0.4973001778125763,
-      "learning_rate": 2.731229597388466e-06,
-      "loss": 0.4404,
-      "step": 251
-    },
-    {
-      "epoch": 0.0164695117966146,
-      "grad_norm": 0.49313315749168396,
-      "learning_rate": 2.7421109902067466e-06,
-      "loss": 0.4513,
-      "step": 252
-    },
-    {
-      "epoch": 0.01653486700215672,
-      "grad_norm": 0.5417724847793579,
-      "learning_rate": 2.752992383025028e-06,
-      "loss": 0.5055,
-      "step": 253
-    },
-    {
-      "epoch": 0.016600222207698842,
-      "grad_norm": 0.4942086935043335,
-      "learning_rate": 2.7638737758433083e-06,
-      "loss": 0.4373,
-      "step": 254
-    },
-    {
-      "epoch": 0.016665577413240963,
-      "grad_norm": 0.548223078250885,
-      "learning_rate": 2.7747551686615887e-06,
-      "loss": 0.5172,
-      "step": 255
-    },
-    {
-      "epoch": 0.016730932618783088,
-      "grad_norm": 0.5284595489501953,
-      "learning_rate": 2.7856365614798696e-06,
-      "loss": 0.4853,
-      "step": 256
-    },
-    {
-      "epoch": 0.01679628782432521,
-      "grad_norm": 0.4953025281429291,
-      "learning_rate": 2.79651795429815e-06,
-      "loss": 0.4207,
-      "step": 257
-    },
-    {
-      "epoch": 0.01686164302986733,
-      "grad_norm": 0.553777813911438,
-      "learning_rate": 2.8073993471164313e-06,
-      "loss": 0.4984,
-      "step": 258
-    },
-    {
-      "epoch": 0.01692699823540945,
-      "grad_norm": 0.588208019733429,
-      "learning_rate": 2.8182807399347118e-06,
-      "loss": 0.5379,
-      "step": 259
-    },
-    {
-      "epoch": 0.016992353440951572,
-      "grad_norm": 0.509104311466217,
-      "learning_rate": 2.8291621327529926e-06,
-      "loss": 0.4606,
-      "step": 260
-    },
-    {
-      "epoch": 0.017057708646493693,
-      "grad_norm": 0.5887159109115601,
-      "learning_rate": 2.8400435255712735e-06,
-      "loss": 0.4976,
-      "step": 261
-    },
-    {
-      "epoch": 0.017123063852035814,
-      "grad_norm": 0.49038127064704895,
-      "learning_rate": 2.8509249183895543e-06,
-      "loss": 0.4299,
-      "step": 262
-    },
-    {
-      "epoch": 0.017188419057577935,
-      "grad_norm": 0.5132138133049011,
-      "learning_rate": 2.8618063112078348e-06,
-      "loss": 0.4689,
-      "step": 263
-    },
-    {
-      "epoch": 0.017253774263120056,
-      "grad_norm": 0.4953780472278595,
-      "learning_rate": 2.872687704026116e-06,
-      "loss": 0.456,
-      "step": 264
-    },
-    {
-      "epoch": 0.017319129468662178,
-      "grad_norm": 0.5370753407478333,
-      "learning_rate": 2.8835690968443965e-06,
-      "loss": 0.5012,
-      "step": 265
-    },
-    {
-      "epoch": 0.0173844846742043,
-      "grad_norm": 0.5061358213424683,
-      "learning_rate": 2.894450489662677e-06,
-      "loss": 0.4574,
-      "step": 266
-    },
-    {
-      "epoch": 0.017449839879746423,
-      "grad_norm": 0.5424850583076477,
-      "learning_rate": 2.9053318824809578e-06,
-      "loss": 0.5142,
-      "step": 267
-    },
-    {
-      "epoch": 0.017515195085288544,
-      "grad_norm": 0.5463417172431946,
-      "learning_rate": 2.9162132752992382e-06,
-      "loss": 0.5358,
-      "step": 268
-    },
-    {
-      "epoch": 0.017580550290830665,
-      "grad_norm": 0.4903034567832947,
-      "learning_rate": 2.9270946681175195e-06,
-      "loss": 0.433,
-      "step": 269
-    },
-    {
-      "epoch": 0.017645905496372787,
-      "grad_norm": 0.6892983913421631,
-      "learning_rate": 2.9379760609358e-06,
-      "loss": 0.5141,
-      "step": 270
-    },
-    {
-      "epoch": 0.017711260701914908,
-      "grad_norm": 0.5312873721122742,
-      "learning_rate": 2.948857453754081e-06,
-      "loss": 0.5043,
-      "step": 271
-    },
-    {
-      "epoch": 0.01777661590745703,
-      "grad_norm": 0.5391185879707336,
-      "learning_rate": 2.9597388465723612e-06,
-      "loss": 0.4861,
-      "step": 272
-    },
-    {
-      "epoch": 0.01784197111299915,
-      "grad_norm": 0.5108627676963806,
-      "learning_rate": 2.9706202393906425e-06,
-      "loss": 0.5066,
-      "step": 273
-    },
-    {
-      "epoch": 0.01790732631854127,
-      "grad_norm": 0.5569199323654175,
-      "learning_rate": 2.981501632208923e-06,
-      "loss": 0.5139,
-      "step": 274
-    },
-    {
-      "epoch": 0.017972681524083392,
-      "grad_norm": 0.5281971096992493,
-      "learning_rate": 2.9923830250272034e-06,
-      "loss": 0.4751,
-      "step": 275
-    },
-    {
-      "epoch": 0.018038036729625513,
-      "grad_norm": 0.538384735584259,
-      "learning_rate": 3.0032644178454847e-06,
-      "loss": 0.495,
-      "step": 276
-    },
-    {
-      "epoch": 0.018103391935167638,
-      "grad_norm": 0.5089460611343384,
-      "learning_rate": 3.014145810663765e-06,
-      "loss": 0.4497,
-      "step": 277
-    },
-    {
-      "epoch": 0.01816874714070976,
-      "grad_norm": 0.5131743550300598,
-      "learning_rate": 3.025027203482046e-06,
-      "loss": 0.3966,
-      "step": 278
-    },
-    {
-      "epoch": 0.01823410234625188,
-      "grad_norm": 0.5346877574920654,
-      "learning_rate": 3.0359085963003264e-06,
-      "loss": 0.4881,
-      "step": 279
-    },
-    {
-      "epoch": 0.018299457551794,
-      "grad_norm": 0.5207666158676147,
-      "learning_rate": 3.0467899891186077e-06,
-      "loss": 0.4874,
-      "step": 280
-    },
-    {
-      "epoch": 0.018364812757336122,
-      "grad_norm": 0.4716075658798218,
-      "learning_rate": 3.057671381936888e-06,
-      "loss": 0.3783,
-      "step": 281
-    },
-    {
-      "epoch": 0.018430167962878243,
-      "grad_norm": 0.5745481848716736,
-      "learning_rate": 3.068552774755169e-06,
-      "loss": 0.5241,
-      "step": 282
-    },
-    {
-      "epoch": 0.018495523168420364,
-      "grad_norm": 0.5004214644432068,
-      "learning_rate": 3.0794341675734494e-06,
-      "loss": 0.4256,
-      "step": 283
-    },
-    {
-      "epoch": 0.018560878373962485,
-      "grad_norm": 0.4955251216888428,
-      "learning_rate": 3.0903155603917307e-06,
-      "loss": 0.4184,
-      "step": 284
-    },
-    {
-      "epoch": 0.018626233579504606,
-      "grad_norm": 0.48548418283462524,
-      "learning_rate": 3.101196953210011e-06,
-      "loss": 0.4214,
-      "step": 285
-    },
-    {
-      "epoch": 0.018691588785046728,
-      "grad_norm": 0.5416601896286011,
-      "learning_rate": 3.1120783460282916e-06,
-      "loss": 0.5082,
-      "step": 286
-    },
-    {
-      "epoch": 0.01875694399058885,
-      "grad_norm": 0.6502732634544373,
-      "learning_rate": 3.122959738846573e-06,
-      "loss": 0.6006,
-      "step": 287
-    },
-    {
-      "epoch": 0.018822299196130973,
-      "grad_norm": 0.5318127870559692,
-      "learning_rate": 3.1338411316648533e-06,
-      "loss": 0.4629,
-      "step": 288
-    },
-    {
-      "epoch": 0.018887654401673094,
-      "grad_norm": 0.5368547439575195,
-      "learning_rate": 3.144722524483134e-06,
-      "loss": 0.4913,
-      "step": 289
-    },
-    {
-      "epoch": 0.018953009607215215,
-      "grad_norm": 0.5342006087303162,
-      "learning_rate": 3.1556039173014146e-06,
-      "loss": 0.5098,
-      "step": 290
-    },
-    {
-      "epoch": 0.019018364812757337,
-      "grad_norm": 0.5251221060752869,
-      "learning_rate": 3.166485310119696e-06,
-      "loss": 0.4688,
-      "step": 291
-    },
-    {
-      "epoch": 0.019083720018299458,
-      "grad_norm": 0.5649088621139526,
-      "learning_rate": 3.1773667029379763e-06,
-      "loss": 0.4969,
-      "step": 292
-    },
-    {
-      "epoch": 0.01914907522384158,
-      "grad_norm": 0.5344524383544922,
-      "learning_rate": 3.188248095756257e-06,
-      "loss": 0.5144,
-      "step": 293
-    },
-    {
-      "epoch": 0.0192144304293837,
-      "grad_norm": 0.5382475852966309,
-      "learning_rate": 3.1991294885745376e-06,
-      "loss": 0.4812,
-      "step": 294
-    },
-    {
-      "epoch": 0.01927978563492582,
-      "grad_norm": 0.5152425765991211,
-      "learning_rate": 3.210010881392819e-06,
-      "loss": 0.4366,
-      "step": 295
-    },
-    {
-      "epoch": 0.019345140840467942,
-      "grad_norm": 0.511518657207489,
-      "learning_rate": 3.2208922742110993e-06,
-      "loss": 0.45,
-      "step": 296
-    },
-    {
-      "epoch": 0.019410496046010063,
-      "grad_norm": 0.6026496291160583,
-      "learning_rate": 3.2317736670293798e-06,
-      "loss": 0.5255,
-      "step": 297
-    },
-    {
-      "epoch": 0.019475851251552188,
-      "grad_norm": 0.5703912973403931,
-      "learning_rate": 3.2426550598476606e-06,
-      "loss": 0.5176,
-      "step": 298
-    },
-    {
-      "epoch": 0.01954120645709431,
-      "grad_norm": 0.5228263735771179,
-      "learning_rate": 3.2535364526659415e-06,
-      "loss": 0.4564,
-      "step": 299
-    },
-    {
-      "epoch": 0.01960656166263643,
-      "grad_norm": 0.5298845171928406,
-      "learning_rate": 3.2644178454842223e-06,
-      "loss": 0.4319,
-      "step": 300
-    },
-    {
-      "epoch": 0.01967191686817855,
-      "grad_norm": 0.5105089545249939,
-      "learning_rate": 3.2752992383025028e-06,
-      "loss": 0.4404,
-      "step": 301
-    },
-    {
-      "epoch": 0.019737272073720672,
-      "grad_norm": 0.5126194357872009,
-      "learning_rate": 3.286180631120784e-06,
-      "loss": 0.4531,
-      "step": 302
-    },
-    {
-      "epoch": 0.019802627279262793,
-      "grad_norm": 0.5196699500083923,
-      "learning_rate": 3.2970620239390645e-06,
-      "loss": 0.4621,
-      "step": 303
-    },
-    {
-      "epoch": 0.019867982484804914,
-      "grad_norm": 0.5301274657249451,
-      "learning_rate": 3.3079434167573454e-06,
-      "loss": 0.4821,
-      "step": 304
-    },
-    {
-      "epoch": 0.019933337690347035,
-      "grad_norm": 0.5426238775253296,
-      "learning_rate": 3.318824809575626e-06,
-      "loss": 0.4118,
-      "step": 305
-    },
-    {
-      "epoch": 0.019998692895889156,
-      "grad_norm": 0.602854311466217,
-      "learning_rate": 3.329706202393907e-06,
-      "loss": 0.5386,
-      "step": 306
-    },
-    {
-      "epoch": 0.020064048101431278,
-      "grad_norm": 0.5767044425010681,
-      "learning_rate": 3.3405875952121875e-06,
-      "loss": 0.5179,
-      "step": 307
-    },
-    {
-      "epoch": 0.020129403306973402,
-      "grad_norm": 0.5042062997817993,
-      "learning_rate": 3.351468988030468e-06,
-      "loss": 0.4432,
-      "step": 308
-    },
-    {
-      "epoch": 0.020194758512515523,
-      "grad_norm": 0.524115800857544,
-      "learning_rate": 3.362350380848749e-06,
-      "loss": 0.4366,
-      "step": 309
-    },
-    {
-      "epoch": 0.020260113718057644,
-      "grad_norm": 0.5185021758079529,
-      "learning_rate": 3.3732317736670297e-06,
-      "loss": 0.4283,
-      "step": 310
-    },
-    {
-      "epoch": 0.020325468923599765,
-      "grad_norm": 0.5096868872642517,
-      "learning_rate": 3.3841131664853105e-06,
-      "loss": 0.456,
-      "step": 311
-    },
-    {
-      "epoch": 0.020390824129141887,
-      "grad_norm": 0.513860821723938,
-      "learning_rate": 3.394994559303591e-06,
-      "loss": 0.4377,
-      "step": 312
-    },
-    {
-      "epoch": 0.020456179334684008,
-      "grad_norm": 0.5447676777839661,
-      "learning_rate": 3.4058759521218722e-06,
-      "loss": 0.4969,
-      "step": 313
-    },
-    {
-      "epoch": 0.02052153454022613,
-      "grad_norm": 0.5675314664840698,
-      "learning_rate": 3.4167573449401527e-06,
-      "loss": 0.4798,
-      "step": 314
-    },
-    {
-      "epoch": 0.02058688974576825,
-      "grad_norm": 0.5476747155189514,
-      "learning_rate": 3.4276387377584335e-06,
-      "loss": 0.473,
-      "step": 315
-    },
-    {
-      "epoch": 0.02065224495131037,
-      "grad_norm": 0.5846548676490784,
-      "learning_rate": 3.438520130576714e-06,
-      "loss": 0.5253,
-      "step": 316
-    },
-    {
-      "epoch": 0.020717600156852492,
-      "grad_norm": 0.5542411804199219,
-      "learning_rate": 3.4494015233949944e-06,
-      "loss": 0.4818,
-      "step": 317
-    },
-    {
-      "epoch": 0.020782955362394613,
-      "grad_norm": 0.48111483454704285,
-      "learning_rate": 3.4602829162132757e-06,
-      "loss": 0.4281,
-      "step": 318
-    },
-    {
-      "epoch": 0.020848310567936738,
-      "grad_norm": 0.5389544367790222,
-      "learning_rate": 3.471164309031556e-06,
-      "loss": 0.4577,
-      "step": 319
-    },
-    {
-      "epoch": 0.02091366577347886,
-      "grad_norm": 0.5208144783973694,
-      "learning_rate": 3.482045701849837e-06,
-      "loss": 0.3836,
-      "step": 320
-    },
-    {
-      "epoch": 0.02097902097902098,
-      "grad_norm": 0.56230229139328,
-      "learning_rate": 3.4929270946681174e-06,
-      "loss": 0.4785,
-      "step": 321
-    },
-    {
-      "epoch": 0.0210443761845631,
-      "grad_norm": 0.5516102313995361,
-      "learning_rate": 3.5038084874863987e-06,
-      "loss": 0.4941,
-      "step": 322
-    },
-    {
-      "epoch": 0.021109731390105222,
-      "grad_norm": 0.561265766620636,
-      "learning_rate": 3.514689880304679e-06,
-      "loss": 0.4969,
-      "step": 323
-    },
-    {
-      "epoch": 0.021175086595647343,
-      "grad_norm": 0.5460501909255981,
-      "learning_rate": 3.52557127312296e-06,
-      "loss": 0.4945,
-      "step": 324
-    },
-    {
-      "epoch": 0.021240441801189464,
-      "grad_norm": 0.6526318192481995,
-      "learning_rate": 3.536452665941241e-06,
-      "loss": 0.49,
-      "step": 325
-    },
-    {
-      "epoch": 0.021305797006731585,
-      "grad_norm": 0.5364881157875061,
-      "learning_rate": 3.5473340587595217e-06,
-      "loss": 0.4776,
-      "step": 326
-    },
-    {
-      "epoch": 0.021371152212273706,
-      "grad_norm": 0.5304608941078186,
-      "learning_rate": 3.558215451577802e-06,
-      "loss": 0.4669,
-      "step": 327
-    },
-    {
-      "epoch": 0.021436507417815828,
-      "grad_norm": 0.5698348879814148,
-      "learning_rate": 3.5690968443960826e-06,
-      "loss": 0.4828,
-      "step": 328
-    },
-    {
-      "epoch": 0.021501862623357952,
-      "grad_norm": 0.5477334260940552,
-      "learning_rate": 3.579978237214364e-06,
-      "loss": 0.4854,
-      "step": 329
-    },
-    {
-      "epoch": 0.021567217828900073,
-      "grad_norm": 0.5482262969017029,
-      "learning_rate": 3.5908596300326443e-06,
-      "loss": 0.4708,
-      "step": 330
-    },
-    {
-      "epoch": 0.021632573034442194,
-      "grad_norm": 0.7397144436836243,
-      "learning_rate": 3.601741022850925e-06,
-      "loss": 0.5223,
-      "step": 331
-    },
-    {
-      "epoch": 0.021697928239984315,
-      "grad_norm": 0.5388163328170776,
-      "learning_rate": 3.6126224156692056e-06,
-      "loss": 0.4929,
-      "step": 332
-    },
-    {
-      "epoch": 0.021763283445526437,
-      "grad_norm": 0.511736273765564,
-      "learning_rate": 3.623503808487487e-06,
-      "loss": 0.4483,
-      "step": 333
-    },
-    {
-      "epoch": 0.021828638651068558,
-      "grad_norm": 0.5120736360549927,
-      "learning_rate": 3.6343852013057673e-06,
-      "loss": 0.46,
-      "step": 334
-    },
-    {
-      "epoch": 0.02189399385661068,
-      "grad_norm": 0.5173547863960266,
-      "learning_rate": 3.645266594124048e-06,
-      "loss": 0.4578,
-      "step": 335
-    },
-    {
-      "epoch": 0.0219593490621528,
-      "grad_norm": 0.5703136324882507,
-      "learning_rate": 3.656147986942329e-06,
-      "loss": 0.4723,
-      "step": 336
-    },
-    {
-      "epoch": 0.02202470426769492,
-      "grad_norm": 0.5174160003662109,
-      "learning_rate": 3.66702937976061e-06,
-      "loss": 0.4545,
-      "step": 337
-    },
-    {
-      "epoch": 0.022090059473237042,
-      "grad_norm": 0.5286480188369751,
-      "learning_rate": 3.6779107725788904e-06,
-      "loss": 0.4637,
-      "step": 338
-    },
-    {
-      "epoch": 0.022155414678779163,
-      "grad_norm": 0.5682716369628906,
-      "learning_rate": 3.688792165397171e-06,
-      "loss": 0.5242,
-      "step": 339
-    },
-    {
-      "epoch": 0.022220769884321288,
-      "grad_norm": 0.5390458106994629,
-      "learning_rate": 3.699673558215452e-06,
-      "loss": 0.4419,
-      "step": 340
-    },
-    {
-      "epoch": 0.02228612508986341,
-      "grad_norm": 0.541253387928009,
-      "learning_rate": 3.7105549510337325e-06,
-      "loss": 0.4825,
-      "step": 341
-    },
-    {
-      "epoch": 0.02235148029540553,
-      "grad_norm": 0.49162402749061584,
-      "learning_rate": 3.7214363438520134e-06,
-      "loss": 0.4254,
-      "step": 342
-    },
-    {
-      "epoch": 0.02241683550094765,
-      "grad_norm": 0.5324705839157104,
-      "learning_rate": 3.732317736670294e-06,
-      "loss": 0.5062,
-      "step": 343
-    },
-    {
-      "epoch": 0.022482190706489772,
-      "grad_norm": 0.5231584310531616,
-      "learning_rate": 3.743199129488575e-06,
-      "loss": 0.4707,
-      "step": 344
-    },
-    {
-      "epoch": 0.022547545912031893,
-      "grad_norm": 0.53677898645401,
-      "learning_rate": 3.7540805223068555e-06,
-      "loss": 0.4996,
-      "step": 345
-    },
-    {
-      "epoch": 0.022612901117574014,
-      "grad_norm": 0.5236653089523315,
-      "learning_rate": 3.7649619151251364e-06,
-      "loss": 0.4294,
-      "step": 346
-    },
-    {
-      "epoch": 0.022678256323116135,
-      "grad_norm": 0.5420531630516052,
-      "learning_rate": 3.775843307943417e-06,
-      "loss": 0.5134,
-      "step": 347
-    },
-    {
-      "epoch": 0.022743611528658256,
-      "grad_norm": 0.5473664999008179,
-      "learning_rate": 3.786724700761698e-06,
-      "loss": 0.4731,
-      "step": 348
-    },
-    {
-      "epoch": 0.022808966734200378,
-      "grad_norm": 0.5433962941169739,
-      "learning_rate": 3.7976060935799785e-06,
-      "loss": 0.5018,
-      "step": 349
-    },
-    {
-      "epoch": 0.022874321939742502,
-      "grad_norm": 0.5301963090896606,
-      "learning_rate": 3.808487486398259e-06,
-      "loss": 0.4671,
-      "step": 350
-    },
-    {
-      "epoch": 0.022939677145284623,
-      "grad_norm": 0.5100705623626709,
-      "learning_rate": 3.81936887921654e-06,
-      "loss": 0.4444,
-      "step": 351
-    },
-    {
-      "epoch": 0.023005032350826744,
-      "grad_norm": 0.5468143224716187,
-      "learning_rate": 3.83025027203482e-06,
-      "loss": 0.5,
-      "step": 352
-    },
-    {
-      "epoch": 0.023070387556368865,
-      "grad_norm": 0.5839720368385315,
-      "learning_rate": 3.841131664853102e-06,
-      "loss": 0.4728,
-      "step": 353
-    },
-    {
-      "epoch": 0.023135742761910986,
-      "grad_norm": 0.5293095707893372,
-      "learning_rate": 3.852013057671382e-06,
-      "loss": 0.47,
-      "step": 354
-    },
-    {
-      "epoch": 0.023201097967453108,
-      "grad_norm": 0.5732521414756775,
-      "learning_rate": 3.862894450489663e-06,
-      "loss": 0.5369,
-      "step": 355
-    },
-    {
-      "epoch": 0.02326645317299523,
-      "grad_norm": 0.5569798946380615,
-      "learning_rate": 3.873775843307944e-06,
-      "loss": 0.5178,
-      "step": 356
-    },
-    {
-      "epoch": 0.02333180837853735,
-      "grad_norm": 0.5082626938819885,
-      "learning_rate": 3.8846572361262246e-06,
-      "loss": 0.4672,
-      "step": 357
-    },
-    {
-      "epoch": 0.02339716358407947,
-      "grad_norm": 0.5630055665969849,
-      "learning_rate": 3.8955386289445054e-06,
-      "loss": 0.4553,
-      "step": 358
-    },
-    {
-      "epoch": 0.023462518789621592,
-      "grad_norm": 0.4950932562351227,
-      "learning_rate": 3.906420021762786e-06,
-      "loss": 0.4705,
-      "step": 359
-    },
-    {
-      "epoch": 0.023527873995163713,
-      "grad_norm": 0.5586927533149719,
-      "learning_rate": 3.917301414581066e-06,
-      "loss": 0.5097,
-      "step": 360
-    },
-    {
-      "epoch": 0.023593229200705838,
-      "grad_norm": 0.5062628388404846,
-      "learning_rate": 3.928182807399347e-06,
-      "loss": 0.4233,
-      "step": 361
-    },
-    {
-      "epoch": 0.02365858440624796,
-      "grad_norm": 0.5195222496986389,
-      "learning_rate": 3.939064200217628e-06,
-      "loss": 0.4026,
-      "step": 362
-    },
-    {
-      "epoch": 0.02372393961179008,
-      "grad_norm": 0.5128687620162964,
-      "learning_rate": 3.949945593035909e-06,
-      "loss": 0.4778,
-      "step": 363
-    },
-    {
-      "epoch": 0.0237892948173322,
-      "grad_norm": 0.5465303063392639,
-      "learning_rate": 3.96082698585419e-06,
-      "loss": 0.4323,
-      "step": 364
-    },
-    {
-      "epoch": 0.023854650022874322,
-      "grad_norm": 0.544765055179596,
-      "learning_rate": 3.971708378672471e-06,
-      "loss": 0.4914,
-      "step": 365
-    },
-    {
-      "epoch": 0.023920005228416443,
-      "grad_norm": 0.5569670796394348,
-      "learning_rate": 3.9825897714907515e-06,
-      "loss": 0.4991,
-      "step": 366
-    },
-    {
-      "epoch": 0.023985360433958564,
-      "grad_norm": 0.5170602798461914,
-      "learning_rate": 3.9934711643090315e-06,
-      "loss": 0.466,
-      "step": 367
-    },
-    {
-      "epoch": 0.024050715639500685,
-      "grad_norm": 0.7524043321609497,
-      "learning_rate": 4.004352557127313e-06,
-      "loss": 0.4417,
-      "step": 368
-    },
-    {
-      "epoch": 0.024116070845042806,
-      "grad_norm": 0.5501721501350403,
-      "learning_rate": 4.015233949945593e-06,
-      "loss": 0.4684,
-      "step": 369
-    },
-    {
-      "epoch": 0.024181426050584928,
-      "grad_norm": 0.566725492477417,
-      "learning_rate": 4.026115342763874e-06,
-      "loss": 0.552,
-      "step": 370
-    },
-    {
-      "epoch": 0.024246781256127052,
-      "grad_norm": 0.563490629196167,
-      "learning_rate": 4.036996735582155e-06,
-      "loss": 0.4769,
-      "step": 371
-    },
-    {
-      "epoch": 0.024312136461669173,
-      "grad_norm": 0.5314151048660278,
-      "learning_rate": 4.047878128400436e-06,
-      "loss": 0.4471,
-      "step": 372
-    },
-    {
-      "epoch": 0.024377491667211294,
-      "grad_norm": 0.5828354954719543,
-      "learning_rate": 4.058759521218717e-06,
-      "loss": 0.4987,
-      "step": 373
-    },
-    {
-      "epoch": 0.024442846872753415,
-      "grad_norm": 0.5326939225196838,
-      "learning_rate": 4.069640914036997e-06,
-      "loss": 0.4299,
-      "step": 374
-    },
-    {
-      "epoch": 0.024508202078295536,
-      "grad_norm": 0.5667662024497986,
-      "learning_rate": 4.080522306855278e-06,
-      "loss": 0.4939,
-      "step": 375
-    },
-    {
-      "epoch": 0.024573557283837658,
-      "grad_norm": 0.5117393136024475,
-      "learning_rate": 4.091403699673558e-06,
-      "loss": 0.441,
-      "step": 376
-    },
-    {
-      "epoch": 0.02463891248937978,
-      "grad_norm": 0.5950685143470764,
-      "learning_rate": 4.102285092491839e-06,
-      "loss": 0.5514,
-      "step": 377
-    },
-    {
-      "epoch": 0.0247042676949219,
-      "grad_norm": 0.5770097970962524,
-      "learning_rate": 4.11316648531012e-06,
-      "loss": 0.493,
-      "step": 378
-    },
-    {
-      "epoch": 0.02476962290046402,
-      "grad_norm": 0.5505401492118835,
-      "learning_rate": 4.124047878128401e-06,
-      "loss": 0.4858,
-      "step": 379
-    },
-    {
-      "epoch": 0.024834978106006142,
-      "grad_norm": 0.48124557733535767,
-      "learning_rate": 4.134929270946682e-06,
-      "loss": 0.3715,
-      "step": 380
-    },
-    {
-      "epoch": 0.024900333311548267,
-      "grad_norm": 0.5477302074432373,
-      "learning_rate": 4.145810663764962e-06,
-      "loss": 0.4675,
-      "step": 381
-    },
-    {
-      "epoch": 0.024965688517090388,
-      "grad_norm": 0.5690323710441589,
-      "learning_rate": 4.156692056583243e-06,
-      "loss": 0.5179,
-      "step": 382
-    },
-    {
-      "epoch": 0.02503104372263251,
-      "grad_norm": 0.5915126800537109,
-      "learning_rate": 4.1675734494015235e-06,
-      "loss": 0.5116,
-      "step": 383
-    },
-    {
-      "epoch": 0.02509639892817463,
-      "grad_norm": 0.5344145894050598,
-      "learning_rate": 4.178454842219804e-06,
-      "loss": 0.4738,
-      "step": 384
-    },
-    {
-      "epoch": 0.02516175413371675,
-      "grad_norm": 0.5669682025909424,
-      "learning_rate": 4.189336235038085e-06,
-      "loss": 0.521,
-      "step": 385
-    },
-    {
-      "epoch": 0.025227109339258872,
-      "grad_norm": 0.5806385278701782,
-      "learning_rate": 4.200217627856366e-06,
-      "loss": 0.4906,
-      "step": 386
-    },
-    {
-      "epoch": 0.025292464544800993,
-      "grad_norm": 0.5490126609802246,
-      "learning_rate": 4.211099020674647e-06,
-      "loss": 0.4858,
-      "step": 387
-    },
-    {
-      "epoch": 0.025357819750343114,
-      "grad_norm": 0.5345961451530457,
-      "learning_rate": 4.221980413492928e-06,
-      "loss": 0.4458,
-      "step": 388
-    },
-    {
-      "epoch": 0.025423174955885235,
-      "grad_norm": 0.5605209469795227,
-      "learning_rate": 4.232861806311208e-06,
-      "loss": 0.4524,
-      "step": 389
-    },
-    {
-      "epoch": 0.025488530161427356,
-      "grad_norm": 0.538071870803833,
-      "learning_rate": 4.2437431991294896e-06,
-      "loss": 0.4865,
-      "step": 390
-    },
-    {
-      "epoch": 0.025553885366969478,
-      "grad_norm": 0.5728001594543457,
-      "learning_rate": 4.2546245919477696e-06,
-      "loss": 0.483,
-      "step": 391
-    },
-    {
-      "epoch": 0.025619240572511602,
-      "grad_norm": 0.5253920555114746,
-      "learning_rate": 4.2655059847660504e-06,
-      "loss": 0.4832,
-      "step": 392
-    },
-    {
-      "epoch": 0.025684595778053723,
-      "grad_norm": 0.6155691146850586,
-      "learning_rate": 4.276387377584331e-06,
-      "loss": 0.5437,
-      "step": 393
-    },
-    {
-      "epoch": 0.025749950983595844,
-      "grad_norm": 0.5745170712471008,
-      "learning_rate": 4.287268770402611e-06,
-      "loss": 0.5444,
-      "step": 394
-    },
-    {
-      "epoch": 0.025815306189137965,
-      "grad_norm": 0.5143162608146667,
-      "learning_rate": 4.298150163220893e-06,
-      "loss": 0.3949,
-      "step": 395
-    },
-    {
-      "epoch": 0.025880661394680086,
-      "grad_norm": 0.5189085602760315,
-      "learning_rate": 4.309031556039173e-06,
-      "loss": 0.4346,
-      "step": 396
-    },
-    {
-      "epoch": 0.025946016600222208,
-      "grad_norm": 0.5619039535522461,
-      "learning_rate": 4.319912948857454e-06,
-      "loss": 0.4984,
-      "step": 397
-    },
-    {
-      "epoch": 0.02601137180576433,
-      "grad_norm": 0.5270082950592041,
-      "learning_rate": 4.330794341675735e-06,
-      "loss": 0.4672,
-      "step": 398
-    },
-    {
-      "epoch": 0.02607672701130645,
-      "grad_norm": 0.5133958458900452,
-      "learning_rate": 4.341675734494016e-06,
-      "loss": 0.4689,
-      "step": 399
-    },
-    {
-      "epoch": 0.02614208221684857,
-      "grad_norm": 0.5218203663825989,
-      "learning_rate": 4.3525571273122965e-06,
-      "loss": 0.4577,
-      "step": 400
-    },
-    {
-      "epoch": 0.026207437422390692,
-      "grad_norm": 0.5627186894416809,
-      "learning_rate": 4.363438520130577e-06,
-      "loss": 0.4454,
-      "step": 401
-    },
-    {
-      "epoch": 0.026272792627932817,
-      "grad_norm": 0.585961639881134,
-      "learning_rate": 4.374319912948858e-06,
-      "loss": 0.4478,
-      "step": 402
-    },
-    {
-      "epoch": 0.026338147833474938,
-      "grad_norm": 0.5784450769424438,
-      "learning_rate": 4.385201305767138e-06,
-      "loss": 0.4511,
-      "step": 403
-    },
-    {
-      "epoch": 0.02640350303901706,
-      "grad_norm": 0.6283280253410339,
-      "learning_rate": 4.396082698585419e-06,
-      "loss": 0.5295,
-      "step": 404
-    },
-    {
-      "epoch": 0.02646885824455918,
-      "grad_norm": 0.5436939001083374,
-      "learning_rate": 4.4069640914037e-06,
-      "loss": 0.4292,
-      "step": 405
-    },
-    {
-      "epoch": 0.0265342134501013,
-      "grad_norm": 0.5761511921882629,
-      "learning_rate": 4.417845484221981e-06,
-      "loss": 0.442,
-      "step": 406
-    },
-    {
-      "epoch": 0.026599568655643422,
-      "grad_norm": 0.5336189866065979,
-      "learning_rate": 4.428726877040262e-06,
-      "loss": 0.4625,
-      "step": 407
-    },
-    {
-      "epoch": 0.026664923861185543,
-      "grad_norm": 0.5750748515129089,
-      "learning_rate": 4.4396082698585425e-06,
-      "loss": 0.5365,
-      "step": 408
-    },
-    {
-      "epoch": 0.026730279066727664,
-      "grad_norm": 0.5452337265014648,
-      "learning_rate": 4.4504896626768225e-06,
-      "loss": 0.4606,
-      "step": 409
-    },
-    {
-      "epoch": 0.026795634272269785,
-      "grad_norm": 0.5602395534515381,
-      "learning_rate": 4.461371055495104e-06,
-      "loss": 0.4824,
-      "step": 410
-    },
-    {
-      "epoch": 0.026860989477811906,
-      "grad_norm": 0.5694287419319153,
-      "learning_rate": 4.472252448313384e-06,
-      "loss": 0.4904,
-      "step": 411
-    },
-    {
-      "epoch": 0.026926344683354028,
-      "grad_norm": 0.5007390975952148,
-      "learning_rate": 4.483133841131665e-06,
-      "loss": 0.4129,
-      "step": 412
-    },
-    {
-      "epoch": 0.026991699888896152,
-      "grad_norm": 0.4548834562301636,
-      "learning_rate": 4.494015233949946e-06,
-      "loss": 0.3564,
-      "step": 413
-    },
-    {
-      "epoch": 0.027057055094438273,
-      "grad_norm": 0.4879645109176636,
-      "learning_rate": 4.504896626768227e-06,
-      "loss": 0.4316,
-      "step": 414
-    },
-    {
-      "epoch": 0.027122410299980394,
-      "grad_norm": 0.5681940913200378,
-      "learning_rate": 4.515778019586508e-06,
-      "loss": 0.4761,
-      "step": 415
-    },
-    {
-      "epoch": 0.027187765505522515,
-      "grad_norm": 0.47913455963134766,
-      "learning_rate": 4.526659412404788e-06,
-      "loss": 0.3899,
-      "step": 416
-    },
-    {
-      "epoch": 0.027253120711064636,
-      "grad_norm": 0.5863394737243652,
-      "learning_rate": 4.537540805223069e-06,
-      "loss": 0.5088,
-      "step": 417
-    },
-    {
-      "epoch": 0.027318475916606758,
-      "grad_norm": 0.5107646584510803,
-      "learning_rate": 4.548422198041349e-06,
-      "loss": 0.3999,
-      "step": 418
-    },
-    {
-      "epoch": 0.02738383112214888,
-      "grad_norm": 0.5345116853713989,
-      "learning_rate": 4.55930359085963e-06,
-      "loss": 0.5046,
-      "step": 419
-    },
-    {
-      "epoch": 0.027449186327691,
-      "grad_norm": 0.5197203755378723,
-      "learning_rate": 4.570184983677911e-06,
-      "loss": 0.4445,
-      "step": 420
-    },
-    {
-      "epoch": 0.02751454153323312,
-      "grad_norm": 0.5551862716674805,
-      "learning_rate": 4.581066376496192e-06,
-      "loss": 0.4799,
-      "step": 421
-    },
-    {
-      "epoch": 0.027579896738775242,
-      "grad_norm": 0.553368330001831,
-      "learning_rate": 4.591947769314473e-06,
-      "loss": 0.4877,
-      "step": 422
-    },
-    {
-      "epoch": 0.027645251944317367,
-      "grad_norm": 0.5523605346679688,
-      "learning_rate": 4.602829162132753e-06,
-      "loss": 0.465,
-      "step": 423
-    },
-    {
-      "epoch": 0.027710607149859488,
-      "grad_norm": 1.0156335830688477,
-      "learning_rate": 4.6137105549510345e-06,
-      "loss": 0.4935,
-      "step": 424
-    },
-    {
-      "epoch": 0.02777596235540161,
-      "grad_norm": 0.5236708521842957,
-      "learning_rate": 4.6245919477693146e-06,
-      "loss": 0.4467,
-      "step": 425
-    },
-    {
-      "epoch": 0.02784131756094373,
-      "grad_norm": 0.5503356456756592,
-      "learning_rate": 4.635473340587595e-06,
-      "loss": 0.5303,
-      "step": 426
-    },
-    {
-      "epoch": 0.02790667276648585,
-      "grad_norm": 0.5287750363349915,
-      "learning_rate": 4.646354733405876e-06,
-      "loss": 0.5106,
-      "step": 427
-    },
-    {
-      "epoch": 0.027972027972027972,
-      "grad_norm": 0.5276519060134888,
-      "learning_rate": 4.657236126224157e-06,
-      "loss": 0.4274,
-      "step": 428
-    },
-    {
-      "epoch": 0.028037383177570093,
-      "grad_norm": 0.5391794443130493,
-      "learning_rate": 4.668117519042438e-06,
-      "loss": 0.4665,
-      "step": 429
-    },
-    {
-      "epoch": 0.028102738383112214,
-      "grad_norm": 0.5067090392112732,
-      "learning_rate": 4.678998911860719e-06,
-      "loss": 0.4665,
-      "step": 430
-    },
-    {
-      "epoch": 0.028168093588654335,
-      "grad_norm": 0.553927481174469,
-      "learning_rate": 4.689880304678999e-06,
-      "loss": 0.4922,
-      "step": 431
-    },
-    {
-      "epoch": 0.028233448794196456,
-      "grad_norm": 0.5183501839637756,
-      "learning_rate": 4.700761697497281e-06,
-      "loss": 0.4454,
-      "step": 432
-    },
-    {
-      "epoch": 0.028298803999738577,
-      "grad_norm": 0.576321005821228,
-      "learning_rate": 4.711643090315561e-06,
-      "loss": 0.5218,
-      "step": 433
-    },
-    {
-      "epoch": 0.028364159205280702,
-      "grad_norm": 0.4934738278388977,
-      "learning_rate": 4.7225244831338415e-06,
-      "loss": 0.4204,
-      "step": 434
-    },
-    {
-      "epoch": 0.028429514410822823,
-      "grad_norm": 0.5205022096633911,
-      "learning_rate": 4.733405875952122e-06,
-      "loss": 0.4427,
-      "step": 435
-    },
-    {
-      "epoch": 0.028494869616364944,
-      "grad_norm": 0.486479789018631,
-      "learning_rate": 4.744287268770403e-06,
-      "loss": 0.395,
-      "step": 436
-    },
-    {
-      "epoch": 0.028560224821907065,
-      "grad_norm": 0.5366991758346558,
-      "learning_rate": 4.755168661588684e-06,
-      "loss": 0.459,
-      "step": 437
-    },
-    {
-      "epoch": 0.028625580027449186,
-      "grad_norm": 0.5452165007591248,
-      "learning_rate": 4.766050054406964e-06,
-      "loss": 0.5097,
-      "step": 438
-    },
-    {
-      "epoch": 0.028690935232991308,
-      "grad_norm": 0.4924579858779907,
-      "learning_rate": 4.776931447225246e-06,
-      "loss": 0.4046,
-      "step": 439
-    },
-    {
-      "epoch": 0.02875629043853343,
-      "grad_norm": 0.5535774827003479,
-      "learning_rate": 4.787812840043526e-06,
-      "loss": 0.438,
-      "step": 440
-    },
-    {
-      "epoch": 0.02882164564407555,
-      "grad_norm": 0.5720954537391663,
-      "learning_rate": 4.798694232861807e-06,
-      "loss": 0.4355,
-      "step": 441
-    },
-    {
-      "epoch": 0.02888700084961767,
-      "grad_norm": 0.5581308603286743,
-      "learning_rate": 4.8095756256800875e-06,
-      "loss": 0.4972,
-      "step": 442
-    },
-    {
-      "epoch": 0.028952356055159792,
-      "grad_norm": 0.5132817625999451,
-      "learning_rate": 4.820457018498368e-06,
-      "loss": 0.4686,
-      "step": 443
-    },
-    {
-      "epoch": 0.029017711260701917,
-      "grad_norm": 0.5478911399841309,
-      "learning_rate": 4.831338411316649e-06,
-      "loss": 0.542,
-      "step": 444
-    },
-    {
-      "epoch": 0.029083066466244038,
-      "grad_norm": 0.5395066738128662,
-      "learning_rate": 4.842219804134929e-06,
-      "loss": 0.4493,
-      "step": 445
-    },
-    {
-      "epoch": 0.02914842167178616,
-      "grad_norm": 0.554943323135376,
-      "learning_rate": 4.85310119695321e-06,
-      "loss": 0.498,
-      "step": 446
-    },
-    {
-      "epoch": 0.02921377687732828,
-      "grad_norm": 0.5648518800735474,
-      "learning_rate": 4.863982589771491e-06,
-      "loss": 0.4738,
-      "step": 447
-    },
-    {
-      "epoch": 0.0292791320828704,
-      "grad_norm": 0.5569779276847839,
-      "learning_rate": 4.874863982589772e-06,
-      "loss": 0.5183,
-      "step": 448
-    },
-    {
-      "epoch": 0.029344487288412522,
-      "grad_norm": 0.5409430265426636,
-      "learning_rate": 4.885745375408053e-06,
-      "loss": 0.4605,
-      "step": 449
-    },
-    {
-      "epoch": 0.029409842493954643,
-      "grad_norm": 0.5412803292274475,
-      "learning_rate": 4.8966267682263335e-06,
-      "loss": 0.4824,
-      "step": 450
-    },
-    {
-      "epoch": 0.029475197699496764,
-      "grad_norm": 0.5730560421943665,
-      "learning_rate": 4.907508161044614e-06,
-      "loss": 0.5014,
-      "step": 451
-    },
-    {
-      "epoch": 0.029540552905038885,
-      "grad_norm": 0.5052904486656189,
-      "learning_rate": 4.918389553862895e-06,
-      "loss": 0.4111,
-      "step": 452
-    },
-    {
-      "epoch": 0.029605908110581006,
-      "grad_norm": 0.536467432975769,
-      "learning_rate": 4.929270946681175e-06,
-      "loss": 0.4557,
-      "step": 453
-    },
-    {
-      "epoch": 0.029671263316123127,
-      "grad_norm": 0.5028022527694702,
-      "learning_rate": 4.940152339499457e-06,
-      "loss": 0.4002,
-      "step": 454
-    },
-    {
-      "epoch": 0.029736618521665252,
-      "grad_norm": 0.5263887643814087,
-      "learning_rate": 4.951033732317737e-06,
-      "loss": 0.4389,
-      "step": 455
-    },
-    {
-      "epoch": 0.029801973727207373,
-      "grad_norm": 0.5358768701553345,
-      "learning_rate": 4.961915125136018e-06,
-      "loss": 0.4788,
-      "step": 456
-    },
-    {
-      "epoch": 0.029867328932749494,
-      "grad_norm": 0.5781591534614563,
-      "learning_rate": 4.972796517954299e-06,
-      "loss": 0.4495,
-      "step": 457
-    },
-    {
-      "epoch": 0.029932684138291615,
-      "grad_norm": 0.5463568568229675,
-      "learning_rate": 4.9836779107725795e-06,
-      "loss": 0.4652,
-      "step": 458
-    },
-    {
-      "epoch": 0.029998039343833736,
-      "grad_norm": 0.5033892393112183,
-      "learning_rate": 4.99455930359086e-06,
-      "loss": 0.4665,
-      "step": 459
-    },
-    {
-      "epoch": 0.030063394549375858,
-      "grad_norm": 0.5578131675720215,
-      "learning_rate": 5.00544069640914e-06,
-      "loss": 0.5104,
-      "step": 460
-    },
-    {
-      "epoch": 0.03012874975491798,
-      "grad_norm": 0.5375128388404846,
-      "learning_rate": 5.016322089227421e-06,
-      "loss": 0.4344,
-      "step": 461
-    },
-    {
-      "epoch": 0.0301941049604601,
-      "grad_norm": 0.5085914731025696,
-      "learning_rate": 5.027203482045703e-06,
-      "loss": 0.4199,
-      "step": 462
-    },
-    {
-      "epoch": 0.03025946016600222,
-      "grad_norm": 0.5424376726150513,
-      "learning_rate": 5.038084874863983e-06,
-      "loss": 0.4852,
-      "step": 463
-    },
-    {
-      "epoch": 0.030324815371544342,
-      "grad_norm": 0.5353248119354248,
-      "learning_rate": 5.048966267682264e-06,
-      "loss": 0.4683,
-      "step": 464
-    },
-    {
-      "epoch": 0.030390170577086466,
-      "grad_norm": 0.5041464567184448,
-      "learning_rate": 5.059847660500544e-06,
-      "loss": 0.4485,
-      "step": 465
-    },
-    {
-      "epoch": 0.030455525782628588,
-      "grad_norm": 0.47988349199295044,
-      "learning_rate": 5.070729053318825e-06,
-      "loss": 0.4278,
-      "step": 466
-    },
-    {
-      "epoch": 0.03052088098817071,
-      "grad_norm": 0.561044454574585,
-      "learning_rate": 5.0816104461371064e-06,
-      "loss": 0.4877,
-      "step": 467
-    },
-    {
-      "epoch": 0.03058623619371283,
-      "grad_norm": 0.5972671508789062,
-      "learning_rate": 5.0924918389553864e-06,
-      "loss": 0.5439,
-      "step": 468
-    },
-    {
-      "epoch": 0.03065159139925495,
-      "grad_norm": 0.5359597206115723,
-      "learning_rate": 5.103373231773667e-06,
-      "loss": 0.4675,
-      "step": 469
-    },
-    {
-      "epoch": 0.030716946604797072,
-      "grad_norm": 0.5351109504699707,
-      "learning_rate": 5.114254624591948e-06,
-      "loss": 0.4827,
-      "step": 470
-    },
-    {
-      "epoch": 0.030782301810339193,
-      "grad_norm": 48.03053283691406,
-      "learning_rate": 5.125136017410229e-06,
-      "loss": 0.3983,
-      "step": 471
-    },
-    {
-      "epoch": 0.030847657015881314,
-      "grad_norm": 0.5735889077186584,
-      "learning_rate": 5.13601741022851e-06,
-      "loss": 0.4391,
-      "step": 472
-    },
-    {
-      "epoch": 0.030913012221423435,
-      "grad_norm": 0.5440057516098022,
-      "learning_rate": 5.146898803046791e-06,
-      "loss": 0.5,
-      "step": 473
-    },
-    {
-      "epoch": 0.030978367426965556,
-      "grad_norm": 0.5572605729103088,
-      "learning_rate": 5.157780195865071e-06,
-      "loss": 0.4396,
-      "step": 474
-    },
-    {
-      "epoch": 0.03104372263250768,
-      "grad_norm": 0.5115422606468201,
-      "learning_rate": 5.1686615886833525e-06,
-      "loss": 0.4114,
-      "step": 475
-    },
-    {
-      "epoch": 0.031109077838049802,
-      "grad_norm": 0.4984517991542816,
-      "learning_rate": 5.179542981501633e-06,
-      "loss": 0.411,
-      "step": 476
-    },
-    {
-      "epoch": 0.031174433043591923,
-      "grad_norm": 0.5517117977142334,
-      "learning_rate": 5.190424374319913e-06,
-      "loss": 0.4754,
-      "step": 477
-    },
-    {
-      "epoch": 0.031239788249134044,
-      "grad_norm": 0.5398489236831665,
-      "learning_rate": 5.201305767138194e-06,
-      "loss": 0.4647,
-      "step": 478
-    },
-    {
-      "epoch": 0.031305143454676165,
-      "grad_norm": 0.5186377763748169,
-      "learning_rate": 5.212187159956474e-06,
-      "loss": 0.4236,
-      "step": 479
-    },
-    {
-      "epoch": 0.031370498660218286,
-      "grad_norm": 0.5774402022361755,
-      "learning_rate": 5.223068552774756e-06,
-      "loss": 0.5579,
-      "step": 480
-    },
-    {
-      "epoch": 0.03143585386576041,
-      "grad_norm": 0.5345587730407715,
-      "learning_rate": 5.233949945593037e-06,
-      "loss": 0.4513,
-      "step": 481
-    },
-    {
-      "epoch": 0.03150120907130253,
-      "grad_norm": 0.5191734433174133,
-      "learning_rate": 5.244831338411317e-06,
-      "loss": 0.4174,
-      "step": 482
-    },
-    {
-      "epoch": 0.03156656427684465,
-      "grad_norm": 0.5381778478622437,
-      "learning_rate": 5.255712731229598e-06,
-      "loss": 0.466,
-      "step": 483
-    },
-    {
-      "epoch": 0.03163191948238677,
-      "grad_norm": 0.5413488745689392,
-      "learning_rate": 5.266594124047879e-06,
-      "loss": 0.4677,
-      "step": 484
-    },
-    {
-      "epoch": 0.03169727468792889,
-      "grad_norm": 0.5327398777008057,
-      "learning_rate": 5.277475516866159e-06,
-      "loss": 0.4372,
-      "step": 485
-    },
-    {
-      "epoch": 0.03176262989347101,
-      "grad_norm": 0.539981484413147,
-      "learning_rate": 5.28835690968444e-06,
-      "loss": 0.4302,
-      "step": 486
-    },
-    {
-      "epoch": 0.031827985099013134,
-      "grad_norm": 0.5105658173561096,
-      "learning_rate": 5.29923830250272e-06,
-      "loss": 0.4502,
-      "step": 487
-    },
-    {
-      "epoch": 0.031893340304555255,
-      "grad_norm": 0.576172411441803,
-      "learning_rate": 5.310119695321001e-06,
-      "loss": 0.5053,
-      "step": 488
-    },
-    {
-      "epoch": 0.031958695510097376,
-      "grad_norm": 0.5153981447219849,
-      "learning_rate": 5.321001088139283e-06,
-      "loss": 0.4475,
-      "step": 489
-    },
-    {
-      "epoch": 0.0320240507156395,
-      "grad_norm": 0.5384749174118042,
-      "learning_rate": 5.331882480957563e-06,
-      "loss": 0.4564,
-      "step": 490
-    },
-    {
-      "epoch": 0.032089405921181625,
-      "grad_norm": 0.5373324155807495,
-      "learning_rate": 5.342763873775844e-06,
-      "loss": 0.4775,
-      "step": 491
-    },
-    {
-      "epoch": 0.03215476112672375,
-      "grad_norm": 0.4921714663505554,
-      "learning_rate": 5.353645266594124e-06,
-      "loss": 0.3957,
-      "step": 492
-    },
-    {
-      "epoch": 0.03222011633226587,
-      "grad_norm": 0.5208613872528076,
-      "learning_rate": 5.364526659412405e-06,
-      "loss": 0.4641,
-      "step": 493
-    },
-    {
-      "epoch": 0.03228547153780799,
-      "grad_norm": 0.5194923877716064,
-      "learning_rate": 5.375408052230686e-06,
-      "loss": 0.4431,
-      "step": 494
-    },
-    {
-      "epoch": 0.03235082674335011,
-      "grad_norm": 1.0010076761245728,
-      "learning_rate": 5.386289445048966e-06,
-      "loss": 0.5236,
-      "step": 495
-    },
-    {
-      "epoch": 0.03241618194889223,
-      "grad_norm": 0.49507319927215576,
-      "learning_rate": 5.397170837867247e-06,
-      "loss": 0.4033,
-      "step": 496
-    },
-    {
-      "epoch": 0.03248153715443435,
-      "grad_norm": 0.5528603792190552,
-      "learning_rate": 5.408052230685528e-06,
-      "loss": 0.4901,
-      "step": 497
-    },
-    {
-      "epoch": 0.03254689235997647,
-      "grad_norm": 0.5661624073982239,
-      "learning_rate": 5.418933623503809e-06,
-      "loss": 0.4757,
-      "step": 498
-    },
-    {
-      "epoch": 0.032612247565518594,
-      "grad_norm": 0.5334492921829224,
-      "learning_rate": 5.42981501632209e-06,
-      "loss": 0.4036,
-      "step": 499
-    },
-    {
-      "epoch": 0.032677602771060715,
-      "grad_norm": 0.5270481705665588,
-      "learning_rate": 5.4406964091403706e-06,
-      "loss": 0.4564,
-      "step": 500
-    },
-    {
-      "epoch": 0.032742957976602836,
-      "grad_norm": 0.5354252457618713,
-      "learning_rate": 5.451577801958651e-06,
-      "loss": 0.4485,
-      "step": 501
-    },
-    {
-      "epoch": 0.03280831318214496,
-      "grad_norm": 0.5266621112823486,
-      "learning_rate": 5.462459194776932e-06,
-      "loss": 0.4308,
-      "step": 502
-    },
-    {
-      "epoch": 0.03287366838768708,
-      "grad_norm": 0.5338118076324463,
-      "learning_rate": 5.473340587595213e-06,
-      "loss": 0.4572,
-      "step": 503
-    },
-    {
-      "epoch": 0.0329390235932292,
-      "grad_norm": 0.5976651310920715,
-      "learning_rate": 5.484221980413493e-06,
-      "loss": 0.4834,
-      "step": 504
-    },
-    {
-      "epoch": 0.03300437879877132,
-      "grad_norm": 0.49956831336021423,
-      "learning_rate": 5.495103373231774e-06,
-      "loss": 0.4135,
-      "step": 505
-    },
-    {
-      "epoch": 0.03306973400431344,
-      "grad_norm": 0.5595198273658752,
-      "learning_rate": 5.505984766050056e-06,
-      "loss": 0.4968,
-      "step": 506
-    },
-    {
-      "epoch": 0.03313508920985556,
-      "grad_norm": 0.491152286529541,
-      "learning_rate": 5.516866158868336e-06,
-      "loss": 0.4026,
-      "step": 507
-    },
-    {
-      "epoch": 0.033200444415397684,
-      "grad_norm": 0.5413976907730103,
-      "learning_rate": 5.527747551686617e-06,
-      "loss": 0.4449,
-      "step": 508
-    },
-    {
-      "epoch": 0.033265799620939805,
-      "grad_norm": 0.5298357605934143,
-      "learning_rate": 5.538628944504897e-06,
-      "loss": 0.4039,
-      "step": 509
-    },
-    {
-      "epoch": 0.033331154826481926,
-      "grad_norm": 0.5238986015319824,
-      "learning_rate": 5.5495103373231775e-06,
-      "loss": 0.4694,
-      "step": 510
-    },
-    {
-      "epoch": 0.03339651003202405,
-      "grad_norm": 0.5359320640563965,
-      "learning_rate": 5.560391730141459e-06,
-      "loss": 0.4702,
-      "step": 511
-    },
-    {
-      "epoch": 0.033461865237566175,
-      "grad_norm": 0.5686987638473511,
-      "learning_rate": 5.571273122959739e-06,
-      "loss": 0.4735,
-      "step": 512
-    },
-    {
-      "epoch": 0.033527220443108297,
-      "grad_norm": 0.5644851326942444,
-      "learning_rate": 5.58215451577802e-06,
-      "loss": 0.489,
-      "step": 513
-    },
-    {
-      "epoch": 0.03359257564865042,
-      "grad_norm": 0.5575340986251831,
-      "learning_rate": 5.5930359085963e-06,
-      "loss": 0.4434,
-      "step": 514
-    },
-    {
-      "epoch": 0.03365793085419254,
-      "grad_norm": 0.5376102924346924,
-      "learning_rate": 5.603917301414582e-06,
-      "loss": 0.4183,
-      "step": 515
-    },
-    {
-      "epoch": 0.03372328605973466,
-      "grad_norm": 0.5130487084388733,
-      "learning_rate": 5.614798694232863e-06,
-      "loss": 0.4094,
-      "step": 516
-    },
-    {
-      "epoch": 0.03378864126527678,
-      "grad_norm": 0.527062714099884,
-      "learning_rate": 5.625680087051143e-06,
-      "loss": 0.4263,
-      "step": 517
-    },
-    {
-      "epoch": 0.0338539964708189,
-      "grad_norm": 0.5213437676429749,
-      "learning_rate": 5.6365614798694235e-06,
-      "loss": 0.4224,
-      "step": 518
-    },
-    {
-      "epoch": 0.03391935167636102,
-      "grad_norm": 0.49539193511009216,
-      "learning_rate": 5.647442872687704e-06,
-      "loss": 0.4028,
-      "step": 519
-    },
-    {
-      "epoch": 0.033984706881903144,
-      "grad_norm": 0.5212374329566956,
-      "learning_rate": 5.658324265505985e-06,
-      "loss": 0.4455,
-      "step": 520
-    },
-    {
-      "epoch": 0.034050062087445265,
-      "grad_norm": 0.5280008316040039,
-      "learning_rate": 5.669205658324266e-06,
-      "loss": 0.4327,
-      "step": 521
-    },
-    {
-      "epoch": 0.034115417292987386,
-      "grad_norm": 0.4981239438056946,
-      "learning_rate": 5.680087051142547e-06,
-      "loss": 0.4224,
-      "step": 522
-    },
-    {
-      "epoch": 0.03418077249852951,
-      "grad_norm": 0.542140781879425,
-      "learning_rate": 5.690968443960827e-06,
-      "loss": 0.4818,
-      "step": 523
-    },
-    {
-      "epoch": 0.03424612770407163,
-      "grad_norm": 0.5307171940803528,
-      "learning_rate": 5.701849836779109e-06,
-      "loss": 0.3948,
-      "step": 524
-    },
-    {
-      "epoch": 0.03431148290961375,
-      "grad_norm": 0.5340259075164795,
-      "learning_rate": 5.7127312295973895e-06,
-      "loss": 0.4668,
-      "step": 525
-    },
-    {
-      "epoch": 0.03437683811515587,
-      "grad_norm": 0.5348265171051025,
-      "learning_rate": 5.7236126224156695e-06,
-      "loss": 0.4919,
-      "step": 526
-    },
-    {
-      "epoch": 0.03444219332069799,
-      "grad_norm": 0.5372627973556519,
-      "learning_rate": 5.73449401523395e-06,
-      "loss": 0.5255,
-      "step": 527
-    },
-    {
-      "epoch": 0.03450754852624011,
-      "grad_norm": 0.5648165345191956,
-      "learning_rate": 5.745375408052232e-06,
-      "loss": 0.4818,
-      "step": 528
-    },
-    {
-      "epoch": 0.034572903731782234,
-      "grad_norm": 0.5220668315887451,
-      "learning_rate": 5.756256800870512e-06,
-      "loss": 0.4706,
-      "step": 529
-    },
-    {
-      "epoch": 0.034638258937324355,
-      "grad_norm": 0.5265078544616699,
-      "learning_rate": 5.767138193688793e-06,
-      "loss": 0.4059,
-      "step": 530
-    },
-    {
-      "epoch": 0.034703614142866476,
-      "grad_norm": 0.5335327386856079,
-      "learning_rate": 5.778019586507073e-06,
-      "loss": 0.4158,
-      "step": 531
-    },
-    {
-      "epoch": 0.0347689693484086,
-      "grad_norm": 0.5514426827430725,
-      "learning_rate": 5.788900979325354e-06,
-      "loss": 0.4555,
-      "step": 532
-    },
-    {
-      "epoch": 0.034834324553950725,
-      "grad_norm": 0.5093376040458679,
-      "learning_rate": 5.7997823721436356e-06,
-      "loss": 0.4121,
-      "step": 533
-    },
-    {
-      "epoch": 0.034899679759492847,
-      "grad_norm": 0.5448405146598816,
-      "learning_rate": 5.8106637649619156e-06,
-      "loss": 0.5008,
-      "step": 534
-    },
-    {
-      "epoch": 0.03496503496503497,
-      "grad_norm": 0.5521230101585388,
-      "learning_rate": 5.8215451577801964e-06,
-      "loss": 0.4608,
-      "step": 535
-    },
-    {
-      "epoch": 0.03503039017057709,
-      "grad_norm": 0.5409108996391296,
-      "learning_rate": 5.8324265505984764e-06,
-      "loss": 0.4605,
-      "step": 536
-    },
-    {
-      "epoch": 0.03509574537611921,
-      "grad_norm": 0.5000602006912231,
-      "learning_rate": 5.843307943416758e-06,
-      "loss": 0.4197,
-      "step": 537
-    },
-    {
-      "epoch": 0.03516110058166133,
-      "grad_norm": 0.5224672555923462,
-      "learning_rate": 5.854189336235039e-06,
-      "loss": 0.4366,
-      "step": 538
-    },
-    {
-      "epoch": 0.03522645578720345,
-      "grad_norm": 0.5698318481445312,
-      "learning_rate": 5.865070729053319e-06,
-      "loss": 0.5485,
-      "step": 539
-    },
-    {
-      "epoch": 0.03529181099274557,
-      "grad_norm": 0.5366117358207703,
-      "learning_rate": 5.8759521218716e-06,
-      "loss": 0.4482,
-      "step": 540
-    },
-    {
-      "epoch": 0.035357166198287694,
-      "grad_norm": 0.54118412733078,
-      "learning_rate": 5.88683351468988e-06,
-      "loss": 0.4476,
-      "step": 541
-    },
-    {
-      "epoch": 0.035422521403829815,
-      "grad_norm": 0.5354444980621338,
-      "learning_rate": 5.897714907508162e-06,
-      "loss": 0.4827,
-      "step": 542
-    },
-    {
-      "epoch": 0.035487876609371936,
-      "grad_norm": 0.523681640625,
-      "learning_rate": 5.9085963003264425e-06,
-      "loss": 0.4486,
-      "step": 543
-    },
-    {
-      "epoch": 0.03555323181491406,
-      "grad_norm": 0.5297341346740723,
-      "learning_rate": 5.9194776931447225e-06,
-      "loss": 0.4525,
-      "step": 544
-    },
-    {
-      "epoch": 0.03561858702045618,
-      "grad_norm": 0.5035794973373413,
-      "learning_rate": 5.930359085963003e-06,
-      "loss": 0.4558,
-      "step": 545
-    },
-    {
-      "epoch": 0.0356839422259983,
-      "grad_norm": 0.5241169929504395,
-      "learning_rate": 5.941240478781285e-06,
-      "loss": 0.4654,
-      "step": 546
-    },
-    {
-      "epoch": 0.03574929743154042,
-      "grad_norm": 0.5188154578208923,
-      "learning_rate": 5.952121871599565e-06,
-      "loss": 0.454,
-      "step": 547
-    },
-    {
-      "epoch": 0.03581465263708254,
-      "grad_norm": 0.4866127073764801,
-      "learning_rate": 5.963003264417846e-06,
-      "loss": 0.411,
-      "step": 548
-    },
-    {
-      "epoch": 0.03588000784262466,
-      "grad_norm": 0.5161880254745483,
-      "learning_rate": 5.973884657236127e-06,
-      "loss": 0.4425,
-      "step": 549
-    },
-    {
-      "epoch": 0.035945363048166784,
-      "grad_norm": 0.5133089423179626,
-      "learning_rate": 5.984766050054407e-06,
-      "loss": 0.4554,
-      "step": 550
-    },
-    {
-      "epoch": 0.036010718253708905,
-      "grad_norm": 0.5831466913223267,
-      "learning_rate": 5.9956474428726885e-06,
-      "loss": 0.5481,
-      "step": 551
-    },
-    {
-      "epoch": 0.036076073459251026,
-      "grad_norm": 0.5680014491081238,
-      "learning_rate": 6.006528835690969e-06,
-      "loss": 0.4916,
-      "step": 552
-    },
-    {
-      "epoch": 0.03614142866479315,
-      "grad_norm": 0.5159114599227905,
-      "learning_rate": 6.017410228509249e-06,
-      "loss": 0.4435,
-      "step": 553
-    },
-    {
-      "epoch": 0.036206783870335275,
-      "grad_norm": 0.5152553915977478,
-      "learning_rate": 6.02829162132753e-06,
-      "loss": 0.418,
-      "step": 554
-    },
-    {
-      "epoch": 0.036272139075877396,
-      "grad_norm": 0.5299906134605408,
-      "learning_rate": 6.039173014145812e-06,
-      "loss": 0.4592,
-      "step": 555
-    },
-    {
-      "epoch": 0.03633749428141952,
-      "grad_norm": 0.506966233253479,
-      "learning_rate": 6.050054406964092e-06,
-      "loss": 0.4238,
-      "step": 556
-    },
-    {
-      "epoch": 0.03640284948696164,
-      "grad_norm": 0.5458035469055176,
-      "learning_rate": 6.060935799782373e-06,
-      "loss": 0.4889,
-      "step": 557
-    },
-    {
-      "epoch": 0.03646820469250376,
-      "grad_norm": 0.5626875758171082,
-      "learning_rate": 6.071817192600653e-06,
-      "loss": 0.4789,
-      "step": 558
-    },
-    {
-      "epoch": 0.03653355989804588,
-      "grad_norm": 0.589255690574646,
-      "learning_rate": 6.0826985854189345e-06,
-      "loss": 0.4852,
-      "step": 559
-    },
-    {
-      "epoch": 0.036598915103588,
-      "grad_norm": 0.5371728539466858,
-      "learning_rate": 6.093579978237215e-06,
-      "loss": 0.4736,
-      "step": 560
-    },
-    {
-      "epoch": 0.03666427030913012,
-      "grad_norm": 0.5571090579032898,
-      "learning_rate": 6.104461371055495e-06,
-      "loss": 0.4766,
-      "step": 561
-    },
-    {
-      "epoch": 0.036729625514672244,
-      "grad_norm": 0.5630400776863098,
-      "learning_rate": 6.115342763873776e-06,
-      "loss": 0.4499,
-      "step": 562
-    },
-    {
-      "epoch": 0.036794980720214365,
-      "grad_norm": 0.5382595062255859,
-      "learning_rate": 6.126224156692056e-06,
-      "loss": 0.4899,
-      "step": 563
-    },
-    {
-      "epoch": 0.036860335925756486,
-      "grad_norm": 0.5315620303153992,
-      "learning_rate": 6.137105549510338e-06,
-      "loss": 0.4521,
-      "step": 564
-    },
-    {
-      "epoch": 0.03692569113129861,
-      "grad_norm": 0.5756667852401733,
-      "learning_rate": 6.147986942328619e-06,
-      "loss": 0.4767,
-      "step": 565
-    },
-    {
-      "epoch": 0.03699104633684073,
-      "grad_norm": 0.509128987789154,
-      "learning_rate": 6.158868335146899e-06,
-      "loss": 0.3923,
-      "step": 566
-    },
-    {
-      "epoch": 0.03705640154238285,
-      "grad_norm": 0.5773864388465881,
-      "learning_rate": 6.16974972796518e-06,
-      "loss": 0.4425,
-      "step": 567
-    },
-    {
-      "epoch": 0.03712175674792497,
-      "grad_norm": 0.5730013251304626,
-      "learning_rate": 6.180631120783461e-06,
-      "loss": 0.4531,
-      "step": 568
-    },
-    {
-      "epoch": 0.03718711195346709,
-      "grad_norm": 0.5450667142868042,
-      "learning_rate": 6.191512513601741e-06,
-      "loss": 0.4685,
-      "step": 569
-    },
-    {
-      "epoch": 0.03725246715900921,
-      "grad_norm": 0.5366420745849609,
-      "learning_rate": 6.202393906420022e-06,
-      "loss": 0.4681,
-      "step": 570
-    },
-    {
-      "epoch": 0.037317822364551334,
-      "grad_norm": 0.5366652011871338,
-      "learning_rate": 6.213275299238303e-06,
-      "loss": 0.3952,
-      "step": 571
-    },
-    {
-      "epoch": 0.037383177570093455,
-      "grad_norm": 0.5294990539550781,
-      "learning_rate": 6.224156692056583e-06,
-      "loss": 0.4675,
-      "step": 572
-    },
-    {
-      "epoch": 0.037448532775635576,
-      "grad_norm": 0.5067282915115356,
-      "learning_rate": 6.235038084874865e-06,
-      "loss": 0.4105,
-      "step": 573
-    },
-    {
-      "epoch": 0.0375138879811777,
-      "grad_norm": 0.5480269193649292,
-      "learning_rate": 6.245919477693146e-06,
-      "loss": 0.4792,
-      "step": 574
-    },
-    {
-      "epoch": 0.037579243186719825,
-      "grad_norm": 0.5152676701545715,
-      "learning_rate": 6.256800870511426e-06,
-      "loss": 0.4149,
-      "step": 575
-    },
-    {
-      "epoch": 0.037644598392261946,
-      "grad_norm": 0.5218726992607117,
-      "learning_rate": 6.267682263329707e-06,
-      "loss": 0.4089,
-      "step": 576
-    },
-    {
-      "epoch": 0.03770995359780407,
-      "grad_norm": 0.5660258531570435,
-      "learning_rate": 6.278563656147988e-06,
-      "loss": 0.4605,
-      "step": 577
-    },
-    {
-      "epoch": 0.03777530880334619,
-      "grad_norm": 0.47741585969924927,
-      "learning_rate": 6.289445048966268e-06,
-      "loss": 0.3852,
-      "step": 578
-    },
-    {
-      "epoch": 0.03784066400888831,
-      "grad_norm": 0.545810341835022,
-      "learning_rate": 6.300326441784549e-06,
-      "loss": 0.4394,
-      "step": 579
-    },
-    {
-      "epoch": 0.03790601921443043,
-      "grad_norm": 0.5607401728630066,
-      "learning_rate": 6.311207834602829e-06,
-      "loss": 0.4964,
-      "step": 580
-    },
-    {
-      "epoch": 0.03797137441997255,
-      "grad_norm": 0.5499708652496338,
-      "learning_rate": 6.32208922742111e-06,
-      "loss": 0.4109,
-      "step": 581
-    },
-    {
-      "epoch": 0.03803672962551467,
-      "grad_norm": 0.5648245811462402,
-      "learning_rate": 6.332970620239392e-06,
-      "loss": 0.5523,
-      "step": 582
-    },
-    {
-      "epoch": 0.038102084831056794,
-      "grad_norm": 0.5284977555274963,
-      "learning_rate": 6.343852013057672e-06,
-      "loss": 0.4311,
-      "step": 583
-    },
-    {
-      "epoch": 0.038167440036598915,
-      "grad_norm": 0.5634139180183411,
-      "learning_rate": 6.354733405875953e-06,
-      "loss": 0.5076,
-      "step": 584
-    },
-    {
-      "epoch": 0.038232795242141036,
-      "grad_norm": 0.5088948607444763,
-      "learning_rate": 6.365614798694233e-06,
-      "loss": 0.4246,
-      "step": 585
-    },
-    {
-      "epoch": 0.03829815044768316,
-      "grad_norm": 0.5613812804222107,
-      "learning_rate": 6.376496191512514e-06,
-      "loss": 0.4443,
-      "step": 586
-    },
-    {
-      "epoch": 0.03836350565322528,
-      "grad_norm": 0.5567344427108765,
-      "learning_rate": 6.387377584330795e-06,
-      "loss": 0.5085,
-      "step": 587
-    },
-    {
-      "epoch": 0.0384288608587674,
-      "grad_norm": 0.49491986632347107,
-      "learning_rate": 6.398258977149075e-06,
-      "loss": 0.3923,
-      "step": 588
-    },
-    {
-      "epoch": 0.03849421606430952,
-      "grad_norm": 0.5684411525726318,
-      "learning_rate": 6.409140369967356e-06,
-      "loss": 0.4444,
-      "step": 589
-    },
-    {
-      "epoch": 0.03855957126985164,
-      "grad_norm": 0.5668994784355164,
-      "learning_rate": 6.420021762785638e-06,
-      "loss": 0.4558,
-      "step": 590
-    },
-    {
-      "epoch": 0.03862492647539376,
-      "grad_norm": 0.5855531096458435,
-      "learning_rate": 6.430903155603918e-06,
-      "loss": 0.4668,
-      "step": 591
-    },
-    {
-      "epoch": 0.038690281680935884,
-      "grad_norm": 0.5380173921585083,
-      "learning_rate": 6.441784548422199e-06,
-      "loss": 0.471,
-      "step": 592
-    },
-    {
-      "epoch": 0.038755636886478005,
-      "grad_norm": 0.5691683292388916,
-      "learning_rate": 6.452665941240479e-06,
-      "loss": 0.4829,
-      "step": 593
-    },
-    {
-      "epoch": 0.038820992092020126,
-      "grad_norm": 0.607672393321991,
-      "learning_rate": 6.4635473340587595e-06,
-      "loss": 0.4983,
-      "step": 594
-    },
-    {
-      "epoch": 0.038886347297562254,
-      "grad_norm": 0.5337852239608765,
-      "learning_rate": 6.474428726877041e-06,
-      "loss": 0.441,
-      "step": 595
-    },
-    {
-      "epoch": 0.038951702503104375,
-      "grad_norm": 0.5541263222694397,
-      "learning_rate": 6.485310119695321e-06,
-      "loss": 0.4492,
-      "step": 596
-    },
-    {
-      "epoch": 0.039017057708646496,
-      "grad_norm": 0.566577672958374,
-      "learning_rate": 6.496191512513602e-06,
-      "loss": 0.5324,
-      "step": 597
-    },
-    {
-      "epoch": 0.03908241291418862,
-      "grad_norm": 0.5563501715660095,
-      "learning_rate": 6.507072905331883e-06,
-      "loss": 0.4868,
-      "step": 598
-    },
-    {
-      "epoch": 0.03914776811973074,
-      "grad_norm": 0.5898980498313904,
-      "learning_rate": 6.517954298150164e-06,
-      "loss": 0.4476,
-      "step": 599
-    },
-    {
-      "epoch": 0.03921312332527286,
-      "grad_norm": 0.5672775506973267,
-      "learning_rate": 6.528835690968445e-06,
-      "loss": 0.4852,
-      "step": 600
-    },
-    {
-      "epoch": 0.03927847853081498,
-      "grad_norm": 0.5669152736663818,
-      "learning_rate": 6.5397170837867255e-06,
-      "loss": 0.4813,
-      "step": 601
-    },
-    {
-      "epoch": 0.0393438337363571,
-      "grad_norm": 0.5214797854423523,
-      "learning_rate": 6.5505984766050056e-06,
-      "loss": 0.4511,
-      "step": 602
-    },
-    {
-      "epoch": 0.03940918894189922,
-      "grad_norm": 0.519077479839325,
-      "learning_rate": 6.561479869423286e-06,
-      "loss": 0.4049,
-      "step": 603
-    },
-    {
-      "epoch": 0.039474544147441344,
-      "grad_norm": 0.5328835844993591,
-      "learning_rate": 6.572361262241568e-06,
-      "loss": 0.4301,
-      "step": 604
-    },
-    {
-      "epoch": 0.039539899352983465,
-      "grad_norm": 0.5590944290161133,
-      "learning_rate": 6.583242655059848e-06,
-      "loss": 0.4755,
-      "step": 605
-    },
-    {
-      "epoch": 0.039605254558525586,
-      "grad_norm": 0.5169242024421692,
-      "learning_rate": 6.594124047878129e-06,
-      "loss": 0.3647,
-      "step": 606
-    },
-    {
-      "epoch": 0.03967060976406771,
-      "grad_norm": 0.5104243159294128,
-      "learning_rate": 6.605005440696409e-06,
-      "loss": 0.4208,
-      "step": 607
-    },
-    {
-      "epoch": 0.03973596496960983,
-      "grad_norm": 0.4833294153213501,
-      "learning_rate": 6.615886833514691e-06,
-      "loss": 0.41,
-      "step": 608
-    },
-    {
-      "epoch": 0.03980132017515195,
-      "grad_norm": 0.5645463466644287,
-      "learning_rate": 6.6267682263329716e-06,
-      "loss": 0.4371,
-      "step": 609
-    },
-    {
-      "epoch": 0.03986667538069407,
-      "grad_norm": 0.4820958375930786,
-      "learning_rate": 6.637649619151252e-06,
-      "loss": 0.3911,
-      "step": 610
-    },
-    {
-      "epoch": 0.03993203058623619,
-      "grad_norm": 0.5233252644538879,
-      "learning_rate": 6.6485310119695324e-06,
-      "loss": 0.4235,
-      "step": 611
-    },
-    {
-      "epoch": 0.03999738579177831,
-      "grad_norm": 0.5318486094474792,
-      "learning_rate": 6.659412404787814e-06,
-      "loss": 0.4399,
-      "step": 612
-    },
-    {
-      "epoch": 0.040062740997320434,
-      "grad_norm": 0.5524714589118958,
-      "learning_rate": 6.670293797606094e-06,
-      "loss": 0.503,
-      "step": 613
-    },
-    {
-      "epoch": 0.040128096202862555,
-      "grad_norm": 0.48769381642341614,
-      "learning_rate": 6.681175190424375e-06,
-      "loss": 0.4067,
-      "step": 614
-    },
-    {
-      "epoch": 0.040193451408404676,
-      "grad_norm": 0.5265127420425415,
-      "learning_rate": 6.692056583242655e-06,
-      "loss": 0.4746,
-      "step": 615
-    },
-    {
-      "epoch": 0.040258806613946804,
-      "grad_norm": 0.49916163086891174,
-      "learning_rate": 6.702937976060936e-06,
-      "loss": 0.3907,
-      "step": 616
-    },
-    {
-      "epoch": 0.040324161819488925,
-      "grad_norm": 0.5203204154968262,
-      "learning_rate": 6.713819368879218e-06,
-      "loss": 0.4469,
-      "step": 617
-    },
-    {
-      "epoch": 0.040389517025031046,
-      "grad_norm": 0.5336718559265137,
-      "learning_rate": 6.724700761697498e-06,
-      "loss": 0.4842,
-      "step": 618
-    },
-    {
-      "epoch": 0.04045487223057317,
-      "grad_norm": 0.5118704438209534,
-      "learning_rate": 6.7355821545157785e-06,
-      "loss": 0.4224,
-      "step": 619
-    },
-    {
-      "epoch": 0.04052022743611529,
-      "grad_norm": 0.5074040293693542,
-      "learning_rate": 6.746463547334059e-06,
-      "loss": 0.4409,
-      "step": 620
-    },
-    {
-      "epoch": 0.04058558264165741,
-      "grad_norm": 0.5172423124313354,
-      "learning_rate": 6.75734494015234e-06,
-      "loss": 0.432,
-      "step": 621
-    },
-    {
-      "epoch": 0.04065093784719953,
-      "grad_norm": 0.5750128030776978,
-      "learning_rate": 6.768226332970621e-06,
-      "loss": 0.5288,
-      "step": 622
-    },
-    {
-      "epoch": 0.04071629305274165,
-      "grad_norm": 0.5487008094787598,
-      "learning_rate": 6.779107725788902e-06,
-      "loss": 0.4392,
-      "step": 623
-    },
-    {
-      "epoch": 0.04078164825828377,
-      "grad_norm": 0.5502815246582031,
-      "learning_rate": 6.789989118607182e-06,
-      "loss": 0.4358,
-      "step": 624
-    },
-    {
-      "epoch": 0.040847003463825894,
-      "grad_norm": 0.5159929394721985,
-      "learning_rate": 6.800870511425463e-06,
-      "loss": 0.4346,
-      "step": 625
-    },
-    {
-      "epoch": 0.040912358669368015,
-      "grad_norm": 0.5169581770896912,
-      "learning_rate": 6.8117519042437445e-06,
-      "loss": 0.4549,
-      "step": 626
-    },
-    {
-      "epoch": 0.040977713874910136,
-      "grad_norm": 0.5334294438362122,
-      "learning_rate": 6.8226332970620245e-06,
-      "loss": 0.4371,
-      "step": 627
-    },
-    {
-      "epoch": 0.04104306908045226,
-      "grad_norm": 0.5245968699455261,
-      "learning_rate": 6.833514689880305e-06,
-      "loss": 0.4437,
-      "step": 628
-    },
-    {
-      "epoch": 0.04110842428599438,
-      "grad_norm": 0.5483084917068481,
-      "learning_rate": 6.844396082698585e-06,
-      "loss": 0.438,
-      "step": 629
-    },
-    {
-      "epoch": 0.0411737794915365,
-      "grad_norm": 0.5519688725471497,
-      "learning_rate": 6.855277475516867e-06,
-      "loss": 0.4163,
-      "step": 630
-    },
-    {
-      "epoch": 0.04123913469707862,
-      "grad_norm": 0.531467080116272,
-      "learning_rate": 6.866158868335148e-06,
-      "loss": 0.4302,
-      "step": 631
-    },
-    {
-      "epoch": 0.04130448990262074,
-      "grad_norm": 0.542911171913147,
-      "learning_rate": 6.877040261153428e-06,
-      "loss": 0.4758,
-      "step": 632
-    },
-    {
-      "epoch": 0.04136984510816286,
-      "grad_norm": 0.514519214630127,
-      "learning_rate": 6.887921653971709e-06,
-      "loss": 0.4265,
-      "step": 633
-    },
-    {
-      "epoch": 0.041435200313704984,
-      "grad_norm": 0.5962108969688416,
-      "learning_rate": 6.898803046789989e-06,
-      "loss": 0.4733,
-      "step": 634
-    },
-    {
-      "epoch": 0.041500555519247105,
-      "grad_norm": 0.5529797077178955,
-      "learning_rate": 6.9096844396082705e-06,
-      "loss": 0.4656,
-      "step": 635
-    },
-    {
-      "epoch": 0.041565910724789226,
-      "grad_norm": 0.4788876175880432,
-      "learning_rate": 6.920565832426551e-06,
-      "loss": 0.3523,
-      "step": 636
-    },
-    {
-      "epoch": 0.041631265930331354,
-      "grad_norm": 0.5504601001739502,
-      "learning_rate": 6.931447225244831e-06,
-      "loss": 0.4427,
-      "step": 637
-    },
-    {
-      "epoch": 0.041696621135873475,
-      "grad_norm": 0.5427254438400269,
-      "learning_rate": 6.942328618063112e-06,
-      "loss": 0.4463,
-      "step": 638
-    },
-    {
-      "epoch": 0.041761976341415596,
-      "grad_norm": 0.47728046774864197,
-      "learning_rate": 6.953210010881394e-06,
-      "loss": 0.4082,
-      "step": 639
-    },
-    {
-      "epoch": 0.04182733154695772,
-      "grad_norm": 0.566616415977478,
-      "learning_rate": 6.964091403699674e-06,
-      "loss": 0.4286,
-      "step": 640
-    },
-    {
-      "epoch": 0.04189268675249984,
-      "grad_norm": 0.5331923365592957,
-      "learning_rate": 6.974972796517955e-06,
-      "loss": 0.4229,
-      "step": 641
-    },
-    {
-      "epoch": 0.04195804195804196,
-      "grad_norm": 0.4955950677394867,
-      "learning_rate": 6.985854189336235e-06,
-      "loss": 0.4257,
-      "step": 642
-    },
-    {
-      "epoch": 0.04202339716358408,
-      "grad_norm": 0.5213703513145447,
-      "learning_rate": 6.9967355821545166e-06,
-      "loss": 0.466,
-      "step": 643
-    },
-    {
-      "epoch": 0.0420887523691262,
-      "grad_norm": 0.5337159037590027,
-      "learning_rate": 7.0076169749727974e-06,
-      "loss": 0.4539,
-      "step": 644
-    },
-    {
-      "epoch": 0.04215410757466832,
-      "grad_norm": 0.5224979519844055,
-      "learning_rate": 7.0184983677910774e-06,
-      "loss": 0.4384,
-      "step": 645
-    },
-    {
-      "epoch": 0.042219462780210444,
-      "grad_norm": 0.5413781404495239,
-      "learning_rate": 7.029379760609358e-06,
-      "loss": 0.4687,
-      "step": 646
-    },
-    {
-      "epoch": 0.042284817985752565,
-      "grad_norm": 0.553650975227356,
-      "learning_rate": 7.040261153427639e-06,
-      "loss": 0.477,
-      "step": 647
-    },
-    {
-      "epoch": 0.042350173191294686,
-      "grad_norm": 0.5719106793403625,
-      "learning_rate": 7.05114254624592e-06,
-      "loss": 0.4478,
-      "step": 648
-    },
-    {
-      "epoch": 0.04241552839683681,
-      "grad_norm": 0.48158884048461914,
-      "learning_rate": 7.062023939064201e-06,
-      "loss": 0.3995,
-      "step": 649
-    },
-    {
-      "epoch": 0.04248088360237893,
-      "grad_norm": 0.5439937710762024,
-      "learning_rate": 7.072905331882482e-06,
-      "loss": 0.4142,
-      "step": 650
-    },
-    {
-      "epoch": 0.04254623880792105,
-      "grad_norm": 0.570808470249176,
-      "learning_rate": 7.083786724700762e-06,
-      "loss": 0.4679,
-      "step": 651
-    },
-    {
-      "epoch": 0.04261159401346317,
-      "grad_norm": 0.5186165571212769,
-      "learning_rate": 7.0946681175190435e-06,
-      "loss": 0.48,
-      "step": 652
-    },
-    {
-      "epoch": 0.04267694921900529,
-      "grad_norm": 0.5627509355545044,
-      "learning_rate": 7.105549510337324e-06,
-      "loss": 0.4676,
-      "step": 653
-    },
-    {
-      "epoch": 0.04274230442454741,
-      "grad_norm": 0.5147085785865784,
-      "learning_rate": 7.116430903155604e-06,
-      "loss": 0.4241,
-      "step": 654
-    },
-    {
-      "epoch": 0.042807659630089534,
-      "grad_norm": 0.5302978754043579,
-      "learning_rate": 7.127312295973885e-06,
-      "loss": 0.4572,
-      "step": 655
-    },
-    {
-      "epoch": 0.042873014835631655,
-      "grad_norm": 0.5276473164558411,
-      "learning_rate": 7.138193688792165e-06,
-      "loss": 0.4091,
-      "step": 656
-    },
-    {
-      "epoch": 0.042938370041173776,
-      "grad_norm": 0.598788321018219,
-      "learning_rate": 7.149075081610447e-06,
-      "loss": 0.4061,
-      "step": 657
-    },
-    {
-      "epoch": 0.043003725246715904,
-      "grad_norm": 0.6199572682380676,
-      "learning_rate": 7.159956474428728e-06,
-      "loss": 0.5387,
-      "step": 658
-    },
-    {
-      "epoch": 0.043069080452258025,
-      "grad_norm": 0.5637344121932983,
-      "learning_rate": 7.170837867247008e-06,
-      "loss": 0.4928,
-      "step": 659
-    },
-    {
-      "epoch": 0.043134435657800146,
-      "grad_norm": 0.5307586789131165,
-      "learning_rate": 7.181719260065289e-06,
-      "loss": 0.438,
-      "step": 660
-    },
-    {
-      "epoch": 0.04319979086334227,
-      "grad_norm": 0.5489634871482849,
-      "learning_rate": 7.19260065288357e-06,
-      "loss": 0.4022,
-      "step": 661
-    },
-    {
-      "epoch": 0.04326514606888439,
-      "grad_norm": 0.5737041234970093,
-      "learning_rate": 7.20348204570185e-06,
-      "loss": 0.4616,
-      "step": 662
-    },
-    {
-      "epoch": 0.04333050127442651,
-      "grad_norm": 0.5683348774909973,
-      "learning_rate": 7.214363438520131e-06,
-      "loss": 0.45,
-      "step": 663
-    },
-    {
-      "epoch": 0.04339585647996863,
-      "grad_norm": 0.5592344403266907,
-      "learning_rate": 7.225244831338411e-06,
-      "loss": 0.4015,
-      "step": 664
-    },
-    {
-      "epoch": 0.04346121168551075,
-      "grad_norm": 0.6093730330467224,
-      "learning_rate": 7.236126224156693e-06,
-      "loss": 0.4693,
-      "step": 665
-    },
-    {
-      "epoch": 0.04352656689105287,
-      "grad_norm": 0.5761213898658752,
-      "learning_rate": 7.247007616974974e-06,
-      "loss": 0.4408,
-      "step": 666
-    },
-    {
-      "epoch": 0.043591922096594994,
-      "grad_norm": 0.5475156307220459,
-      "learning_rate": 7.257889009793254e-06,
-      "loss": 0.481,
-      "step": 667
-    },
-    {
-      "epoch": 0.043657277302137115,
-      "grad_norm": 0.5181083083152771,
-      "learning_rate": 7.268770402611535e-06,
-      "loss": 0.4391,
-      "step": 668
-    },
-    {
-      "epoch": 0.043722632507679236,
-      "grad_norm": 0.574974000453949,
-      "learning_rate": 7.2796517954298155e-06,
-      "loss": 0.4821,
-      "step": 669
-    },
-    {
-      "epoch": 0.04378798771322136,
-      "grad_norm": 0.6030741930007935,
-      "learning_rate": 7.290533188248096e-06,
-      "loss": 0.457,
-      "step": 670
-    },
-    {
-      "epoch": 0.04385334291876348,
-      "grad_norm": 0.5326198935508728,
-      "learning_rate": 7.301414581066377e-06,
-      "loss": 0.4408,
-      "step": 671
-    },
-    {
-      "epoch": 0.0439186981243056,
-      "grad_norm": 0.5482437610626221,
-      "learning_rate": 7.312295973884658e-06,
-      "loss": 0.4234,
-      "step": 672
-    },
-    {
-      "epoch": 0.04398405332984772,
-      "grad_norm": 0.5907866954803467,
-      "learning_rate": 7.323177366702938e-06,
-      "loss": 0.4332,
-      "step": 673
-    },
-    {
-      "epoch": 0.04404940853538984,
-      "grad_norm": 0.5776271820068359,
-      "learning_rate": 7.33405875952122e-06,
-      "loss": 0.4587,
-      "step": 674
-    },
-    {
-      "epoch": 0.04411476374093196,
-      "grad_norm": 0.5511412620544434,
-      "learning_rate": 7.344940152339501e-06,
-      "loss": 0.451,
-      "step": 675
-    },
-    {
-      "epoch": 0.044180118946474084,
-      "grad_norm": 0.5105332136154175,
-      "learning_rate": 7.355821545157781e-06,
-      "loss": 0.4066,
-      "step": 676
-    },
-    {
-      "epoch": 0.044245474152016205,
-      "grad_norm": 0.5682520270347595,
-      "learning_rate": 7.3667029379760616e-06,
-      "loss": 0.4715,
-      "step": 677
-    },
-    {
-      "epoch": 0.044310829357558326,
-      "grad_norm": 0.6172433495521545,
-      "learning_rate": 7.377584330794342e-06,
-      "loss": 0.53,
-      "step": 678
-    },
-    {
-      "epoch": 0.044376184563100454,
-      "grad_norm": 0.5491930842399597,
-      "learning_rate": 7.388465723612623e-06,
-      "loss": 0.4559,
-      "step": 679
-    },
-    {
-      "epoch": 0.044441539768642575,
-      "grad_norm": 0.5460163950920105,
-      "learning_rate": 7.399347116430904e-06,
-      "loss": 0.4509,
-      "step": 680
-    },
-    {
-      "epoch": 0.044506894974184696,
-      "grad_norm": 0.5838077664375305,
-      "learning_rate": 7.410228509249184e-06,
-      "loss": 0.5506,
-      "step": 681
-    },
-    {
-      "epoch": 0.04457225017972682,
-      "grad_norm": 0.5743003487586975,
-      "learning_rate": 7.421109902067465e-06,
-      "loss": 0.4691,
-      "step": 682
-    },
-    {
-      "epoch": 0.04463760538526894,
-      "grad_norm": 0.507209837436676,
-      "learning_rate": 7.431991294885747e-06,
-      "loss": 0.3838,
-      "step": 683
-    },
-    {
-      "epoch": 0.04470296059081106,
-      "grad_norm": 0.5423256158828735,
-      "learning_rate": 7.442872687704027e-06,
-      "loss": 0.4069,
-      "step": 684
-    },
-    {
-      "epoch": 0.04476831579635318,
-      "grad_norm": 0.5516565442085266,
-      "learning_rate": 7.453754080522308e-06,
-      "loss": 0.4655,
-      "step": 685
-    },
-    {
-      "epoch": 0.0448336710018953,
-      "grad_norm": 0.5188184380531311,
-      "learning_rate": 7.464635473340588e-06,
-      "loss": 0.4598,
-      "step": 686
-    },
-    {
-      "epoch": 0.04489902620743742,
-      "grad_norm": 0.5910035967826843,
-      "learning_rate": 7.4755168661588685e-06,
-      "loss": 0.48,
-      "step": 687
-    },
-    {
-      "epoch": 0.044964381412979544,
-      "grad_norm": 0.5000728964805603,
-      "learning_rate": 7.48639825897715e-06,
-      "loss": 0.3985,
-      "step": 688
-    },
-    {
-      "epoch": 0.045029736618521665,
-      "grad_norm": 1.4668203592300415,
-      "learning_rate": 7.49727965179543e-06,
-      "loss": 0.5042,
-      "step": 689
-    },
-    {
-      "epoch": 0.045095091824063786,
-      "grad_norm": 0.5863484144210815,
-      "learning_rate": 7.508161044613711e-06,
-      "loss": 0.4646,
-      "step": 690
-    },
-    {
-      "epoch": 0.04516044702960591,
-      "grad_norm": 0.5586270093917847,
-      "learning_rate": 7.519042437431991e-06,
-      "loss": 0.3939,
-      "step": 691
-    },
-    {
-      "epoch": 0.04522580223514803,
-      "grad_norm": 0.5796319842338562,
-      "learning_rate": 7.529923830250273e-06,
-      "loss": 0.471,
-      "step": 692
-    },
-    {
-      "epoch": 0.04529115744069015,
-      "grad_norm": 0.5738232731819153,
-      "learning_rate": 7.540805223068554e-06,
-      "loss": 0.4838,
-      "step": 693
-    },
-    {
-      "epoch": 0.04535651264623227,
-      "grad_norm": 0.531330943107605,
-      "learning_rate": 7.551686615886834e-06,
-      "loss": 0.4456,
-      "step": 694
-    },
-    {
-      "epoch": 0.04542186785177439,
-      "grad_norm": 0.5848109126091003,
-      "learning_rate": 7.5625680087051145e-06,
-      "loss": 0.4741,
-      "step": 695
-    },
-    {
-      "epoch": 0.04548722305731651,
-      "grad_norm": 0.5010263919830322,
-      "learning_rate": 7.573449401523396e-06,
-      "loss": 0.459,
-      "step": 696
-    },
-    {
-      "epoch": 0.045552578262858634,
-      "grad_norm": 0.5514805316925049,
-      "learning_rate": 7.584330794341676e-06,
-      "loss": 0.4871,
-      "step": 697
-    },
-    {
-      "epoch": 0.045617933468400755,
-      "grad_norm": 0.5259074568748474,
-      "learning_rate": 7.595212187159957e-06,
-      "loss": 0.4667,
-      "step": 698
-    },
-    {
-      "epoch": 0.045683288673942876,
-      "grad_norm": 0.5424180030822754,
-      "learning_rate": 7.606093579978238e-06,
-      "loss": 0.457,
-      "step": 699
-    },
-    {
-      "epoch": 0.045748643879485004,
-      "grad_norm": 0.5486298203468323,
-      "learning_rate": 7.616974972796518e-06,
-      "loss": 0.5152,
-      "step": 700
-    },
-    {
-      "epoch": 0.045813999085027125,
-      "grad_norm": 0.5700769424438477,
-      "learning_rate": 7.6278563656148e-06,
-      "loss": 0.5096,
-      "step": 701
-    },
-    {
-      "epoch": 0.045879354290569246,
-      "grad_norm": 0.5538732409477234,
-      "learning_rate": 7.63873775843308e-06,
-      "loss": 0.4997,
-      "step": 702
-    },
-    {
-      "epoch": 0.04594470949611137,
-      "grad_norm": 0.5293434262275696,
-      "learning_rate": 7.64961915125136e-06,
-      "loss": 0.4286,
-      "step": 703
-    },
-    {
-      "epoch": 0.04601006470165349,
-      "grad_norm": 0.5087040662765503,
-      "learning_rate": 7.66050054406964e-06,
-      "loss": 0.4463,
-      "step": 704
-    },
-    {
-      "epoch": 0.04607541990719561,
-      "grad_norm": 0.5482538938522339,
-      "learning_rate": 7.671381936887922e-06,
-      "loss": 0.4553,
-      "step": 705
-    },
-    {
-      "epoch": 0.04614077511273773,
-      "grad_norm": 0.5640394687652588,
-      "learning_rate": 7.682263329706204e-06,
-      "loss": 0.4935,
-      "step": 706
-    },
-    {
-      "epoch": 0.04620613031827985,
-      "grad_norm": 0.49894604086875916,
-      "learning_rate": 7.693144722524484e-06,
-      "loss": 0.4091,
-      "step": 707
-    },
-    {
-      "epoch": 0.04627148552382197,
-      "grad_norm": 0.5695271492004395,
-      "learning_rate": 7.704026115342764e-06,
-      "loss": 0.4704,
-      "step": 708
-    },
-    {
-      "epoch": 0.046336840729364094,
-      "grad_norm": 0.5633416771888733,
-      "learning_rate": 7.714907508161044e-06,
-      "loss": 0.4555,
-      "step": 709
-    },
-    {
-      "epoch": 0.046402195934906215,
-      "grad_norm": 0.6127053499221802,
-      "learning_rate": 7.725788900979326e-06,
-      "loss": 0.4554,
-      "step": 710
-    },
-    {
-      "epoch": 0.046467551140448336,
-      "grad_norm": 0.4968777298927307,
-      "learning_rate": 7.736670293797607e-06,
-      "loss": 0.3906,
-      "step": 711
-    },
-    {
-      "epoch": 0.04653290634599046,
-      "grad_norm": 0.5375694632530212,
-      "learning_rate": 7.747551686615887e-06,
-      "loss": 0.4089,
-      "step": 712
-    },
-    {
-      "epoch": 0.04659826155153258,
-      "grad_norm": 0.6223379373550415,
-      "learning_rate": 7.758433079434167e-06,
-      "loss": 0.4875,
-      "step": 713
-    },
-    {
-      "epoch": 0.0466636167570747,
-      "grad_norm": 0.48704859614372253,
-      "learning_rate": 7.769314472252449e-06,
-      "loss": 0.3397,
-      "step": 714
-    },
-    {
-      "epoch": 0.04672897196261682,
-      "grad_norm": 0.5273993611335754,
-      "learning_rate": 7.780195865070729e-06,
-      "loss": 0.4646,
-      "step": 715
-    },
-    {
-      "epoch": 0.04679432716815894,
-      "grad_norm": 0.563279390335083,
-      "learning_rate": 7.791077257889011e-06,
-      "loss": 0.514,
-      "step": 716
-    },
-    {
-      "epoch": 0.04685968237370106,
-      "grad_norm": 0.5412565469741821,
-      "learning_rate": 7.801958650707291e-06,
-      "loss": 0.4696,
-      "step": 717
-    },
-    {
-      "epoch": 0.046925037579243184,
-      "grad_norm": 0.5288301706314087,
-      "learning_rate": 7.812840043525573e-06,
-      "loss": 0.4215,
-      "step": 718
-    },
-    {
-      "epoch": 0.046990392784785305,
-      "grad_norm": 0.4975229501724243,
-      "learning_rate": 7.823721436343853e-06,
-      "loss": 0.4005,
-      "step": 719
-    },
-    {
-      "epoch": 0.047055747990327426,
-      "grad_norm": 0.5747808814048767,
-      "learning_rate": 7.834602829162133e-06,
-      "loss": 0.469,
-      "step": 720
-    },
-    {
-      "epoch": 0.047121103195869554,
-      "grad_norm": 0.5444490909576416,
-      "learning_rate": 7.845484221980414e-06,
-      "loss": 0.489,
-      "step": 721
-    },
-    {
-      "epoch": 0.047186458401411675,
-      "grad_norm": 0.5794644951820374,
-      "learning_rate": 7.856365614798694e-06,
-      "loss": 0.4856,
-      "step": 722
-    },
-    {
-      "epoch": 0.047251813606953796,
-      "grad_norm": 0.5320571064949036,
-      "learning_rate": 7.867247007616976e-06,
-      "loss": 0.4433,
-      "step": 723
-    },
-    {
-      "epoch": 0.04731716881249592,
-      "grad_norm": 0.501530647277832,
-      "learning_rate": 7.878128400435256e-06,
-      "loss": 0.412,
-      "step": 724
-    },
-    {
-      "epoch": 0.04738252401803804,
-      "grad_norm": 0.5136852860450745,
-      "learning_rate": 7.889009793253538e-06,
-      "loss": 0.435,
-      "step": 725
-    },
-    {
-      "epoch": 0.04744787922358016,
-      "grad_norm": 0.5464115142822266,
-      "learning_rate": 7.899891186071818e-06,
-      "loss": 0.4377,
-      "step": 726
-    },
-    {
-      "epoch": 0.04751323442912228,
-      "grad_norm": 0.5368839502334595,
-      "learning_rate": 7.9107725788901e-06,
-      "loss": 0.4023,
-      "step": 727
-    },
-    {
-      "epoch": 0.0475785896346644,
-      "grad_norm": 0.5113937854766846,
-      "learning_rate": 7.92165397170838e-06,
-      "loss": 0.4257,
-      "step": 728
-    },
-    {
-      "epoch": 0.04764394484020652,
-      "grad_norm": 0.5719674229621887,
-      "learning_rate": 7.93253536452666e-06,
-      "loss": 0.4748,
-      "step": 729
-    },
-    {
-      "epoch": 0.047709300045748644,
-      "grad_norm": 0.5421920418739319,
-      "learning_rate": 7.943416757344941e-06,
-      "loss": 0.4914,
-      "step": 730
-    },
-    {
-      "epoch": 0.047774655251290765,
-      "grad_norm": 0.5723292231559753,
-      "learning_rate": 7.954298150163221e-06,
-      "loss": 0.4412,
-      "step": 731
-    },
-    {
-      "epoch": 0.047840010456832886,
-      "grad_norm": 0.5472431778907776,
-      "learning_rate": 7.965179542981503e-06,
-      "loss": 0.4433,
-      "step": 732
-    },
-    {
-      "epoch": 0.04790536566237501,
-      "grad_norm": 0.600566565990448,
-      "learning_rate": 7.976060935799783e-06,
-      "loss": 0.5535,
-      "step": 733
-    },
-    {
-      "epoch": 0.04797072086791713,
-      "grad_norm": 0.5550974011421204,
-      "learning_rate": 7.986942328618063e-06,
-      "loss": 0.4749,
-      "step": 734
-    },
-    {
-      "epoch": 0.04803607607345925,
-      "grad_norm": 0.5856187343597412,
-      "learning_rate": 7.997823721436345e-06,
-      "loss": 0.5042,
-      "step": 735
-    },
-    {
-      "epoch": 0.04810143127900137,
-      "grad_norm": 0.5473790168762207,
-      "learning_rate": 8.008705114254626e-06,
-      "loss": 0.487,
-      "step": 736
-    },
-    {
-      "epoch": 0.04816678648454349,
-      "grad_norm": 0.549591064453125,
-      "learning_rate": 8.019586507072906e-06,
-      "loss": 0.4657,
-      "step": 737
-    },
-    {
-      "epoch": 0.04823214169008561,
-      "grad_norm": 0.5372916460037231,
-      "learning_rate": 8.030467899891186e-06,
-      "loss": 0.4445,
-      "step": 738
-    },
-    {
-      "epoch": 0.048297496895627734,
-      "grad_norm": 0.5481677651405334,
-      "learning_rate": 8.041349292709466e-06,
-      "loss": 0.4527,
-      "step": 739
-    },
-    {
-      "epoch": 0.048362852101169855,
-      "grad_norm": 0.5766710638999939,
-      "learning_rate": 8.052230685527748e-06,
-      "loss": 0.4772,
-      "step": 740
-    },
-    {
-      "epoch": 0.048428207306711976,
-      "grad_norm": 0.5556133985519409,
-      "learning_rate": 8.06311207834603e-06,
-      "loss": 0.3829,
-      "step": 741
-    },
-    {
-      "epoch": 0.048493562512254104,
-      "grad_norm": 0.5223917365074158,
-      "learning_rate": 8.07399347116431e-06,
-      "loss": 0.4482,
-      "step": 742
-    },
-    {
-      "epoch": 0.048558917717796225,
-      "grad_norm": 0.5459646582603455,
-      "learning_rate": 8.08487486398259e-06,
-      "loss": 0.4289,
-      "step": 743
-    },
-    {
-      "epoch": 0.048624272923338346,
-      "grad_norm": 0.5293027758598328,
-      "learning_rate": 8.095756256800872e-06,
-      "loss": 0.4199,
-      "step": 744
-    },
-    {
-      "epoch": 0.04868962812888047,
-      "grad_norm": 0.5380204319953918,
-      "learning_rate": 8.106637649619152e-06,
-      "loss": 0.4665,
-      "step": 745
-    },
-    {
-      "epoch": 0.04875498333442259,
-      "grad_norm": 0.5344138145446777,
-      "learning_rate": 8.117519042437433e-06,
-      "loss": 0.4722,
-      "step": 746
-    },
-    {
-      "epoch": 0.04882033853996471,
-      "grad_norm": 0.518205463886261,
-      "learning_rate": 8.128400435255713e-06,
-      "loss": 0.4425,
-      "step": 747
-    },
-    {
-      "epoch": 0.04888569374550683,
-      "grad_norm": 0.5475545525550842,
-      "learning_rate": 8.139281828073993e-06,
-      "loss": 0.4778,
-      "step": 748
-    },
-    {
-      "epoch": 0.04895104895104895,
-      "grad_norm": 0.57927006483078,
-      "learning_rate": 8.150163220892275e-06,
-      "loss": 0.4909,
-      "step": 749
-    },
-    {
-      "epoch": 0.04901640415659107,
-      "grad_norm": 0.5073854923248291,
-      "learning_rate": 8.161044613710557e-06,
-      "loss": 0.4322,
-      "step": 750
-    },
-    {
-      "epoch": 0.049081759362133194,
-      "grad_norm": 0.5383774638175964,
-      "learning_rate": 8.171926006528837e-06,
-      "loss": 0.4194,
-      "step": 751
-    },
-    {
-      "epoch": 0.049147114567675315,
-      "grad_norm": 0.5598601698875427,
-      "learning_rate": 8.182807399347117e-06,
-      "loss": 0.4452,
-      "step": 752
-    },
-    {
-      "epoch": 0.049212469773217436,
-      "grad_norm": 0.5418949127197266,
-      "learning_rate": 8.193688792165397e-06,
-      "loss": 0.4533,
-      "step": 753
-    },
-    {
-      "epoch": 0.04927782497875956,
-      "grad_norm": 0.5240333080291748,
-      "learning_rate": 8.204570184983678e-06,
-      "loss": 0.4535,
-      "step": 754
-    },
-    {
-      "epoch": 0.04934318018430168,
-      "grad_norm": 0.6091196537017822,
-      "learning_rate": 8.21545157780196e-06,
-      "loss": 0.4184,
-      "step": 755
-    },
-    {
-      "epoch": 0.0494085353898438,
-      "grad_norm": 0.5421688556671143,
-      "learning_rate": 8.22633297062024e-06,
-      "loss": 0.4393,
-      "step": 756
-    },
-    {
-      "epoch": 0.04947389059538592,
-      "grad_norm": 0.5601967573165894,
-      "learning_rate": 8.23721436343852e-06,
-      "loss": 0.4746,
-      "step": 757
-    },
-    {
-      "epoch": 0.04953924580092804,
-      "grad_norm": 0.5984824895858765,
-      "learning_rate": 8.248095756256802e-06,
-      "loss": 0.5617,
-      "step": 758
-    },
-    {
-      "epoch": 0.04960460100647016,
-      "grad_norm": 0.5614069104194641,
-      "learning_rate": 8.258977149075082e-06,
-      "loss": 0.5006,
-      "step": 759
-    },
-    {
-      "epoch": 0.049669956212012284,
-      "grad_norm": 0.559607207775116,
-      "learning_rate": 8.269858541893364e-06,
-      "loss": 0.4674,
-      "step": 760
-    },
-    {
-      "epoch": 0.049735311417554405,
-      "grad_norm": 0.5390970706939697,
-      "learning_rate": 8.280739934711644e-06,
-      "loss": 0.4281,
-      "step": 761
-    },
-    {
-      "epoch": 0.04980066662309653,
-      "grad_norm": 0.5136409997940063,
-      "learning_rate": 8.291621327529924e-06,
-      "loss": 0.4185,
-      "step": 762
-    },
-    {
-      "epoch": 0.049866021828638654,
-      "grad_norm": 0.506865918636322,
-      "learning_rate": 8.302502720348205e-06,
-      "loss": 0.4401,
-      "step": 763
-    },
-    {
-      "epoch": 0.049931377034180775,
-      "grad_norm": 0.5017704367637634,
-      "learning_rate": 8.313384113166485e-06,
-      "loss": 0.3789,
-      "step": 764
-    },
-    {
-      "epoch": 0.049996732239722896,
-      "grad_norm": 0.5536909699440002,
-      "learning_rate": 8.324265505984767e-06,
-      "loss": 0.4543,
-      "step": 765
-    },
-    {
-      "epoch": 0.05006208744526502,
-      "grad_norm": 0.5310081839561462,
-      "learning_rate": 8.335146898803047e-06,
-      "loss": 0.4019,
-      "step": 766
-    },
-    {
-      "epoch": 0.05012744265080714,
-      "grad_norm": 0.5387538075447083,
-      "learning_rate": 8.346028291621329e-06,
-      "loss": 0.4151,
-      "step": 767
-    },
-    {
-      "epoch": 0.05019279785634926,
-      "grad_norm": 0.5525701642036438,
-      "learning_rate": 8.356909684439609e-06,
-      "loss": 0.4788,
-      "step": 768
-    },
-    {
-      "epoch": 0.05025815306189138,
-      "grad_norm": 0.557262659072876,
-      "learning_rate": 8.367791077257889e-06,
-      "loss": 0.447,
-      "step": 769
-    },
-    {
-      "epoch": 0.0503235082674335,
-      "grad_norm": 0.5567659735679626,
-      "learning_rate": 8.37867247007617e-06,
-      "loss": 0.4929,
-      "step": 770
-    },
-    {
-      "epoch": 0.05038886347297562,
-      "grad_norm": 0.5414775013923645,
-      "learning_rate": 8.389553862894452e-06,
-      "loss": 0.4282,
-      "step": 771
-    },
-    {
-      "epoch": 0.050454218678517744,
-      "grad_norm": 0.6173386573791504,
-      "learning_rate": 8.400435255712732e-06,
-      "loss": 0.5012,
-      "step": 772
-    },
-    {
-      "epoch": 0.050519573884059865,
-      "grad_norm": 0.5573265552520752,
-      "learning_rate": 8.411316648531012e-06,
-      "loss": 0.4603,
-      "step": 773
-    },
-    {
-      "epoch": 0.050584929089601986,
-      "grad_norm": 0.5525687336921692,
-      "learning_rate": 8.422198041349294e-06,
-      "loss": 0.4163,
-      "step": 774
-    },
-    {
-      "epoch": 0.05065028429514411,
-      "grad_norm": 0.5632034540176392,
-      "learning_rate": 8.433079434167574e-06,
-      "loss": 0.4471,
-      "step": 775
-    },
-    {
-      "epoch": 0.05071563950068623,
-      "grad_norm": 0.4934619665145874,
-      "learning_rate": 8.443960826985856e-06,
-      "loss": 0.3795,
-      "step": 776
-    },
-    {
-      "epoch": 0.05078099470622835,
-      "grad_norm": 0.5690104961395264,
-      "learning_rate": 8.454842219804136e-06,
-      "loss": 0.4735,
-      "step": 777
-    },
-    {
-      "epoch": 0.05084634991177047,
-      "grad_norm": 0.5554296970367432,
-      "learning_rate": 8.465723612622416e-06,
-      "loss": 0.474,
-      "step": 778
-    },
-    {
-      "epoch": 0.05091170511731259,
-      "grad_norm": 0.5207845568656921,
-      "learning_rate": 8.476605005440697e-06,
-      "loss": 0.3919,
-      "step": 779
-    },
-    {
-      "epoch": 0.05097706032285471,
-      "grad_norm": 0.5561384558677673,
-      "learning_rate": 8.487486398258979e-06,
-      "loss": 0.5127,
-      "step": 780
-    },
-    {
-      "epoch": 0.051042415528396834,
-      "grad_norm": 0.516481339931488,
-      "learning_rate": 8.498367791077259e-06,
-      "loss": 0.4004,
-      "step": 781
-    },
-    {
-      "epoch": 0.051107770733938955,
-      "grad_norm": 0.538422167301178,
-      "learning_rate": 8.509249183895539e-06,
-      "loss": 0.4719,
-      "step": 782
-    },
-    {
-      "epoch": 0.05117312593948108,
-      "grad_norm": 0.5367478132247925,
-      "learning_rate": 8.520130576713819e-06,
-      "loss": 0.4235,
-      "step": 783
-    },
-    {
-      "epoch": 0.051238481145023204,
-      "grad_norm": 0.5337499380111694,
-      "learning_rate": 8.531011969532101e-06,
-      "loss": 0.4185,
-      "step": 784
-    },
-    {
-      "epoch": 0.051303836350565325,
-      "grad_norm": 0.5809076428413391,
-      "learning_rate": 8.541893362350383e-06,
-      "loss": 0.5318,
-      "step": 785
-    },
-    {
-      "epoch": 0.051369191556107446,
-      "grad_norm": 0.5143489241600037,
-      "learning_rate": 8.552774755168663e-06,
-      "loss": 0.4697,
-      "step": 786
-    },
-    {
-      "epoch": 0.05143454676164957,
-      "grad_norm": 0.5569016337394714,
-      "learning_rate": 8.563656147986943e-06,
-      "loss": 0.4646,
-      "step": 787
-    },
-    {
-      "epoch": 0.05149990196719169,
-      "grad_norm": 0.5275742411613464,
-      "learning_rate": 8.574537540805223e-06,
-      "loss": 0.4293,
-      "step": 788
-    },
-    {
-      "epoch": 0.05156525717273381,
-      "grad_norm": 0.5588119029998779,
-      "learning_rate": 8.585418933623504e-06,
-      "loss": 0.3993,
-      "step": 789
-    },
-    {
-      "epoch": 0.05163061237827593,
-      "grad_norm": 0.5158429741859436,
-      "learning_rate": 8.596300326441786e-06,
-      "loss": 0.468,
-      "step": 790
-    },
-    {
-      "epoch": 0.05169596758381805,
-      "grad_norm": 0.5410779118537903,
-      "learning_rate": 8.607181719260066e-06,
-      "loss": 0.5098,
-      "step": 791
-    },
-    {
-      "epoch": 0.05176132278936017,
-      "grad_norm": 0.5506214499473572,
-      "learning_rate": 8.618063112078346e-06,
-      "loss": 0.4229,
-      "step": 792
-    },
-    {
-      "epoch": 0.051826677994902294,
-      "grad_norm": 0.5361695885658264,
-      "learning_rate": 8.628944504896628e-06,
-      "loss": 0.4313,
-      "step": 793
-    },
-    {
-      "epoch": 0.051892033200444415,
-      "grad_norm": 0.5345651507377625,
-      "learning_rate": 8.639825897714908e-06,
-      "loss": 0.4366,
-      "step": 794
-    },
-    {
-      "epoch": 0.051957388405986536,
-      "grad_norm": 0.5610102415084839,
-      "learning_rate": 8.65070729053319e-06,
-      "loss": 0.4524,
-      "step": 795
-    },
-    {
-      "epoch": 0.05202274361152866,
-      "grad_norm": 0.5412043333053589,
-      "learning_rate": 8.66158868335147e-06,
-      "loss": 0.427,
-      "step": 796
-    },
-    {
-      "epoch": 0.05208809881707078,
-      "grad_norm": 0.5883796811103821,
-      "learning_rate": 8.67247007616975e-06,
-      "loss": 0.4357,
-      "step": 797
-    },
-    {
-      "epoch": 0.0521534540226129,
-      "grad_norm": 0.527977705001831,
-      "learning_rate": 8.683351468988031e-06,
-      "loss": 0.4123,
-      "step": 798
-    },
-    {
-      "epoch": 0.05221880922815502,
-      "grad_norm": 0.5837652683258057,
-      "learning_rate": 8.694232861806313e-06,
-      "loss": 0.4355,
-      "step": 799
-    },
-    {
-      "epoch": 0.05228416443369714,
-      "grad_norm": 0.5441904664039612,
-      "learning_rate": 8.705114254624593e-06,
-      "loss": 0.4301,
-      "step": 800
-    },
-    {
-      "epoch": 0.05234951963923926,
-      "grad_norm": 0.5277956128120422,
-      "learning_rate": 8.715995647442873e-06,
-      "loss": 0.4592,
-      "step": 801
-    },
-    {
-      "epoch": 0.052414874844781384,
-      "grad_norm": 0.5981716513633728,
-      "learning_rate": 8.726877040261155e-06,
-      "loss": 0.4253,
-      "step": 802
-    },
-    {
-      "epoch": 0.052480230050323505,
-      "grad_norm": 0.5430685877799988,
-      "learning_rate": 8.737758433079435e-06,
-      "loss": 0.456,
-      "step": 803
-    },
-    {
-      "epoch": 0.05254558525586563,
-      "grad_norm": 0.5334436893463135,
-      "learning_rate": 8.748639825897716e-06,
-      "loss": 0.3857,
-      "step": 804
-    },
-    {
-      "epoch": 0.052610940461407754,
-      "grad_norm": 0.5099507570266724,
-      "learning_rate": 8.759521218715996e-06,
-      "loss": 0.4,
-      "step": 805
-    },
-    {
-      "epoch": 0.052676295666949875,
-      "grad_norm": 0.5122255086898804,
-      "learning_rate": 8.770402611534276e-06,
-      "loss": 0.4687,
-      "step": 806
-    },
-    {
-      "epoch": 0.052741650872491996,
-      "grad_norm": 0.5638749599456787,
-      "learning_rate": 8.781284004352558e-06,
-      "loss": 0.4455,
-      "step": 807
-    },
-    {
-      "epoch": 0.05280700607803412,
-      "grad_norm": 0.7972688674926758,
-      "learning_rate": 8.792165397170838e-06,
-      "loss": 0.4383,
-      "step": 808
-    },
-    {
-      "epoch": 0.05287236128357624,
-      "grad_norm": 0.49088436365127563,
-      "learning_rate": 8.80304678998912e-06,
-      "loss": 0.3938,
-      "step": 809
-    },
-    {
-      "epoch": 0.05293771648911836,
-      "grad_norm": 0.5083533525466919,
-      "learning_rate": 8.8139281828074e-06,
-      "loss": 0.3863,
-      "step": 810
-    },
-    {
-      "epoch": 0.05300307169466048,
-      "grad_norm": 0.5804579257965088,
-      "learning_rate": 8.824809575625682e-06,
-      "loss": 0.5243,
-      "step": 811
-    },
-    {
-      "epoch": 0.0530684269002026,
-      "grad_norm": 0.517282247543335,
-      "learning_rate": 8.835690968443962e-06,
-      "loss": 0.4416,
-      "step": 812
-    },
-    {
-      "epoch": 0.05313378210574472,
-      "grad_norm": 0.579800009727478,
-      "learning_rate": 8.846572361262242e-06,
-      "loss": 0.5161,
-      "step": 813
-    },
-    {
-      "epoch": 0.053199137311286844,
-      "grad_norm": 0.5247344970703125,
-      "learning_rate": 8.857453754080523e-06,
-      "loss": 0.4162,
-      "step": 814
-    },
-    {
-      "epoch": 0.053264492516828965,
-      "grad_norm": 0.5368344187736511,
-      "learning_rate": 8.868335146898803e-06,
-      "loss": 0.4507,
-      "step": 815
-    },
-    {
-      "epoch": 0.053329847722371086,
-      "grad_norm": 0.558159589767456,
-      "learning_rate": 8.879216539717085e-06,
-      "loss": 0.5208,
-      "step": 816
-    },
-    {
-      "epoch": 0.05339520292791321,
-      "grad_norm": 0.5349926948547363,
-      "learning_rate": 8.890097932535365e-06,
-      "loss": 0.4552,
-      "step": 817
-    },
-    {
-      "epoch": 0.05346055813345533,
-      "grad_norm": 0.5398790836334229,
-      "learning_rate": 8.900979325353645e-06,
-      "loss": 0.4633,
-      "step": 818
-    },
-    {
-      "epoch": 0.05352591333899745,
-      "grad_norm": 0.5353681445121765,
-      "learning_rate": 8.911860718171927e-06,
-      "loss": 0.4302,
-      "step": 819
-    },
-    {
-      "epoch": 0.05359126854453957,
-      "grad_norm": 0.5524716973304749,
-      "learning_rate": 8.922742110990208e-06,
-      "loss": 0.4916,
-      "step": 820
-    },
-    {
-      "epoch": 0.05365662375008169,
-      "grad_norm": 0.49374276399612427,
-      "learning_rate": 8.933623503808488e-06,
-      "loss": 0.3643,
-      "step": 821
-    },
-    {
-      "epoch": 0.05372197895562381,
-      "grad_norm": 0.5321981906890869,
-      "learning_rate": 8.944504896626768e-06,
-      "loss": 0.4435,
-      "step": 822
-    },
-    {
-      "epoch": 0.053787334161165934,
-      "grad_norm": 0.5989148020744324,
-      "learning_rate": 8.95538628944505e-06,
-      "loss": 0.4757,
-      "step": 823
-    },
-    {
-      "epoch": 0.053852689366708055,
-      "grad_norm": 0.5643342137336731,
-      "learning_rate": 8.96626768226333e-06,
-      "loss": 0.5149,
-      "step": 824
-    },
-    {
-      "epoch": 0.05391804457225018,
-      "grad_norm": 0.5305259227752686,
-      "learning_rate": 8.977149075081612e-06,
-      "loss": 0.4443,
-      "step": 825
-    },
-    {
-      "epoch": 0.053983399777792304,
-      "grad_norm": 0.5195196270942688,
-      "learning_rate": 8.988030467899892e-06,
-      "loss": 0.4115,
-      "step": 826
-    },
-    {
-      "epoch": 0.054048754983334425,
-      "grad_norm": 0.5913254618644714,
-      "learning_rate": 8.998911860718172e-06,
-      "loss": 0.4783,
-      "step": 827
-    },
-    {
-      "epoch": 0.054114110188876546,
-      "grad_norm": 0.553651750087738,
-      "learning_rate": 9.009793253536454e-06,
-      "loss": 0.4455,
-      "step": 828
-    },
-    {
-      "epoch": 0.05417946539441867,
-      "grad_norm": 0.5101826190948486,
-      "learning_rate": 9.020674646354735e-06,
-      "loss": 0.358,
-      "step": 829
-    },
-    {
-      "epoch": 0.05424482059996079,
-      "grad_norm": 0.5553346872329712,
-      "learning_rate": 9.031556039173015e-06,
-      "loss": 0.4284,
-      "step": 830
-    },
-    {
-      "epoch": 0.05431017580550291,
-      "grad_norm": 0.5712149739265442,
-      "learning_rate": 9.042437431991295e-06,
-      "loss": 0.485,
-      "step": 831
-    },
-    {
-      "epoch": 0.05437553101104503,
-      "grad_norm": 0.5390961170196533,
-      "learning_rate": 9.053318824809575e-06,
-      "loss": 0.4059,
-      "step": 832
-    },
-    {
-      "epoch": 0.05444088621658715,
-      "grad_norm": 0.5263479351997375,
-      "learning_rate": 9.064200217627857e-06,
-      "loss": 0.4044,
-      "step": 833
-    },
-    {
-      "epoch": 0.05450624142212927,
-      "grad_norm": 0.5573816895484924,
-      "learning_rate": 9.075081610446139e-06,
-      "loss": 0.475,
-      "step": 834
-    },
-    {
-      "epoch": 0.054571596627671394,
-      "grad_norm": 0.5833741426467896,
-      "learning_rate": 9.085963003264419e-06,
-      "loss": 0.4946,
-      "step": 835
-    },
-    {
-      "epoch": 0.054636951833213515,
-      "grad_norm": 0.5497047901153564,
-      "learning_rate": 9.096844396082699e-06,
-      "loss": 0.4481,
-      "step": 836
-    },
-    {
-      "epoch": 0.054702307038755636,
-      "grad_norm": 0.5038496255874634,
-      "learning_rate": 9.107725788900979e-06,
-      "loss": 0.4085,
-      "step": 837
-    },
-    {
-      "epoch": 0.05476766224429776,
-      "grad_norm": 0.5255208015441895,
-      "learning_rate": 9.11860718171926e-06,
-      "loss": 0.4543,
-      "step": 838
-    },
-    {
-      "epoch": 0.05483301744983988,
-      "grad_norm": 0.4954850673675537,
-      "learning_rate": 9.129488574537542e-06,
-      "loss": 0.3709,
-      "step": 839
-    },
-    {
-      "epoch": 0.054898372655382,
-      "grad_norm": 0.5355923175811768,
-      "learning_rate": 9.140369967355822e-06,
-      "loss": 0.4283,
-      "step": 840
-    },
-    {
-      "epoch": 0.05496372786092412,
-      "grad_norm": 0.6211481094360352,
-      "learning_rate": 9.151251360174102e-06,
-      "loss": 0.5203,
-      "step": 841
-    },
-    {
-      "epoch": 0.05502908306646624,
-      "grad_norm": 0.566294252872467,
-      "learning_rate": 9.162132752992384e-06,
-      "loss": 0.4664,
-      "step": 842
-    },
-    {
-      "epoch": 0.05509443827200836,
-      "grad_norm": 0.557698667049408,
-      "learning_rate": 9.173014145810664e-06,
-      "loss": 0.4644,
-      "step": 843
-    },
-    {
-      "epoch": 0.055159793477550484,
-      "grad_norm": 0.5852855443954468,
-      "learning_rate": 9.183895538628946e-06,
-      "loss": 0.4967,
-      "step": 844
-    },
-    {
-      "epoch": 0.055225148683092605,
-      "grad_norm": 0.5442548990249634,
-      "learning_rate": 9.194776931447226e-06,
-      "loss": 0.4765,
-      "step": 845
-    },
-    {
-      "epoch": 0.05529050388863473,
-      "grad_norm": 0.591444730758667,
-      "learning_rate": 9.205658324265506e-06,
-      "loss": 0.4683,
-      "step": 846
-    },
-    {
-      "epoch": 0.055355859094176854,
-      "grad_norm": 0.5280168652534485,
-      "learning_rate": 9.216539717083787e-06,
-      "loss": 0.396,
-      "step": 847
-    },
-    {
-      "epoch": 0.055421214299718975,
-      "grad_norm": 0.5102829337120056,
-      "learning_rate": 9.227421109902069e-06,
-      "loss": 0.4289,
-      "step": 848
-    },
-    {
-      "epoch": 0.055486569505261096,
-      "grad_norm": 0.581202507019043,
-      "learning_rate": 9.238302502720349e-06,
-      "loss": 0.4502,
-      "step": 849
-    },
-    {
-      "epoch": 0.05555192471080322,
-      "grad_norm": 0.5247049331665039,
-      "learning_rate": 9.249183895538629e-06,
-      "loss": 0.4348,
-      "step": 850
-    },
-    {
-      "epoch": 0.05561727991634534,
-      "grad_norm": 0.5914639234542847,
-      "learning_rate": 9.26006528835691e-06,
-      "loss": 0.5089,
-      "step": 851
-    },
-    {
-      "epoch": 0.05568263512188746,
-      "grad_norm": 0.532891035079956,
-      "learning_rate": 9.27094668117519e-06,
-      "loss": 0.4431,
-      "step": 852
-    },
-    {
-      "epoch": 0.05574799032742958,
-      "grad_norm": 0.5694881677627563,
-      "learning_rate": 9.281828073993473e-06,
-      "loss": 0.4512,
-      "step": 853
-    },
-    {
-      "epoch": 0.0558133455329717,
-      "grad_norm": 0.596929669380188,
-      "learning_rate": 9.292709466811753e-06,
-      "loss": 0.4639,
-      "step": 854
-    },
-    {
-      "epoch": 0.05587870073851382,
-      "grad_norm": 0.5148254632949829,
-      "learning_rate": 9.303590859630034e-06,
-      "loss": 0.4262,
-      "step": 855
-    },
-    {
-      "epoch": 0.055944055944055944,
-      "grad_norm": 0.5276803970336914,
-      "learning_rate": 9.314472252448314e-06,
-      "loss": 0.4571,
-      "step": 856
-    },
-    {
-      "epoch": 0.056009411149598065,
-      "grad_norm": 0.553871214389801,
-      "learning_rate": 9.325353645266594e-06,
-      "loss": 0.4855,
-      "step": 857
-    },
-    {
-      "epoch": 0.056074766355140186,
-      "grad_norm": 0.5498103499412537,
-      "learning_rate": 9.336235038084876e-06,
-      "loss": 0.4508,
-      "step": 858
-    },
-    {
-      "epoch": 0.05614012156068231,
-      "grad_norm": 0.571270763874054,
-      "learning_rate": 9.347116430903156e-06,
-      "loss": 0.4735,
-      "step": 859
-    },
-    {
-      "epoch": 0.05620547676622443,
-      "grad_norm": 0.48032382130622864,
-      "learning_rate": 9.357997823721438e-06,
-      "loss": 0.3663,
-      "step": 860
-    },
-    {
-      "epoch": 0.05627083197176655,
-      "grad_norm": 0.5407636761665344,
-      "learning_rate": 9.368879216539718e-06,
-      "loss": 0.4516,
-      "step": 861
-    },
-    {
-      "epoch": 0.05633618717730867,
-      "grad_norm": 0.5646923780441284,
-      "learning_rate": 9.379760609357998e-06,
-      "loss": 0.4501,
-      "step": 862
-    },
-    {
-      "epoch": 0.05640154238285079,
-      "grad_norm": 0.5731817483901978,
-      "learning_rate": 9.39064200217628e-06,
-      "loss": 0.4508,
-      "step": 863
-    },
-    {
-      "epoch": 0.05646689758839291,
-      "grad_norm": 0.6380143761634827,
-      "learning_rate": 9.401523394994561e-06,
-      "loss": 0.468,
-      "step": 864
-    },
-    {
-      "epoch": 0.056532252793935034,
-      "grad_norm": 0.5581998229026794,
-      "learning_rate": 9.412404787812841e-06,
-      "loss": 0.4683,
-      "step": 865
-    },
-    {
-      "epoch": 0.056597607999477155,
-      "grad_norm": 0.5660893321037292,
-      "learning_rate": 9.423286180631121e-06,
-      "loss": 0.4792,
-      "step": 866
-    },
-    {
-      "epoch": 0.05666296320501928,
-      "grad_norm": 0.46633392572402954,
-      "learning_rate": 9.434167573449401e-06,
-      "loss": 0.3853,
-      "step": 867
-    },
-    {
-      "epoch": 0.056728318410561404,
-      "grad_norm": 0.6074538230895996,
-      "learning_rate": 9.445048966267683e-06,
-      "loss": 0.5162,
-      "step": 868
-    },
-    {
-      "epoch": 0.056793673616103525,
-      "grad_norm": 0.5075995326042175,
-      "learning_rate": 9.455930359085965e-06,
-      "loss": 0.4245,
-      "step": 869
-    },
-    {
-      "epoch": 0.056859028821645646,
-      "grad_norm": 0.5416386723518372,
-      "learning_rate": 9.466811751904245e-06,
-      "loss": 0.4191,
-      "step": 870
-    },
-    {
-      "epoch": 0.05692438402718777,
-      "grad_norm": 0.5612762570381165,
-      "learning_rate": 9.477693144722525e-06,
-      "loss": 0.4899,
-      "step": 871
-    },
-    {
-      "epoch": 0.05698973923272989,
-      "grad_norm": 0.5579902529716492,
-      "learning_rate": 9.488574537540806e-06,
-      "loss": 0.4449,
-      "step": 872
-    },
-    {
-      "epoch": 0.05705509443827201,
-      "grad_norm": 0.6033218502998352,
-      "learning_rate": 9.499455930359086e-06,
-      "loss": 0.5391,
-      "step": 873
-    },
-    {
-      "epoch": 0.05712044964381413,
-      "grad_norm": 0.5937187075614929,
-      "learning_rate": 9.510337323177368e-06,
-      "loss": 0.4332,
-      "step": 874
-    },
-    {
-      "epoch": 0.05718580484935625,
-      "grad_norm": 0.5878888368606567,
-      "learning_rate": 9.521218715995648e-06,
-      "loss": 0.5223,
-      "step": 875
-    },
-    {
-      "epoch": 0.05725116005489837,
-      "grad_norm": 0.5516737103462219,
-      "learning_rate": 9.532100108813928e-06,
-      "loss": 0.4321,
-      "step": 876
-    },
-    {
-      "epoch": 0.057316515260440494,
-      "grad_norm": 0.5144123435020447,
-      "learning_rate": 9.54298150163221e-06,
-      "loss": 0.352,
-      "step": 877
-    },
-    {
-      "epoch": 0.057381870465982615,
-      "grad_norm": 0.5788304209709167,
-      "learning_rate": 9.553862894450491e-06,
-      "loss": 0.4312,
-      "step": 878
-    },
-    {
-      "epoch": 0.057447225671524736,
-      "grad_norm": 0.558205783367157,
-      "learning_rate": 9.564744287268772e-06,
-      "loss": 0.4623,
-      "step": 879
-    },
-    {
-      "epoch": 0.05751258087706686,
-      "grad_norm": 0.564784824848175,
-      "learning_rate": 9.575625680087052e-06,
-      "loss": 0.484,
-      "step": 880
-    },
-    {
-      "epoch": 0.05757793608260898,
-      "grad_norm": 0.5264977812767029,
-      "learning_rate": 9.586507072905332e-06,
-      "loss": 0.4197,
-      "step": 881
-    },
-    {
-      "epoch": 0.0576432912881511,
-      "grad_norm": 0.5363552570343018,
-      "learning_rate": 9.597388465723613e-06,
-      "loss": 0.4211,
-      "step": 882
-    },
-    {
-      "epoch": 0.05770864649369322,
-      "grad_norm": 0.5306923985481262,
-      "learning_rate": 9.608269858541895e-06,
-      "loss": 0.4587,
-      "step": 883
-    },
-    {
-      "epoch": 0.05777400169923534,
-      "grad_norm": 0.5312590003013611,
-      "learning_rate": 9.619151251360175e-06,
-      "loss": 0.4374,
-      "step": 884
-    },
-    {
-      "epoch": 0.05783935690477746,
-      "grad_norm": 0.5317679643630981,
-      "learning_rate": 9.630032644178455e-06,
-      "loss": 0.469,
-      "step": 885
-    },
-    {
-      "epoch": 0.057904712110319584,
-      "grad_norm": 0.5046166181564331,
-      "learning_rate": 9.640914036996737e-06,
-      "loss": 0.4133,
-      "step": 886
-    },
-    {
-      "epoch": 0.057970067315861705,
-      "grad_norm": 0.6376034021377563,
-      "learning_rate": 9.651795429815017e-06,
-      "loss": 0.4476,
-      "step": 887
-    },
-    {
-      "epoch": 0.05803542252140383,
-      "grad_norm": 0.5270693302154541,
-      "learning_rate": 9.662676822633298e-06,
-      "loss": 0.4211,
-      "step": 888
-    },
-    {
-      "epoch": 0.058100777726945954,
-      "grad_norm": 0.5376937389373779,
-      "learning_rate": 9.673558215451578e-06,
-      "loss": 0.4576,
-      "step": 889
-    },
-    {
-      "epoch": 0.058166132932488075,
-      "grad_norm": 0.5243609547615051,
-      "learning_rate": 9.684439608269858e-06,
-      "loss": 0.4567,
-      "step": 890
-    },
-    {
-      "epoch": 0.058231488138030196,
-      "grad_norm": 0.5273962020874023,
-      "learning_rate": 9.69532100108814e-06,
-      "loss": 0.4256,
-      "step": 891
-    },
-    {
-      "epoch": 0.05829684334357232,
-      "grad_norm": 0.5864324569702148,
-      "learning_rate": 9.70620239390642e-06,
-      "loss": 0.4907,
-      "step": 892
-    },
-    {
-      "epoch": 0.05836219854911444,
-      "grad_norm": 0.5216048359870911,
-      "learning_rate": 9.717083786724702e-06,
-      "loss": 0.4245,
-      "step": 893
-    },
-    {
-      "epoch": 0.05842755375465656,
-      "grad_norm": 0.48350924253463745,
-      "learning_rate": 9.727965179542982e-06,
-      "loss": 0.388,
-      "step": 894
-    },
-    {
-      "epoch": 0.05849290896019868,
-      "grad_norm": 0.5730065107345581,
-      "learning_rate": 9.738846572361264e-06,
-      "loss": 0.4364,
-      "step": 895
-    },
-    {
-      "epoch": 0.0585582641657408,
-      "grad_norm": 0.5184619426727295,
-      "learning_rate": 9.749727965179544e-06,
-      "loss": 0.4087,
-      "step": 896
-    },
-    {
-      "epoch": 0.05862361937128292,
-      "grad_norm": 0.5821964740753174,
-      "learning_rate": 9.760609357997825e-06,
-      "loss": 0.5004,
-      "step": 897
-    },
-    {
-      "epoch": 0.058688974576825044,
-      "grad_norm": 0.5208947062492371,
-      "learning_rate": 9.771490750816105e-06,
-      "loss": 0.3913,
-      "step": 898
-    },
-    {
-      "epoch": 0.058754329782367165,
-      "grad_norm": 0.5657762885093689,
-      "learning_rate": 9.782372143634385e-06,
-      "loss": 0.454,
-      "step": 899
-    },
-    {
-      "epoch": 0.058819684987909286,
-      "grad_norm": 0.5364660620689392,
-      "learning_rate": 9.793253536452667e-06,
-      "loss": 0.4305,
-      "step": 900
-    },
-    {
-      "epoch": 0.05888504019345141,
-      "grad_norm": 0.510415256023407,
-      "learning_rate": 9.804134929270947e-06,
-      "loss": 0.4463,
-      "step": 901
-    },
-    {
-      "epoch": 0.05895039539899353,
-      "grad_norm": 0.6170061826705933,
-      "learning_rate": 9.815016322089229e-06,
-      "loss": 0.4997,
-      "step": 902
-    },
-    {
-      "epoch": 0.05901575060453565,
-      "grad_norm": 0.6070553064346313,
-      "learning_rate": 9.825897714907509e-06,
-      "loss": 0.4979,
-      "step": 903
-    },
-    {
-      "epoch": 0.05908110581007777,
-      "grad_norm": 0.5229129791259766,
-      "learning_rate": 9.83677910772579e-06,
-      "loss": 0.3985,
-      "step": 904
-    },
-    {
-      "epoch": 0.05914646101561989,
-      "grad_norm": 0.5174643397331238,
-      "learning_rate": 9.84766050054407e-06,
-      "loss": 0.4483,
-      "step": 905
-    },
-    {
-      "epoch": 0.05921181622116201,
-      "grad_norm": 0.5256576538085938,
-      "learning_rate": 9.85854189336235e-06,
-      "loss": 0.3863,
-      "step": 906
-    },
-    {
-      "epoch": 0.059277171426704134,
-      "grad_norm": 0.5165674686431885,
-      "learning_rate": 9.869423286180632e-06,
-      "loss": 0.354,
-      "step": 907
-    },
-    {
-      "epoch": 0.059342526632246255,
-      "grad_norm": 0.5451213717460632,
-      "learning_rate": 9.880304678998914e-06,
-      "loss": 0.4456,
-      "step": 908
-    },
-    {
-      "epoch": 0.05940788183778838,
-      "grad_norm": 0.4831288158893585,
-      "learning_rate": 9.891186071817194e-06,
-      "loss": 0.3507,
-      "step": 909
-    },
-    {
-      "epoch": 0.059473237043330504,
-      "grad_norm": 0.510681688785553,
-      "learning_rate": 9.902067464635474e-06,
-      "loss": 0.3763,
-      "step": 910
-    },
-    {
-      "epoch": 0.059538592248872625,
-      "grad_norm": 0.6625702977180481,
-      "learning_rate": 9.912948857453754e-06,
-      "loss": 0.4688,
-      "step": 911
-    },
-    {
-      "epoch": 0.059603947454414746,
-      "grad_norm": 0.5980280637741089,
-      "learning_rate": 9.923830250272036e-06,
-      "loss": 0.4215,
-      "step": 912
-    },
-    {
-      "epoch": 0.05966930265995687,
-      "grad_norm": 0.6070137023925781,
-      "learning_rate": 9.934711643090317e-06,
-      "loss": 0.4185,
-      "step": 913
-    },
-    {
-      "epoch": 0.05973465786549899,
-      "grad_norm": 0.6734384298324585,
-      "learning_rate": 9.945593035908597e-06,
-      "loss": 0.5606,
-      "step": 914
-    },
-    {
-      "epoch": 0.05980001307104111,
-      "grad_norm": 0.574552595615387,
-      "learning_rate": 9.956474428726877e-06,
-      "loss": 0.5411,
-      "step": 915
-    },
-    {
-      "epoch": 0.05986536827658323,
-      "grad_norm": 0.5196554660797119,
-      "learning_rate": 9.967355821545159e-06,
-      "loss": 0.4218,
-      "step": 916
-    },
-    {
-      "epoch": 0.05993072348212535,
-      "grad_norm": 0.5171263813972473,
-      "learning_rate": 9.978237214363439e-06,
-      "loss": 0.4259,
-      "step": 917
-    },
-    {
-      "epoch": 0.05999607868766747,
-      "grad_norm": 0.6005182862281799,
-      "learning_rate": 9.98911860718172e-06,
-      "loss": 0.5567,
-      "step": 918
-    },
-    {
-      "epoch": 0.060061433893209594,
-      "grad_norm": 0.5370460748672485,
-      "learning_rate": 1e-05,
-      "loss": 0.4488,
-      "step": 919
-    },
-    {
-      "epoch": 0.060126789098751715,
-      "grad_norm": 0.5299703478813171,
-      "learning_rate": 9.999999987806635e-06,
-      "loss": 0.4235,
-      "step": 920
-    },
-    {
-      "epoch": 0.060192144304293836,
-      "grad_norm": 0.556443989276886,
-      "learning_rate": 9.999999951226536e-06,
-      "loss": 0.4769,
-      "step": 921
-    },
-    {
-      "epoch": 0.06025749950983596,
-      "grad_norm": 0.5349219441413879,
-      "learning_rate": 9.999999890259706e-06,
-      "loss": 0.4508,
-      "step": 922
-    },
-    {
-      "epoch": 0.06032285471537808,
-      "grad_norm": 0.548328161239624,
-      "learning_rate": 9.999999804906145e-06,
-      "loss": 0.5043,
-      "step": 923
-    },
-    {
-      "epoch": 0.0603882099209202,
-      "grad_norm": 0.5339406728744507,
-      "learning_rate": 9.999999695165852e-06,
-      "loss": 0.4141,
-      "step": 924
-    },
-    {
-      "epoch": 0.06045356512646232,
-      "grad_norm": 0.5387172698974609,
-      "learning_rate": 9.999999561038828e-06,
-      "loss": 0.4016,
-      "step": 925
-    },
-    {
-      "epoch": 0.06051892033200444,
-      "grad_norm": 0.5516391396522522,
-      "learning_rate": 9.999999402525074e-06,
-      "loss": 0.4303,
-      "step": 926
-    },
-    {
-      "epoch": 0.06058427553754656,
-      "grad_norm": 0.5340887308120728,
-      "learning_rate": 9.999999219624593e-06,
-      "loss": 0.4463,
-      "step": 927
-    },
-    {
-      "epoch": 0.060649630743088684,
-      "grad_norm": 0.5485219359397888,
-      "learning_rate": 9.99999901233738e-06,
-      "loss": 0.4164,
-      "step": 928
-    },
-    {
-      "epoch": 0.06071498594863081,
-      "grad_norm": 0.5041017532348633,
-      "learning_rate": 9.999998780663442e-06,
-      "loss": 0.3712,
-      "step": 929
-    },
-    {
-      "epoch": 0.06078034115417293,
-      "grad_norm": 0.577510416507721,
-      "learning_rate": 9.999998524602777e-06,
-      "loss": 0.4896,
-      "step": 930
-    },
-    {
-      "epoch": 0.060845696359715054,
-      "grad_norm": 0.5734471082687378,
-      "learning_rate": 9.999998244155387e-06,
-      "loss": 0.4917,
-      "step": 931
-    },
-    {
-      "epoch": 0.060911051565257175,
-      "grad_norm": 0.5420089960098267,
-      "learning_rate": 9.999997939321274e-06,
-      "loss": 0.456,
-      "step": 932
-    },
-    {
-      "epoch": 0.060976406770799296,
-      "grad_norm": 0.5438268184661865,
-      "learning_rate": 9.999997610100438e-06,
-      "loss": 0.4542,
-      "step": 933
-    },
-    {
-      "epoch": 0.06104176197634142,
-      "grad_norm": 0.5289359092712402,
-      "learning_rate": 9.999997256492882e-06,
-      "loss": 0.4261,
-      "step": 934
-    },
-    {
-      "epoch": 0.06110711718188354,
-      "grad_norm": 0.530704915523529,
-      "learning_rate": 9.999996878498607e-06,
-      "loss": 0.444,
-      "step": 935
-    },
-    {
-      "epoch": 0.06117247238742566,
-      "grad_norm": 0.5149810314178467,
-      "learning_rate": 9.999996476117614e-06,
-      "loss": 0.4656,
-      "step": 936
-    },
-    {
-      "epoch": 0.06123782759296778,
-      "grad_norm": 0.5710325837135315,
-      "learning_rate": 9.99999604934991e-06,
-      "loss": 0.4992,
-      "step": 937
-    },
-    {
-      "epoch": 0.0613031827985099,
-      "grad_norm": 0.5574349164962769,
-      "learning_rate": 9.99999559819549e-06,
-      "loss": 0.4507,
-      "step": 938
-    },
-    {
-      "epoch": 0.06136853800405202,
-      "grad_norm": 0.5291350483894348,
-      "learning_rate": 9.999995122654357e-06,
-      "loss": 0.4293,
-      "step": 939
-    },
-    {
-      "epoch": 0.061433893209594144,
-      "grad_norm": 0.5154761075973511,
-      "learning_rate": 9.99999462272652e-06,
-      "loss": 0.417,
-      "step": 940
-    },
-    {
-      "epoch": 0.061499248415136265,
-      "grad_norm": 0.5109313130378723,
-      "learning_rate": 9.999994098411975e-06,
-      "loss": 0.4051,
-      "step": 941
-    },
-    {
-      "epoch": 0.061564603620678386,
-      "grad_norm": 0.5403725504875183,
-      "learning_rate": 9.999993549710727e-06,
-      "loss": 0.4648,
-      "step": 942
-    },
-    {
-      "epoch": 0.06162995882622051,
-      "grad_norm": 0.5102765560150146,
-      "learning_rate": 9.999992976622778e-06,
-      "loss": 0.4253,
-      "step": 943
-    },
-    {
-      "epoch": 0.06169531403176263,
-      "grad_norm": 0.562269389629364,
-      "learning_rate": 9.999992379148131e-06,
-      "loss": 0.431,
-      "step": 944
-    },
-    {
-      "epoch": 0.06176066923730475,
-      "grad_norm": 0.5346286296844482,
-      "learning_rate": 9.99999175728679e-06,
-      "loss": 0.4257,
-      "step": 945
-    },
-    {
-      "epoch": 0.06182602444284687,
-      "grad_norm": 0.5671541094779968,
-      "learning_rate": 9.999991111038756e-06,
-      "loss": 0.4605,
-      "step": 946
-    },
-    {
-      "epoch": 0.06189137964838899,
-      "grad_norm": 0.5588353276252747,
-      "learning_rate": 9.999990440404034e-06,
-      "loss": 0.4266,
-      "step": 947
-    },
-    {
-      "epoch": 0.06195673485393111,
-      "grad_norm": 0.51191645860672,
-      "learning_rate": 9.999989745382626e-06,
-      "loss": 0.4281,
-      "step": 948
-    },
-    {
-      "epoch": 0.062022090059473234,
-      "grad_norm": 0.5754305124282837,
-      "learning_rate": 9.999989025974535e-06,
-      "loss": 0.4728,
-      "step": 949
-    },
-    {
-      "epoch": 0.06208744526501536,
-      "grad_norm": 0.5369451642036438,
-      "learning_rate": 9.999988282179766e-06,
-      "loss": 0.4651,
-      "step": 950
-    },
-    {
-      "epoch": 0.06215280047055748,
-      "grad_norm": 0.5306262969970703,
-      "learning_rate": 9.999987513998324e-06,
-      "loss": 0.4238,
-      "step": 951
-    },
-    {
-      "epoch": 0.062218155676099604,
-      "grad_norm": 0.5413644909858704,
-      "learning_rate": 9.999986721430208e-06,
-      "loss": 0.3978,
-      "step": 952
-    },
-    {
-      "epoch": 0.062283510881641725,
-      "grad_norm": 0.49585309624671936,
-      "learning_rate": 9.999985904475427e-06,
-      "loss": 0.4195,
-      "step": 953
-    },
-    {
-      "epoch": 0.062348866087183846,
-      "grad_norm": 0.5562087893486023,
-      "learning_rate": 9.99998506313398e-06,
-      "loss": 0.4443,
-      "step": 954
-    },
-    {
-      "epoch": 0.06241422129272597,
-      "grad_norm": 0.5618133544921875,
-      "learning_rate": 9.999984197405874e-06,
-      "loss": 0.4596,
-      "step": 955
-    },
-    {
-      "epoch": 0.06247957649826809,
-      "grad_norm": 0.5170841813087463,
-      "learning_rate": 9.999983307291115e-06,
-      "loss": 0.4443,
-      "step": 956
-    },
-    {
-      "epoch": 0.0625449317038102,
-      "grad_norm": 0.5080602169036865,
-      "learning_rate": 9.999982392789703e-06,
-      "loss": 0.409,
-      "step": 957
-    },
-    {
-      "epoch": 0.06261028690935233,
-      "grad_norm": 0.5565418601036072,
-      "learning_rate": 9.999981453901647e-06,
-      "loss": 0.4677,
-      "step": 958
-    },
-    {
-      "epoch": 0.06267564211489444,
-      "grad_norm": 0.5918434262275696,
-      "learning_rate": 9.999980490626948e-06,
-      "loss": 0.4442,
-      "step": 959
-    },
-    {
-      "epoch": 0.06274099732043657,
-      "grad_norm": 0.5190476775169373,
-      "learning_rate": 9.999979502965611e-06,
-      "loss": 0.462,
-      "step": 960
-    },
-    {
-      "epoch": 0.0628063525259787,
-      "grad_norm": 0.47892889380455017,
-      "learning_rate": 9.999978490917644e-06,
-      "loss": 0.3883,
-      "step": 961
-    },
-    {
-      "epoch": 0.06287170773152082,
-      "grad_norm": 0.579393208026886,
-      "learning_rate": 9.999977454483047e-06,
-      "loss": 0.5126,
-      "step": 962
-    },
-    {
-      "epoch": 0.06293706293706294,
-      "grad_norm": 0.58144211769104,
-      "learning_rate": 9.99997639366183e-06,
-      "loss": 0.4811,
-      "step": 963
-    },
-    {
-      "epoch": 0.06300241814260506,
-      "grad_norm": 0.5610965490341187,
-      "learning_rate": 9.999975308453996e-06,
-      "loss": 0.4737,
-      "step": 964
-    },
-    {
-      "epoch": 0.06306777334814719,
-      "grad_norm": 0.500460684299469,
-      "learning_rate": 9.999974198859548e-06,
-      "loss": 0.4236,
-      "step": 965
-    },
-    {
-      "epoch": 0.0631331285536893,
-      "grad_norm": 0.569346010684967,
-      "learning_rate": 9.999973064878496e-06,
-      "loss": 0.5006,
-      "step": 966
-    },
-    {
-      "epoch": 0.06319848375923143,
-      "grad_norm": 0.49624568223953247,
-      "learning_rate": 9.999971906510842e-06,
-      "loss": 0.3761,
-      "step": 967
-    },
-    {
-      "epoch": 0.06326383896477354,
-      "grad_norm": 0.5685256719589233,
-      "learning_rate": 9.999970723756594e-06,
-      "loss": 0.4643,
-      "step": 968
-    },
-    {
-      "epoch": 0.06332919417031567,
-      "grad_norm": 0.5329216122627258,
-      "learning_rate": 9.999969516615755e-06,
-      "loss": 0.4207,
-      "step": 969
-    },
-    {
-      "epoch": 0.06339454937585778,
-      "grad_norm": 0.5951482057571411,
-      "learning_rate": 9.999968285088332e-06,
-      "loss": 0.417,
-      "step": 970
-    },
-    {
-      "epoch": 0.06345990458139991,
-      "grad_norm": 0.5494505167007446,
-      "learning_rate": 9.999967029174334e-06,
-      "loss": 0.5063,
-      "step": 971
-    },
-    {
-      "epoch": 0.06352525978694203,
-      "grad_norm": 0.5310516953468323,
-      "learning_rate": 9.999965748873763e-06,
-      "loss": 0.4259,
-      "step": 972
-    },
-    {
-      "epoch": 0.06359061499248415,
-      "grad_norm": 0.5489716529846191,
-      "learning_rate": 9.999964444186628e-06,
-      "loss": 0.4494,
-      "step": 973
-    },
-    {
-      "epoch": 0.06365597019802627,
-      "grad_norm": 0.5342676639556885,
-      "learning_rate": 9.999963115112934e-06,
-      "loss": 0.4479,
-      "step": 974
-    },
-    {
-      "epoch": 0.0637213254035684,
-      "grad_norm": 0.4992792308330536,
-      "learning_rate": 9.999961761652688e-06,
-      "loss": 0.4007,
-      "step": 975
-    },
-    {
-      "epoch": 0.06378668060911051,
-      "grad_norm": 0.5222305059432983,
-      "learning_rate": 9.999960383805895e-06,
-      "loss": 0.3886,
-      "step": 976
-    },
-    {
-      "epoch": 0.06385203581465264,
-      "grad_norm": 0.5601535439491272,
-      "learning_rate": 9.999958981572565e-06,
-      "loss": 0.4434,
-      "step": 977
-    },
-    {
-      "epoch": 0.06391739102019475,
-      "grad_norm": 0.7607196569442749,
-      "learning_rate": 9.999957554952702e-06,
-      "loss": 0.4415,
-      "step": 978
-    },
-    {
-      "epoch": 0.06398274622573688,
-      "grad_norm": 0.5980501770973206,
-      "learning_rate": 9.999956103946313e-06,
-      "loss": 0.5057,
-      "step": 979
-    },
-    {
-      "epoch": 0.064048101431279,
-      "grad_norm": 0.5119470357894897,
-      "learning_rate": 9.999954628553406e-06,
-      "loss": 0.4124,
-      "step": 980
-    },
-    {
-      "epoch": 0.06411345663682112,
-      "grad_norm": 0.5557889938354492,
-      "learning_rate": 9.99995312877399e-06,
-      "loss": 0.4192,
-      "step": 981
-    },
-    {
-      "epoch": 0.06417881184236325,
-      "grad_norm": 0.6168763041496277,
-      "learning_rate": 9.999951604608067e-06,
-      "loss": 0.5102,
-      "step": 982
-    },
-    {
-      "epoch": 0.06424416704790537,
-      "grad_norm": 0.5247033834457397,
-      "learning_rate": 9.99995005605565e-06,
-      "loss": 0.427,
-      "step": 983
-    },
-    {
-      "epoch": 0.0643095222534475,
-      "grad_norm": 0.4901919960975647,
-      "learning_rate": 9.999948483116746e-06,
-      "loss": 0.391,
-      "step": 984
-    },
-    {
-      "epoch": 0.06437487745898961,
-      "grad_norm": 0.537260890007019,
-      "learning_rate": 9.999946885791359e-06,
-      "loss": 0.436,
-      "step": 985
-    },
-    {
-      "epoch": 0.06444023266453174,
-      "grad_norm": 0.5723612904548645,
-      "learning_rate": 9.9999452640795e-06,
-      "loss": 0.4787,
-      "step": 986
-    },
-    {
-      "epoch": 0.06450558787007385,
-      "grad_norm": 0.5198825001716614,
-      "learning_rate": 9.999943617981174e-06,
-      "loss": 0.4308,
-      "step": 987
-    },
-    {
-      "epoch": 0.06457094307561598,
-      "grad_norm": 0.519920289516449,
-      "learning_rate": 9.999941947496392e-06,
-      "loss": 0.4615,
-      "step": 988
-    },
-    {
-      "epoch": 0.06463629828115809,
-      "grad_norm": 0.5049760341644287,
-      "learning_rate": 9.99994025262516e-06,
-      "loss": 0.4347,
-      "step": 989
-    },
-    {
-      "epoch": 0.06470165348670022,
-      "grad_norm": 0.5089125633239746,
-      "learning_rate": 9.99993853336749e-06,
-      "loss": 0.4497,
-      "step": 990
-    },
-    {
-      "epoch": 0.06476700869224233,
-      "grad_norm": 0.537655234336853,
-      "learning_rate": 9.999936789723385e-06,
-      "loss": 0.4734,
-      "step": 991
-    },
-    {
-      "epoch": 0.06483236389778446,
-      "grad_norm": 0.5458921790122986,
-      "learning_rate": 9.999935021692857e-06,
-      "loss": 0.4165,
-      "step": 992
-    },
-    {
-      "epoch": 0.06489771910332658,
-      "grad_norm": 0.5852230787277222,
-      "learning_rate": 9.999933229275912e-06,
-      "loss": 0.5647,
-      "step": 993
-    },
-    {
-      "epoch": 0.0649630743088687,
-      "grad_norm": 0.5207486748695374,
-      "learning_rate": 9.999931412472564e-06,
-      "loss": 0.4477,
-      "step": 994
-    },
-    {
-      "epoch": 0.06502842951441082,
-      "grad_norm": 0.7490776181221008,
-      "learning_rate": 9.999929571282816e-06,
-      "loss": 0.4672,
-      "step": 995
-    },
-    {
-      "epoch": 0.06509378471995295,
-      "grad_norm": 0.5382635593414307,
-      "learning_rate": 9.99992770570668e-06,
-      "loss": 0.4679,
-      "step": 996
-    },
-    {
-      "epoch": 0.06515913992549506,
-      "grad_norm": 0.5522093772888184,
-      "learning_rate": 9.999925815744164e-06,
-      "loss": 0.4407,
-      "step": 997
-    },
-    {
-      "epoch": 0.06522449513103719,
-      "grad_norm": 0.5206857323646545,
-      "learning_rate": 9.999923901395278e-06,
-      "loss": 0.4224,
-      "step": 998
-    },
-    {
-      "epoch": 0.0652898503365793,
-      "grad_norm": 0.5740740299224854,
-      "learning_rate": 9.999921962660032e-06,
-      "loss": 0.5181,
-      "step": 999
-    },
-    {
-      "epoch": 0.06535520554212143,
-      "grad_norm": 0.5404792428016663,
-      "learning_rate": 9.999919999538433e-06,
-      "loss": 0.4804,
-      "step": 1000
-    },
-    {
-      "epoch": 0.06542056074766354,
-      "grad_norm": 0.5321018695831299,
-      "learning_rate": 9.999918012030493e-06,
-      "loss": 0.4637,
-      "step": 1001
-    },
-    {
-      "epoch": 0.06548591595320567,
-      "grad_norm": 0.49648821353912354,
-      "learning_rate": 9.999916000136221e-06,
-      "loss": 0.383,
-      "step": 1002
-    },
-    {
-      "epoch": 0.0655512711587478,
-      "grad_norm": 0.5232150554656982,
-      "learning_rate": 9.999913963855626e-06,
-      "loss": 0.4104,
-      "step": 1003
-    },
-    {
-      "epoch": 0.06561662636428992,
-      "grad_norm": 0.5995215177536011,
-      "learning_rate": 9.999911903188717e-06,
-      "loss": 0.5065,
-      "step": 1004
-    },
-    {
-      "epoch": 0.06568198156983204,
-      "grad_norm": 0.5301916003227234,
-      "learning_rate": 9.999909818135507e-06,
-      "loss": 0.4998,
-      "step": 1005
-    },
-    {
-      "epoch": 0.06574733677537416,
-      "grad_norm": 0.5459743738174438,
-      "learning_rate": 9.999907708696004e-06,
-      "loss": 0.4283,
-      "step": 1006
-    },
-    {
-      "epoch": 0.06581269198091629,
-      "grad_norm": 0.5282205939292908,
-      "learning_rate": 9.999905574870219e-06,
-      "loss": 0.4219,
-      "step": 1007
-    },
-    {
-      "epoch": 0.0658780471864584,
-      "grad_norm": 0.5390070080757141,
-      "learning_rate": 9.999903416658164e-06,
-      "loss": 0.4787,
-      "step": 1008
-    },
-    {
-      "epoch": 0.06594340239200053,
-      "grad_norm": 0.5368457436561584,
-      "learning_rate": 9.999901234059845e-06,
-      "loss": 0.4519,
-      "step": 1009
-    },
-    {
-      "epoch": 0.06600875759754264,
-      "grad_norm": 0.5622742176055908,
-      "learning_rate": 9.999899027075279e-06,
-      "loss": 0.4821,
-      "step": 1010
-    },
-    {
-      "epoch": 0.06607411280308477,
-      "grad_norm": 0.502387285232544,
-      "learning_rate": 9.999896795704471e-06,
-      "loss": 0.413,
-      "step": 1011
-    },
-    {
-      "epoch": 0.06613946800862688,
-      "grad_norm": 0.5299873948097229,
-      "learning_rate": 9.999894539947435e-06,
-      "loss": 0.4521,
-      "step": 1012
-    },
-    {
-      "epoch": 0.06620482321416901,
-      "grad_norm": 0.5077336430549622,
-      "learning_rate": 9.99989225980418e-06,
-      "loss": 0.4049,
-      "step": 1013
-    },
-    {
-      "epoch": 0.06627017841971113,
-      "grad_norm": 0.5375440716743469,
-      "learning_rate": 9.999889955274719e-06,
-      "loss": 0.4666,
-      "step": 1014
-    },
-    {
-      "epoch": 0.06633553362525325,
-      "grad_norm": 0.5278374552726746,
-      "learning_rate": 9.999887626359064e-06,
-      "loss": 0.4591,
-      "step": 1015
-    },
-    {
-      "epoch": 0.06640088883079537,
-      "grad_norm": 0.5720160603523254,
-      "learning_rate": 9.999885273057223e-06,
-      "loss": 0.4545,
-      "step": 1016
-    },
-    {
-      "epoch": 0.0664662440363375,
-      "grad_norm": 0.5435892939567566,
-      "learning_rate": 9.99988289536921e-06,
-      "loss": 0.4822,
-      "step": 1017
-    },
-    {
-      "epoch": 0.06653159924187961,
-      "grad_norm": 0.5654741525650024,
-      "learning_rate": 9.999880493295035e-06,
-      "loss": 0.4699,
-      "step": 1018
-    },
-    {
-      "epoch": 0.06659695444742174,
-      "grad_norm": 0.569327712059021,
-      "learning_rate": 9.999878066834713e-06,
-      "loss": 0.4985,
-      "step": 1019
-    },
-    {
-      "epoch": 0.06666230965296385,
-      "grad_norm": 0.5556838512420654,
-      "learning_rate": 9.999875615988252e-06,
-      "loss": 0.4365,
-      "step": 1020
-    },
-    {
-      "epoch": 0.06672766485850598,
-      "grad_norm": 0.5812886953353882,
-      "learning_rate": 9.999873140755666e-06,
-      "loss": 0.4512,
-      "step": 1021
-    },
-    {
-      "epoch": 0.0667930200640481,
-      "grad_norm": 0.5536554455757141,
-      "learning_rate": 9.999870641136966e-06,
-      "loss": 0.4766,
-      "step": 1022
-    },
-    {
-      "epoch": 0.06685837526959022,
-      "grad_norm": 0.5084673166275024,
-      "learning_rate": 9.999868117132166e-06,
-      "loss": 0.3906,
-      "step": 1023
-    },
-    {
-      "epoch": 0.06692373047513235,
-      "grad_norm": 0.5405479073524475,
-      "learning_rate": 9.999865568741275e-06,
-      "loss": 0.4192,
-      "step": 1024
-    },
-    {
-      "epoch": 0.06698908568067447,
-      "grad_norm": 0.5325936079025269,
-      "learning_rate": 9.99986299596431e-06,
-      "loss": 0.3922,
-      "step": 1025
-    },
-    {
-      "epoch": 0.06705444088621659,
-      "grad_norm": 0.522232174873352,
-      "learning_rate": 9.99986039880128e-06,
-      "loss": 0.4321,
-      "step": 1026
-    },
-    {
-      "epoch": 0.06711979609175871,
-      "grad_norm": 0.5085806846618652,
-      "learning_rate": 9.999857777252198e-06,
-      "loss": 0.4419,
-      "step": 1027
-    },
-    {
-      "epoch": 0.06718515129730084,
-      "grad_norm": 0.5148561596870422,
-      "learning_rate": 9.999855131317077e-06,
-      "loss": 0.4104,
-      "step": 1028
-    },
-    {
-      "epoch": 0.06725050650284295,
-      "grad_norm": 0.5366589426994324,
-      "learning_rate": 9.999852460995933e-06,
-      "loss": 0.413,
-      "step": 1029
-    },
-    {
-      "epoch": 0.06731586170838508,
-      "grad_norm": 0.5244190096855164,
-      "learning_rate": 9.999849766288774e-06,
-      "loss": 0.4195,
-      "step": 1030
-    },
-    {
-      "epoch": 0.06738121691392719,
-      "grad_norm": 0.4705215096473694,
-      "learning_rate": 9.999847047195616e-06,
-      "loss": 0.3667,
-      "step": 1031
-    },
-    {
-      "epoch": 0.06744657211946932,
-      "grad_norm": 0.5042105913162231,
-      "learning_rate": 9.999844303716473e-06,
-      "loss": 0.4562,
-      "step": 1032
-    },
-    {
-      "epoch": 0.06751192732501143,
-      "grad_norm": 0.5559691190719604,
-      "learning_rate": 9.999841535851356e-06,
-      "loss": 0.4836,
-      "step": 1033
-    },
-    {
-      "epoch": 0.06757728253055356,
-      "grad_norm": 0.5235845446586609,
-      "learning_rate": 9.999838743600281e-06,
-      "loss": 0.3812,
-      "step": 1034
-    },
-    {
-      "epoch": 0.06764263773609568,
-      "grad_norm": 0.5642328858375549,
-      "learning_rate": 9.99983592696326e-06,
-      "loss": 0.512,
-      "step": 1035
-    },
-    {
-      "epoch": 0.0677079929416378,
-      "grad_norm": 0.5955402851104736,
-      "learning_rate": 9.999833085940306e-06,
-      "loss": 0.4639,
-      "step": 1036
-    },
-    {
-      "epoch": 0.06777334814717992,
-      "grad_norm": 0.5364554524421692,
-      "learning_rate": 9.999830220531434e-06,
-      "loss": 0.5009,
-      "step": 1037
-    },
-    {
-      "epoch": 0.06783870335272205,
-      "grad_norm": 0.5376973748207092,
-      "learning_rate": 9.99982733073666e-06,
-      "loss": 0.4239,
-      "step": 1038
-    },
-    {
-      "epoch": 0.06790405855826416,
-      "grad_norm": 0.5236703157424927,
-      "learning_rate": 9.999824416555993e-06,
-      "loss": 0.4436,
-      "step": 1039
-    },
-    {
-      "epoch": 0.06796941376380629,
-      "grad_norm": 0.5033944249153137,
-      "learning_rate": 9.999821477989452e-06,
-      "loss": 0.4452,
-      "step": 1040
-    },
-    {
-      "epoch": 0.0680347689693484,
-      "grad_norm": 0.5482272505760193,
-      "learning_rate": 9.99981851503705e-06,
-      "loss": 0.4499,
-      "step": 1041
-    },
-    {
-      "epoch": 0.06810012417489053,
-      "grad_norm": 0.5539483428001404,
-      "learning_rate": 9.9998155276988e-06,
-      "loss": 0.4985,
-      "step": 1042
-    },
-    {
-      "epoch": 0.06816547938043264,
-      "grad_norm": 0.528601884841919,
-      "learning_rate": 9.999812515974717e-06,
-      "loss": 0.478,
-      "step": 1043
-    },
-    {
-      "epoch": 0.06823083458597477,
-      "grad_norm": 0.4756832718849182,
-      "learning_rate": 9.999809479864817e-06,
-      "loss": 0.3917,
-      "step": 1044
-    },
-    {
-      "epoch": 0.0682961897915169,
-      "grad_norm": 0.5553827881813049,
-      "learning_rate": 9.999806419369114e-06,
-      "loss": 0.4666,
-      "step": 1045
-    },
-    {
-      "epoch": 0.06836154499705901,
-      "grad_norm": 0.5052905082702637,
-      "learning_rate": 9.999803334487624e-06,
-      "loss": 0.3979,
-      "step": 1046
-    },
-    {
-      "epoch": 0.06842690020260114,
-      "grad_norm": 0.5354941487312317,
-      "learning_rate": 9.999800225220359e-06,
-      "loss": 0.4174,
-      "step": 1047
-    },
-    {
-      "epoch": 0.06849225540814326,
-      "grad_norm": 0.4951237142086029,
-      "learning_rate": 9.999797091567339e-06,
-      "loss": 0.4191,
-      "step": 1048
-    },
-    {
-      "epoch": 0.06855761061368539,
-      "grad_norm": 0.5031898021697998,
-      "learning_rate": 9.999793933528575e-06,
-      "loss": 0.3946,
-      "step": 1049
-    },
-    {
-      "epoch": 0.0686229658192275,
-      "grad_norm": 0.5494363307952881,
-      "learning_rate": 9.999790751104082e-06,
-      "loss": 0.4309,
-      "step": 1050
-    },
-    {
-      "epoch": 0.06868832102476963,
-      "grad_norm": 0.5712085962295532,
-      "learning_rate": 9.99978754429388e-06,
-      "loss": 0.5081,
-      "step": 1051
-    },
-    {
-      "epoch": 0.06875367623031174,
-      "grad_norm": 0.5110986232757568,
-      "learning_rate": 9.99978431309798e-06,
-      "loss": 0.4063,
-      "step": 1052
-    },
-    {
-      "epoch": 0.06881903143585387,
-      "grad_norm": 0.5322535037994385,
-      "learning_rate": 9.999781057516402e-06,
-      "loss": 0.3905,
-      "step": 1053
-    },
-    {
-      "epoch": 0.06888438664139598,
-      "grad_norm": 0.5490505695343018,
-      "learning_rate": 9.999777777549158e-06,
-      "loss": 0.4842,
-      "step": 1054
-    },
-    {
-      "epoch": 0.06894974184693811,
-      "grad_norm": 0.5415787100791931,
-      "learning_rate": 9.999774473196266e-06,
-      "loss": 0.4193,
-      "step": 1055
-    },
-    {
-      "epoch": 0.06901509705248023,
-      "grad_norm": 0.5278641581535339,
-      "learning_rate": 9.999771144457743e-06,
-      "loss": 0.466,
-      "step": 1056
-    },
-    {
-      "epoch": 0.06908045225802235,
-      "grad_norm": 0.5006137490272522,
-      "learning_rate": 9.999767791333604e-06,
-      "loss": 0.4058,
-      "step": 1057
-    },
-    {
-      "epoch": 0.06914580746356447,
-      "grad_norm": 0.5041103363037109,
-      "learning_rate": 9.999764413823864e-06,
-      "loss": 0.3976,
-      "step": 1058
-    },
-    {
-      "epoch": 0.0692111626691066,
-      "grad_norm": 0.5599545240402222,
-      "learning_rate": 9.999761011928542e-06,
-      "loss": 0.4106,
-      "step": 1059
-    },
-    {
-      "epoch": 0.06927651787464871,
-      "grad_norm": 0.529897928237915,
-      "learning_rate": 9.999757585647653e-06,
-      "loss": 0.4151,
-      "step": 1060
-    },
-    {
-      "epoch": 0.06934187308019084,
-      "grad_norm": 0.5626737475395203,
-      "learning_rate": 9.999754134981215e-06,
-      "loss": 0.349,
-      "step": 1061
-    },
-    {
-      "epoch": 0.06940722828573295,
-      "grad_norm": 0.5403541922569275,
-      "learning_rate": 9.999750659929241e-06,
-      "loss": 0.4523,
-      "step": 1062
-    },
-    {
-      "epoch": 0.06947258349127508,
-      "grad_norm": 0.5121879577636719,
-      "learning_rate": 9.999747160491754e-06,
-      "loss": 0.3941,
-      "step": 1063
-    },
-    {
-      "epoch": 0.0695379386968172,
-      "grad_norm": 0.5438024997711182,
-      "learning_rate": 9.999743636668767e-06,
-      "loss": 0.4034,
-      "step": 1064
-    },
-    {
-      "epoch": 0.06960329390235932,
-      "grad_norm": 0.5025306344032288,
-      "learning_rate": 9.999740088460299e-06,
-      "loss": 0.3851,
-      "step": 1065
-    },
-    {
-      "epoch": 0.06966864910790145,
-      "grad_norm": 0.5248860716819763,
-      "learning_rate": 9.999736515866365e-06,
-      "loss": 0.4613,
-      "step": 1066
-    },
-    {
-      "epoch": 0.06973400431344356,
-      "grad_norm": 0.5184089541435242,
-      "learning_rate": 9.999732918886985e-06,
-      "loss": 0.3957,
-      "step": 1067
-    },
-    {
-      "epoch": 0.06979935951898569,
-      "grad_norm": 0.5219573974609375,
-      "learning_rate": 9.999729297522176e-06,
-      "loss": 0.433,
-      "step": 1068
-    },
-    {
-      "epoch": 0.06986471472452781,
-      "grad_norm": 0.5599725842475891,
-      "learning_rate": 9.999725651771955e-06,
-      "loss": 0.4545,
-      "step": 1069
-    },
-    {
-      "epoch": 0.06993006993006994,
-      "grad_norm": 0.558887243270874,
-      "learning_rate": 9.99972198163634e-06,
-      "loss": 0.484,
-      "step": 1070
-    },
-    {
-      "epoch": 0.06999542513561205,
-      "grad_norm": 0.5470594167709351,
-      "learning_rate": 9.999718287115346e-06,
-      "loss": 0.5024,
-      "step": 1071
-    },
-    {
-      "epoch": 0.07006078034115418,
-      "grad_norm": 0.6109849214553833,
-      "learning_rate": 9.999714568208997e-06,
-      "loss": 0.4765,
-      "step": 1072
-    },
-    {
-      "epoch": 0.07012613554669629,
-      "grad_norm": 0.47927340865135193,
-      "learning_rate": 9.999710824917306e-06,
-      "loss": 0.399,
-      "step": 1073
-    },
-    {
-      "epoch": 0.07019149075223842,
-      "grad_norm": 0.5048418045043945,
-      "learning_rate": 9.999707057240294e-06,
-      "loss": 0.3745,
-      "step": 1074
-    },
-    {
-      "epoch": 0.07025684595778053,
-      "grad_norm": 0.5382951498031616,
-      "learning_rate": 9.999703265177979e-06,
-      "loss": 0.4251,
-      "step": 1075
-    },
-    {
-      "epoch": 0.07032220116332266,
-      "grad_norm": 0.4840947985649109,
-      "learning_rate": 9.99969944873038e-06,
-      "loss": 0.4107,
-      "step": 1076
-    },
-    {
-      "epoch": 0.07038755636886478,
-      "grad_norm": 0.5169443488121033,
-      "learning_rate": 9.999695607897513e-06,
-      "loss": 0.4057,
-      "step": 1077
-    },
-    {
-      "epoch": 0.0704529115744069,
-      "grad_norm": 0.6006619334220886,
-      "learning_rate": 9.999691742679398e-06,
-      "loss": 0.5285,
-      "step": 1078
-    },
-    {
-      "epoch": 0.07051826677994902,
-      "grad_norm": 0.5611095428466797,
-      "learning_rate": 9.999687853076056e-06,
-      "loss": 0.4978,
-      "step": 1079
-    },
-    {
-      "epoch": 0.07058362198549115,
-      "grad_norm": 0.538235604763031,
-      "learning_rate": 9.999683939087504e-06,
-      "loss": 0.4559,
-      "step": 1080
-    },
-    {
-      "epoch": 0.07064897719103326,
-      "grad_norm": 0.58662348985672,
-      "learning_rate": 9.999680000713761e-06,
-      "loss": 0.5676,
-      "step": 1081
-    },
-    {
-      "epoch": 0.07071433239657539,
-      "grad_norm": 0.5555466413497925,
-      "learning_rate": 9.999676037954846e-06,
-      "loss": 0.3845,
-      "step": 1082
-    },
-    {
-      "epoch": 0.0707796876021175,
-      "grad_norm": 0.4942343533039093,
-      "learning_rate": 9.999672050810781e-06,
-      "loss": 0.4048,
-      "step": 1083
-    },
-    {
-      "epoch": 0.07084504280765963,
-      "grad_norm": 0.5459291338920593,
-      "learning_rate": 9.99966803928158e-06,
-      "loss": 0.4352,
-      "step": 1084
-    },
-    {
-      "epoch": 0.07091039801320174,
-      "grad_norm": 0.5130406022071838,
-      "learning_rate": 9.99966400336727e-06,
-      "loss": 0.4111,
-      "step": 1085
-    },
-    {
-      "epoch": 0.07097575321874387,
-      "grad_norm": 0.4856773912906647,
-      "learning_rate": 9.999659943067864e-06,
-      "loss": 0.3968,
-      "step": 1086
-    },
-    {
-      "epoch": 0.071041108424286,
-      "grad_norm": 0.5157826542854309,
-      "learning_rate": 9.999655858383384e-06,
-      "loss": 0.4324,
-      "step": 1087
-    },
-    {
-      "epoch": 0.07110646362982811,
-      "grad_norm": 0.5272077322006226,
-      "learning_rate": 9.999651749313852e-06,
-      "loss": 0.4589,
-      "step": 1088
-    },
-    {
-      "epoch": 0.07117181883537024,
-      "grad_norm": 0.49379098415374756,
-      "learning_rate": 9.999647615859284e-06,
-      "loss": 0.4067,
-      "step": 1089
-    },
-    {
-      "epoch": 0.07123717404091236,
-      "grad_norm": 0.5427277088165283,
-      "learning_rate": 9.999643458019706e-06,
-      "loss": 0.4638,
-      "step": 1090
-    },
-    {
-      "epoch": 0.07130252924645449,
-      "grad_norm": 0.49519068002700806,
-      "learning_rate": 9.999639275795132e-06,
-      "loss": 0.4352,
-      "step": 1091
-    },
-    {
-      "epoch": 0.0713678844519966,
-      "grad_norm": 0.506629228591919,
-      "learning_rate": 9.999635069185587e-06,
-      "loss": 0.4753,
-      "step": 1092
-    },
-    {
-      "epoch": 0.07143323965753873,
-      "grad_norm": 0.46790847182273865,
-      "learning_rate": 9.999630838191087e-06,
-      "loss": 0.4048,
-      "step": 1093
-    },
-    {
-      "epoch": 0.07149859486308084,
-      "grad_norm": 0.581870436668396,
-      "learning_rate": 9.99962658281166e-06,
-      "loss": 0.4858,
-      "step": 1094
-    },
-    {
-      "epoch": 0.07156395006862297,
-      "grad_norm": 0.5799008011817932,
-      "learning_rate": 9.999622303047318e-06,
-      "loss": 0.5081,
-      "step": 1095
-    },
-    {
-      "epoch": 0.07162930527416508,
-      "grad_norm": 0.5335209369659424,
-      "learning_rate": 9.999617998898087e-06,
-      "loss": 0.4161,
-      "step": 1096
-    },
-    {
-      "epoch": 0.07169466047970721,
-      "grad_norm": 0.5431477427482605,
-      "learning_rate": 9.999613670363988e-06,
-      "loss": 0.4598,
-      "step": 1097
-    },
-    {
-      "epoch": 0.07176001568524933,
-      "grad_norm": 0.5139455795288086,
-      "learning_rate": 9.999609317445041e-06,
-      "loss": 0.4032,
-      "step": 1098
-    },
-    {
-      "epoch": 0.07182537089079145,
-      "grad_norm": 0.5462816953659058,
-      "learning_rate": 9.999604940141266e-06,
-      "loss": 0.4114,
-      "step": 1099
-    },
-    {
-      "epoch": 0.07189072609633357,
-      "grad_norm": 0.5116475820541382,
-      "learning_rate": 9.999600538452687e-06,
-      "loss": 0.425,
-      "step": 1100
-    },
-    {
-      "epoch": 0.0719560813018757,
-      "grad_norm": 0.552571713924408,
-      "learning_rate": 9.999596112379323e-06,
-      "loss": 0.4446,
-      "step": 1101
-    },
-    {
-      "epoch": 0.07202143650741781,
-      "grad_norm": 0.5241992473602295,
-      "learning_rate": 9.999591661921197e-06,
-      "loss": 0.438,
-      "step": 1102
-    },
-    {
-      "epoch": 0.07208679171295994,
-      "grad_norm": 0.5425822138786316,
-      "learning_rate": 9.99958718707833e-06,
-      "loss": 0.4437,
-      "step": 1103
-    },
-    {
-      "epoch": 0.07215214691850205,
-      "grad_norm": 0.5181475877761841,
-      "learning_rate": 9.999582687850746e-06,
-      "loss": 0.4483,
-      "step": 1104
-    },
-    {
-      "epoch": 0.07221750212404418,
-      "grad_norm": 0.49834105372428894,
-      "learning_rate": 9.999578164238463e-06,
-      "loss": 0.4055,
-      "step": 1105
-    },
-    {
-      "epoch": 0.0722828573295863,
-      "grad_norm": 0.5015610456466675,
-      "learning_rate": 9.999573616241509e-06,
-      "loss": 0.424,
-      "step": 1106
-    },
-    {
-      "epoch": 0.07234821253512842,
-      "grad_norm": 0.5378239154815674,
-      "learning_rate": 9.999569043859898e-06,
-      "loss": 0.4475,
-      "step": 1107
-    },
-    {
-      "epoch": 0.07241356774067055,
-      "grad_norm": 0.5474442839622498,
-      "learning_rate": 9.99956444709366e-06,
-      "loss": 0.4708,
-      "step": 1108
-    },
-    {
-      "epoch": 0.07247892294621266,
-      "grad_norm": 0.5020264387130737,
-      "learning_rate": 9.999559825942812e-06,
-      "loss": 0.4158,
-      "step": 1109
-    },
-    {
-      "epoch": 0.07254427815175479,
-      "grad_norm": 0.5356208086013794,
-      "learning_rate": 9.99955518040738e-06,
-      "loss": 0.383,
-      "step": 1110
-    },
-    {
-      "epoch": 0.07260963335729691,
-      "grad_norm": 0.5045859217643738,
-      "learning_rate": 9.999550510487385e-06,
-      "loss": 0.3824,
-      "step": 1111
-    },
-    {
-      "epoch": 0.07267498856283904,
-      "grad_norm": 0.596736490726471,
-      "learning_rate": 9.99954581618285e-06,
-      "loss": 0.4847,
-      "step": 1112
-    },
-    {
-      "epoch": 0.07274034376838115,
-      "grad_norm": 0.5553385615348816,
-      "learning_rate": 9.999541097493799e-06,
-      "loss": 0.4628,
-      "step": 1113
-    },
-    {
-      "epoch": 0.07280569897392328,
-      "grad_norm": 0.5719872713088989,
-      "learning_rate": 9.999536354420252e-06,
-      "loss": 0.4532,
-      "step": 1114
-    },
-    {
-      "epoch": 0.07287105417946539,
-      "grad_norm": 0.5502328872680664,
-      "learning_rate": 9.999531586962236e-06,
-      "loss": 0.4702,
-      "step": 1115
-    },
-    {
-      "epoch": 0.07293640938500752,
-      "grad_norm": 0.60325688123703,
-      "learning_rate": 9.999526795119771e-06,
-      "loss": 0.4916,
-      "step": 1116
-    },
-    {
-      "epoch": 0.07300176459054963,
-      "grad_norm": 0.5491304993629456,
-      "learning_rate": 9.999521978892882e-06,
-      "loss": 0.466,
-      "step": 1117
-    },
-    {
-      "epoch": 0.07306711979609176,
-      "grad_norm": 0.5431767702102661,
-      "learning_rate": 9.999517138281594e-06,
-      "loss": 0.4389,
-      "step": 1118
-    },
-    {
-      "epoch": 0.07313247500163388,
-      "grad_norm": 0.5361014008522034,
-      "learning_rate": 9.999512273285928e-06,
-      "loss": 0.4183,
-      "step": 1119
-    },
-    {
-      "epoch": 0.073197830207176,
-      "grad_norm": 0.5825538635253906,
-      "learning_rate": 9.999507383905908e-06,
-      "loss": 0.4194,
-      "step": 1120
-    },
-    {
-      "epoch": 0.07326318541271812,
-      "grad_norm": 0.5627398490905762,
-      "learning_rate": 9.99950247014156e-06,
-      "loss": 0.4284,
-      "step": 1121
-    },
-    {
-      "epoch": 0.07332854061826025,
-      "grad_norm": 0.547063946723938,
-      "learning_rate": 9.999497531992905e-06,
-      "loss": 0.4409,
-      "step": 1122
-    },
-    {
-      "epoch": 0.07339389582380236,
-      "grad_norm": 0.50401771068573,
-      "learning_rate": 9.99949256945997e-06,
-      "loss": 0.3774,
-      "step": 1123
-    },
-    {
-      "epoch": 0.07345925102934449,
-      "grad_norm": 0.6625709533691406,
-      "learning_rate": 9.999487582542777e-06,
-      "loss": 0.5254,
-      "step": 1124
-    },
-    {
-      "epoch": 0.0735246062348866,
-      "grad_norm": 0.5538679361343384,
-      "learning_rate": 9.99948257124135e-06,
-      "loss": 0.4445,
-      "step": 1125
-    },
-    {
-      "epoch": 0.07358996144042873,
-      "grad_norm": 0.5207232236862183,
-      "learning_rate": 9.999477535555716e-06,
-      "loss": 0.414,
-      "step": 1126
-    },
-    {
-      "epoch": 0.07365531664597084,
-      "grad_norm": 0.5595287084579468,
-      "learning_rate": 9.999472475485897e-06,
-      "loss": 0.5615,
-      "step": 1127
-    },
-    {
-      "epoch": 0.07372067185151297,
-      "grad_norm": 0.4857555031776428,
-      "learning_rate": 9.999467391031918e-06,
-      "loss": 0.4213,
-      "step": 1128
-    },
-    {
-      "epoch": 0.0737860270570551,
-      "grad_norm": 0.5126360654830933,
-      "learning_rate": 9.999462282193808e-06,
-      "loss": 0.4042,
-      "step": 1129
-    },
-    {
-      "epoch": 0.07385138226259721,
-      "grad_norm": 0.5549189448356628,
-      "learning_rate": 9.999457148971585e-06,
-      "loss": 0.4648,
-      "step": 1130
-    },
-    {
-      "epoch": 0.07391673746813934,
-      "grad_norm": 0.5788109302520752,
-      "learning_rate": 9.999451991365278e-06,
-      "loss": 0.458,
-      "step": 1131
-    },
-    {
-      "epoch": 0.07398209267368146,
-      "grad_norm": 0.580055296421051,
-      "learning_rate": 9.999446809374913e-06,
-      "loss": 0.4423,
-      "step": 1132
-    },
-    {
-      "epoch": 0.07404744787922359,
-      "grad_norm": 0.5349094867706299,
-      "learning_rate": 9.999441603000514e-06,
-      "loss": 0.4635,
-      "step": 1133
-    },
-    {
-      "epoch": 0.0741128030847657,
-      "grad_norm": 0.5247857570648193,
-      "learning_rate": 9.999436372242106e-06,
-      "loss": 0.4573,
-      "step": 1134
-    },
-    {
-      "epoch": 0.07417815829030783,
-      "grad_norm": 0.5215403437614441,
-      "learning_rate": 9.999431117099714e-06,
-      "loss": 0.4152,
-      "step": 1135
-    },
-    {
-      "epoch": 0.07424351349584994,
-      "grad_norm": 0.5917163491249084,
-      "learning_rate": 9.999425837573364e-06,
-      "loss": 0.4539,
-      "step": 1136
-    },
-    {
-      "epoch": 0.07430886870139207,
-      "grad_norm": 0.5871264934539795,
-      "learning_rate": 9.999420533663084e-06,
-      "loss": 0.5734,
-      "step": 1137
-    },
-    {
-      "epoch": 0.07437422390693418,
-      "grad_norm": 0.5448279976844788,
-      "learning_rate": 9.999415205368897e-06,
-      "loss": 0.4242,
-      "step": 1138
-    },
-    {
-      "epoch": 0.07443957911247631,
-      "grad_norm": 0.4966379404067993,
-      "learning_rate": 9.999409852690832e-06,
-      "loss": 0.3862,
-      "step": 1139
-    },
-    {
-      "epoch": 0.07450493431801843,
-      "grad_norm": 0.5236524343490601,
-      "learning_rate": 9.99940447562891e-06,
-      "loss": 0.4625,
-      "step": 1140
-    },
-    {
-      "epoch": 0.07457028952356055,
-      "grad_norm": 0.5394020080566406,
-      "learning_rate": 9.999399074183163e-06,
-      "loss": 0.4387,
-      "step": 1141
-    },
-    {
-      "epoch": 0.07463564472910267,
-      "grad_norm": 0.5543898940086365,
-      "learning_rate": 9.999393648353613e-06,
-      "loss": 0.4771,
-      "step": 1142
-    },
-    {
-      "epoch": 0.0747009999346448,
-      "grad_norm": 0.5211477279663086,
-      "learning_rate": 9.99938819814029e-06,
-      "loss": 0.3871,
-      "step": 1143
-    },
-    {
-      "epoch": 0.07476635514018691,
-      "grad_norm": 0.7291384935379028,
-      "learning_rate": 9.999382723543216e-06,
-      "loss": 0.3779,
-      "step": 1144
-    },
-    {
-      "epoch": 0.07483171034572904,
-      "grad_norm": 0.5049942135810852,
-      "learning_rate": 9.999377224562424e-06,
-      "loss": 0.4263,
-      "step": 1145
-    },
-    {
-      "epoch": 0.07489706555127115,
-      "grad_norm": 0.5275691151618958,
-      "learning_rate": 9.999371701197935e-06,
-      "loss": 0.4243,
-      "step": 1146
-    },
-    {
-      "epoch": 0.07496242075681328,
-      "grad_norm": 0.5381787419319153,
-      "learning_rate": 9.99936615344978e-06,
-      "loss": 0.4225,
-      "step": 1147
-    },
-    {
-      "epoch": 0.0750277759623554,
-      "grad_norm": 0.5490167140960693,
-      "learning_rate": 9.999360581317982e-06,
-      "loss": 0.4843,
-      "step": 1148
-    },
-    {
-      "epoch": 0.07509313116789752,
-      "grad_norm": 0.5006473660469055,
-      "learning_rate": 9.999354984802572e-06,
-      "loss": 0.4086,
-      "step": 1149
-    },
-    {
-      "epoch": 0.07515848637343965,
-      "grad_norm": 0.5244328379631042,
-      "learning_rate": 9.999349363903574e-06,
-      "loss": 0.4037,
-      "step": 1150
-    },
-    {
-      "epoch": 0.07522384157898176,
-      "grad_norm": 0.527870774269104,
-      "learning_rate": 9.99934371862102e-06,
-      "loss": 0.3991,
-      "step": 1151
-    },
-    {
-      "epoch": 0.07528919678452389,
-      "grad_norm": 0.5216260552406311,
-      "learning_rate": 9.999338048954933e-06,
-      "loss": 0.357,
-      "step": 1152
-    },
-    {
-      "epoch": 0.07535455199006601,
-      "grad_norm": 0.531283974647522,
-      "learning_rate": 9.999332354905343e-06,
-      "loss": 0.4682,
-      "step": 1153
-    },
-    {
-      "epoch": 0.07541990719560814,
-      "grad_norm": 0.5901212692260742,
-      "learning_rate": 9.999326636472278e-06,
-      "loss": 0.4902,
-      "step": 1154
-    },
-    {
-      "epoch": 0.07548526240115025,
-      "grad_norm": 0.5611442923545837,
-      "learning_rate": 9.999320893655762e-06,
-      "loss": 0.4659,
-      "step": 1155
-    },
-    {
-      "epoch": 0.07555061760669238,
-      "grad_norm": 0.5276128053665161,
-      "learning_rate": 9.99931512645583e-06,
-      "loss": 0.4918,
-      "step": 1156
-    },
-    {
-      "epoch": 0.07561597281223449,
-      "grad_norm": 0.5524461269378662,
-      "learning_rate": 9.999309334872503e-06,
-      "loss": 0.4832,
-      "step": 1157
-    },
-    {
-      "epoch": 0.07568132801777662,
-      "grad_norm": 0.5486389398574829,
-      "learning_rate": 9.999303518905815e-06,
-      "loss": 0.4823,
-      "step": 1158
-    },
-    {
-      "epoch": 0.07574668322331873,
-      "grad_norm": 0.562732994556427,
-      "learning_rate": 9.99929767855579e-06,
-      "loss": 0.4375,
-      "step": 1159
-    },
-    {
-      "epoch": 0.07581203842886086,
-      "grad_norm": 0.49552303552627563,
-      "learning_rate": 9.999291813822459e-06,
-      "loss": 0.3802,
-      "step": 1160
-    },
-    {
-      "epoch": 0.07587739363440298,
-      "grad_norm": 0.505366325378418,
-      "learning_rate": 9.99928592470585e-06,
-      "loss": 0.4133,
-      "step": 1161
-    },
-    {
-      "epoch": 0.0759427488399451,
-      "grad_norm": 0.4455549716949463,
-      "learning_rate": 9.999280011205991e-06,
-      "loss": 0.3705,
-      "step": 1162
-    },
-    {
-      "epoch": 0.07600810404548722,
-      "grad_norm": 0.4988388419151306,
-      "learning_rate": 9.99927407332291e-06,
-      "loss": 0.4121,
-      "step": 1163
-    },
-    {
-      "epoch": 0.07607345925102935,
-      "grad_norm": 0.5100632905960083,
-      "learning_rate": 9.999268111056641e-06,
-      "loss": 0.4365,
-      "step": 1164
-    },
-    {
-      "epoch": 0.07613881445657146,
-      "grad_norm": 0.5309730768203735,
-      "learning_rate": 9.999262124407207e-06,
-      "loss": 0.4792,
-      "step": 1165
-    },
-    {
-      "epoch": 0.07620416966211359,
-      "grad_norm": 0.5231695771217346,
-      "learning_rate": 9.99925611337464e-06,
-      "loss": 0.4703,
-      "step": 1166
-    },
-    {
-      "epoch": 0.0762695248676557,
-      "grad_norm": 0.5190871953964233,
-      "learning_rate": 9.99925007795897e-06,
-      "loss": 0.4615,
-      "step": 1167
-    },
-    {
-      "epoch": 0.07633488007319783,
-      "grad_norm": 0.5430537462234497,
-      "learning_rate": 9.999244018160225e-06,
-      "loss": 0.4867,
-      "step": 1168
-    },
-    {
-      "epoch": 0.07640023527873996,
-      "grad_norm": 0.5315860509872437,
-      "learning_rate": 9.999237933978437e-06,
-      "loss": 0.4364,
-      "step": 1169
-    },
-    {
-      "epoch": 0.07646559048428207,
-      "grad_norm": 0.553806722164154,
-      "learning_rate": 9.999231825413631e-06,
-      "loss": 0.4627,
-      "step": 1170
-    },
-    {
-      "epoch": 0.0765309456898242,
-      "grad_norm": 0.5221887826919556,
-      "learning_rate": 9.999225692465839e-06,
-      "loss": 0.3913,
-      "step": 1171
-    },
-    {
-      "epoch": 0.07659630089536631,
-      "grad_norm": 0.5462474226951599,
-      "learning_rate": 9.999219535135093e-06,
-      "loss": 0.4811,
-      "step": 1172
-    },
-    {
-      "epoch": 0.07666165610090844,
-      "grad_norm": 0.496337890625,
-      "learning_rate": 9.999213353421422e-06,
-      "loss": 0.3781,
-      "step": 1173
-    },
-    {
-      "epoch": 0.07672701130645056,
-      "grad_norm": 0.5239957571029663,
-      "learning_rate": 9.999207147324854e-06,
-      "loss": 0.4518,
-      "step": 1174
-    },
-    {
-      "epoch": 0.07679236651199269,
-      "grad_norm": 0.5142859816551208,
-      "learning_rate": 9.999200916845422e-06,
-      "loss": 0.446,
-      "step": 1175
-    },
-    {
-      "epoch": 0.0768577217175348,
-      "grad_norm": 0.5262611508369446,
-      "learning_rate": 9.999194661983154e-06,
-      "loss": 0.4095,
-      "step": 1176
-    },
-    {
-      "epoch": 0.07692307692307693,
-      "grad_norm": 0.5234584808349609,
-      "learning_rate": 9.999188382738083e-06,
-      "loss": 0.44,
-      "step": 1177
-    },
-    {
-      "epoch": 0.07698843212861904,
-      "grad_norm": 0.5316318869590759,
-      "learning_rate": 9.999182079110238e-06,
-      "loss": 0.3961,
-      "step": 1178
-    },
-    {
-      "epoch": 0.07705378733416117,
-      "grad_norm": 0.553196370601654,
-      "learning_rate": 9.99917575109965e-06,
-      "loss": 0.4897,
-      "step": 1179
-    },
-    {
-      "epoch": 0.07711914253970328,
-      "grad_norm": 0.6131114959716797,
-      "learning_rate": 9.99916939870635e-06,
-      "loss": 0.4655,
-      "step": 1180
-    },
-    {
-      "epoch": 0.07718449774524541,
-      "grad_norm": 0.5405822992324829,
-      "learning_rate": 9.999163021930369e-06,
-      "loss": 0.4634,
-      "step": 1181
-    },
-    {
-      "epoch": 0.07724985295078753,
-      "grad_norm": 0.4962118864059448,
-      "learning_rate": 9.999156620771736e-06,
-      "loss": 0.4538,
-      "step": 1182
-    },
-    {
-      "epoch": 0.07731520815632965,
-      "grad_norm": 0.5419692397117615,
-      "learning_rate": 9.999150195230487e-06,
-      "loss": 0.4292,
-      "step": 1183
-    },
-    {
-      "epoch": 0.07738056336187177,
-      "grad_norm": 0.5354037880897522,
-      "learning_rate": 9.99914374530665e-06,
-      "loss": 0.4915,
-      "step": 1184
-    },
-    {
-      "epoch": 0.0774459185674139,
-      "grad_norm": 0.6750656962394714,
-      "learning_rate": 9.999137271000258e-06,
-      "loss": 0.4394,
-      "step": 1185
-    },
-    {
-      "epoch": 0.07751127377295601,
-      "grad_norm": 0.6161524057388306,
-      "learning_rate": 9.99913077231134e-06,
-      "loss": 0.4917,
-      "step": 1186
-    },
-    {
-      "epoch": 0.07757662897849814,
-      "grad_norm": 0.5116806030273438,
-      "learning_rate": 9.99912424923993e-06,
-      "loss": 0.4292,
-      "step": 1187
-    },
-    {
-      "epoch": 0.07764198418404025,
-      "grad_norm": 0.5059065818786621,
-      "learning_rate": 9.999117701786059e-06,
-      "loss": 0.4539,
-      "step": 1188
-    },
-    {
-      "epoch": 0.07770733938958238,
-      "grad_norm": 0.548248291015625,
-      "learning_rate": 9.999111129949759e-06,
-      "loss": 0.4325,
-      "step": 1189
-    },
-    {
-      "epoch": 0.07777269459512451,
-      "grad_norm": 0.5650060176849365,
-      "learning_rate": 9.999104533731064e-06,
-      "loss": 0.4258,
-      "step": 1190
-    },
-    {
-      "epoch": 0.07783804980066662,
-      "grad_norm": 0.5023185014724731,
-      "learning_rate": 9.999097913130002e-06,
-      "loss": 0.455,
-      "step": 1191
-    },
-    {
-      "epoch": 0.07790340500620875,
-      "grad_norm": 0.5457863211631775,
-      "learning_rate": 9.999091268146608e-06,
-      "loss": 0.4893,
-      "step": 1192
-    },
-    {
-      "epoch": 0.07796876021175086,
-      "grad_norm": 0.5315849184989929,
-      "learning_rate": 9.999084598780914e-06,
-      "loss": 0.4525,
-      "step": 1193
-    },
-    {
-      "epoch": 0.07803411541729299,
-      "grad_norm": 0.513178288936615,
-      "learning_rate": 9.999077905032953e-06,
-      "loss": 0.4396,
-      "step": 1194
-    },
-    {
-      "epoch": 0.07809947062283511,
-      "grad_norm": 0.525798499584198,
-      "learning_rate": 9.999071186902758e-06,
-      "loss": 0.4627,
-      "step": 1195
-    },
-    {
-      "epoch": 0.07816482582837724,
-      "grad_norm": 0.5288268327713013,
-      "learning_rate": 9.999064444390361e-06,
-      "loss": 0.4314,
-      "step": 1196
-    },
-    {
-      "epoch": 0.07823018103391935,
-      "grad_norm": 0.48024097084999084,
-      "learning_rate": 9.999057677495794e-06,
-      "loss": 0.4085,
-      "step": 1197
-    },
-    {
-      "epoch": 0.07829553623946148,
-      "grad_norm": 0.5265238881111145,
-      "learning_rate": 9.99905088621909e-06,
-      "loss": 0.4385,
-      "step": 1198
-    },
-    {
-      "epoch": 0.07836089144500359,
-      "grad_norm": 0.5488736033439636,
-      "learning_rate": 9.999044070560285e-06,
-      "loss": 0.4245,
-      "step": 1199
-    },
-    {
-      "epoch": 0.07842624665054572,
-      "grad_norm": 0.5518878698348999,
-      "learning_rate": 9.999037230519408e-06,
-      "loss": 0.4259,
-      "step": 1200
-    },
-    {
-      "epoch": 0.07849160185608783,
-      "grad_norm": 0.49042201042175293,
-      "learning_rate": 9.999030366096495e-06,
-      "loss": 0.4178,
-      "step": 1201
-    },
-    {
-      "epoch": 0.07855695706162996,
-      "grad_norm": 0.5355582237243652,
-      "learning_rate": 9.99902347729158e-06,
-      "loss": 0.4266,
-      "step": 1202
-    },
-    {
-      "epoch": 0.07862231226717208,
-      "grad_norm": 0.6018792986869812,
-      "learning_rate": 9.999016564104696e-06,
-      "loss": 0.5454,
-      "step": 1203
-    },
-    {
-      "epoch": 0.0786876674727142,
-      "grad_norm": 0.504183828830719,
-      "learning_rate": 9.999009626535877e-06,
-      "loss": 0.398,
-      "step": 1204
-    },
-    {
-      "epoch": 0.07875302267825632,
-      "grad_norm": 0.5036664009094238,
-      "learning_rate": 9.999002664585153e-06,
-      "loss": 0.4108,
-      "step": 1205
-    },
-    {
-      "epoch": 0.07881837788379845,
-      "grad_norm": 0.4950101673603058,
-      "learning_rate": 9.998995678252564e-06,
-      "loss": 0.4001,
-      "step": 1206
-    },
-    {
-      "epoch": 0.07888373308934056,
-      "grad_norm": 0.5225205421447754,
-      "learning_rate": 9.99898866753814e-06,
-      "loss": 0.4438,
-      "step": 1207
-    },
-    {
-      "epoch": 0.07894908829488269,
-      "grad_norm": 0.5419541001319885,
-      "learning_rate": 9.998981632441917e-06,
-      "loss": 0.4618,
-      "step": 1208
-    },
-    {
-      "epoch": 0.0790144435004248,
-      "grad_norm": 0.47289782762527466,
-      "learning_rate": 9.998974572963929e-06,
-      "loss": 0.4279,
-      "step": 1209
-    },
-    {
-      "epoch": 0.07907979870596693,
-      "grad_norm": 0.4971350133419037,
-      "learning_rate": 9.99896748910421e-06,
-      "loss": 0.4121,
-      "step": 1210
-    },
-    {
-      "epoch": 0.07914515391150906,
-      "grad_norm": 0.48509129881858826,
-      "learning_rate": 9.998960380862794e-06,
-      "loss": 0.3895,
-      "step": 1211
-    },
-    {
-      "epoch": 0.07921050911705117,
-      "grad_norm": 0.5354259014129639,
-      "learning_rate": 9.998953248239717e-06,
-      "loss": 0.4433,
-      "step": 1212
-    },
-    {
-      "epoch": 0.0792758643225933,
-      "grad_norm": 0.49842292070388794,
-      "learning_rate": 9.998946091235014e-06,
-      "loss": 0.4251,
-      "step": 1213
-    },
-    {
-      "epoch": 0.07934121952813541,
-      "grad_norm": 0.5182493925094604,
-      "learning_rate": 9.998938909848718e-06,
-      "loss": 0.4526,
-      "step": 1214
-    },
-    {
-      "epoch": 0.07940657473367754,
-      "grad_norm": 0.48397088050842285,
-      "learning_rate": 9.998931704080867e-06,
-      "loss": 0.3854,
-      "step": 1215
-    },
-    {
-      "epoch": 0.07947192993921966,
-      "grad_norm": 0.5367823839187622,
-      "learning_rate": 9.998924473931493e-06,
-      "loss": 0.4529,
-      "step": 1216
-    },
-    {
-      "epoch": 0.07953728514476179,
-      "grad_norm": 0.5396056175231934,
-      "learning_rate": 9.998917219400632e-06,
-      "loss": 0.464,
-      "step": 1217
-    },
-    {
-      "epoch": 0.0796026403503039,
-      "grad_norm": 0.5314146280288696,
-      "learning_rate": 9.99890994048832e-06,
-      "loss": 0.441,
-      "step": 1218
-    },
-    {
-      "epoch": 0.07966799555584603,
-      "grad_norm": 0.50490403175354,
-      "learning_rate": 9.998902637194593e-06,
-      "loss": 0.4734,
-      "step": 1219
-    },
-    {
-      "epoch": 0.07973335076138814,
-      "grad_norm": 0.5450233817100525,
-      "learning_rate": 9.998895309519484e-06,
-      "loss": 0.4509,
-      "step": 1220
-    },
-    {
-      "epoch": 0.07979870596693027,
-      "grad_norm": 0.5138707756996155,
-      "learning_rate": 9.998887957463034e-06,
-      "loss": 0.4645,
-      "step": 1221
-    },
-    {
-      "epoch": 0.07986406117247238,
-      "grad_norm": 0.5474032163619995,
-      "learning_rate": 9.998880581025274e-06,
-      "loss": 0.4259,
-      "step": 1222
-    },
-    {
-      "epoch": 0.07992941637801451,
-      "grad_norm": 0.4801062345504761,
-      "learning_rate": 9.998873180206242e-06,
-      "loss": 0.4245,
-      "step": 1223
-    },
-    {
-      "epoch": 0.07999477158355663,
-      "grad_norm": 0.5042070746421814,
-      "learning_rate": 9.998865755005973e-06,
-      "loss": 0.4157,
-      "step": 1224
-    },
-    {
-      "epoch": 0.08006012678909875,
-      "grad_norm": 0.49243617057800293,
-      "learning_rate": 9.998858305424506e-06,
-      "loss": 0.4372,
-      "step": 1225
-    },
-    {
-      "epoch": 0.08012548199464087,
-      "grad_norm": 0.5416117906570435,
-      "learning_rate": 9.998850831461873e-06,
-      "loss": 0.4519,
-      "step": 1226
-    },
-    {
-      "epoch": 0.080190837200183,
-      "grad_norm": 0.528499186038971,
-      "learning_rate": 9.998843333118113e-06,
-      "loss": 0.4893,
-      "step": 1227
-    },
-    {
-      "epoch": 0.08025619240572511,
-      "grad_norm": 0.4788937270641327,
-      "learning_rate": 9.998835810393264e-06,
-      "loss": 0.4147,
-      "step": 1228
-    },
-    {
-      "epoch": 0.08032154761126724,
-      "grad_norm": 0.5041825771331787,
-      "learning_rate": 9.998828263287359e-06,
-      "loss": 0.4216,
-      "step": 1229
-    },
-    {
-      "epoch": 0.08038690281680935,
-      "grad_norm": 0.5103921890258789,
-      "learning_rate": 9.998820691800439e-06,
-      "loss": 0.4718,
-      "step": 1230
-    },
-    {
-      "epoch": 0.08045225802235148,
-      "grad_norm": 0.4930843114852905,
-      "learning_rate": 9.998813095932536e-06,
-      "loss": 0.4359,
-      "step": 1231
-    },
-    {
-      "epoch": 0.08051761322789361,
-      "grad_norm": 0.46574723720550537,
-      "learning_rate": 9.998805475683691e-06,
-      "loss": 0.3872,
-      "step": 1232
-    },
-    {
-      "epoch": 0.08058296843343572,
-      "grad_norm": 0.5203606486320496,
-      "learning_rate": 9.998797831053942e-06,
-      "loss": 0.4558,
-      "step": 1233
-    },
-    {
-      "epoch": 0.08064832363897785,
-      "grad_norm": 0.5298285484313965,
-      "learning_rate": 9.998790162043321e-06,
-      "loss": 0.4588,
-      "step": 1234
-    },
-    {
-      "epoch": 0.08071367884451996,
-      "grad_norm": 0.5079378485679626,
-      "learning_rate": 9.99878246865187e-06,
-      "loss": 0.4715,
-      "step": 1235
-    },
-    {
-      "epoch": 0.08077903405006209,
-      "grad_norm": 0.4804714620113373,
-      "learning_rate": 9.998774750879626e-06,
-      "loss": 0.3643,
-      "step": 1236
-    },
-    {
-      "epoch": 0.08084438925560421,
-      "grad_norm": 0.5205805897712708,
-      "learning_rate": 9.998767008726624e-06,
-      "loss": 0.3987,
-      "step": 1237
-    },
-    {
-      "epoch": 0.08090974446114634,
-      "grad_norm": 0.5701265931129456,
-      "learning_rate": 9.998759242192904e-06,
-      "loss": 0.4508,
-      "step": 1238
-    },
-    {
-      "epoch": 0.08097509966668845,
-      "grad_norm": 0.4982219934463501,
-      "learning_rate": 9.998751451278504e-06,
-      "loss": 0.4393,
-      "step": 1239
-    },
-    {
-      "epoch": 0.08104045487223058,
-      "grad_norm": 0.5158557891845703,
-      "learning_rate": 9.99874363598346e-06,
-      "loss": 0.3905,
-      "step": 1240
-    },
-    {
-      "epoch": 0.08110581007777269,
-      "grad_norm": 0.5369464755058289,
-      "learning_rate": 9.998735796307815e-06,
-      "loss": 0.4403,
-      "step": 1241
-    },
-    {
-      "epoch": 0.08117116528331482,
-      "grad_norm": 0.515998125076294,
-      "learning_rate": 9.998727932251602e-06,
-      "loss": 0.409,
-      "step": 1242
-    },
-    {
-      "epoch": 0.08123652048885693,
-      "grad_norm": 0.5507062673568726,
-      "learning_rate": 9.99872004381486e-06,
-      "loss": 0.4413,
-      "step": 1243
-    },
-    {
-      "epoch": 0.08130187569439906,
-      "grad_norm": 0.5182384848594666,
-      "learning_rate": 9.99871213099763e-06,
-      "loss": 0.4661,
-      "step": 1244
-    },
-    {
-      "epoch": 0.08136723089994118,
-      "grad_norm": 0.510947585105896,
-      "learning_rate": 9.998704193799948e-06,
-      "loss": 0.4406,
-      "step": 1245
-    },
-    {
-      "epoch": 0.0814325861054833,
-      "grad_norm": 0.5021692514419556,
-      "learning_rate": 9.998696232221854e-06,
-      "loss": 0.4215,
-      "step": 1246
-    },
-    {
-      "epoch": 0.08149794131102542,
-      "grad_norm": 0.49621596932411194,
-      "learning_rate": 9.998688246263388e-06,
-      "loss": 0.4088,
-      "step": 1247
-    },
-    {
-      "epoch": 0.08156329651656755,
-      "grad_norm": 0.528205931186676,
-      "learning_rate": 9.998680235924587e-06,
-      "loss": 0.4198,
-      "step": 1248
-    },
-    {
-      "epoch": 0.08162865172210966,
-      "grad_norm": 0.46504709124565125,
-      "learning_rate": 9.99867220120549e-06,
-      "loss": 0.3673,
-      "step": 1249
-    },
-    {
-      "epoch": 0.08169400692765179,
-      "grad_norm": 0.5134182572364807,
-      "learning_rate": 9.998664142106138e-06,
-      "loss": 0.4325,
-      "step": 1250
-    },
-    {
-      "epoch": 0.0817593621331939,
-      "grad_norm": 0.5338544845581055,
-      "learning_rate": 9.99865605862657e-06,
-      "loss": 0.4973,
-      "step": 1251
-    },
-    {
-      "epoch": 0.08182471733873603,
-      "grad_norm": 0.4876263439655304,
-      "learning_rate": 9.998647950766824e-06,
-      "loss": 0.4073,
-      "step": 1252
-    },
-    {
-      "epoch": 0.08189007254427816,
-      "grad_norm": 0.4959527850151062,
-      "learning_rate": 9.998639818526939e-06,
-      "loss": 0.4325,
-      "step": 1253
-    },
-    {
-      "epoch": 0.08195542774982027,
-      "grad_norm": 0.5380318760871887,
-      "learning_rate": 9.998631661906957e-06,
-      "loss": 0.447,
-      "step": 1254
-    },
-    {
-      "epoch": 0.0820207829553624,
-      "grad_norm": 0.5196751952171326,
-      "learning_rate": 9.998623480906917e-06,
-      "loss": 0.471,
-      "step": 1255
-    },
-    {
-      "epoch": 0.08208613816090451,
-      "grad_norm": 0.49254852533340454,
-      "learning_rate": 9.998615275526859e-06,
-      "loss": 0.4241,
-      "step": 1256
-    },
-    {
-      "epoch": 0.08215149336644664,
-      "grad_norm": 0.5189266800880432,
-      "learning_rate": 9.998607045766822e-06,
-      "loss": 0.4352,
-      "step": 1257
-    },
-    {
-      "epoch": 0.08221684857198876,
-      "grad_norm": 0.531071662902832,
-      "learning_rate": 9.998598791626846e-06,
-      "loss": 0.48,
-      "step": 1258
-    },
-    {
-      "epoch": 0.08228220377753089,
-      "grad_norm": 0.49283671379089355,
-      "learning_rate": 9.998590513106973e-06,
-      "loss": 0.3903,
-      "step": 1259
-    },
-    {
-      "epoch": 0.082347558983073,
-      "grad_norm": 0.47999781370162964,
-      "learning_rate": 9.998582210207242e-06,
-      "loss": 0.4123,
-      "step": 1260
-    },
-    {
-      "epoch": 0.08241291418861513,
-      "grad_norm": 0.477071613073349,
-      "learning_rate": 9.998573882927694e-06,
-      "loss": 0.3761,
-      "step": 1261
-    },
-    {
-      "epoch": 0.08247826939415724,
-      "grad_norm": 0.5222316980361938,
-      "learning_rate": 9.998565531268369e-06,
-      "loss": 0.4376,
-      "step": 1262
-    },
-    {
-      "epoch": 0.08254362459969937,
-      "grad_norm": 0.516411304473877,
-      "learning_rate": 9.998557155229308e-06,
-      "loss": 0.4561,
-      "step": 1263
-    },
-    {
-      "epoch": 0.08260897980524148,
-      "grad_norm": 0.5287901163101196,
-      "learning_rate": 9.998548754810553e-06,
-      "loss": 0.4359,
-      "step": 1264
-    },
-    {
-      "epoch": 0.08267433501078361,
-      "grad_norm": 0.5043032169342041,
-      "learning_rate": 9.998540330012143e-06,
-      "loss": 0.44,
-      "step": 1265
-    },
-    {
-      "epoch": 0.08273969021632573,
-      "grad_norm": 0.5158609747886658,
-      "learning_rate": 9.998531880834121e-06,
-      "loss": 0.4713,
-      "step": 1266
-    },
-    {
-      "epoch": 0.08280504542186785,
-      "grad_norm": 0.5348635315895081,
-      "learning_rate": 9.998523407276528e-06,
-      "loss": 0.4345,
-      "step": 1267
-    },
-    {
-      "epoch": 0.08287040062740997,
-      "grad_norm": 0.500033438205719,
-      "learning_rate": 9.998514909339404e-06,
-      "loss": 0.416,
-      "step": 1268
-    },
-    {
-      "epoch": 0.0829357558329521,
-      "grad_norm": 0.5261061191558838,
-      "learning_rate": 9.99850638702279e-06,
-      "loss": 0.4685,
-      "step": 1269
-    },
-    {
-      "epoch": 0.08300111103849421,
-      "grad_norm": 0.5408302545547485,
-      "learning_rate": 9.998497840326731e-06,
-      "loss": 0.4679,
-      "step": 1270
-    },
-    {
-      "epoch": 0.08306646624403634,
-      "grad_norm": 0.5662031173706055,
-      "learning_rate": 9.998489269251266e-06,
-      "loss": 0.4473,
-      "step": 1271
-    },
-    {
-      "epoch": 0.08313182144957845,
-      "grad_norm": 0.5220599174499512,
-      "learning_rate": 9.998480673796435e-06,
-      "loss": 0.4566,
-      "step": 1272
-    },
-    {
-      "epoch": 0.08319717665512058,
-      "grad_norm": 0.5670110583305359,
-      "learning_rate": 9.998472053962285e-06,
-      "loss": 0.5111,
-      "step": 1273
-    },
-    {
-      "epoch": 0.08326253186066271,
-      "grad_norm": 0.5320886373519897,
-      "learning_rate": 9.998463409748852e-06,
-      "loss": 0.4318,
-      "step": 1274
-    },
-    {
-      "epoch": 0.08332788706620482,
-      "grad_norm": 0.5130778551101685,
-      "learning_rate": 9.998454741156184e-06,
-      "loss": 0.3946,
-      "step": 1275
-    },
-    {
-      "epoch": 0.08339324227174695,
-      "grad_norm": 0.4906269311904907,
-      "learning_rate": 9.99844604818432e-06,
-      "loss": 0.414,
-      "step": 1276
-    },
-    {
-      "epoch": 0.08345859747728906,
-      "grad_norm": 0.5453106760978699,
-      "learning_rate": 9.998437330833302e-06,
-      "loss": 0.5047,
-      "step": 1277
-    },
-    {
-      "epoch": 0.08352395268283119,
-      "grad_norm": 0.5290608406066895,
-      "learning_rate": 9.998428589103174e-06,
-      "loss": 0.4424,
-      "step": 1278
-    },
-    {
-      "epoch": 0.08358930788837331,
-      "grad_norm": 0.5562750697135925,
-      "learning_rate": 9.998419822993979e-06,
-      "loss": 0.4446,
-      "step": 1279
-    },
-    {
-      "epoch": 0.08365466309391544,
-      "grad_norm": 0.47817593812942505,
-      "learning_rate": 9.998411032505758e-06,
-      "loss": 0.4054,
-      "step": 1280
-    },
-    {
-      "epoch": 0.08372001829945755,
-      "grad_norm": 0.5376811027526855,
-      "learning_rate": 9.998402217638554e-06,
-      "loss": 0.4594,
-      "step": 1281
-    },
-    {
-      "epoch": 0.08378537350499968,
-      "grad_norm": 0.5015135407447815,
-      "learning_rate": 9.998393378392413e-06,
-      "loss": 0.3934,
-      "step": 1282
-    },
-    {
-      "epoch": 0.08385072871054179,
-      "grad_norm": 0.5472721457481384,
-      "learning_rate": 9.998384514767374e-06,
-      "loss": 0.4683,
-      "step": 1283
-    },
-    {
-      "epoch": 0.08391608391608392,
-      "grad_norm": 0.47711455821990967,
-      "learning_rate": 9.998375626763482e-06,
-      "loss": 0.4335,
-      "step": 1284
-    },
-    {
-      "epoch": 0.08398143912162603,
-      "grad_norm": 0.5403627753257751,
-      "learning_rate": 9.99836671438078e-06,
-      "loss": 0.4436,
-      "step": 1285
-    },
-    {
-      "epoch": 0.08404679432716816,
-      "grad_norm": 0.47652876377105713,
-      "learning_rate": 9.998357777619314e-06,
-      "loss": 0.4103,
-      "step": 1286
-    },
-    {
-      "epoch": 0.08411214953271028,
-      "grad_norm": 0.5195986032485962,
-      "learning_rate": 9.998348816479124e-06,
-      "loss": 0.4508,
-      "step": 1287
-    },
-    {
-      "epoch": 0.0841775047382524,
-      "grad_norm": 0.4842926561832428,
-      "learning_rate": 9.998339830960257e-06,
-      "loss": 0.4039,
-      "step": 1288
-    },
-    {
-      "epoch": 0.08424285994379452,
-      "grad_norm": 0.5277332067489624,
-      "learning_rate": 9.998330821062754e-06,
-      "loss": 0.4339,
-      "step": 1289
-    },
-    {
-      "epoch": 0.08430821514933665,
-      "grad_norm": 0.5088818073272705,
-      "learning_rate": 9.99832178678666e-06,
-      "loss": 0.4918,
-      "step": 1290
-    },
-    {
-      "epoch": 0.08437357035487876,
-      "grad_norm": 0.5293411016464233,
-      "learning_rate": 9.998312728132019e-06,
-      "loss": 0.4452,
-      "step": 1291
-    },
-    {
-      "epoch": 0.08443892556042089,
-      "grad_norm": 0.4776824116706848,
-      "learning_rate": 9.998303645098875e-06,
-      "loss": 0.3591,
-      "step": 1292
-    },
-    {
-      "epoch": 0.084504280765963,
-      "grad_norm": 0.5139163136482239,
-      "learning_rate": 9.998294537687273e-06,
-      "loss": 0.4308,
-      "step": 1293
-    },
-    {
-      "epoch": 0.08456963597150513,
-      "grad_norm": 0.48579853773117065,
-      "learning_rate": 9.998285405897256e-06,
-      "loss": 0.3964,
-      "step": 1294
-    },
-    {
-      "epoch": 0.08463499117704726,
-      "grad_norm": 0.5369120836257935,
-      "learning_rate": 9.99827624972887e-06,
-      "loss": 0.4503,
-      "step": 1295
-    },
-    {
-      "epoch": 0.08470034638258937,
-      "grad_norm": 0.5286991000175476,
-      "learning_rate": 9.99826706918216e-06,
-      "loss": 0.4463,
-      "step": 1296
-    },
-    {
-      "epoch": 0.0847657015881315,
-      "grad_norm": 0.5808084607124329,
-      "learning_rate": 9.998257864257169e-06,
-      "loss": 0.4637,
-      "step": 1297
-    },
-    {
-      "epoch": 0.08483105679367361,
-      "grad_norm": 0.474880576133728,
-      "learning_rate": 9.998248634953942e-06,
-      "loss": 0.3975,
-      "step": 1298
-    },
-    {
-      "epoch": 0.08489641199921574,
-      "grad_norm": 0.5344276428222656,
-      "learning_rate": 9.998239381272527e-06,
-      "loss": 0.4606,
-      "step": 1299
-    },
-    {
-      "epoch": 0.08496176720475786,
-      "grad_norm": 0.5275298357009888,
-      "learning_rate": 9.998230103212966e-06,
-      "loss": 0.4647,
-      "step": 1300
-    },
-    {
-      "epoch": 0.08502712241029999,
-      "grad_norm": 0.5716614723205566,
-      "learning_rate": 9.998220800775304e-06,
-      "loss": 0.4029,
-      "step": 1301
-    },
-    {
-      "epoch": 0.0850924776158421,
-      "grad_norm": 0.5430360436439514,
-      "learning_rate": 9.99821147395959e-06,
-      "loss": 0.4885,
-      "step": 1302
-    },
-    {
-      "epoch": 0.08515783282138423,
-      "grad_norm": 0.5336328148841858,
-      "learning_rate": 9.998202122765866e-06,
-      "loss": 0.4458,
-      "step": 1303
-    },
-    {
-      "epoch": 0.08522318802692634,
-      "grad_norm": 0.5014939904212952,
-      "learning_rate": 9.998192747194178e-06,
-      "loss": 0.4085,
-      "step": 1304
-    },
-    {
-      "epoch": 0.08528854323246847,
-      "grad_norm": 0.49508243799209595,
-      "learning_rate": 9.998183347244574e-06,
-      "loss": 0.3786,
-      "step": 1305
-    },
-    {
-      "epoch": 0.08535389843801058,
-      "grad_norm": 0.5499253273010254,
-      "learning_rate": 9.998173922917096e-06,
-      "loss": 0.4776,
-      "step": 1306
-    },
-    {
-      "epoch": 0.08541925364355271,
-      "grad_norm": 0.5309222340583801,
-      "learning_rate": 9.998164474211794e-06,
-      "loss": 0.4511,
-      "step": 1307
-    },
-    {
-      "epoch": 0.08548460884909483,
-      "grad_norm": 0.5119657516479492,
-      "learning_rate": 9.998155001128713e-06,
-      "loss": 0.3875,
-      "step": 1308
-    },
-    {
-      "epoch": 0.08554996405463695,
-      "grad_norm": 0.5707727074623108,
-      "learning_rate": 9.998145503667896e-06,
-      "loss": 0.4268,
-      "step": 1309
-    },
-    {
-      "epoch": 0.08561531926017907,
-      "grad_norm": 0.5600884556770325,
-      "learning_rate": 9.998135981829393e-06,
-      "loss": 0.5066,
-      "step": 1310
-    },
-    {
-      "epoch": 0.0856806744657212,
-      "grad_norm": 0.5002754330635071,
-      "learning_rate": 9.99812643561325e-06,
-      "loss": 0.3898,
-      "step": 1311
-    },
-    {
-      "epoch": 0.08574602967126331,
-      "grad_norm": 0.5500643253326416,
-      "learning_rate": 9.998116865019513e-06,
-      "loss": 0.4632,
-      "step": 1312
-    },
-    {
-      "epoch": 0.08581138487680544,
-      "grad_norm": 0.4922448992729187,
-      "learning_rate": 9.998107270048228e-06,
-      "loss": 0.4229,
-      "step": 1313
-    },
-    {
-      "epoch": 0.08587674008234755,
-      "grad_norm": 0.5603806376457214,
-      "learning_rate": 9.998097650699441e-06,
-      "loss": 0.4493,
-      "step": 1314
-    },
-    {
-      "epoch": 0.08594209528788968,
-      "grad_norm": 0.604393720626831,
-      "learning_rate": 9.998088006973203e-06,
-      "loss": 0.3981,
-      "step": 1315
-    },
-    {
-      "epoch": 0.08600745049343181,
-      "grad_norm": 0.51221764087677,
-      "learning_rate": 9.998078338869557e-06,
-      "loss": 0.4158,
-      "step": 1316
-    },
-    {
-      "epoch": 0.08607280569897392,
-      "grad_norm": 0.49882960319519043,
-      "learning_rate": 9.998068646388551e-06,
-      "loss": 0.3689,
-      "step": 1317
-    },
-    {
-      "epoch": 0.08613816090451605,
-      "grad_norm": 0.48062098026275635,
-      "learning_rate": 9.998058929530233e-06,
-      "loss": 0.4114,
-      "step": 1318
-    },
-    {
-      "epoch": 0.08620351611005816,
-      "grad_norm": 0.5422084927558899,
-      "learning_rate": 9.998049188294649e-06,
-      "loss": 0.3996,
-      "step": 1319
-    },
-    {
-      "epoch": 0.08626887131560029,
-      "grad_norm": 0.46148210763931274,
-      "learning_rate": 9.99803942268185e-06,
-      "loss": 0.34,
-      "step": 1320
-    },
-    {
-      "epoch": 0.08633422652114241,
-      "grad_norm": 0.5104997754096985,
-      "learning_rate": 9.998029632691879e-06,
-      "loss": 0.4344,
-      "step": 1321
-    },
-    {
-      "epoch": 0.08639958172668454,
-      "grad_norm": 0.49040332436561584,
-      "learning_rate": 9.998019818324787e-06,
-      "loss": 0.4351,
-      "step": 1322
-    },
-    {
-      "epoch": 0.08646493693222665,
-      "grad_norm": 0.471892386674881,
-      "learning_rate": 9.998009979580621e-06,
-      "loss": 0.3806,
-      "step": 1323
-    },
-    {
-      "epoch": 0.08653029213776878,
-      "grad_norm": 0.5132998824119568,
-      "learning_rate": 9.998000116459429e-06,
-      "loss": 0.4145,
-      "step": 1324
-    },
-    {
-      "epoch": 0.08659564734331089,
-      "grad_norm": 0.7055545449256897,
-      "learning_rate": 9.997990228961258e-06,
-      "loss": 0.4662,
-      "step": 1325
-    },
-    {
-      "epoch": 0.08666100254885302,
-      "grad_norm": 0.49374476075172424,
-      "learning_rate": 9.997980317086157e-06,
-      "loss": 0.4424,
-      "step": 1326
-    },
-    {
-      "epoch": 0.08672635775439513,
-      "grad_norm": 0.4715639054775238,
-      "learning_rate": 9.997970380834176e-06,
-      "loss": 0.3937,
-      "step": 1327
-    },
-    {
-      "epoch": 0.08679171295993726,
-      "grad_norm": 0.4831608831882477,
-      "learning_rate": 9.997960420205361e-06,
-      "loss": 0.3895,
-      "step": 1328
-    },
-    {
-      "epoch": 0.08685706816547938,
-      "grad_norm": 0.5228745937347412,
-      "learning_rate": 9.997950435199763e-06,
-      "loss": 0.4129,
-      "step": 1329
-    },
-    {
-      "epoch": 0.0869224233710215,
-      "grad_norm": 0.5391446352005005,
-      "learning_rate": 9.997940425817427e-06,
-      "loss": 0.4552,
-      "step": 1330
-    },
-    {
-      "epoch": 0.08698777857656362,
-      "grad_norm": 0.4888913333415985,
-      "learning_rate": 9.997930392058405e-06,
-      "loss": 0.4266,
-      "step": 1331
-    },
-    {
-      "epoch": 0.08705313378210575,
-      "grad_norm": 0.4974735379219055,
-      "learning_rate": 9.997920333922745e-06,
-      "loss": 0.4104,
-      "step": 1332
-    },
-    {
-      "epoch": 0.08711848898764786,
-      "grad_norm": 0.5001447200775146,
-      "learning_rate": 9.997910251410499e-06,
-      "loss": 0.4628,
-      "step": 1333
-    },
-    {
-      "epoch": 0.08718384419318999,
-      "grad_norm": 0.5306075215339661,
-      "learning_rate": 9.99790014452171e-06,
-      "loss": 0.4202,
-      "step": 1334
-    },
-    {
-      "epoch": 0.0872491993987321,
-      "grad_norm": 0.4916905164718628,
-      "learning_rate": 9.99789001325643e-06,
-      "loss": 0.4299,
-      "step": 1335
-    },
-    {
-      "epoch": 0.08731455460427423,
-      "grad_norm": 0.5258871912956238,
-      "learning_rate": 9.99787985761471e-06,
-      "loss": 0.4328,
-      "step": 1336
-    },
-    {
-      "epoch": 0.08737990980981636,
-      "grad_norm": 0.4798186421394348,
-      "learning_rate": 9.9978696775966e-06,
-      "loss": 0.4385,
-      "step": 1337
-    },
-    {
-      "epoch": 0.08744526501535847,
-      "grad_norm": 0.48297154903411865,
-      "learning_rate": 9.997859473202146e-06,
-      "loss": 0.3989,
-      "step": 1338
-    },
-    {
-      "epoch": 0.0875106202209006,
-      "grad_norm": 0.501555860042572,
-      "learning_rate": 9.997849244431401e-06,
-      "loss": 0.4328,
-      "step": 1339
-    },
-    {
-      "epoch": 0.08757597542644271,
-      "grad_norm": 0.5189911723136902,
-      "learning_rate": 9.997838991284415e-06,
-      "loss": 0.4502,
-      "step": 1340
-    },
-    {
-      "epoch": 0.08764133063198484,
-      "grad_norm": 0.5261041522026062,
-      "learning_rate": 9.997828713761233e-06,
-      "loss": 0.4627,
-      "step": 1341
-    },
-    {
-      "epoch": 0.08770668583752696,
-      "grad_norm": 0.495754212141037,
-      "learning_rate": 9.997818411861912e-06,
-      "loss": 0.4418,
-      "step": 1342
-    },
-    {
-      "epoch": 0.08777204104306909,
-      "grad_norm": 0.4789591133594513,
-      "learning_rate": 9.997808085586499e-06,
-      "loss": 0.3847,
-      "step": 1343
-    },
-    {
-      "epoch": 0.0878373962486112,
-      "grad_norm": 0.48939236998558044,
-      "learning_rate": 9.997797734935045e-06,
-      "loss": 0.4056,
-      "step": 1344
-    },
-    {
-      "epoch": 0.08790275145415333,
-      "grad_norm": 0.4681905210018158,
-      "learning_rate": 9.997787359907598e-06,
-      "loss": 0.3849,
-      "step": 1345
-    },
-    {
-      "epoch": 0.08796810665969544,
-      "grad_norm": 0.5467929244041443,
-      "learning_rate": 9.997776960504212e-06,
-      "loss": 0.4675,
-      "step": 1346
-    },
-    {
-      "epoch": 0.08803346186523757,
-      "grad_norm": 0.5142377614974976,
-      "learning_rate": 9.997766536724936e-06,
-      "loss": 0.4462,
-      "step": 1347
-    },
-    {
-      "epoch": 0.08809881707077968,
-      "grad_norm": 0.5171924233436584,
-      "learning_rate": 9.997756088569821e-06,
-      "loss": 0.471,
-      "step": 1348
-    },
-    {
-      "epoch": 0.08816417227632181,
-      "grad_norm": 0.5064006447792053,
-      "learning_rate": 9.997745616038918e-06,
-      "loss": 0.4007,
-      "step": 1349
-    },
-    {
-      "epoch": 0.08822952748186393,
-      "grad_norm": 0.5326420068740845,
-      "learning_rate": 9.997735119132279e-06,
-      "loss": 0.5005,
-      "step": 1350
-    },
-    {
-      "epoch": 0.08829488268740605,
-      "grad_norm": 0.5336724519729614,
-      "learning_rate": 9.997724597849955e-06,
-      "loss": 0.4846,
-      "step": 1351
-    },
-    {
-      "epoch": 0.08836023789294817,
-      "grad_norm": 0.5144051909446716,
-      "learning_rate": 9.997714052191996e-06,
-      "loss": 0.4372,
-      "step": 1352
-    },
-    {
-      "epoch": 0.0884255930984903,
-      "grad_norm": 0.5164487957954407,
-      "learning_rate": 9.997703482158454e-06,
-      "loss": 0.4418,
-      "step": 1353
-    },
-    {
-      "epoch": 0.08849094830403241,
-      "grad_norm": 0.4768177270889282,
-      "learning_rate": 9.997692887749381e-06,
-      "loss": 0.4048,
-      "step": 1354
-    },
-    {
-      "epoch": 0.08855630350957454,
-      "grad_norm": 0.5341006517410278,
-      "learning_rate": 9.997682268964828e-06,
-      "loss": 0.4612,
-      "step": 1355
-    },
-    {
-      "epoch": 0.08862165871511665,
-      "grad_norm": 0.5181215405464172,
-      "learning_rate": 9.997671625804848e-06,
-      "loss": 0.4461,
-      "step": 1356
-    },
-    {
-      "epoch": 0.08868701392065878,
-      "grad_norm": 0.47202199697494507,
-      "learning_rate": 9.997660958269491e-06,
-      "loss": 0.3583,
-      "step": 1357
-    },
-    {
-      "epoch": 0.08875236912620091,
-      "grad_norm": 0.4560900926589966,
-      "learning_rate": 9.997650266358811e-06,
-      "loss": 0.3716,
-      "step": 1358
-    },
-    {
-      "epoch": 0.08881772433174302,
-      "grad_norm": 0.5311445593833923,
-      "learning_rate": 9.99763955007286e-06,
-      "loss": 0.4473,
-      "step": 1359
-    },
-    {
-      "epoch": 0.08888307953728515,
-      "grad_norm": 0.4800487756729126,
-      "learning_rate": 9.997628809411688e-06,
-      "loss": 0.4024,
-      "step": 1360
-    },
-    {
-      "epoch": 0.08894843474282726,
-      "grad_norm": 0.5415546894073486,
-      "learning_rate": 9.99761804437535e-06,
-      "loss": 0.4829,
-      "step": 1361
-    },
-    {
-      "epoch": 0.08901378994836939,
-      "grad_norm": 0.47608330845832825,
-      "learning_rate": 9.997607254963896e-06,
-      "loss": 0.3918,
-      "step": 1362
-    },
-    {
-      "epoch": 0.08907914515391151,
-      "grad_norm": 0.490690678358078,
-      "learning_rate": 9.997596441177381e-06,
-      "loss": 0.4053,
-      "step": 1363
-    },
-    {
-      "epoch": 0.08914450035945364,
-      "grad_norm": 0.4878924489021301,
-      "learning_rate": 9.997585603015858e-06,
-      "loss": 0.4084,
-      "step": 1364
-    },
-    {
-      "epoch": 0.08920985556499575,
-      "grad_norm": 0.5493990182876587,
-      "learning_rate": 9.997574740479377e-06,
-      "loss": 0.4568,
-      "step": 1365
-    },
-    {
-      "epoch": 0.08927521077053788,
-      "grad_norm": 0.5351925492286682,
-      "learning_rate": 9.997563853567994e-06,
-      "loss": 0.4328,
-      "step": 1366
-    },
-    {
-      "epoch": 0.08934056597607999,
-      "grad_norm": 0.4934317171573639,
-      "learning_rate": 9.997552942281759e-06,
-      "loss": 0.3862,
-      "step": 1367
-    },
-    {
-      "epoch": 0.08940592118162212,
-      "grad_norm": 0.5227459073066711,
-      "learning_rate": 9.997542006620728e-06,
-      "loss": 0.4412,
-      "step": 1368
-    },
-    {
-      "epoch": 0.08947127638716423,
-      "grad_norm": 0.5125664472579956,
-      "learning_rate": 9.997531046584954e-06,
-      "loss": 0.4295,
-      "step": 1369
-    },
-    {
-      "epoch": 0.08953663159270636,
-      "grad_norm": 0.5206894278526306,
-      "learning_rate": 9.99752006217449e-06,
-      "loss": 0.4631,
-      "step": 1370
-    },
-    {
-      "epoch": 0.08960198679824848,
-      "grad_norm": 0.5265061855316162,
-      "learning_rate": 9.997509053389386e-06,
-      "loss": 0.4301,
-      "step": 1371
-    },
-    {
-      "epoch": 0.0896673420037906,
-      "grad_norm": 0.4820745885372162,
-      "learning_rate": 9.997498020229703e-06,
-      "loss": 0.4047,
-      "step": 1372
-    },
-    {
-      "epoch": 0.08973269720933272,
-      "grad_norm": 0.4692222476005554,
-      "learning_rate": 9.99748696269549e-06,
-      "loss": 0.3754,
-      "step": 1373
-    },
-    {
-      "epoch": 0.08979805241487485,
-      "grad_norm": 0.5081862211227417,
-      "learning_rate": 9.9974758807868e-06,
-      "loss": 0.4253,
-      "step": 1374
-    },
-    {
-      "epoch": 0.08986340762041696,
-      "grad_norm": 0.5094119310379028,
-      "learning_rate": 9.997464774503691e-06,
-      "loss": 0.4216,
-      "step": 1375
-    },
-    {
-      "epoch": 0.08992876282595909,
-      "grad_norm": 0.5726935267448425,
-      "learning_rate": 9.997453643846213e-06,
-      "loss": 0.4377,
-      "step": 1376
-    },
-    {
-      "epoch": 0.0899941180315012,
-      "grad_norm": 0.5270561575889587,
-      "learning_rate": 9.997442488814423e-06,
-      "loss": 0.4702,
-      "step": 1377
-    },
-    {
-      "epoch": 0.09005947323704333,
-      "grad_norm": 0.5101881623268127,
-      "learning_rate": 9.997431309408376e-06,
-      "loss": 0.4717,
-      "step": 1378
-    },
-    {
-      "epoch": 0.09012482844258546,
-      "grad_norm": 0.5228449106216431,
-      "learning_rate": 9.997420105628124e-06,
-      "loss": 0.4525,
-      "step": 1379
-    },
-    {
-      "epoch": 0.09019018364812757,
-      "grad_norm": 0.5499043464660645,
-      "learning_rate": 9.997408877473724e-06,
-      "loss": 0.4226,
-      "step": 1380
-    },
-    {
-      "epoch": 0.0902555388536697,
-      "grad_norm": 0.5792656540870667,
-      "learning_rate": 9.997397624945229e-06,
-      "loss": 0.49,
-      "step": 1381
-    },
-    {
-      "epoch": 0.09032089405921181,
-      "grad_norm": 0.49885094165802,
-      "learning_rate": 9.997386348042694e-06,
-      "loss": 0.4134,
-      "step": 1382
-    },
-    {
-      "epoch": 0.09038624926475394,
-      "grad_norm": 0.5309830904006958,
-      "learning_rate": 9.997375046766175e-06,
-      "loss": 0.4073,
-      "step": 1383
-    },
-    {
-      "epoch": 0.09045160447029606,
-      "grad_norm": 0.4986577332019806,
-      "learning_rate": 9.997363721115725e-06,
-      "loss": 0.4107,
-      "step": 1384
-    },
-    {
-      "epoch": 0.09051695967583819,
-      "grad_norm": 0.4967551827430725,
-      "learning_rate": 9.997352371091403e-06,
-      "loss": 0.3932,
-      "step": 1385
-    },
-    {
-      "epoch": 0.0905823148813803,
-      "grad_norm": 0.4970942735671997,
-      "learning_rate": 9.997340996693262e-06,
-      "loss": 0.4114,
-      "step": 1386
-    },
-    {
-      "epoch": 0.09064767008692243,
-      "grad_norm": 0.4806903302669525,
-      "learning_rate": 9.997329597921356e-06,
-      "loss": 0.4308,
-      "step": 1387
-    },
-    {
-      "epoch": 0.09071302529246454,
-      "grad_norm": 0.47248023748397827,
-      "learning_rate": 9.997318174775746e-06,
-      "loss": 0.3902,
-      "step": 1388
-    },
-    {
-      "epoch": 0.09077838049800667,
-      "grad_norm": 0.5082437992095947,
-      "learning_rate": 9.997306727256481e-06,
-      "loss": 0.4391,
-      "step": 1389
-    },
-    {
-      "epoch": 0.09084373570354878,
-      "grad_norm": 0.4743526577949524,
-      "learning_rate": 9.99729525536362e-06,
-      "loss": 0.3933,
-      "step": 1390
-    },
-    {
-      "epoch": 0.09090909090909091,
-      "grad_norm": 0.5254323482513428,
-      "learning_rate": 9.997283759097219e-06,
-      "loss": 0.4292,
-      "step": 1391
-    },
-    {
-      "epoch": 0.09097444611463303,
-      "grad_norm": 0.5116239786148071,
-      "learning_rate": 9.997272238457334e-06,
-      "loss": 0.4309,
-      "step": 1392
-    },
-    {
-      "epoch": 0.09103980132017515,
-      "grad_norm": 0.5199630856513977,
-      "learning_rate": 9.997260693444023e-06,
-      "loss": 0.4408,
-      "step": 1393
-    },
-    {
-      "epoch": 0.09110515652571727,
-      "grad_norm": 0.533906102180481,
-      "learning_rate": 9.997249124057337e-06,
-      "loss": 0.5071,
-      "step": 1394
-    },
-    {
-      "epoch": 0.0911705117312594,
-      "grad_norm": 0.5672672986984253,
-      "learning_rate": 9.997237530297338e-06,
-      "loss": 0.4823,
-      "step": 1395
-    },
-    {
-      "epoch": 0.09123586693680151,
-      "grad_norm": 0.5255613327026367,
-      "learning_rate": 9.997225912164078e-06,
-      "loss": 0.428,
-      "step": 1396
-    },
-    {
-      "epoch": 0.09130122214234364,
-      "grad_norm": 0.536078155040741,
-      "learning_rate": 9.99721426965762e-06,
-      "loss": 0.4634,
-      "step": 1397
-    },
-    {
-      "epoch": 0.09136657734788575,
-      "grad_norm": 0.48807597160339355,
-      "learning_rate": 9.997202602778014e-06,
-      "loss": 0.3934,
-      "step": 1398
-    },
-    {
-      "epoch": 0.09143193255342788,
-      "grad_norm": 0.4788658320903778,
-      "learning_rate": 9.99719091152532e-06,
-      "loss": 0.4111,
-      "step": 1399
-    },
-    {
-      "epoch": 0.09149728775897001,
-      "grad_norm": 0.47326433658599854,
-      "learning_rate": 9.997179195899595e-06,
-      "loss": 0.3763,
-      "step": 1400
-    },
-    {
-      "epoch": 0.09156264296451212,
-      "grad_norm": 0.5255954265594482,
-      "learning_rate": 9.997167455900896e-06,
-      "loss": 0.4695,
-      "step": 1401
-    },
-    {
-      "epoch": 0.09162799817005425,
-      "grad_norm": 0.474807471036911,
-      "learning_rate": 9.99715569152928e-06,
-      "loss": 0.3514,
-      "step": 1402
-    },
-    {
-      "epoch": 0.09169335337559636,
-      "grad_norm": 0.5895363092422485,
-      "learning_rate": 9.997143902784805e-06,
-      "loss": 0.5219,
-      "step": 1403
-    },
-    {
-      "epoch": 0.09175870858113849,
-      "grad_norm": 0.4866838753223419,
-      "learning_rate": 9.997132089667527e-06,
-      "loss": 0.3969,
-      "step": 1404
-    },
-    {
-      "epoch": 0.09182406378668061,
-      "grad_norm": 0.8142039179801941,
-      "learning_rate": 9.997120252177507e-06,
-      "loss": 0.5273,
-      "step": 1405
-    },
-    {
-      "epoch": 0.09188941899222274,
-      "grad_norm": 0.49827301502227783,
-      "learning_rate": 9.997108390314798e-06,
-      "loss": 0.4436,
-      "step": 1406
-    },
-    {
-      "epoch": 0.09195477419776485,
-      "grad_norm": 0.46191340684890747,
-      "learning_rate": 9.99709650407946e-06,
-      "loss": 0.357,
-      "step": 1407
-    },
-    {
-      "epoch": 0.09202012940330698,
-      "grad_norm": 0.5786252021789551,
-      "learning_rate": 9.997084593471552e-06,
-      "loss": 0.5216,
-      "step": 1408
-    },
-    {
-      "epoch": 0.09208548460884909,
-      "grad_norm": 0.5046772360801697,
-      "learning_rate": 9.997072658491131e-06,
-      "loss": 0.4555,
-      "step": 1409
-    },
-    {
-      "epoch": 0.09215083981439122,
-      "grad_norm": 0.524642288684845,
-      "learning_rate": 9.997060699138255e-06,
-      "loss": 0.4381,
-      "step": 1410
-    },
-    {
-      "epoch": 0.09221619501993333,
-      "grad_norm": 0.4975273311138153,
-      "learning_rate": 9.997048715412984e-06,
-      "loss": 0.3946,
-      "step": 1411
-    },
-    {
-      "epoch": 0.09228155022547546,
-      "grad_norm": 0.5561621785163879,
-      "learning_rate": 9.997036707315375e-06,
-      "loss": 0.4494,
-      "step": 1412
-    },
-    {
-      "epoch": 0.09234690543101758,
-      "grad_norm": 0.48719069361686707,
-      "learning_rate": 9.997024674845488e-06,
-      "loss": 0.4129,
-      "step": 1413
-    },
-    {
-      "epoch": 0.0924122606365597,
-      "grad_norm": 0.5011395812034607,
-      "learning_rate": 9.997012618003379e-06,
-      "loss": 0.4148,
-      "step": 1414
-    },
-    {
-      "epoch": 0.09247761584210182,
-      "grad_norm": 0.5327781438827515,
-      "learning_rate": 9.997000536789108e-06,
-      "loss": 0.4556,
-      "step": 1415
-    },
-    {
-      "epoch": 0.09254297104764395,
-      "grad_norm": 0.475110799074173,
-      "learning_rate": 9.996988431202735e-06,
-      "loss": 0.3896,
-      "step": 1416
-    },
-    {
-      "epoch": 0.09260832625318606,
-      "grad_norm": 0.5412651300430298,
-      "learning_rate": 9.996976301244317e-06,
-      "loss": 0.4329,
-      "step": 1417
-    },
-    {
-      "epoch": 0.09267368145872819,
-      "grad_norm": 0.4974592924118042,
-      "learning_rate": 9.996964146913917e-06,
-      "loss": 0.3904,
-      "step": 1418
-    },
-    {
-      "epoch": 0.0927390366642703,
-      "grad_norm": 0.5360898375511169,
-      "learning_rate": 9.99695196821159e-06,
-      "loss": 0.4746,
-      "step": 1419
-    },
-    {
-      "epoch": 0.09280439186981243,
-      "grad_norm": 0.5064204335212708,
-      "learning_rate": 9.996939765137396e-06,
-      "loss": 0.3793,
-      "step": 1420
-    },
-    {
-      "epoch": 0.09286974707535456,
-      "grad_norm": 0.47146132588386536,
-      "learning_rate": 9.996927537691398e-06,
-      "loss": 0.3923,
-      "step": 1421
-    },
-    {
-      "epoch": 0.09293510228089667,
-      "grad_norm": 0.5135983824729919,
-      "learning_rate": 9.996915285873652e-06,
-      "loss": 0.4106,
-      "step": 1422
-    },
-    {
-      "epoch": 0.0930004574864388,
-      "grad_norm": 0.4887148439884186,
-      "learning_rate": 9.99690300968422e-06,
-      "loss": 0.4121,
-      "step": 1423
-    },
-    {
-      "epoch": 0.09306581269198091,
-      "grad_norm": 0.5050368905067444,
-      "learning_rate": 9.996890709123161e-06,
-      "loss": 0.459,
-      "step": 1424
-    },
-    {
-      "epoch": 0.09313116789752304,
-      "grad_norm": 0.494385689496994,
-      "learning_rate": 9.996878384190534e-06,
-      "loss": 0.3678,
-      "step": 1425
-    },
-    {
-      "epoch": 0.09319652310306516,
-      "grad_norm": 0.5028787851333618,
-      "learning_rate": 9.9968660348864e-06,
-      "loss": 0.4066,
-      "step": 1426
-    },
-    {
-      "epoch": 0.09326187830860729,
-      "grad_norm": 0.4854961633682251,
-      "learning_rate": 9.99685366121082e-06,
-      "loss": 0.4324,
-      "step": 1427
-    },
-    {
-      "epoch": 0.0933272335141494,
-      "grad_norm": 0.5298318266868591,
-      "learning_rate": 9.996841263163853e-06,
-      "loss": 0.4301,
-      "step": 1428
-    },
-    {
-      "epoch": 0.09339258871969153,
-      "grad_norm": 0.5543140769004822,
-      "learning_rate": 9.996828840745561e-06,
-      "loss": 0.5094,
-      "step": 1429
-    },
-    {
-      "epoch": 0.09345794392523364,
-      "grad_norm": 0.4847946763038635,
-      "learning_rate": 9.996816393956002e-06,
-      "loss": 0.4099,
-      "step": 1430
-    },
-    {
-      "epoch": 0.09352329913077577,
-      "grad_norm": 0.5146633386611938,
-      "learning_rate": 9.996803922795239e-06,
-      "loss": 0.4173,
-      "step": 1431
-    },
-    {
-      "epoch": 0.09358865433631788,
-      "grad_norm": 0.5044020414352417,
-      "learning_rate": 9.996791427263333e-06,
-      "loss": 0.4084,
-      "step": 1432
-    },
-    {
-      "epoch": 0.09365400954186001,
-      "grad_norm": 0.49075740575790405,
-      "learning_rate": 9.996778907360343e-06,
-      "loss": 0.3819,
-      "step": 1433
-    },
-    {
-      "epoch": 0.09371936474740213,
-      "grad_norm": 0.4910449683666229,
-      "learning_rate": 9.996766363086332e-06,
-      "loss": 0.3806,
-      "step": 1434
-    },
-    {
-      "epoch": 0.09378471995294425,
-      "grad_norm": 0.48591721057891846,
-      "learning_rate": 9.996753794441361e-06,
-      "loss": 0.4651,
-      "step": 1435
-    },
-    {
-      "epoch": 0.09385007515848637,
-      "grad_norm": 0.5201261639595032,
-      "learning_rate": 9.996741201425491e-06,
-      "loss": 0.4427,
-      "step": 1436
-    },
-    {
-      "epoch": 0.0939154303640285,
-      "grad_norm": 0.48716381192207336,
-      "learning_rate": 9.996728584038782e-06,
-      "loss": 0.3618,
-      "step": 1437
-    },
-    {
-      "epoch": 0.09398078556957061,
-      "grad_norm": 0.4862307906150818,
-      "learning_rate": 9.996715942281297e-06,
-      "loss": 0.4023,
-      "step": 1438
-    },
-    {
-      "epoch": 0.09404614077511274,
-      "grad_norm": 0.48926007747650146,
-      "learning_rate": 9.996703276153095e-06,
-      "loss": 0.3815,
-      "step": 1439
-    },
-    {
-      "epoch": 0.09411149598065485,
-      "grad_norm": 1.492929220199585,
-      "learning_rate": 9.996690585654243e-06,
-      "loss": 0.4893,
-      "step": 1440
-    },
-    {
-      "epoch": 0.09417685118619698,
-      "grad_norm": 0.5764843821525574,
-      "learning_rate": 9.996677870784799e-06,
-      "loss": 0.4688,
-      "step": 1441
-    },
-    {
-      "epoch": 0.09424220639173911,
-      "grad_norm": 0.6438168287277222,
-      "learning_rate": 9.996665131544828e-06,
-      "loss": 0.4721,
-      "step": 1442
-    },
-    {
-      "epoch": 0.09430756159728122,
-      "grad_norm": 0.5278419256210327,
-      "learning_rate": 9.996652367934388e-06,
-      "loss": 0.4687,
-      "step": 1443
-    },
-    {
-      "epoch": 0.09437291680282335,
-      "grad_norm": 0.5648612976074219,
-      "learning_rate": 9.996639579953542e-06,
-      "loss": 0.4598,
-      "step": 1444
-    },
-    {
-      "epoch": 0.09443827200836546,
-      "grad_norm": 0.5199219584465027,
-      "learning_rate": 9.996626767602356e-06,
-      "loss": 0.4317,
-      "step": 1445
-    },
-    {
-      "epoch": 0.09450362721390759,
-      "grad_norm": 0.4909183382987976,
-      "learning_rate": 9.99661393088089e-06,
-      "loss": 0.4075,
-      "step": 1446
-    },
-    {
-      "epoch": 0.09456898241944971,
-      "grad_norm": 0.5209808945655823,
-      "learning_rate": 9.996601069789207e-06,
-      "loss": 0.4602,
-      "step": 1447
-    },
-    {
-      "epoch": 0.09463433762499183,
-      "grad_norm": 0.5387190580368042,
-      "learning_rate": 9.996588184327369e-06,
-      "loss": 0.4483,
-      "step": 1448
-    },
-    {
-      "epoch": 0.09469969283053395,
-      "grad_norm": 0.5025069117546082,
-      "learning_rate": 9.99657527449544e-06,
-      "loss": 0.403,
-      "step": 1449
-    },
-    {
-      "epoch": 0.09476504803607608,
-      "grad_norm": 0.5117092132568359,
-      "learning_rate": 9.996562340293482e-06,
-      "loss": 0.4055,
-      "step": 1450
-    },
-    {
-      "epoch": 0.09483040324161819,
-      "grad_norm": 0.45879194140434265,
-      "learning_rate": 9.996549381721558e-06,
-      "loss": 0.3907,
-      "step": 1451
-    },
-    {
-      "epoch": 0.09489575844716032,
-      "grad_norm": 0.5125434994697571,
-      "learning_rate": 9.996536398779732e-06,
-      "loss": 0.4107,
-      "step": 1452
-    },
-    {
-      "epoch": 0.09496111365270243,
-      "grad_norm": 0.5189753770828247,
-      "learning_rate": 9.996523391468068e-06,
-      "loss": 0.4015,
-      "step": 1453
-    },
-    {
-      "epoch": 0.09502646885824456,
-      "grad_norm": 0.5442360043525696,
-      "learning_rate": 9.996510359786628e-06,
-      "loss": 0.4504,
-      "step": 1454
-    },
-    {
-      "epoch": 0.09509182406378668,
-      "grad_norm": 0.500811755657196,
-      "learning_rate": 9.996497303735474e-06,
-      "loss": 0.4213,
-      "step": 1455
-    },
-    {
-      "epoch": 0.0951571792693288,
-      "grad_norm": 0.5382258296012878,
-      "learning_rate": 9.996484223314676e-06,
-      "loss": 0.4811,
-      "step": 1456
-    },
-    {
-      "epoch": 0.09522253447487092,
-      "grad_norm": 0.48647505044937134,
-      "learning_rate": 9.996471118524291e-06,
-      "loss": 0.377,
-      "step": 1457
-    },
-    {
-      "epoch": 0.09528788968041305,
-      "grad_norm": 0.5107442140579224,
-      "learning_rate": 9.996457989364385e-06,
-      "loss": 0.4103,
-      "step": 1458
-    },
-    {
-      "epoch": 0.09535324488595516,
-      "grad_norm": 0.4937625229358673,
-      "learning_rate": 9.996444835835023e-06,
-      "loss": 0.3966,
-      "step": 1459
-    },
-    {
-      "epoch": 0.09541860009149729,
-      "grad_norm": 0.5010054111480713,
-      "learning_rate": 9.996431657936267e-06,
-      "loss": 0.3988,
-      "step": 1460
-    },
-    {
-      "epoch": 0.0954839552970394,
-      "grad_norm": 0.5201773643493652,
-      "learning_rate": 9.996418455668185e-06,
-      "loss": 0.4288,
-      "step": 1461
-    },
-    {
-      "epoch": 0.09554931050258153,
-      "grad_norm": 0.5089337229728699,
-      "learning_rate": 9.99640522903084e-06,
-      "loss": 0.3942,
-      "step": 1462
-    },
-    {
-      "epoch": 0.09561466570812366,
-      "grad_norm": 0.5113603472709656,
-      "learning_rate": 9.996391978024294e-06,
-      "loss": 0.4371,
-      "step": 1463
-    },
-    {
-      "epoch": 0.09568002091366577,
-      "grad_norm": 0.4873717129230499,
-      "learning_rate": 9.996378702648612e-06,
-      "loss": 0.4065,
-      "step": 1464
-    },
-    {
-      "epoch": 0.0957453761192079,
-      "grad_norm": 0.4917776584625244,
-      "learning_rate": 9.996365402903863e-06,
-      "loss": 0.4379,
-      "step": 1465
-    },
-    {
-      "epoch": 0.09581073132475001,
-      "grad_norm": 0.5205647945404053,
-      "learning_rate": 9.996352078790109e-06,
-      "loss": 0.4396,
-      "step": 1466
-    },
-    {
-      "epoch": 0.09587608653029214,
-      "grad_norm": 0.496063232421875,
-      "learning_rate": 9.996338730307413e-06,
-      "loss": 0.3942,
-      "step": 1467
-    },
-    {
-      "epoch": 0.09594144173583426,
-      "grad_norm": 0.5094549059867859,
-      "learning_rate": 9.996325357455843e-06,
-      "loss": 0.3956,
-      "step": 1468
-    },
-    {
-      "epoch": 0.09600679694137638,
-      "grad_norm": 0.5358335375785828,
-      "learning_rate": 9.996311960235463e-06,
-      "loss": 0.4488,
-      "step": 1469
-    },
-    {
-      "epoch": 0.0960721521469185,
-      "grad_norm": 0.5215104222297668,
-      "learning_rate": 9.99629853864634e-06,
-      "loss": 0.4823,
-      "step": 1470
-    },
-    {
-      "epoch": 0.09613750735246063,
-      "grad_norm": 0.5060777068138123,
-      "learning_rate": 9.996285092688537e-06,
-      "loss": 0.4378,
-      "step": 1471
-    },
-    {
-      "epoch": 0.09620286255800274,
-      "grad_norm": 0.5364307165145874,
-      "learning_rate": 9.99627162236212e-06,
-      "loss": 0.4764,
-      "step": 1472
-    },
-    {
-      "epoch": 0.09626821776354487,
-      "grad_norm": 0.5011062622070312,
-      "learning_rate": 9.996258127667158e-06,
-      "loss": 0.4395,
-      "step": 1473
-    },
-    {
-      "epoch": 0.09633357296908698,
-      "grad_norm": 0.5110766887664795,
-      "learning_rate": 9.996244608603711e-06,
-      "loss": 0.3971,
-      "step": 1474
-    },
-    {
-      "epoch": 0.09639892817462911,
-      "grad_norm": 0.5151036977767944,
-      "learning_rate": 9.99623106517185e-06,
-      "loss": 0.4513,
-      "step": 1475
-    },
-    {
-      "epoch": 0.09646428338017123,
-      "grad_norm": 0.5112361907958984,
-      "learning_rate": 9.996217497371639e-06,
-      "loss": 0.446,
-      "step": 1476
-    },
-    {
-      "epoch": 0.09652963858571335,
-      "grad_norm": 0.4761430323123932,
-      "learning_rate": 9.996203905203146e-06,
-      "loss": 0.3894,
-      "step": 1477
-    },
-    {
-      "epoch": 0.09659499379125547,
-      "grad_norm": 0.5030077695846558,
-      "learning_rate": 9.996190288666433e-06,
-      "loss": 0.4398,
-      "step": 1478
-    },
-    {
-      "epoch": 0.0966603489967976,
-      "grad_norm": 0.507631778717041,
-      "learning_rate": 9.99617664776157e-06,
-      "loss": 0.4237,
-      "step": 1479
-    },
-    {
-      "epoch": 0.09672570420233971,
-      "grad_norm": 0.518007755279541,
-      "learning_rate": 9.996162982488624e-06,
-      "loss": 0.4235,
-      "step": 1480
-    },
-    {
-      "epoch": 0.09679105940788184,
-      "grad_norm": 0.5226354002952576,
-      "learning_rate": 9.996149292847658e-06,
-      "loss": 0.466,
-      "step": 1481
-    },
-    {
-      "epoch": 0.09685641461342395,
-      "grad_norm": 0.5401211977005005,
-      "learning_rate": 9.996135578838743e-06,
-      "loss": 0.4778,
-      "step": 1482
-    },
-    {
-      "epoch": 0.09692176981896608,
-      "grad_norm": 0.5578910112380981,
-      "learning_rate": 9.996121840461943e-06,
-      "loss": 0.4761,
-      "step": 1483
-    },
-    {
-      "epoch": 0.09698712502450821,
-      "grad_norm": 0.5571273565292358,
-      "learning_rate": 9.996108077717325e-06,
-      "loss": 0.4708,
-      "step": 1484
-    },
-    {
-      "epoch": 0.09705248023005032,
-      "grad_norm": 0.5445072650909424,
-      "learning_rate": 9.99609429060496e-06,
-      "loss": 0.4494,
-      "step": 1485
-    },
-    {
-      "epoch": 0.09711783543559245,
-      "grad_norm": 0.5183945298194885,
-      "learning_rate": 9.99608047912491e-06,
-      "loss": 0.4311,
-      "step": 1486
-    },
-    {
-      "epoch": 0.09718319064113456,
-      "grad_norm": 0.5346569418907166,
-      "learning_rate": 9.996066643277245e-06,
-      "loss": 0.3996,
-      "step": 1487
-    },
-    {
-      "epoch": 0.09724854584667669,
-      "grad_norm": 0.5302724838256836,
-      "learning_rate": 9.996052783062034e-06,
-      "loss": 0.4387,
-      "step": 1488
-    },
-    {
-      "epoch": 0.0973139010522188,
-      "grad_norm": 0.489535927772522,
-      "learning_rate": 9.99603889847934e-06,
-      "loss": 0.3866,
-      "step": 1489
-    },
-    {
-      "epoch": 0.09737925625776093,
-      "grad_norm": 0.5626073479652405,
-      "learning_rate": 9.996024989529235e-06,
-      "loss": 0.4795,
-      "step": 1490
-    },
-    {
-      "epoch": 0.09744461146330305,
-      "grad_norm": 0.5010583400726318,
-      "learning_rate": 9.996011056211785e-06,
-      "loss": 0.4482,
-      "step": 1491
-    },
-    {
-      "epoch": 0.09750996666884518,
-      "grad_norm": 0.49008432030677795,
-      "learning_rate": 9.995997098527058e-06,
-      "loss": 0.416,
-      "step": 1492
-    },
-    {
-      "epoch": 0.09757532187438729,
-      "grad_norm": 0.5158562660217285,
-      "learning_rate": 9.995983116475123e-06,
-      "loss": 0.4335,
-      "step": 1493
-    },
-    {
-      "epoch": 0.09764067707992942,
-      "grad_norm": 0.5218483209609985,
-      "learning_rate": 9.995969110056047e-06,
-      "loss": 0.4127,
-      "step": 1494
-    },
-    {
-      "epoch": 0.09770603228547153,
-      "grad_norm": 0.5236784815788269,
-      "learning_rate": 9.9959550792699e-06,
-      "loss": 0.4047,
-      "step": 1495
-    },
-    {
-      "epoch": 0.09777138749101366,
-      "grad_norm": 0.5455120801925659,
-      "learning_rate": 9.995941024116748e-06,
-      "loss": 0.451,
-      "step": 1496
-    },
-    {
-      "epoch": 0.09783674269655578,
-      "grad_norm": 0.5234445929527283,
-      "learning_rate": 9.995926944596662e-06,
-      "loss": 0.4518,
-      "step": 1497
-    },
-    {
-      "epoch": 0.0979020979020979,
-      "grad_norm": 0.5312226414680481,
-      "learning_rate": 9.995912840709709e-06,
-      "loss": 0.4351,
-      "step": 1498
-    },
-    {
-      "epoch": 0.09796745310764002,
-      "grad_norm": 0.5053860545158386,
-      "learning_rate": 9.995898712455958e-06,
-      "loss": 0.4531,
-      "step": 1499
-    },
-    {
-      "epoch": 0.09803280831318215,
-      "grad_norm": 0.5232927203178406,
-      "learning_rate": 9.995884559835478e-06,
-      "loss": 0.4266,
-      "step": 1500
-    },
-    {
-      "epoch": 0.09809816351872426,
-      "grad_norm": 0.4847826063632965,
-      "learning_rate": 9.995870382848338e-06,
-      "loss": 0.419,
-      "step": 1501
-    },
-    {
-      "epoch": 0.09816351872426639,
-      "grad_norm": 0.5197350978851318,
-      "learning_rate": 9.99585618149461e-06,
-      "loss": 0.4592,
-      "step": 1502
-    },
-    {
-      "epoch": 0.09822887392980852,
-      "grad_norm": 0.5199298858642578,
-      "learning_rate": 9.995841955774358e-06,
-      "loss": 0.4474,
-      "step": 1503
-    },
-    {
-      "epoch": 0.09829422913535063,
-      "grad_norm": 0.5503425598144531,
-      "learning_rate": 9.995827705687655e-06,
-      "loss": 0.5241,
-      "step": 1504
-    },
-    {
-      "epoch": 0.09835958434089276,
-      "grad_norm": 0.48477134108543396,
-      "learning_rate": 9.995813431234569e-06,
-      "loss": 0.3931,
-      "step": 1505
-    },
-    {
-      "epoch": 0.09842493954643487,
-      "grad_norm": 0.516323983669281,
-      "learning_rate": 9.995799132415172e-06,
-      "loss": 0.4527,
-      "step": 1506
-    },
-    {
-      "epoch": 0.098490294751977,
-      "grad_norm": 0.46379178762435913,
-      "learning_rate": 9.99578480922953e-06,
-      "loss": 0.3489,
-      "step": 1507
-    },
-    {
-      "epoch": 0.09855564995751911,
-      "grad_norm": 0.4947770833969116,
-      "learning_rate": 9.995770461677715e-06,
-      "loss": 0.4138,
-      "step": 1508
-    },
-    {
-      "epoch": 0.09862100516306124,
-      "grad_norm": 0.5108930468559265,
-      "learning_rate": 9.995756089759797e-06,
-      "loss": 0.4641,
-      "step": 1509
-    },
-    {
-      "epoch": 0.09868636036860336,
-      "grad_norm": 0.469390332698822,
-      "learning_rate": 9.995741693475846e-06,
-      "loss": 0.3713,
-      "step": 1510
-    },
-    {
-      "epoch": 0.09875171557414548,
-      "grad_norm": 0.5141362547874451,
-      "learning_rate": 9.995727272825931e-06,
-      "loss": 0.4445,
-      "step": 1511
-    },
-    {
-      "epoch": 0.0988170707796876,
-      "grad_norm": 1.0825378894805908,
-      "learning_rate": 9.995712827810125e-06,
-      "loss": 0.4494,
-      "step": 1512
-    },
-    {
-      "epoch": 0.09888242598522973,
-      "grad_norm": 0.4801233410835266,
-      "learning_rate": 9.995698358428496e-06,
-      "loss": 0.3884,
-      "step": 1513
-    },
-    {
-      "epoch": 0.09894778119077184,
-      "grad_norm": 0.5004074573516846,
-      "learning_rate": 9.995683864681116e-06,
-      "loss": 0.4609,
-      "step": 1514
-    },
-    {
-      "epoch": 0.09901313639631397,
-      "grad_norm": 0.5093396306037903,
-      "learning_rate": 9.995669346568055e-06,
-      "loss": 0.4095,
-      "step": 1515
-    },
-    {
-      "epoch": 0.09907849160185608,
-      "grad_norm": 0.5307527780532837,
-      "learning_rate": 9.995654804089384e-06,
-      "loss": 0.4027,
-      "step": 1516
-    },
-    {
-      "epoch": 0.09914384680739821,
-      "grad_norm": 0.5089740753173828,
-      "learning_rate": 9.995640237245173e-06,
-      "loss": 0.4296,
-      "step": 1517
-    },
-    {
-      "epoch": 0.09920920201294033,
-      "grad_norm": 0.48025357723236084,
-      "learning_rate": 9.995625646035495e-06,
-      "loss": 0.3968,
-      "step": 1518
-    },
-    {
-      "epoch": 0.09927455721848245,
-      "grad_norm": 0.4784804582595825,
-      "learning_rate": 9.99561103046042e-06,
-      "loss": 0.3747,
-      "step": 1519
-    },
-    {
-      "epoch": 0.09933991242402457,
-      "grad_norm": 0.5449445247650146,
-      "learning_rate": 9.995596390520018e-06,
-      "loss": 0.4331,
-      "step": 1520
-    },
-    {
-      "epoch": 0.0994052676295667,
-      "grad_norm": 0.488118439912796,
-      "learning_rate": 9.995581726214362e-06,
-      "loss": 0.4057,
-      "step": 1521
-    },
-    {
-      "epoch": 0.09947062283510881,
-      "grad_norm": 0.5430318117141724,
-      "learning_rate": 9.995567037543525e-06,
-      "loss": 0.4705,
-      "step": 1522
-    },
-    {
-      "epoch": 0.09953597804065094,
-      "grad_norm": 0.6241530179977417,
-      "learning_rate": 9.995552324507575e-06,
-      "loss": 0.5303,
-      "step": 1523
-    },
-    {
-      "epoch": 0.09960133324619307,
-      "grad_norm": 0.4596569240093231,
-      "learning_rate": 9.995537587106587e-06,
-      "loss": 0.3809,
-      "step": 1524
-    },
-    {
-      "epoch": 0.09966668845173518,
-      "grad_norm": 0.5403998494148254,
-      "learning_rate": 9.995522825340631e-06,
-      "loss": 0.4972,
-      "step": 1525
-    },
-    {
-      "epoch": 0.09973204365727731,
-      "grad_norm": 0.5301600098609924,
-      "learning_rate": 9.99550803920978e-06,
-      "loss": 0.3813,
-      "step": 1526
-    },
-    {
-      "epoch": 0.09979739886281942,
-      "grad_norm": 0.5359787344932556,
-      "learning_rate": 9.995493228714105e-06,
-      "loss": 0.4376,
-      "step": 1527
-    },
-    {
-      "epoch": 0.09986275406836155,
-      "grad_norm": 0.5042316317558289,
-      "learning_rate": 9.995478393853679e-06,
-      "loss": 0.433,
-      "step": 1528
-    },
-    {
-      "epoch": 0.09992810927390366,
-      "grad_norm": 0.5169788599014282,
-      "learning_rate": 9.995463534628575e-06,
-      "loss": 0.4544,
-      "step": 1529
-    },
-    {
-      "epoch": 0.09999346447944579,
-      "grad_norm": 0.497503399848938,
-      "learning_rate": 9.995448651038865e-06,
-      "loss": 0.4213,
-      "step": 1530
-    },
-    {
-      "epoch": 0.1000588196849879,
-      "grad_norm": 0.47543859481811523,
-      "learning_rate": 9.99543374308462e-06,
-      "loss": 0.3667,
-      "step": 1531
-    },
-    {
-      "epoch": 0.10012417489053003,
-      "grad_norm": 0.5248241424560547,
-      "learning_rate": 9.995418810765915e-06,
-      "loss": 0.3839,
-      "step": 1532
-    },
-    {
-      "epoch": 0.10018953009607215,
-      "grad_norm": 0.5335806012153625,
-      "learning_rate": 9.995403854082822e-06,
-      "loss": 0.4861,
-      "step": 1533
-    },
-    {
-      "epoch": 0.10025488530161428,
-      "grad_norm": 0.5063391923904419,
-      "learning_rate": 9.995388873035415e-06,
-      "loss": 0.472,
-      "step": 1534
-    },
-    {
-      "epoch": 0.10032024050715639,
-      "grad_norm": 0.46077948808670044,
-      "learning_rate": 9.995373867623763e-06,
-      "loss": 0.3847,
-      "step": 1535
-    },
-    {
-      "epoch": 0.10038559571269852,
-      "grad_norm": 0.48244237899780273,
-      "learning_rate": 9.995358837847943e-06,
-      "loss": 0.4041,
-      "step": 1536
-    },
-    {
-      "epoch": 0.10045095091824063,
-      "grad_norm": 0.5086256861686707,
-      "learning_rate": 9.995343783708029e-06,
-      "loss": 0.4282,
-      "step": 1537
-    },
-    {
-      "epoch": 0.10051630612378276,
-      "grad_norm": 0.5021991729736328,
-      "learning_rate": 9.995328705204091e-06,
-      "loss": 0.3936,
-      "step": 1538
-    },
-    {
-      "epoch": 0.10058166132932488,
-      "grad_norm": 0.5289314389228821,
-      "learning_rate": 9.995313602336206e-06,
-      "loss": 0.4739,
-      "step": 1539
-    },
-    {
-      "epoch": 0.100647016534867,
-      "grad_norm": 0.5517817139625549,
-      "learning_rate": 9.995298475104443e-06,
-      "loss": 0.4827,
-      "step": 1540
-    },
-    {
-      "epoch": 0.10071237174040912,
-      "grad_norm": 0.5402710437774658,
-      "learning_rate": 9.995283323508882e-06,
-      "loss": 0.444,
-      "step": 1541
-    },
-    {
-      "epoch": 0.10077772694595125,
-      "grad_norm": 0.4955950081348419,
-      "learning_rate": 9.995268147549594e-06,
-      "loss": 0.4211,
-      "step": 1542
-    },
-    {
-      "epoch": 0.10084308215149336,
-      "grad_norm": 0.46556901931762695,
-      "learning_rate": 9.99525294722665e-06,
-      "loss": 0.3613,
-      "step": 1543
-    },
-    {
-      "epoch": 0.10090843735703549,
-      "grad_norm": 0.5250805020332336,
-      "learning_rate": 9.995237722540127e-06,
-      "loss": 0.4193,
-      "step": 1544
-    },
-    {
-      "epoch": 0.10097379256257762,
-      "grad_norm": 0.574885904788971,
-      "learning_rate": 9.9952224734901e-06,
-      "loss": 0.4805,
-      "step": 1545
-    },
-    {
-      "epoch": 0.10103914776811973,
-      "grad_norm": 0.5209019780158997,
-      "learning_rate": 9.995207200076642e-06,
-      "loss": 0.4788,
-      "step": 1546
-    },
-    {
-      "epoch": 0.10110450297366186,
-      "grad_norm": 0.5465680956840515,
-      "learning_rate": 9.99519190229983e-06,
-      "loss": 0.5027,
-      "step": 1547
-    },
-    {
-      "epoch": 0.10116985817920397,
-      "grad_norm": 0.5158794522285461,
-      "learning_rate": 9.995176580159735e-06,
-      "loss": 0.4429,
-      "step": 1548
-    },
-    {
-      "epoch": 0.1012352133847461,
-      "grad_norm": 0.5002278089523315,
-      "learning_rate": 9.995161233656434e-06,
-      "loss": 0.417,
-      "step": 1549
-    },
-    {
-      "epoch": 0.10130056859028821,
-      "grad_norm": 0.4885588586330414,
-      "learning_rate": 9.995145862790001e-06,
-      "loss": 0.4068,
-      "step": 1550
-    },
-    {
-      "epoch": 0.10136592379583034,
-      "grad_norm": 0.5092244148254395,
-      "learning_rate": 9.99513046756051e-06,
-      "loss": 0.429,
-      "step": 1551
-    },
-    {
-      "epoch": 0.10143127900137246,
-      "grad_norm": 0.49464020133018494,
-      "learning_rate": 9.995115047968038e-06,
-      "loss": 0.4077,
-      "step": 1552
-    },
-    {
-      "epoch": 0.10149663420691458,
-      "grad_norm": 0.488465279340744,
-      "learning_rate": 9.99509960401266e-06,
-      "loss": 0.389,
-      "step": 1553
-    },
-    {
-      "epoch": 0.1015619894124567,
-      "grad_norm": 0.514087975025177,
-      "learning_rate": 9.995084135694451e-06,
-      "loss": 0.4167,
-      "step": 1554
-    },
-    {
-      "epoch": 0.10162734461799883,
-      "grad_norm": 0.47458693385124207,
-      "learning_rate": 9.995068643013486e-06,
-      "loss": 0.4161,
-      "step": 1555
-    },
-    {
-      "epoch": 0.10169269982354094,
-      "grad_norm": 0.4821431636810303,
-      "learning_rate": 9.995053125969839e-06,
-      "loss": 0.3993,
-      "step": 1556
-    },
-    {
-      "epoch": 0.10175805502908307,
-      "grad_norm": 0.5431994795799255,
-      "learning_rate": 9.99503758456359e-06,
-      "loss": 0.4565,
-      "step": 1557
-    },
-    {
-      "epoch": 0.10182341023462518,
-      "grad_norm": 0.5592271685600281,
-      "learning_rate": 9.99502201879481e-06,
-      "loss": 0.5253,
-      "step": 1558
-    },
-    {
-      "epoch": 0.10188876544016731,
-      "grad_norm": 0.48341649770736694,
-      "learning_rate": 9.99500642866358e-06,
-      "loss": 0.3993,
-      "step": 1559
-    },
-    {
-      "epoch": 0.10195412064570943,
-      "grad_norm": 0.5405024886131287,
-      "learning_rate": 9.994990814169969e-06,
-      "loss": 0.4567,
-      "step": 1560
-    },
-    {
-      "epoch": 0.10201947585125155,
-      "grad_norm": 0.49808263778686523,
-      "learning_rate": 9.994975175314061e-06,
-      "loss": 0.4328,
-      "step": 1561
-    },
-    {
-      "epoch": 0.10208483105679367,
-      "grad_norm": 0.5058544278144836,
-      "learning_rate": 9.994959512095928e-06,
-      "loss": 0.4123,
-      "step": 1562
-    },
-    {
-      "epoch": 0.1021501862623358,
-      "grad_norm": 0.5445604920387268,
-      "learning_rate": 9.994943824515644e-06,
-      "loss": 0.4722,
-      "step": 1563
-    },
-    {
-      "epoch": 0.10221554146787791,
-      "grad_norm": 0.5007298588752747,
-      "learning_rate": 9.994928112573292e-06,
-      "loss": 0.4426,
-      "step": 1564
-    },
-    {
-      "epoch": 0.10228089667342004,
-      "grad_norm": 0.5044405460357666,
-      "learning_rate": 9.994912376268943e-06,
-      "loss": 0.442,
-      "step": 1565
-    },
-    {
-      "epoch": 0.10234625187896217,
-      "grad_norm": 0.4776204526424408,
-      "learning_rate": 9.994896615602679e-06,
-      "loss": 0.3454,
-      "step": 1566
-    },
-    {
-      "epoch": 0.10241160708450428,
-      "grad_norm": 0.4716126620769501,
-      "learning_rate": 9.994880830574572e-06,
-      "loss": 0.3728,
-      "step": 1567
-    },
-    {
-      "epoch": 0.10247696229004641,
-      "grad_norm": 0.5031254291534424,
-      "learning_rate": 9.994865021184701e-06,
-      "loss": 0.4132,
-      "step": 1568
-    },
-    {
-      "epoch": 0.10254231749558852,
-      "grad_norm": 0.4964865446090698,
-      "learning_rate": 9.994849187433142e-06,
-      "loss": 0.4052,
-      "step": 1569
-    },
-    {
-      "epoch": 0.10260767270113065,
-      "grad_norm": 0.5192766785621643,
-      "learning_rate": 9.994833329319972e-06,
-      "loss": 0.4655,
-      "step": 1570
-    },
-    {
-      "epoch": 0.10267302790667276,
-      "grad_norm": 0.5177010297775269,
-      "learning_rate": 9.994817446845273e-06,
-      "loss": 0.4839,
-      "step": 1571
-    },
-    {
-      "epoch": 0.10273838311221489,
-      "grad_norm": 0.5169263482093811,
-      "learning_rate": 9.994801540009117e-06,
-      "loss": 0.4489,
-      "step": 1572
-    },
-    {
-      "epoch": 0.102803738317757,
-      "grad_norm": 0.5630887746810913,
-      "learning_rate": 9.994785608811584e-06,
-      "loss": 0.4681,
-      "step": 1573
-    },
-    {
-      "epoch": 0.10286909352329913,
-      "grad_norm": 0.4649583399295807,
-      "learning_rate": 9.99476965325275e-06,
-      "loss": 0.3973,
-      "step": 1574
-    },
-    {
-      "epoch": 0.10293444872884125,
-      "grad_norm": 0.5167850852012634,
-      "learning_rate": 9.994753673332696e-06,
-      "loss": 0.4593,
-      "step": 1575
-    },
-    {
-      "epoch": 0.10299980393438338,
-      "grad_norm": 0.45438265800476074,
-      "learning_rate": 9.994737669051497e-06,
-      "loss": 0.3794,
-      "step": 1576
-    },
-    {
-      "epoch": 0.10306515913992549,
-      "grad_norm": 0.534008800983429,
-      "learning_rate": 9.994721640409231e-06,
-      "loss": 0.4541,
-      "step": 1577
-    },
-    {
-      "epoch": 0.10313051434546762,
-      "grad_norm": 0.5064383745193481,
-      "learning_rate": 9.99470558740598e-06,
-      "loss": 0.4248,
-      "step": 1578
-    },
-    {
-      "epoch": 0.10319586955100973,
-      "grad_norm": 0.46664220094680786,
-      "learning_rate": 9.994689510041817e-06,
-      "loss": 0.3905,
-      "step": 1579
-    },
-    {
-      "epoch": 0.10326122475655186,
-      "grad_norm": 0.5122666358947754,
-      "learning_rate": 9.994673408316824e-06,
-      "loss": 0.4329,
-      "step": 1580
-    },
-    {
-      "epoch": 0.10332657996209398,
-      "grad_norm": 0.5435193777084351,
-      "learning_rate": 9.994657282231077e-06,
-      "loss": 0.4561,
-      "step": 1581
-    },
-    {
-      "epoch": 0.1033919351676361,
-      "grad_norm": 0.4839650094509125,
-      "learning_rate": 9.994641131784657e-06,
-      "loss": 0.4042,
-      "step": 1582
-    },
-    {
-      "epoch": 0.10345729037317822,
-      "grad_norm": 0.5004949569702148,
-      "learning_rate": 9.994624956977644e-06,
-      "loss": 0.3976,
-      "step": 1583
-    },
-    {
-      "epoch": 0.10352264557872035,
-      "grad_norm": 0.5046452283859253,
-      "learning_rate": 9.994608757810111e-06,
-      "loss": 0.4224,
-      "step": 1584
-    },
-    {
-      "epoch": 0.10358800078426246,
-      "grad_norm": 0.45456764101982117,
-      "learning_rate": 9.994592534282144e-06,
-      "loss": 0.3623,
-      "step": 1585
-    },
-    {
-      "epoch": 0.10365335598980459,
-      "grad_norm": 0.5273374319076538,
-      "learning_rate": 9.994576286393817e-06,
-      "loss": 0.4477,
-      "step": 1586
-    },
-    {
-      "epoch": 0.10371871119534672,
-      "grad_norm": 0.5582617521286011,
-      "learning_rate": 9.994560014145211e-06,
-      "loss": 0.5149,
-      "step": 1587
-    },
-    {
-      "epoch": 0.10378406640088883,
-      "grad_norm": 0.4669838547706604,
-      "learning_rate": 9.994543717536407e-06,
-      "loss": 0.3804,
-      "step": 1588
-    },
-    {
-      "epoch": 0.10384942160643096,
-      "grad_norm": 0.5422541499137878,
-      "learning_rate": 9.994527396567483e-06,
-      "loss": 0.4509,
-      "step": 1589
-    },
-    {
-      "epoch": 0.10391477681197307,
-      "grad_norm": 0.5106235146522522,
-      "learning_rate": 9.994511051238518e-06,
-      "loss": 0.4067,
-      "step": 1590
-    },
-    {
-      "epoch": 0.1039801320175152,
-      "grad_norm": 0.5292513966560364,
-      "learning_rate": 9.994494681549592e-06,
-      "loss": 0.4103,
-      "step": 1591
-    },
-    {
-      "epoch": 0.10404548722305731,
-      "grad_norm": 0.5347157716751099,
-      "learning_rate": 9.994478287500786e-06,
-      "loss": 0.44,
-      "step": 1592
-    },
-    {
-      "epoch": 0.10411084242859944,
-      "grad_norm": 0.5292321443557739,
-      "learning_rate": 9.994461869092178e-06,
-      "loss": 0.4746,
-      "step": 1593
-    },
-    {
-      "epoch": 0.10417619763414156,
-      "grad_norm": 0.5306785106658936,
-      "learning_rate": 9.99444542632385e-06,
-      "loss": 0.4639,
-      "step": 1594
-    },
-    {
-      "epoch": 0.10424155283968368,
-      "grad_norm": 0.5565119385719299,
-      "learning_rate": 9.994428959195882e-06,
-      "loss": 0.4482,
-      "step": 1595
-    },
-    {
-      "epoch": 0.1043069080452258,
-      "grad_norm": 0.4725090265274048,
-      "learning_rate": 9.994412467708352e-06,
-      "loss": 0.3954,
-      "step": 1596
-    },
-    {
-      "epoch": 0.10437226325076793,
-      "grad_norm": 0.5077757835388184,
-      "learning_rate": 9.994395951861343e-06,
-      "loss": 0.443,
-      "step": 1597
-    },
-    {
-      "epoch": 0.10443761845631004,
-      "grad_norm": 0.6092401742935181,
-      "learning_rate": 9.994379411654936e-06,
-      "loss": 0.4611,
-      "step": 1598
-    },
-    {
-      "epoch": 0.10450297366185217,
-      "grad_norm": 0.5067799091339111,
-      "learning_rate": 9.994362847089207e-06,
-      "loss": 0.4469,
-      "step": 1599
-    },
-    {
-      "epoch": 0.10456832886739428,
-      "grad_norm": 0.5002574324607849,
-      "learning_rate": 9.994346258164244e-06,
-      "loss": 0.4265,
-      "step": 1600
-    },
-    {
-      "epoch": 0.10463368407293641,
-      "grad_norm": 0.5395341515541077,
-      "learning_rate": 9.994329644880121e-06,
-      "loss": 0.4755,
-      "step": 1601
-    },
-    {
-      "epoch": 0.10469903927847853,
-      "grad_norm": 0.47599154710769653,
-      "learning_rate": 9.994313007236925e-06,
-      "loss": 0.376,
-      "step": 1602
-    },
-    {
-      "epoch": 0.10476439448402065,
-      "grad_norm": 0.5089483857154846,
-      "learning_rate": 9.994296345234733e-06,
-      "loss": 0.4201,
-      "step": 1603
-    },
-    {
-      "epoch": 0.10482974968956277,
-      "grad_norm": 0.5276210904121399,
-      "learning_rate": 9.994279658873627e-06,
-      "loss": 0.4472,
-      "step": 1604
-    },
-    {
-      "epoch": 0.1048951048951049,
-      "grad_norm": 0.4988803565502167,
-      "learning_rate": 9.99426294815369e-06,
-      "loss": 0.4022,
-      "step": 1605
-    },
-    {
-      "epoch": 0.10496046010064701,
-      "grad_norm": 0.5025919079780579,
-      "learning_rate": 9.994246213075001e-06,
-      "loss": 0.4693,
-      "step": 1606
-    },
-    {
-      "epoch": 0.10502581530618914,
-      "grad_norm": 0.5430751442909241,
-      "learning_rate": 9.994229453637644e-06,
-      "loss": 0.4879,
-      "step": 1607
-    },
-    {
-      "epoch": 0.10509117051173127,
-      "grad_norm": 0.4761691689491272,
-      "learning_rate": 9.9942126698417e-06,
-      "loss": 0.3696,
-      "step": 1608
-    },
-    {
-      "epoch": 0.10515652571727338,
-      "grad_norm": 0.4970346689224243,
-      "learning_rate": 9.99419586168725e-06,
-      "loss": 0.4098,
-      "step": 1609
-    },
-    {
-      "epoch": 0.10522188092281551,
-      "grad_norm": 0.5387327075004578,
-      "learning_rate": 9.994179029174377e-06,
-      "loss": 0.3672,
-      "step": 1610
-    },
-    {
-      "epoch": 0.10528723612835762,
-      "grad_norm": 0.4932538568973541,
-      "learning_rate": 9.994162172303162e-06,
-      "loss": 0.3763,
-      "step": 1611
-    },
-    {
-      "epoch": 0.10535259133389975,
-      "grad_norm": 0.5018151998519897,
-      "learning_rate": 9.994145291073688e-06,
-      "loss": 0.477,
-      "step": 1612
-    },
-    {
-      "epoch": 0.10541794653944186,
-      "grad_norm": 0.5169281363487244,
-      "learning_rate": 9.994128385486039e-06,
-      "loss": 0.4181,
-      "step": 1613
-    },
-    {
-      "epoch": 0.10548330174498399,
-      "grad_norm": 0.5027089715003967,
-      "learning_rate": 9.994111455540294e-06,
-      "loss": 0.4196,
-      "step": 1614
-    },
-    {
-      "epoch": 0.1055486569505261,
-      "grad_norm": 0.5028634071350098,
-      "learning_rate": 9.994094501236537e-06,
-      "loss": 0.4098,
-      "step": 1615
-    },
-    {
-      "epoch": 0.10561401215606823,
-      "grad_norm": 0.5034271478652954,
-      "learning_rate": 9.994077522574853e-06,
-      "loss": 0.4586,
-      "step": 1616
-    },
-    {
-      "epoch": 0.10567936736161035,
-      "grad_norm": 0.5230409502983093,
-      "learning_rate": 9.994060519555323e-06,
-      "loss": 0.4776,
-      "step": 1617
-    },
-    {
-      "epoch": 0.10574472256715248,
-      "grad_norm": 0.5798277854919434,
-      "learning_rate": 9.994043492178027e-06,
-      "loss": 0.493,
-      "step": 1618
-    },
-    {
-      "epoch": 0.10581007777269459,
-      "grad_norm": 0.4700954556465149,
-      "learning_rate": 9.994026440443052e-06,
-      "loss": 0.3865,
-      "step": 1619
-    },
-    {
-      "epoch": 0.10587543297823672,
-      "grad_norm": 0.5281357765197754,
-      "learning_rate": 9.994009364350481e-06,
-      "loss": 0.4346,
-      "step": 1620
-    },
-    {
-      "epoch": 0.10594078818377883,
-      "grad_norm": 0.5067729353904724,
-      "learning_rate": 9.993992263900395e-06,
-      "loss": 0.4011,
-      "step": 1621
-    },
-    {
-      "epoch": 0.10600614338932096,
-      "grad_norm": 0.4662727117538452,
-      "learning_rate": 9.993975139092879e-06,
-      "loss": 0.3957,
-      "step": 1622
-    },
-    {
-      "epoch": 0.10607149859486308,
-      "grad_norm": 0.48641085624694824,
-      "learning_rate": 9.993957989928016e-06,
-      "loss": 0.4027,
-      "step": 1623
-    },
-    {
-      "epoch": 0.1061368538004052,
-      "grad_norm": 0.5185033679008484,
-      "learning_rate": 9.99394081640589e-06,
-      "loss": 0.4105,
-      "step": 1624
-    },
-    {
-      "epoch": 0.10620220900594732,
-      "grad_norm": 0.5151177048683167,
-      "learning_rate": 9.993923618526584e-06,
-      "loss": 0.4104,
-      "step": 1625
-    },
-    {
-      "epoch": 0.10626756421148945,
-      "grad_norm": 0.5204619765281677,
-      "learning_rate": 9.993906396290184e-06,
-      "loss": 0.43,
-      "step": 1626
-    },
-    {
-      "epoch": 0.10633291941703156,
-      "grad_norm": 0.5157848596572876,
-      "learning_rate": 9.993889149696772e-06,
-      "loss": 0.4142,
-      "step": 1627
-    },
-    {
-      "epoch": 0.10639827462257369,
-      "grad_norm": 0.5414742231369019,
-      "learning_rate": 9.993871878746432e-06,
-      "loss": 0.4357,
-      "step": 1628
-    },
-    {
-      "epoch": 0.10646362982811582,
-      "grad_norm": 0.5015614628791809,
-      "learning_rate": 9.993854583439249e-06,
-      "loss": 0.4337,
-      "step": 1629
-    },
-    {
-      "epoch": 0.10652898503365793,
-      "grad_norm": 0.5032528638839722,
-      "learning_rate": 9.993837263775306e-06,
-      "loss": 0.4211,
-      "step": 1630
-    },
-    {
-      "epoch": 0.10659434023920006,
-      "grad_norm": 0.5219725370407104,
-      "learning_rate": 9.993819919754691e-06,
-      "loss": 0.4959,
-      "step": 1631
-    },
-    {
-      "epoch": 0.10665969544474217,
-      "grad_norm": 0.4600250720977783,
-      "learning_rate": 9.993802551377484e-06,
-      "loss": 0.3833,
-      "step": 1632
-    },
-    {
-      "epoch": 0.1067250506502843,
-      "grad_norm": 0.4900052547454834,
-      "learning_rate": 9.993785158643774e-06,
-      "loss": 0.4029,
-      "step": 1633
-    },
-    {
-      "epoch": 0.10679040585582641,
-      "grad_norm": 0.48988422751426697,
-      "learning_rate": 9.99376774155364e-06,
-      "loss": 0.4248,
-      "step": 1634
-    },
-    {
-      "epoch": 0.10685576106136854,
-      "grad_norm": 0.4932857155799866,
-      "learning_rate": 9.993750300107174e-06,
-      "loss": 0.396,
-      "step": 1635
-    },
-    {
-      "epoch": 0.10692111626691066,
-      "grad_norm": 0.5180819630622864,
-      "learning_rate": 9.993732834304458e-06,
-      "loss": 0.4724,
-      "step": 1636
-    },
-    {
-      "epoch": 0.10698647147245278,
-      "grad_norm": 0.5097213983535767,
-      "learning_rate": 9.993715344145576e-06,
-      "loss": 0.4581,
-      "step": 1637
-    },
-    {
-      "epoch": 0.1070518266779949,
-      "grad_norm": 0.48215875029563904,
-      "learning_rate": 9.993697829630614e-06,
-      "loss": 0.4307,
-      "step": 1638
-    },
-    {
-      "epoch": 0.10711718188353703,
-      "grad_norm": 0.5152288675308228,
-      "learning_rate": 9.993680290759657e-06,
-      "loss": 0.4411,
-      "step": 1639
-    },
-    {
-      "epoch": 0.10718253708907914,
-      "grad_norm": 0.5181601643562317,
-      "learning_rate": 9.993662727532794e-06,
-      "loss": 0.4685,
-      "step": 1640
-    },
-    {
-      "epoch": 0.10724789229462127,
-      "grad_norm": 0.5137525200843811,
-      "learning_rate": 9.993645139950106e-06,
-      "loss": 0.4556,
-      "step": 1641
-    },
-    {
-      "epoch": 0.10731324750016338,
-      "grad_norm": 0.5311994552612305,
-      "learning_rate": 9.99362752801168e-06,
-      "loss": 0.4174,
-      "step": 1642
-    },
-    {
-      "epoch": 0.10737860270570551,
-      "grad_norm": 0.5230871438980103,
-      "learning_rate": 9.993609891717605e-06,
-      "loss": 0.4633,
-      "step": 1643
-    },
-    {
-      "epoch": 0.10744395791124763,
-      "grad_norm": 0.48344889283180237,
-      "learning_rate": 9.993592231067962e-06,
-      "loss": 0.3779,
-      "step": 1644
-    },
-    {
-      "epoch": 0.10750931311678975,
-      "grad_norm": 0.5213485360145569,
-      "learning_rate": 9.993574546062841e-06,
-      "loss": 0.4961,
-      "step": 1645
-    },
-    {
-      "epoch": 0.10757466832233187,
-      "grad_norm": 0.5260332822799683,
-      "learning_rate": 9.993556836702327e-06,
-      "loss": 0.4409,
-      "step": 1646
-    },
-    {
-      "epoch": 0.107640023527874,
-      "grad_norm": 0.5283243060112,
-      "learning_rate": 9.993539102986506e-06,
-      "loss": 0.4591,
-      "step": 1647
-    },
-    {
-      "epoch": 0.10770537873341611,
-      "grad_norm": 0.49721431732177734,
-      "learning_rate": 9.993521344915464e-06,
-      "loss": 0.4099,
-      "step": 1648
-    },
-    {
-      "epoch": 0.10777073393895824,
-      "grad_norm": 0.5064494013786316,
-      "learning_rate": 9.993503562489291e-06,
-      "loss": 0.3882,
-      "step": 1649
-    },
-    {
-      "epoch": 0.10783608914450037,
-      "grad_norm": 0.5435827970504761,
-      "learning_rate": 9.99348575570807e-06,
-      "loss": 0.4477,
-      "step": 1650
-    },
-    {
-      "epoch": 0.10790144435004248,
-      "grad_norm": 0.5121777057647705,
-      "learning_rate": 9.993467924571888e-06,
-      "loss": 0.4361,
-      "step": 1651
-    },
-    {
-      "epoch": 0.10796679955558461,
-      "grad_norm": 0.5017474889755249,
-      "learning_rate": 9.993450069080834e-06,
-      "loss": 0.4572,
-      "step": 1652
-    },
-    {
-      "epoch": 0.10803215476112672,
-      "grad_norm": 0.4521902799606323,
-      "learning_rate": 9.993432189234995e-06,
-      "loss": 0.3896,
-      "step": 1653
-    },
-    {
-      "epoch": 0.10809750996666885,
-      "grad_norm": 0.5386385321617126,
-      "learning_rate": 9.993414285034456e-06,
-      "loss": 0.4219,
-      "step": 1654
-    },
-    {
-      "epoch": 0.10816286517221096,
-      "grad_norm": 0.5199062824249268,
-      "learning_rate": 9.993396356479305e-06,
-      "loss": 0.4084,
-      "step": 1655
-    },
-    {
-      "epoch": 0.10822822037775309,
-      "grad_norm": 0.5277518033981323,
-      "learning_rate": 9.993378403569632e-06,
-      "loss": 0.4688,
-      "step": 1656
-    },
-    {
-      "epoch": 0.1082935755832952,
-      "grad_norm": 0.5194827914237976,
-      "learning_rate": 9.99336042630552e-06,
-      "loss": 0.4943,
-      "step": 1657
-    },
-    {
-      "epoch": 0.10835893078883733,
-      "grad_norm": 0.5181694626808167,
-      "learning_rate": 9.993342424687063e-06,
-      "loss": 0.4819,
-      "step": 1658
-    },
-    {
-      "epoch": 0.10842428599437945,
-      "grad_norm": 0.5547198057174683,
-      "learning_rate": 9.993324398714341e-06,
-      "loss": 0.4555,
-      "step": 1659
-    },
-    {
-      "epoch": 0.10848964119992158,
-      "grad_norm": 0.5099554657936096,
-      "learning_rate": 9.99330634838745e-06,
-      "loss": 0.4147,
-      "step": 1660
-    },
-    {
-      "epoch": 0.10855499640546369,
-      "grad_norm": 0.4664912223815918,
-      "learning_rate": 9.993288273706471e-06,
-      "loss": 0.3636,
-      "step": 1661
-    },
-    {
-      "epoch": 0.10862035161100582,
-      "grad_norm": 0.5194811224937439,
-      "learning_rate": 9.993270174671496e-06,
-      "loss": 0.4497,
-      "step": 1662
-    },
-    {
-      "epoch": 0.10868570681654793,
-      "grad_norm": 0.47649380564689636,
-      "learning_rate": 9.993252051282614e-06,
-      "loss": 0.395,
-      "step": 1663
-    },
-    {
-      "epoch": 0.10875106202209006,
-      "grad_norm": 0.5036858320236206,
-      "learning_rate": 9.993233903539913e-06,
-      "loss": 0.4341,
-      "step": 1664
-    },
-    {
-      "epoch": 0.10881641722763218,
-      "grad_norm": 0.5305114984512329,
-      "learning_rate": 9.993215731443476e-06,
-      "loss": 0.5104,
-      "step": 1665
-    },
-    {
-      "epoch": 0.1088817724331743,
-      "grad_norm": 0.4840755760669708,
-      "learning_rate": 9.9931975349934e-06,
-      "loss": 0.4561,
-      "step": 1666
-    },
-    {
-      "epoch": 0.10894712763871642,
-      "grad_norm": 0.5001997947692871,
-      "learning_rate": 9.993179314189767e-06,
-      "loss": 0.4431,
-      "step": 1667
-    },
-    {
-      "epoch": 0.10901248284425855,
-      "grad_norm": 0.5084608793258667,
-      "learning_rate": 9.993161069032671e-06,
-      "loss": 0.3941,
-      "step": 1668
-    },
-    {
-      "epoch": 0.10907783804980066,
-      "grad_norm": 0.5229108333587646,
-      "learning_rate": 9.993142799522198e-06,
-      "loss": 0.4406,
-      "step": 1669
-    },
-    {
-      "epoch": 0.10914319325534279,
-      "grad_norm": 0.5073447227478027,
-      "learning_rate": 9.993124505658435e-06,
-      "loss": 0.4173,
-      "step": 1670
-    },
-    {
-      "epoch": 0.10920854846088492,
-      "grad_norm": 0.5383942723274231,
-      "learning_rate": 9.993106187441477e-06,
-      "loss": 0.4033,
-      "step": 1671
-    },
-    {
-      "epoch": 0.10927390366642703,
-      "grad_norm": 0.5326029062271118,
-      "learning_rate": 9.99308784487141e-06,
-      "loss": 0.4214,
-      "step": 1672
-    },
-    {
-      "epoch": 0.10933925887196916,
-      "grad_norm": 0.5241849422454834,
-      "learning_rate": 9.993069477948325e-06,
-      "loss": 0.4317,
-      "step": 1673
-    },
-    {
-      "epoch": 0.10940461407751127,
-      "grad_norm": 0.5135782361030579,
-      "learning_rate": 9.993051086672308e-06,
-      "loss": 0.4172,
-      "step": 1674
-    },
-    {
-      "epoch": 0.1094699692830534,
-      "grad_norm": 0.4599984288215637,
-      "learning_rate": 9.993032671043451e-06,
-      "loss": 0.3514,
-      "step": 1675
-    },
-    {
-      "epoch": 0.10953532448859551,
-      "grad_norm": 0.4994356036186218,
-      "learning_rate": 9.993014231061846e-06,
-      "loss": 0.4263,
-      "step": 1676
-    },
-    {
-      "epoch": 0.10960067969413764,
-      "grad_norm": 0.5081284046173096,
-      "learning_rate": 9.99299576672758e-06,
-      "loss": 0.4016,
-      "step": 1677
-    },
-    {
-      "epoch": 0.10966603489967976,
-      "grad_norm": 0.5670870542526245,
-      "learning_rate": 9.992977278040743e-06,
-      "loss": 0.4764,
-      "step": 1678
-    },
-    {
-      "epoch": 0.10973139010522188,
-      "grad_norm": 0.5047267079353333,
-      "learning_rate": 9.992958765001427e-06,
-      "loss": 0.4452,
-      "step": 1679
-    },
-    {
-      "epoch": 0.109796745310764,
-      "grad_norm": 0.5062739849090576,
-      "learning_rate": 9.992940227609721e-06,
-      "loss": 0.4526,
-      "step": 1680
-    },
-    {
-      "epoch": 0.10986210051630613,
-      "grad_norm": 0.5351704955101013,
-      "learning_rate": 9.992921665865717e-06,
-      "loss": 0.5013,
-      "step": 1681
-    },
-    {
-      "epoch": 0.10992745572184824,
-      "grad_norm": 0.4982771575450897,
-      "learning_rate": 9.992903079769503e-06,
-      "loss": 0.4476,
-      "step": 1682
-    },
-    {
-      "epoch": 0.10999281092739037,
-      "grad_norm": 0.4960046410560608,
-      "learning_rate": 9.992884469321171e-06,
-      "loss": 0.4369,
-      "step": 1683
-    },
-    {
-      "epoch": 0.11005816613293248,
-      "grad_norm": 0.4791679084300995,
-      "learning_rate": 9.99286583452081e-06,
-      "loss": 0.391,
-      "step": 1684
-    },
-    {
-      "epoch": 0.11012352133847461,
-      "grad_norm": 0.5098387002944946,
-      "learning_rate": 9.992847175368516e-06,
-      "loss": 0.4256,
-      "step": 1685
-    },
-    {
-      "epoch": 0.11018887654401673,
-      "grad_norm": 0.5323256254196167,
-      "learning_rate": 9.992828491864375e-06,
-      "loss": 0.4628,
-      "step": 1686
-    },
-    {
-      "epoch": 0.11025423174955885,
-      "grad_norm": 0.48827052116394043,
-      "learning_rate": 9.99280978400848e-06,
-      "loss": 0.4276,
-      "step": 1687
-    },
-    {
-      "epoch": 0.11031958695510097,
-      "grad_norm": 0.5217145085334778,
-      "learning_rate": 9.992791051800922e-06,
-      "loss": 0.457,
-      "step": 1688
-    },
-    {
-      "epoch": 0.1103849421606431,
-      "grad_norm": 0.48917123675346375,
-      "learning_rate": 9.992772295241793e-06,
-      "loss": 0.3973,
-      "step": 1689
-    },
-    {
-      "epoch": 0.11045029736618521,
-      "grad_norm": 0.5238152742385864,
-      "learning_rate": 9.992753514331184e-06,
-      "loss": 0.475,
-      "step": 1690
-    },
-    {
-      "epoch": 0.11051565257172734,
-      "grad_norm": 0.5162878632545471,
-      "learning_rate": 9.992734709069184e-06,
-      "loss": 0.4535,
-      "step": 1691
-    },
-    {
-      "epoch": 0.11058100777726947,
-      "grad_norm": 0.5076547861099243,
-      "learning_rate": 9.99271587945589e-06,
-      "loss": 0.4339,
-      "step": 1692
-    },
-    {
-      "epoch": 0.11064636298281158,
-      "grad_norm": 0.53127521276474,
-      "learning_rate": 9.992697025491389e-06,
-      "loss": 0.4814,
-      "step": 1693
-    },
-    {
-      "epoch": 0.11071171818835371,
-      "grad_norm": 0.4735592305660248,
-      "learning_rate": 9.992678147175776e-06,
-      "loss": 0.3961,
-      "step": 1694
-    },
-    {
-      "epoch": 0.11077707339389582,
-      "grad_norm": 0.49765831232070923,
-      "learning_rate": 9.992659244509141e-06,
-      "loss": 0.4069,
-      "step": 1695
-    },
-    {
-      "epoch": 0.11084242859943795,
-      "grad_norm": 0.5660755038261414,
-      "learning_rate": 9.99264031749158e-06,
-      "loss": 0.499,
-      "step": 1696
-    },
-    {
-      "epoch": 0.11090778380498006,
-      "grad_norm": 0.5559900403022766,
-      "learning_rate": 9.99262136612318e-06,
-      "loss": 0.489,
-      "step": 1697
-    },
-    {
-      "epoch": 0.11097313901052219,
-      "grad_norm": 0.5337640047073364,
-      "learning_rate": 9.992602390404037e-06,
-      "loss": 0.4689,
-      "step": 1698
-    },
-    {
-      "epoch": 0.1110384942160643,
-      "grad_norm": 0.5071684122085571,
-      "learning_rate": 9.992583390334243e-06,
-      "loss": 0.4457,
-      "step": 1699
-    },
-    {
-      "epoch": 0.11110384942160643,
-      "grad_norm": 0.5411041975021362,
-      "learning_rate": 9.992564365913888e-06,
-      "loss": 0.5051,
-      "step": 1700
-    },
-    {
-      "epoch": 0.11116920462714855,
-      "grad_norm": 0.530849039554596,
-      "learning_rate": 9.99254531714307e-06,
-      "loss": 0.4164,
-      "step": 1701
-    },
-    {
-      "epoch": 0.11123455983269068,
-      "grad_norm": 0.5025213956832886,
-      "learning_rate": 9.992526244021877e-06,
-      "loss": 0.4171,
-      "step": 1702
-    },
-    {
-      "epoch": 0.11129991503823279,
-      "grad_norm": 0.5115631222724915,
-      "learning_rate": 9.992507146550404e-06,
-      "loss": 0.4399,
-      "step": 1703
-    },
-    {
-      "epoch": 0.11136527024377492,
-      "grad_norm": 0.5250508785247803,
-      "learning_rate": 9.992488024728744e-06,
-      "loss": 0.4049,
-      "step": 1704
-    },
-    {
-      "epoch": 0.11143062544931703,
-      "grad_norm": 0.4998297095298767,
-      "learning_rate": 9.992468878556992e-06,
-      "loss": 0.3713,
-      "step": 1705
-    },
-    {
-      "epoch": 0.11149598065485916,
-      "grad_norm": 0.5240525007247925,
-      "learning_rate": 9.992449708035237e-06,
-      "loss": 0.4612,
-      "step": 1706
-    },
-    {
-      "epoch": 0.11156133586040128,
-      "grad_norm": 0.538304328918457,
-      "learning_rate": 9.992430513163578e-06,
-      "loss": 0.4479,
-      "step": 1707
-    },
-    {
-      "epoch": 0.1116266910659434,
-      "grad_norm": 0.4934563934803009,
-      "learning_rate": 9.992411293942104e-06,
-      "loss": 0.3545,
-      "step": 1708
-    },
-    {
-      "epoch": 0.11169204627148552,
-      "grad_norm": 0.5011223554611206,
-      "learning_rate": 9.99239205037091e-06,
-      "loss": 0.3981,
-      "step": 1709
-    },
-    {
-      "epoch": 0.11175740147702765,
-      "grad_norm": 0.4800606071949005,
-      "learning_rate": 9.992372782450091e-06,
-      "loss": 0.3889,
-      "step": 1710
-    },
-    {
-      "epoch": 0.11182275668256976,
-      "grad_norm": 0.565067708492279,
-      "learning_rate": 9.992353490179741e-06,
-      "loss": 0.5346,
-      "step": 1711
-    },
-    {
-      "epoch": 0.11188811188811189,
-      "grad_norm": 0.5308020710945129,
-      "learning_rate": 9.992334173559955e-06,
-      "loss": 0.456,
-      "step": 1712
-    },
-    {
-      "epoch": 0.11195346709365402,
-      "grad_norm": 0.5349618196487427,
-      "learning_rate": 9.992314832590823e-06,
-      "loss": 0.4232,
-      "step": 1713
-    },
-    {
-      "epoch": 0.11201882229919613,
-      "grad_norm": 0.5136187076568604,
-      "learning_rate": 9.992295467272445e-06,
-      "loss": 0.3996,
-      "step": 1714
-    },
-    {
-      "epoch": 0.11208417750473826,
-      "grad_norm": 0.4795161485671997,
-      "learning_rate": 9.99227607760491e-06,
-      "loss": 0.404,
-      "step": 1715
-    },
-    {
-      "epoch": 0.11214953271028037,
-      "grad_norm": 0.5381472706794739,
-      "learning_rate": 9.992256663588315e-06,
-      "loss": 0.3986,
-      "step": 1716
-    },
-    {
-      "epoch": 0.1122148879158225,
-      "grad_norm": 0.5232036113739014,
-      "learning_rate": 9.992237225222756e-06,
-      "loss": 0.4077,
-      "step": 1717
-    },
-    {
-      "epoch": 0.11228024312136461,
-      "grad_norm": 0.5185948610305786,
-      "learning_rate": 9.992217762508324e-06,
-      "loss": 0.441,
-      "step": 1718
-    },
-    {
-      "epoch": 0.11234559832690674,
-      "grad_norm": 0.5535147190093994,
-      "learning_rate": 9.99219827544512e-06,
-      "loss": 0.3869,
-      "step": 1719
-    },
-    {
-      "epoch": 0.11241095353244886,
-      "grad_norm": 0.48375648260116577,
-      "learning_rate": 9.992178764033234e-06,
-      "loss": 0.3639,
-      "step": 1720
-    },
-    {
-      "epoch": 0.11247630873799098,
-      "grad_norm": 0.5208756923675537,
-      "learning_rate": 9.992159228272764e-06,
-      "loss": 0.394,
-      "step": 1721
-    },
-    {
-      "epoch": 0.1125416639435331,
-      "grad_norm": 0.5771656632423401,
-      "learning_rate": 9.992139668163803e-06,
-      "loss": 0.5047,
-      "step": 1722
-    },
-    {
-      "epoch": 0.11260701914907523,
-      "grad_norm": 0.5113723874092102,
-      "learning_rate": 9.992120083706447e-06,
-      "loss": 0.4335,
-      "step": 1723
-    },
-    {
-      "epoch": 0.11267237435461734,
-      "grad_norm": 0.530390202999115,
-      "learning_rate": 9.992100474900793e-06,
-      "loss": 0.4678,
-      "step": 1724
-    },
-    {
-      "epoch": 0.11273772956015947,
-      "grad_norm": 0.49649545550346375,
-      "learning_rate": 9.992080841746934e-06,
-      "loss": 0.4306,
-      "step": 1725
-    },
-    {
-      "epoch": 0.11280308476570158,
-      "grad_norm": 0.49351996183395386,
-      "learning_rate": 9.992061184244967e-06,
-      "loss": 0.401,
-      "step": 1726
-    },
-    {
-      "epoch": 0.11286843997124371,
-      "grad_norm": 0.501213788986206,
-      "learning_rate": 9.99204150239499e-06,
-      "loss": 0.4023,
-      "step": 1727
-    },
-    {
-      "epoch": 0.11293379517678583,
-      "grad_norm": 0.4906255304813385,
-      "learning_rate": 9.992021796197095e-06,
-      "loss": 0.3941,
-      "step": 1728
-    },
-    {
-      "epoch": 0.11299915038232795,
-      "grad_norm": 0.49137502908706665,
-      "learning_rate": 9.992002065651383e-06,
-      "loss": 0.4386,
-      "step": 1729
-    },
-    {
-      "epoch": 0.11306450558787007,
-      "grad_norm": 0.5452741980552673,
-      "learning_rate": 9.991982310757946e-06,
-      "loss": 0.4914,
-      "step": 1730
-    },
-    {
-      "epoch": 0.1131298607934122,
-      "grad_norm": 0.49915504455566406,
-      "learning_rate": 9.991962531516882e-06,
-      "loss": 0.4056,
-      "step": 1731
-    },
-    {
-      "epoch": 0.11319521599895431,
-      "grad_norm": 0.49487340450286865,
-      "learning_rate": 9.991942727928288e-06,
-      "loss": 0.4123,
-      "step": 1732
-    },
-    {
-      "epoch": 0.11326057120449644,
-      "grad_norm": 0.4953140616416931,
-      "learning_rate": 9.99192289999226e-06,
-      "loss": 0.4098,
-      "step": 1733
-    },
-    {
-      "epoch": 0.11332592641003857,
-      "grad_norm": 0.4806705117225647,
-      "learning_rate": 9.991903047708893e-06,
-      "loss": 0.4107,
-      "step": 1734
-    },
-    {
-      "epoch": 0.11339128161558068,
-      "grad_norm": 0.5293604135513306,
-      "learning_rate": 9.991883171078287e-06,
-      "loss": 0.4342,
-      "step": 1735
-    },
-    {
-      "epoch": 0.11345663682112281,
-      "grad_norm": 0.5292914509773254,
-      "learning_rate": 9.991863270100537e-06,
-      "loss": 0.4091,
-      "step": 1736
-    },
-    {
-      "epoch": 0.11352199202666492,
-      "grad_norm": 0.4674372375011444,
-      "learning_rate": 9.99184334477574e-06,
-      "loss": 0.3822,
-      "step": 1737
-    },
-    {
-      "epoch": 0.11358734723220705,
-      "grad_norm": 0.53412264585495,
-      "learning_rate": 9.991823395103995e-06,
-      "loss": 0.474,
-      "step": 1738
-    },
-    {
-      "epoch": 0.11365270243774916,
-      "grad_norm": 0.49650678038597107,
-      "learning_rate": 9.991803421085397e-06,
-      "loss": 0.3835,
-      "step": 1739
-    },
-    {
-      "epoch": 0.11371805764329129,
-      "grad_norm": 0.49791911244392395,
-      "learning_rate": 9.991783422720046e-06,
-      "loss": 0.4151,
-      "step": 1740
-    },
-    {
-      "epoch": 0.1137834128488334,
-      "grad_norm": 0.510292649269104,
-      "learning_rate": 9.991763400008035e-06,
-      "loss": 0.4251,
-      "step": 1741
-    },
-    {
-      "epoch": 0.11384876805437553,
-      "grad_norm": 0.5056973099708557,
-      "learning_rate": 9.991743352949466e-06,
-      "loss": 0.4595,
-      "step": 1742
-    },
-    {
-      "epoch": 0.11391412325991765,
-      "grad_norm": 0.5104454755783081,
-      "learning_rate": 9.991723281544433e-06,
-      "loss": 0.4174,
-      "step": 1743
-    },
-    {
-      "epoch": 0.11397947846545978,
-      "grad_norm": 0.5372150540351868,
-      "learning_rate": 9.991703185793041e-06,
-      "loss": 0.3844,
-      "step": 1744
-    },
-    {
-      "epoch": 0.11404483367100189,
-      "grad_norm": 0.5111712217330933,
-      "learning_rate": 9.99168306569538e-06,
-      "loss": 0.4219,
-      "step": 1745
-    },
-    {
-      "epoch": 0.11411018887654402,
-      "grad_norm": 0.4930039346218109,
-      "learning_rate": 9.991662921251552e-06,
-      "loss": 0.3793,
-      "step": 1746
-    },
-    {
-      "epoch": 0.11417554408208613,
-      "grad_norm": 0.49626004695892334,
-      "learning_rate": 9.991642752461657e-06,
-      "loss": 0.4392,
-      "step": 1747
-    },
-    {
-      "epoch": 0.11424089928762826,
-      "grad_norm": 0.5289170145988464,
-      "learning_rate": 9.991622559325787e-06,
-      "loss": 0.4399,
-      "step": 1748
-    },
-    {
-      "epoch": 0.11430625449317038,
-      "grad_norm": 0.5389708280563354,
-      "learning_rate": 9.991602341844047e-06,
-      "loss": 0.4415,
-      "step": 1749
-    },
-    {
-      "epoch": 0.1143716096987125,
-      "grad_norm": 0.46755653619766235,
-      "learning_rate": 9.991582100016532e-06,
-      "loss": 0.3867,
-      "step": 1750
-    },
-    {
-      "epoch": 0.11443696490425462,
-      "grad_norm": 0.4930843710899353,
-      "learning_rate": 9.991561833843344e-06,
-      "loss": 0.4016,
-      "step": 1751
-    },
-    {
-      "epoch": 0.11450232010979675,
-      "grad_norm": 0.4829428195953369,
-      "learning_rate": 9.991541543324578e-06,
-      "loss": 0.4276,
-      "step": 1752
-    },
-    {
-      "epoch": 0.11456767531533886,
-      "grad_norm": 0.4952201247215271,
-      "learning_rate": 9.991521228460334e-06,
-      "loss": 0.4201,
-      "step": 1753
-    },
-    {
-      "epoch": 0.11463303052088099,
-      "grad_norm": 0.5221691131591797,
-      "learning_rate": 9.991500889250713e-06,
-      "loss": 0.4356,
-      "step": 1754
-    },
-    {
-      "epoch": 0.11469838572642312,
-      "grad_norm": 0.5008527636528015,
-      "learning_rate": 9.991480525695813e-06,
-      "loss": 0.4374,
-      "step": 1755
-    },
-    {
-      "epoch": 0.11476374093196523,
-      "grad_norm": 0.5217990279197693,
-      "learning_rate": 9.991460137795733e-06,
-      "loss": 0.4545,
-      "step": 1756
-    },
-    {
-      "epoch": 0.11482909613750736,
-      "grad_norm": 0.5186170339584351,
-      "learning_rate": 9.991439725550571e-06,
-      "loss": 0.4272,
-      "step": 1757
-    },
-    {
-      "epoch": 0.11489445134304947,
-      "grad_norm": 0.4857713580131531,
-      "learning_rate": 9.99141928896043e-06,
-      "loss": 0.3829,
-      "step": 1758
-    },
-    {
-      "epoch": 0.1149598065485916,
-      "grad_norm": 0.5543569326400757,
-      "learning_rate": 9.991398828025408e-06,
-      "loss": 0.4596,
-      "step": 1759
-    },
-    {
-      "epoch": 0.11502516175413371,
-      "grad_norm": 0.5568326711654663,
-      "learning_rate": 9.991378342745604e-06,
-      "loss": 0.4903,
-      "step": 1760
-    },
-    {
-      "epoch": 0.11509051695967584,
-      "grad_norm": 0.472256064414978,
-      "learning_rate": 9.991357833121119e-06,
-      "loss": 0.4053,
-      "step": 1761
-    },
-    {
-      "epoch": 0.11515587216521796,
-      "grad_norm": 0.5433557629585266,
-      "learning_rate": 9.991337299152054e-06,
-      "loss": 0.4756,
-      "step": 1762
-    },
-    {
-      "epoch": 0.11522122737076008,
-      "grad_norm": 0.5174272060394287,
-      "learning_rate": 9.991316740838506e-06,
-      "loss": 0.3913,
-      "step": 1763
-    },
-    {
-      "epoch": 0.1152865825763022,
-      "grad_norm": 0.500368595123291,
-      "learning_rate": 9.991296158180577e-06,
-      "loss": 0.4013,
-      "step": 1764
-    },
-    {
-      "epoch": 0.11535193778184433,
-      "grad_norm": 0.5240638256072998,
-      "learning_rate": 9.991275551178368e-06,
-      "loss": 0.4406,
-      "step": 1765
-    },
-    {
-      "epoch": 0.11541729298738644,
-      "grad_norm": 0.5398024320602417,
-      "learning_rate": 9.99125491983198e-06,
-      "loss": 0.4668,
-      "step": 1766
-    },
-    {
-      "epoch": 0.11548264819292857,
-      "grad_norm": 0.5100404620170593,
-      "learning_rate": 9.991234264141512e-06,
-      "loss": 0.3996,
-      "step": 1767
-    },
-    {
-      "epoch": 0.11554800339847068,
-      "grad_norm": 0.5085362195968628,
-      "learning_rate": 9.991213584107065e-06,
-      "loss": 0.4481,
-      "step": 1768
-    },
-    {
-      "epoch": 0.11561335860401281,
-      "grad_norm": 0.619003176689148,
-      "learning_rate": 9.991192879728739e-06,
-      "loss": 0.4742,
-      "step": 1769
-    },
-    {
-      "epoch": 0.11567871380955493,
-      "grad_norm": 0.5342796444892883,
-      "learning_rate": 9.991172151006639e-06,
-      "loss": 0.4228,
-      "step": 1770
-    },
-    {
-      "epoch": 0.11574406901509705,
-      "grad_norm": 0.4943518340587616,
-      "learning_rate": 9.991151397940862e-06,
-      "loss": 0.3999,
-      "step": 1771
-    },
-    {
-      "epoch": 0.11580942422063917,
-      "grad_norm": 0.5110808610916138,
-      "learning_rate": 9.99113062053151e-06,
-      "loss": 0.4131,
-      "step": 1772
-    },
-    {
-      "epoch": 0.1158747794261813,
-      "grad_norm": 0.4850325286388397,
-      "learning_rate": 9.991109818778686e-06,
-      "loss": 0.3905,
-      "step": 1773
-    },
-    {
-      "epoch": 0.11594013463172341,
-      "grad_norm": 0.5520676374435425,
-      "learning_rate": 9.991088992682489e-06,
-      "loss": 0.4437,
-      "step": 1774
-    },
-    {
-      "epoch": 0.11600548983726554,
-      "grad_norm": 0.5156357288360596,
-      "learning_rate": 9.991068142243021e-06,
-      "loss": 0.4561,
-      "step": 1775
-    },
-    {
-      "epoch": 0.11607084504280767,
-      "grad_norm": 0.505234956741333,
-      "learning_rate": 9.991047267460387e-06,
-      "loss": 0.3864,
-      "step": 1776
-    },
-    {
-      "epoch": 0.11613620024834978,
-      "grad_norm": 0.5520933866500854,
-      "learning_rate": 9.991026368334685e-06,
-      "loss": 0.4962,
-      "step": 1777
-    },
-    {
-      "epoch": 0.11620155545389191,
-      "grad_norm": 0.5127909779548645,
-      "learning_rate": 9.991005444866019e-06,
-      "loss": 0.4412,
-      "step": 1778
-    },
-    {
-      "epoch": 0.11626691065943402,
-      "grad_norm": 0.47656968235969543,
-      "learning_rate": 9.99098449705449e-06,
-      "loss": 0.3944,
-      "step": 1779
-    },
-    {
-      "epoch": 0.11633226586497615,
-      "grad_norm": 0.4825184941291809,
-      "learning_rate": 9.990963524900202e-06,
-      "loss": 0.4123,
-      "step": 1780
-    },
-    {
-      "epoch": 0.11639762107051826,
-      "grad_norm": 0.4872205853462219,
-      "learning_rate": 9.990942528403253e-06,
-      "loss": 0.3892,
-      "step": 1781
-    },
-    {
-      "epoch": 0.11646297627606039,
-      "grad_norm": 0.644323468208313,
-      "learning_rate": 9.990921507563752e-06,
-      "loss": 0.4625,
-      "step": 1782
-    },
-    {
-      "epoch": 0.1165283314816025,
-      "grad_norm": 0.5240007042884827,
-      "learning_rate": 9.990900462381794e-06,
-      "loss": 0.4141,
-      "step": 1783
-    },
-    {
-      "epoch": 0.11659368668714463,
-      "grad_norm": 0.4936656951904297,
-      "learning_rate": 9.990879392857486e-06,
-      "loss": 0.4686,
-      "step": 1784
-    },
-    {
-      "epoch": 0.11665904189268675,
-      "grad_norm": 0.5088244080543518,
-      "learning_rate": 9.990858298990932e-06,
-      "loss": 0.4375,
-      "step": 1785
-    },
-    {
-      "epoch": 0.11672439709822888,
-      "grad_norm": 0.5279545187950134,
-      "learning_rate": 9.990837180782234e-06,
-      "loss": 0.4559,
-      "step": 1786
-    },
-    {
-      "epoch": 0.11678975230377099,
-      "grad_norm": 0.512206494808197,
-      "learning_rate": 9.99081603823149e-06,
-      "loss": 0.4568,
-      "step": 1787
-    },
-    {
-      "epoch": 0.11685510750931312,
-      "grad_norm": 0.4780323803424835,
-      "learning_rate": 9.990794871338811e-06,
-      "loss": 0.389,
-      "step": 1788
-    },
-    {
-      "epoch": 0.11692046271485523,
-      "grad_norm": 0.48555147647857666,
-      "learning_rate": 9.990773680104296e-06,
-      "loss": 0.4388,
-      "step": 1789
-    },
-    {
-      "epoch": 0.11698581792039736,
-      "grad_norm": 0.4882522225379944,
-      "learning_rate": 9.990752464528047e-06,
-      "loss": 0.4183,
-      "step": 1790
-    },
-    {
-      "epoch": 0.11705117312593948,
-      "grad_norm": 0.5026500225067139,
-      "learning_rate": 9.99073122461017e-06,
-      "loss": 0.4497,
-      "step": 1791
-    },
-    {
-      "epoch": 0.1171165283314816,
-      "grad_norm": 0.4988664984703064,
-      "learning_rate": 9.990709960350769e-06,
-      "loss": 0.415,
-      "step": 1792
-    },
-    {
-      "epoch": 0.11718188353702372,
-      "grad_norm": 0.581355094909668,
-      "learning_rate": 9.990688671749944e-06,
-      "loss": 0.3973,
-      "step": 1793
-    },
-    {
-      "epoch": 0.11724723874256585,
-      "grad_norm": 0.4597133696079254,
-      "learning_rate": 9.990667358807804e-06,
-      "loss": 0.4133,
-      "step": 1794
-    },
-    {
-      "epoch": 0.11731259394810796,
-      "grad_norm": 0.5237943530082703,
-      "learning_rate": 9.990646021524449e-06,
-      "loss": 0.4414,
-      "step": 1795
-    },
-    {
-      "epoch": 0.11737794915365009,
-      "grad_norm": 0.549954891204834,
-      "learning_rate": 9.990624659899986e-06,
-      "loss": 0.5089,
-      "step": 1796
-    },
-    {
-      "epoch": 0.11744330435919222,
-      "grad_norm": 0.45088574290275574,
-      "learning_rate": 9.990603273934516e-06,
-      "loss": 0.3577,
-      "step": 1797
-    },
-    {
-      "epoch": 0.11750865956473433,
-      "grad_norm": 0.4860559403896332,
-      "learning_rate": 9.990581863628144e-06,
-      "loss": 0.4324,
-      "step": 1798
-    },
-    {
-      "epoch": 0.11757401477027646,
-      "grad_norm": 0.5469778776168823,
-      "learning_rate": 9.990560428980977e-06,
-      "loss": 0.4381,
-      "step": 1799
-    },
-    {
-      "epoch": 0.11763936997581857,
-      "grad_norm": 0.49300289154052734,
-      "learning_rate": 9.990538969993118e-06,
-      "loss": 0.4192,
-      "step": 1800
-    },
-    {
-      "epoch": 0.1177047251813607,
-      "grad_norm": 0.5044761896133423,
-      "learning_rate": 9.99051748666467e-06,
-      "loss": 0.4707,
-      "step": 1801
-    },
-    {
-      "epoch": 0.11777008038690281,
-      "grad_norm": 0.4667385220527649,
-      "learning_rate": 9.99049597899574e-06,
-      "loss": 0.4064,
-      "step": 1802
-    },
-    {
-      "epoch": 0.11783543559244494,
-      "grad_norm": 0.4754604399204254,
-      "learning_rate": 9.990474446986433e-06,
-      "loss": 0.4128,
-      "step": 1803
-    },
-    {
-      "epoch": 0.11790079079798706,
-      "grad_norm": 0.4578009247779846,
-      "learning_rate": 9.990452890636856e-06,
-      "loss": 0.38,
-      "step": 1804
-    },
-    {
-      "epoch": 0.11796614600352918,
-      "grad_norm": 0.4580478370189667,
-      "learning_rate": 9.990431309947106e-06,
-      "loss": 0.3957,
-      "step": 1805
-    },
-    {
-      "epoch": 0.1180315012090713,
-      "grad_norm": 0.4998745620250702,
-      "learning_rate": 9.990409704917297e-06,
-      "loss": 0.4659,
-      "step": 1806
-    },
-    {
-      "epoch": 0.11809685641461343,
-      "grad_norm": 0.5275945663452148,
-      "learning_rate": 9.99038807554753e-06,
-      "loss": 0.488,
-      "step": 1807
-    },
-    {
-      "epoch": 0.11816221162015554,
-      "grad_norm": 0.4460282623767853,
-      "learning_rate": 9.990366421837912e-06,
-      "loss": 0.3598,
-      "step": 1808
-    },
-    {
-      "epoch": 0.11822756682569767,
-      "grad_norm": 0.5143075585365295,
-      "learning_rate": 9.990344743788547e-06,
-      "loss": 0.4938,
-      "step": 1809
-    },
-    {
-      "epoch": 0.11829292203123978,
-      "grad_norm": 0.5240001678466797,
-      "learning_rate": 9.990323041399543e-06,
-      "loss": 0.4981,
-      "step": 1810
-    },
-    {
-      "epoch": 0.11835827723678191,
-      "grad_norm": 0.5387542843818665,
-      "learning_rate": 9.990301314671003e-06,
-      "loss": 0.4879,
-      "step": 1811
-    },
-    {
-      "epoch": 0.11842363244232403,
-      "grad_norm": 0.4960077106952667,
-      "learning_rate": 9.990279563603035e-06,
-      "loss": 0.4133,
-      "step": 1812
-    },
-    {
-      "epoch": 0.11848898764786615,
-      "grad_norm": 0.5138953328132629,
-      "learning_rate": 9.990257788195747e-06,
-      "loss": 0.4514,
-      "step": 1813
-    },
-    {
-      "epoch": 0.11855434285340827,
-      "grad_norm": 0.4938754141330719,
-      "learning_rate": 9.990235988449242e-06,
-      "loss": 0.4345,
-      "step": 1814
-    },
-    {
-      "epoch": 0.1186196980589504,
-      "grad_norm": 0.4556517004966736,
-      "learning_rate": 9.990214164363628e-06,
-      "loss": 0.3922,
-      "step": 1815
-    },
-    {
-      "epoch": 0.11868505326449251,
-      "grad_norm": 0.5381069779396057,
-      "learning_rate": 9.99019231593901e-06,
-      "loss": 0.467,
-      "step": 1816
-    },
-    {
-      "epoch": 0.11875040847003464,
-      "grad_norm": 0.4847467243671417,
-      "learning_rate": 9.990170443175493e-06,
-      "loss": 0.4205,
-      "step": 1817
-    },
-    {
-      "epoch": 0.11881576367557677,
-      "grad_norm": 0.48896893858909607,
-      "learning_rate": 9.990148546073189e-06,
-      "loss": 0.394,
-      "step": 1818
-    },
-    {
-      "epoch": 0.11888111888111888,
-      "grad_norm": 0.5556392669677734,
-      "learning_rate": 9.990126624632201e-06,
-      "loss": 0.4897,
-      "step": 1819
-    },
-    {
-      "epoch": 0.11894647408666101,
-      "grad_norm": 0.555661678314209,
-      "learning_rate": 9.990104678852635e-06,
-      "loss": 0.4079,
-      "step": 1820
-    },
-    {
-      "epoch": 0.11901182929220312,
-      "grad_norm": 0.5040169954299927,
-      "learning_rate": 9.990082708734602e-06,
-      "loss": 0.4017,
-      "step": 1821
-    },
-    {
-      "epoch": 0.11907718449774525,
-      "grad_norm": 0.5507213473320007,
-      "learning_rate": 9.990060714278207e-06,
-      "loss": 0.5192,
-      "step": 1822
-    },
-    {
-      "epoch": 0.11914253970328736,
-      "grad_norm": 0.5001083612442017,
-      "learning_rate": 9.990038695483555e-06,
-      "loss": 0.459,
-      "step": 1823
-    },
-    {
-      "epoch": 0.11920789490882949,
-      "grad_norm": 0.4889615476131439,
-      "learning_rate": 9.990016652350759e-06,
-      "loss": 0.4397,
-      "step": 1824
-    },
-    {
-      "epoch": 0.1192732501143716,
-      "grad_norm": 0.47661325335502625,
-      "learning_rate": 9.989994584879919e-06,
-      "loss": 0.3961,
-      "step": 1825
-    },
-    {
-      "epoch": 0.11933860531991373,
-      "grad_norm": 0.49876272678375244,
-      "learning_rate": 9.98997249307115e-06,
-      "loss": 0.4079,
-      "step": 1826
-    },
-    {
-      "epoch": 0.11940396052545585,
-      "grad_norm": 0.46574166417121887,
-      "learning_rate": 9.989950376924555e-06,
-      "loss": 0.3667,
-      "step": 1827
-    },
-    {
-      "epoch": 0.11946931573099798,
-      "grad_norm": 0.5499573349952698,
-      "learning_rate": 9.989928236440242e-06,
-      "loss": 0.4348,
-      "step": 1828
-    },
-    {
-      "epoch": 0.11953467093654009,
-      "grad_norm": 0.5078558325767517,
-      "learning_rate": 9.989906071618323e-06,
-      "loss": 0.4414,
-      "step": 1829
-    },
-    {
-      "epoch": 0.11960002614208222,
-      "grad_norm": 0.47681233286857605,
-      "learning_rate": 9.989883882458902e-06,
-      "loss": 0.3838,
-      "step": 1830
-    },
-    {
-      "epoch": 0.11966538134762433,
-      "grad_norm": 0.4673955738544464,
-      "learning_rate": 9.989861668962089e-06,
-      "loss": 0.4117,
-      "step": 1831
-    },
-    {
-      "epoch": 0.11973073655316646,
-      "grad_norm": 0.4810030460357666,
-      "learning_rate": 9.989839431127992e-06,
-      "loss": 0.4078,
-      "step": 1832
-    },
-    {
-      "epoch": 0.11979609175870858,
-      "grad_norm": 0.4352864623069763,
-      "learning_rate": 9.989817168956719e-06,
-      "loss": 0.3383,
-      "step": 1833
-    },
-    {
-      "epoch": 0.1198614469642507,
-      "grad_norm": 0.5169110894203186,
-      "learning_rate": 9.989794882448378e-06,
-      "loss": 0.4659,
-      "step": 1834
-    },
-    {
-      "epoch": 0.11992680216979282,
-      "grad_norm": 0.5480664372444153,
-      "learning_rate": 9.98977257160308e-06,
-      "loss": 0.4267,
-      "step": 1835
-    },
-    {
-      "epoch": 0.11999215737533495,
-      "grad_norm": 0.48018908500671387,
-      "learning_rate": 9.989750236420933e-06,
-      "loss": 0.3706,
-      "step": 1836
-    },
-    {
-      "epoch": 0.12005751258087707,
-      "grad_norm": 0.5030679702758789,
-      "learning_rate": 9.989727876902044e-06,
-      "loss": 0.4464,
-      "step": 1837
-    },
-    {
-      "epoch": 0.12012286778641919,
-      "grad_norm": 0.566403329372406,
-      "learning_rate": 9.989705493046527e-06,
-      "loss": 0.4736,
-      "step": 1838
-    },
-    {
-      "epoch": 0.12018822299196132,
-      "grad_norm": 0.4993888735771179,
-      "learning_rate": 9.989683084854484e-06,
-      "loss": 0.4262,
-      "step": 1839
-    },
-    {
-      "epoch": 0.12025357819750343,
-      "grad_norm": 0.4965766966342926,
-      "learning_rate": 9.98966065232603e-06,
-      "loss": 0.4548,
-      "step": 1840
-    },
-    {
-      "epoch": 0.12031893340304556,
-      "grad_norm": 0.6620581746101379,
-      "learning_rate": 9.989638195461271e-06,
-      "loss": 0.4863,
-      "step": 1841
-    },
-    {
-      "epoch": 0.12038428860858767,
-      "grad_norm": 0.48163750767707825,
-      "learning_rate": 9.989615714260319e-06,
-      "loss": 0.3745,
-      "step": 1842
-    },
-    {
-      "epoch": 0.1204496438141298,
-      "grad_norm": 0.5864192843437195,
-      "learning_rate": 9.989593208723283e-06,
-      "loss": 0.5025,
-      "step": 1843
-    },
-    {
-      "epoch": 0.12051499901967191,
-      "grad_norm": 0.5082445740699768,
-      "learning_rate": 9.989570678850271e-06,
-      "loss": 0.41,
-      "step": 1844
-    },
-    {
-      "epoch": 0.12058035422521404,
-      "grad_norm": 0.47217845916748047,
-      "learning_rate": 9.989548124641396e-06,
-      "loss": 0.3884,
-      "step": 1845
-    },
-    {
-      "epoch": 0.12064570943075616,
-      "grad_norm": 0.5516869425773621,
-      "learning_rate": 9.989525546096762e-06,
-      "loss": 0.4281,
-      "step": 1846
-    },
-    {
-      "epoch": 0.12071106463629828,
-      "grad_norm": 0.5288026332855225,
-      "learning_rate": 9.989502943216488e-06,
-      "loss": 0.4794,
-      "step": 1847
-    },
-    {
-      "epoch": 0.1207764198418404,
-      "grad_norm": 0.5199207067489624,
-      "learning_rate": 9.989480316000678e-06,
-      "loss": 0.3937,
-      "step": 1848
-    },
-    {
-      "epoch": 0.12084177504738253,
-      "grad_norm": 0.5149030089378357,
-      "learning_rate": 9.989457664449445e-06,
-      "loss": 0.411,
-      "step": 1849
-    },
-    {
-      "epoch": 0.12090713025292464,
-      "grad_norm": 0.5265056490898132,
-      "learning_rate": 9.989434988562896e-06,
-      "loss": 0.4697,
-      "step": 1850
-    },
-    {
-      "epoch": 0.12097248545846677,
-      "grad_norm": 0.5422550439834595,
-      "learning_rate": 9.989412288341148e-06,
-      "loss": 0.4513,
-      "step": 1851
-    },
-    {
-      "epoch": 0.12103784066400888,
-      "grad_norm": 0.4541454613208771,
-      "learning_rate": 9.989389563784304e-06,
-      "loss": 0.3599,
-      "step": 1852
-    },
-    {
-      "epoch": 0.12110319586955101,
-      "grad_norm": 0.46409621834754944,
-      "learning_rate": 9.98936681489248e-06,
-      "loss": 0.3761,
-      "step": 1853
-    },
-    {
-      "epoch": 0.12116855107509313,
-      "grad_norm": 0.4885619878768921,
-      "learning_rate": 9.989344041665784e-06,
-      "loss": 0.4299,
-      "step": 1854
-    },
-    {
-      "epoch": 0.12123390628063525,
-      "grad_norm": 0.5073054432868958,
-      "learning_rate": 9.989321244104331e-06,
-      "loss": 0.3957,
-      "step": 1855
-    },
-    {
-      "epoch": 0.12129926148617737,
-      "grad_norm": 0.5251143574714661,
-      "learning_rate": 9.989298422208228e-06,
-      "loss": 0.4334,
-      "step": 1856
-    },
-    {
-      "epoch": 0.1213646166917195,
-      "grad_norm": 0.4727812707424164,
-      "learning_rate": 9.98927557597759e-06,
-      "loss": 0.3721,
-      "step": 1857
-    },
-    {
-      "epoch": 0.12142997189726162,
-      "grad_norm": 0.5348283052444458,
-      "learning_rate": 9.989252705412526e-06,
-      "loss": 0.4731,
-      "step": 1858
-    },
-    {
-      "epoch": 0.12149532710280374,
-      "grad_norm": 0.513326108455658,
-      "learning_rate": 9.989229810513147e-06,
-      "loss": 0.4111,
-      "step": 1859
-    },
-    {
-      "epoch": 0.12156068230834587,
-      "grad_norm": 0.4820597171783447,
-      "learning_rate": 9.989206891279569e-06,
-      "loss": 0.3684,
-      "step": 1860
-    },
-    {
-      "epoch": 0.12162603751388798,
-      "grad_norm": 0.5179558992385864,
-      "learning_rate": 9.989183947711898e-06,
-      "loss": 0.4668,
-      "step": 1861
-    },
-    {
-      "epoch": 0.12169139271943011,
-      "grad_norm": 0.4776163101196289,
-      "learning_rate": 9.989160979810247e-06,
-      "loss": 0.4098,
-      "step": 1862
-    },
-    {
-      "epoch": 0.12175674792497222,
-      "grad_norm": 0.5698267221450806,
-      "learning_rate": 9.989137987574731e-06,
-      "loss": 0.5449,
-      "step": 1863
-    },
-    {
-      "epoch": 0.12182210313051435,
-      "grad_norm": 0.5245393514633179,
-      "learning_rate": 9.98911497100546e-06,
-      "loss": 0.43,
-      "step": 1864
-    },
-    {
-      "epoch": 0.12188745833605646,
-      "grad_norm": 0.48276931047439575,
-      "learning_rate": 9.989091930102549e-06,
-      "loss": 0.3807,
-      "step": 1865
-    },
-    {
-      "epoch": 0.12195281354159859,
-      "grad_norm": 0.5482274889945984,
-      "learning_rate": 9.989068864866108e-06,
-      "loss": 0.4946,
-      "step": 1866
-    },
-    {
-      "epoch": 0.1220181687471407,
-      "grad_norm": 0.5497584939002991,
-      "learning_rate": 9.989045775296247e-06,
-      "loss": 0.5072,
-      "step": 1867
-    },
-    {
-      "epoch": 0.12208352395268283,
-      "grad_norm": 0.5020398497581482,
-      "learning_rate": 9.989022661393084e-06,
-      "loss": 0.4395,
-      "step": 1868
-    },
-    {
-      "epoch": 0.12214887915822495,
-      "grad_norm": 0.5637296438217163,
-      "learning_rate": 9.988999523156728e-06,
-      "loss": 0.4452,
-      "step": 1869
-    },
-    {
-      "epoch": 0.12221423436376708,
-      "grad_norm": 0.5545478463172913,
-      "learning_rate": 9.988976360587292e-06,
-      "loss": 0.4496,
-      "step": 1870
-    },
-    {
-      "epoch": 0.12227958956930919,
-      "grad_norm": 0.5133147835731506,
-      "learning_rate": 9.988953173684892e-06,
-      "loss": 0.4046,
-      "step": 1871
-    },
-    {
-      "epoch": 0.12234494477485132,
-      "grad_norm": 0.4993395507335663,
-      "learning_rate": 9.988929962449638e-06,
-      "loss": 0.4166,
-      "step": 1872
-    },
-    {
-      "epoch": 0.12241029998039343,
-      "grad_norm": 0.5099334120750427,
-      "learning_rate": 9.988906726881644e-06,
-      "loss": 0.4395,
-      "step": 1873
-    },
-    {
-      "epoch": 0.12247565518593556,
-      "grad_norm": 0.5003343224525452,
-      "learning_rate": 9.988883466981024e-06,
-      "loss": 0.4326,
-      "step": 1874
-    },
-    {
-      "epoch": 0.12254101039147768,
-      "grad_norm": 0.5290389060974121,
-      "learning_rate": 9.988860182747891e-06,
-      "loss": 0.4164,
-      "step": 1875
-    },
-    {
-      "epoch": 0.1226063655970198,
-      "grad_norm": 0.48433294892311096,
-      "learning_rate": 9.988836874182359e-06,
-      "loss": 0.406,
-      "step": 1876
-    },
-    {
-      "epoch": 0.12267172080256192,
-      "grad_norm": 0.4913772940635681,
-      "learning_rate": 9.98881354128454e-06,
-      "loss": 0.424,
-      "step": 1877
-    },
-    {
-      "epoch": 0.12273707600810405,
-      "grad_norm": 0.515835165977478,
-      "learning_rate": 9.988790184054551e-06,
-      "loss": 0.4607,
-      "step": 1878
-    },
-    {
-      "epoch": 0.12280243121364617,
-      "grad_norm": 0.5250944495201111,
-      "learning_rate": 9.988766802492503e-06,
-      "loss": 0.4528,
-      "step": 1879
-    },
-    {
-      "epoch": 0.12286778641918829,
-      "grad_norm": 0.5235453248023987,
-      "learning_rate": 9.988743396598511e-06,
-      "loss": 0.4573,
-      "step": 1880
-    },
-    {
-      "epoch": 0.12293314162473042,
-      "grad_norm": 0.5616452097892761,
-      "learning_rate": 9.988719966372688e-06,
-      "loss": 0.4835,
-      "step": 1881
-    },
-    {
-      "epoch": 0.12299849683027253,
-      "grad_norm": 0.534363865852356,
-      "learning_rate": 9.988696511815151e-06,
-      "loss": 0.4628,
-      "step": 1882
-    },
-    {
-      "epoch": 0.12306385203581466,
-      "grad_norm": 0.5168138146400452,
-      "learning_rate": 9.988673032926011e-06,
-      "loss": 0.4264,
-      "step": 1883
-    },
-    {
-      "epoch": 0.12312920724135677,
-      "grad_norm": 0.4910268187522888,
-      "learning_rate": 9.988649529705386e-06,
-      "loss": 0.4386,
-      "step": 1884
-    },
-    {
-      "epoch": 0.1231945624468989,
-      "grad_norm": 0.4870971441268921,
-      "learning_rate": 9.98862600215339e-06,
-      "loss": 0.3803,
-      "step": 1885
-    },
-    {
-      "epoch": 0.12325991765244101,
-      "grad_norm": 0.47671326994895935,
-      "learning_rate": 9.988602450270135e-06,
-      "loss": 0.4337,
-      "step": 1886
-    },
-    {
-      "epoch": 0.12332527285798314,
-      "grad_norm": 0.498994380235672,
-      "learning_rate": 9.988578874055739e-06,
-      "loss": 0.4714,
-      "step": 1887
-    },
-    {
-      "epoch": 0.12339062806352526,
-      "grad_norm": 0.5405935049057007,
-      "learning_rate": 9.988555273510315e-06,
-      "loss": 0.4635,
-      "step": 1888
-    },
-    {
-      "epoch": 0.12345598326906738,
-      "grad_norm": 0.49952763319015503,
-      "learning_rate": 9.988531648633976e-06,
-      "loss": 0.4473,
-      "step": 1889
-    },
-    {
-      "epoch": 0.1235213384746095,
-      "grad_norm": 0.5323501825332642,
-      "learning_rate": 9.988507999426845e-06,
-      "loss": 0.4491,
-      "step": 1890
-    },
-    {
-      "epoch": 0.12358669368015163,
-      "grad_norm": 0.4978667199611664,
-      "learning_rate": 9.988484325889029e-06,
-      "loss": 0.4162,
-      "step": 1891
-    },
-    {
-      "epoch": 0.12365204888569374,
-      "grad_norm": 0.4646618366241455,
-      "learning_rate": 9.988460628020646e-06,
-      "loss": 0.3946,
-      "step": 1892
-    },
-    {
-      "epoch": 0.12371740409123587,
-      "grad_norm": 0.4838745594024658,
-      "learning_rate": 9.988436905821814e-06,
-      "loss": 0.4222,
-      "step": 1893
-    },
-    {
-      "epoch": 0.12378275929677798,
-      "grad_norm": 0.478985458612442,
-      "learning_rate": 9.988413159292648e-06,
-      "loss": 0.3889,
-      "step": 1894
-    },
-    {
-      "epoch": 0.12384811450232011,
-      "grad_norm": 0.5021668076515198,
-      "learning_rate": 9.988389388433262e-06,
-      "loss": 0.4015,
-      "step": 1895
-    },
-    {
-      "epoch": 0.12391346970786223,
-      "grad_norm": 0.48870745301246643,
-      "learning_rate": 9.988365593243772e-06,
-      "loss": 0.4003,
-      "step": 1896
-    },
-    {
-      "epoch": 0.12397882491340435,
-      "grad_norm": 0.4673478305339813,
-      "learning_rate": 9.988341773724297e-06,
-      "loss": 0.3946,
-      "step": 1897
-    },
-    {
-      "epoch": 0.12404418011894647,
-      "grad_norm": 0.5073935985565186,
-      "learning_rate": 9.988317929874948e-06,
-      "loss": 0.4193,
-      "step": 1898
-    },
-    {
-      "epoch": 0.1241095353244886,
-      "grad_norm": 0.44659894704818726,
-      "learning_rate": 9.988294061695846e-06,
-      "loss": 0.3657,
-      "step": 1899
-    },
-    {
-      "epoch": 0.12417489053003072,
-      "grad_norm": 0.5482346415519714,
-      "learning_rate": 9.988270169187106e-06,
-      "loss": 0.5001,
-      "step": 1900
-    },
-    {
-      "epoch": 0.12424024573557284,
-      "grad_norm": 0.514901340007782,
-      "learning_rate": 9.988246252348843e-06,
-      "loss": 0.4342,
-      "step": 1901
-    },
-    {
-      "epoch": 0.12430560094111497,
-      "grad_norm": 0.4485880434513092,
-      "learning_rate": 9.988222311181177e-06,
-      "loss": 0.3543,
-      "step": 1902
-    },
-    {
-      "epoch": 0.12437095614665708,
-      "grad_norm": 0.5502244830131531,
-      "learning_rate": 9.988198345684222e-06,
-      "loss": 0.4872,
-      "step": 1903
-    },
-    {
-      "epoch": 0.12443631135219921,
-      "grad_norm": 0.5377838015556335,
-      "learning_rate": 9.988174355858093e-06,
-      "loss": 0.4766,
-      "step": 1904
-    },
-    {
-      "epoch": 0.12450166655774132,
-      "grad_norm": 0.520308792591095,
-      "learning_rate": 9.988150341702913e-06,
-      "loss": 0.4778,
-      "step": 1905
-    },
-    {
-      "epoch": 0.12456702176328345,
-      "grad_norm": 0.5440343618392944,
-      "learning_rate": 9.988126303218794e-06,
-      "loss": 0.4564,
-      "step": 1906
-    },
-    {
-      "epoch": 0.12463237696882556,
-      "grad_norm": 0.5795683264732361,
-      "learning_rate": 9.988102240405856e-06,
-      "loss": 0.4957,
-      "step": 1907
-    },
-    {
-      "epoch": 0.12469773217436769,
-      "grad_norm": 0.5030964016914368,
-      "learning_rate": 9.988078153264215e-06,
-      "loss": 0.4178,
-      "step": 1908
-    },
-    {
-      "epoch": 0.1247630873799098,
-      "grad_norm": 0.4966350197792053,
-      "learning_rate": 9.988054041793989e-06,
-      "loss": 0.4039,
-      "step": 1909
-    },
-    {
-      "epoch": 0.12482844258545193,
-      "grad_norm": 0.5520350933074951,
-      "learning_rate": 9.988029905995293e-06,
-      "loss": 0.4348,
-      "step": 1910
-    },
-    {
-      "epoch": 0.12489379779099405,
-      "grad_norm": 0.48109182715415955,
-      "learning_rate": 9.98800574586825e-06,
-      "loss": 0.3874,
-      "step": 1911
-    },
-    {
-      "epoch": 0.12495915299653618,
-      "grad_norm": 0.48294177651405334,
-      "learning_rate": 9.987981561412975e-06,
-      "loss": 0.4106,
-      "step": 1912
-    },
-    {
-      "epoch": 0.1250245082020783,
-      "grad_norm": 0.5936354994773865,
-      "learning_rate": 9.987957352629585e-06,
-      "loss": 0.4344,
-      "step": 1913
-    },
-    {
-      "epoch": 0.1250898634076204,
-      "grad_norm": 0.6715177893638611,
-      "learning_rate": 9.987933119518199e-06,
-      "loss": 0.4559,
-      "step": 1914
-    },
-    {
-      "epoch": 0.12515521861316253,
-      "grad_norm": 0.5224996209144592,
-      "learning_rate": 9.987908862078934e-06,
-      "loss": 0.4576,
-      "step": 1915
-    },
-    {
-      "epoch": 0.12522057381870466,
-      "grad_norm": 0.5017551779747009,
-      "learning_rate": 9.98788458031191e-06,
-      "loss": 0.3905,
-      "step": 1916
-    },
-    {
-      "epoch": 0.1252859290242468,
-      "grad_norm": 0.5825735926628113,
-      "learning_rate": 9.987860274217247e-06,
-      "loss": 0.4431,
-      "step": 1917
-    },
-    {
-      "epoch": 0.1253512842297889,
-      "grad_norm": 0.5097066760063171,
-      "learning_rate": 9.987835943795059e-06,
-      "loss": 0.4021,
-      "step": 1918
-    },
-    {
-      "epoch": 0.12541663943533102,
-      "grad_norm": 0.5344629883766174,
-      "learning_rate": 9.987811589045468e-06,
-      "loss": 0.4483,
-      "step": 1919
-    },
-    {
-      "epoch": 0.12548199464087315,
-      "grad_norm": 0.5319956541061401,
-      "learning_rate": 9.987787209968594e-06,
-      "loss": 0.4149,
-      "step": 1920
-    },
-    {
-      "epoch": 0.12554734984641527,
-      "grad_norm": 0.5144539475440979,
-      "learning_rate": 9.987762806564551e-06,
-      "loss": 0.4404,
-      "step": 1921
-    },
-    {
-      "epoch": 0.1256127050519574,
-      "grad_norm": 0.5102441310882568,
-      "learning_rate": 9.987738378833463e-06,
-      "loss": 0.4205,
-      "step": 1922
-    },
-    {
-      "epoch": 0.1256780602574995,
-      "grad_norm": 0.4958871304988861,
-      "learning_rate": 9.987713926775444e-06,
-      "loss": 0.3938,
-      "step": 1923
-    },
-    {
-      "epoch": 0.12574341546304163,
-      "grad_norm": 0.48275959491729736,
-      "learning_rate": 9.987689450390619e-06,
-      "loss": 0.3981,
-      "step": 1924
-    },
-    {
-      "epoch": 0.12580877066858376,
-      "grad_norm": 0.5233222246170044,
-      "learning_rate": 9.987664949679103e-06,
-      "loss": 0.4472,
-      "step": 1925
-    },
-    {
-      "epoch": 0.1258741258741259,
-      "grad_norm": 0.48806333541870117,
-      "learning_rate": 9.987640424641018e-06,
-      "loss": 0.3999,
-      "step": 1926
-    },
-    {
-      "epoch": 0.125939481079668,
-      "grad_norm": 0.5304328203201294,
-      "learning_rate": 9.987615875276483e-06,
-      "loss": 0.4717,
-      "step": 1927
-    },
-    {
-      "epoch": 0.12600483628521011,
-      "grad_norm": 0.5369510650634766,
-      "learning_rate": 9.987591301585618e-06,
-      "loss": 0.475,
-      "step": 1928
-    },
-    {
-      "epoch": 0.12607019149075224,
-      "grad_norm": 0.4829438626766205,
-      "learning_rate": 9.98756670356854e-06,
-      "loss": 0.413,
-      "step": 1929
-    },
-    {
-      "epoch": 0.12613554669629437,
-      "grad_norm": 0.4957874119281769,
-      "learning_rate": 9.987542081225374e-06,
-      "loss": 0.4514,
-      "step": 1930
-    },
-    {
-      "epoch": 0.12620090190183647,
-      "grad_norm": 0.49196264147758484,
-      "learning_rate": 9.987517434556237e-06,
-      "loss": 0.4087,
-      "step": 1931
-    },
-    {
-      "epoch": 0.1262662571073786,
-      "grad_norm": 0.4845946729183197,
-      "learning_rate": 9.987492763561249e-06,
-      "loss": 0.4006,
-      "step": 1932
-    },
-    {
-      "epoch": 0.12633161231292073,
-      "grad_norm": 0.5046796202659607,
-      "learning_rate": 9.98746806824053e-06,
-      "loss": 0.418,
-      "step": 1933
-    },
-    {
-      "epoch": 0.12639696751846285,
-      "grad_norm": 0.4948212504386902,
-      "learning_rate": 9.987443348594202e-06,
-      "loss": 0.4302,
-      "step": 1934
-    },
-    {
-      "epoch": 0.12646232272400496,
-      "grad_norm": 0.45483171939849854,
-      "learning_rate": 9.987418604622385e-06,
-      "loss": 0.3677,
-      "step": 1935
-    },
-    {
-      "epoch": 0.12652767792954708,
-      "grad_norm": 0.5139275789260864,
-      "learning_rate": 9.987393836325202e-06,
-      "loss": 0.4894,
-      "step": 1936
-    },
-    {
-      "epoch": 0.1265930331350892,
-      "grad_norm": 0.5322288274765015,
-      "learning_rate": 9.987369043702769e-06,
-      "loss": 0.4111,
-      "step": 1937
-    },
-    {
-      "epoch": 0.12665838834063134,
-      "grad_norm": 0.47072285413742065,
-      "learning_rate": 9.98734422675521e-06,
-      "loss": 0.3524,
-      "step": 1938
-    },
-    {
-      "epoch": 0.12672374354617344,
-      "grad_norm": 0.4729613959789276,
-      "learning_rate": 9.987319385482643e-06,
-      "loss": 0.4062,
-      "step": 1939
-    },
-    {
-      "epoch": 0.12678909875171557,
-      "grad_norm": 0.5595700740814209,
-      "learning_rate": 9.987294519885195e-06,
-      "loss": 0.4698,
-      "step": 1940
-    },
-    {
-      "epoch": 0.1268544539572577,
-      "grad_norm": 0.45723602175712585,
-      "learning_rate": 9.987269629962982e-06,
-      "loss": 0.3809,
-      "step": 1941
-    },
-    {
-      "epoch": 0.12691980916279982,
-      "grad_norm": 0.5328608155250549,
-      "learning_rate": 9.987244715716129e-06,
-      "loss": 0.4729,
-      "step": 1942
-    },
-    {
-      "epoch": 0.12698516436834195,
-      "grad_norm": 0.4701862633228302,
-      "learning_rate": 9.987219777144754e-06,
-      "loss": 0.3939,
-      "step": 1943
-    },
-    {
-      "epoch": 0.12705051957388405,
-      "grad_norm": 0.518118679523468,
-      "learning_rate": 9.987194814248981e-06,
-      "loss": 0.4676,
-      "step": 1944
-    },
-    {
-      "epoch": 0.12711587477942618,
-      "grad_norm": 0.5471842288970947,
-      "learning_rate": 9.987169827028931e-06,
-      "loss": 0.5107,
-      "step": 1945
-    },
-    {
-      "epoch": 0.1271812299849683,
-      "grad_norm": 0.4979475736618042,
-      "learning_rate": 9.987144815484726e-06,
-      "loss": 0.4176,
-      "step": 1946
-    },
-    {
-      "epoch": 0.12724658519051044,
-      "grad_norm": 0.4850045442581177,
-      "learning_rate": 9.987119779616489e-06,
-      "loss": 0.4047,
-      "step": 1947
-    },
-    {
-      "epoch": 0.12731194039605254,
-      "grad_norm": 0.4952496290206909,
-      "learning_rate": 9.98709471942434e-06,
-      "loss": 0.4438,
-      "step": 1948
-    },
-    {
-      "epoch": 0.12737729560159466,
-      "grad_norm": 0.5202100872993469,
-      "learning_rate": 9.987069634908402e-06,
-      "loss": 0.4774,
-      "step": 1949
-    },
-    {
-      "epoch": 0.1274426508071368,
-      "grad_norm": 0.5104197263717651,
-      "learning_rate": 9.987044526068799e-06,
-      "loss": 0.4328,
-      "step": 1950
-    },
-    {
-      "epoch": 0.12750800601267892,
-      "grad_norm": 0.473518431186676,
-      "learning_rate": 9.987019392905653e-06,
-      "loss": 0.3709,
-      "step": 1951
-    },
-    {
-      "epoch": 0.12757336121822102,
-      "grad_norm": 0.5406576991081238,
-      "learning_rate": 9.986994235419084e-06,
-      "loss": 0.4342,
-      "step": 1952
-    },
-    {
-      "epoch": 0.12763871642376315,
-      "grad_norm": 0.5023006200790405,
-      "learning_rate": 9.986969053609216e-06,
-      "loss": 0.4567,
-      "step": 1953
-    },
-    {
-      "epoch": 0.12770407162930528,
-      "grad_norm": 0.46537768840789795,
-      "learning_rate": 9.986943847476174e-06,
-      "loss": 0.3792,
-      "step": 1954
-    },
-    {
-      "epoch": 0.1277694268348474,
-      "grad_norm": 0.5310960412025452,
-      "learning_rate": 9.986918617020078e-06,
-      "loss": 0.4643,
-      "step": 1955
-    },
-    {
-      "epoch": 0.1278347820403895,
-      "grad_norm": 0.5471760034561157,
-      "learning_rate": 9.986893362241053e-06,
-      "loss": 0.5443,
-      "step": 1956
-    },
-    {
-      "epoch": 0.12790013724593163,
-      "grad_norm": 0.47290247678756714,
-      "learning_rate": 9.986868083139221e-06,
-      "loss": 0.3673,
-      "step": 1957
-    },
-    {
-      "epoch": 0.12796549245147376,
-      "grad_norm": 0.5044741034507751,
-      "learning_rate": 9.986842779714704e-06,
-      "loss": 0.4415,
-      "step": 1958
-    },
-    {
-      "epoch": 0.1280308476570159,
-      "grad_norm": 0.4717695415019989,
-      "learning_rate": 9.98681745196763e-06,
-      "loss": 0.3531,
-      "step": 1959
-    },
-    {
-      "epoch": 0.128096202862558,
-      "grad_norm": 0.49389272928237915,
-      "learning_rate": 9.98679209989812e-06,
-      "loss": 0.4117,
-      "step": 1960
-    },
-    {
-      "epoch": 0.12816155806810012,
-      "grad_norm": 0.5558046102523804,
-      "learning_rate": 9.986766723506295e-06,
-      "loss": 0.4613,
-      "step": 1961
-    },
-    {
-      "epoch": 0.12822691327364225,
-      "grad_norm": 0.4360634982585907,
-      "learning_rate": 9.986741322792282e-06,
-      "loss": 0.3352,
-      "step": 1962
-    },
-    {
-      "epoch": 0.12829226847918437,
-      "grad_norm": 0.5748230814933777,
-      "learning_rate": 9.986715897756206e-06,
-      "loss": 0.4892,
-      "step": 1963
-    },
-    {
-      "epoch": 0.1283576236847265,
-      "grad_norm": 0.46454548835754395,
-      "learning_rate": 9.986690448398185e-06,
-      "loss": 0.3994,
-      "step": 1964
-    },
-    {
-      "epoch": 0.1284229788902686,
-      "grad_norm": 0.5058770775794983,
-      "learning_rate": 9.98666497471835e-06,
-      "loss": 0.4254,
-      "step": 1965
-    },
-    {
-      "epoch": 0.12848833409581073,
-      "grad_norm": 0.5321208238601685,
-      "learning_rate": 9.986639476716821e-06,
-      "loss": 0.4693,
-      "step": 1966
-    },
-    {
-      "epoch": 0.12855368930135286,
-      "grad_norm": 0.5063179731369019,
-      "learning_rate": 9.986613954393725e-06,
-      "loss": 0.4256,
-      "step": 1967
-    },
-    {
-      "epoch": 0.128619044506895,
-      "grad_norm": 0.4850414991378784,
-      "learning_rate": 9.986588407749185e-06,
-      "loss": 0.3837,
-      "step": 1968
-    },
-    {
-      "epoch": 0.1286843997124371,
-      "grad_norm": 0.49293845891952515,
-      "learning_rate": 9.986562836783325e-06,
-      "loss": 0.4171,
-      "step": 1969
-    },
-    {
-      "epoch": 0.12874975491797921,
-      "grad_norm": 0.46116194128990173,
-      "learning_rate": 9.98653724149627e-06,
-      "loss": 0.3743,
-      "step": 1970
-    },
-    {
-      "epoch": 0.12881511012352134,
-      "grad_norm": 0.49085626006126404,
-      "learning_rate": 9.986511621888146e-06,
-      "loss": 0.3887,
-      "step": 1971
-    },
-    {
-      "epoch": 0.12888046532906347,
-      "grad_norm": 0.5477619171142578,
-      "learning_rate": 9.986485977959078e-06,
-      "loss": 0.4911,
-      "step": 1972
-    },
-    {
-      "epoch": 0.12894582053460557,
-      "grad_norm": 0.4762897193431854,
-      "learning_rate": 9.98646030970919e-06,
-      "loss": 0.3835,
-      "step": 1973
-    },
-    {
-      "epoch": 0.1290111757401477,
-      "grad_norm": 0.5051640868186951,
-      "learning_rate": 9.986434617138608e-06,
-      "loss": 0.3974,
-      "step": 1974
-    },
-    {
-      "epoch": 0.12907653094568983,
-      "grad_norm": 0.5479174256324768,
-      "learning_rate": 9.986408900247457e-06,
-      "loss": 0.4457,
-      "step": 1975
-    },
-    {
-      "epoch": 0.12914188615123195,
-      "grad_norm": 0.5653102993965149,
-      "learning_rate": 9.986383159035862e-06,
-      "loss": 0.491,
-      "step": 1976
-    },
-    {
-      "epoch": 0.12920724135677406,
-      "grad_norm": 0.4818381369113922,
-      "learning_rate": 9.986357393503947e-06,
-      "loss": 0.3849,
-      "step": 1977
-    },
-    {
-      "epoch": 0.12927259656231618,
-      "grad_norm": 0.48640090227127075,
-      "learning_rate": 9.986331603651843e-06,
-      "loss": 0.3895,
-      "step": 1978
-    },
-    {
-      "epoch": 0.1293379517678583,
-      "grad_norm": 0.5210935473442078,
-      "learning_rate": 9.986305789479669e-06,
-      "loss": 0.4264,
-      "step": 1979
-    },
-    {
-      "epoch": 0.12940330697340044,
-      "grad_norm": 0.5012984275817871,
-      "learning_rate": 9.986279950987556e-06,
-      "loss": 0.4618,
-      "step": 1980
-    },
-    {
-      "epoch": 0.12946866217894254,
-      "grad_norm": 0.504041314125061,
-      "learning_rate": 9.986254088175629e-06,
-      "loss": 0.4376,
-      "step": 1981
-    },
-    {
-      "epoch": 0.12953401738448467,
-      "grad_norm": 0.5589451789855957,
-      "learning_rate": 9.986228201044013e-06,
-      "loss": 0.4753,
-      "step": 1982
-    },
-    {
-      "epoch": 0.1295993725900268,
-      "grad_norm": 0.5148593187332153,
-      "learning_rate": 9.986202289592833e-06,
-      "loss": 0.4215,
-      "step": 1983
-    },
-    {
-      "epoch": 0.12966472779556892,
-      "grad_norm": 0.5945443511009216,
-      "learning_rate": 9.986176353822219e-06,
-      "loss": 0.4762,
-      "step": 1984
-    },
-    {
-      "epoch": 0.12973008300111105,
-      "grad_norm": 0.522632896900177,
-      "learning_rate": 9.986150393732294e-06,
-      "loss": 0.4608,
-      "step": 1985
-    },
-    {
-      "epoch": 0.12979543820665315,
-      "grad_norm": 0.5101238489151001,
-      "learning_rate": 9.986124409323188e-06,
-      "loss": 0.4402,
-      "step": 1986
-    },
-    {
-      "epoch": 0.12986079341219528,
-      "grad_norm": 0.5400654673576355,
-      "learning_rate": 9.986098400595024e-06,
-      "loss": 0.4693,
-      "step": 1987
-    },
-    {
-      "epoch": 0.1299261486177374,
-      "grad_norm": 0.5147498250007629,
-      "learning_rate": 9.986072367547932e-06,
-      "loss": 0.4267,
-      "step": 1988
-    },
-    {
-      "epoch": 0.12999150382327954,
-      "grad_norm": 0.5210046768188477,
-      "learning_rate": 9.986046310182037e-06,
-      "loss": 0.4559,
-      "step": 1989
-    },
-    {
-      "epoch": 0.13005685902882164,
-      "grad_norm": 0.488017737865448,
-      "learning_rate": 9.986020228497467e-06,
-      "loss": 0.3971,
-      "step": 1990
-    },
-    {
-      "epoch": 0.13012221423436376,
-      "grad_norm": 0.5373337864875793,
-      "learning_rate": 9.98599412249435e-06,
-      "loss": 0.4133,
-      "step": 1991
-    },
-    {
-      "epoch": 0.1301875694399059,
-      "grad_norm": 0.4798405170440674,
-      "learning_rate": 9.985967992172812e-06,
-      "loss": 0.388,
-      "step": 1992
-    },
-    {
-      "epoch": 0.13025292464544802,
-      "grad_norm": 0.5257364511489868,
-      "learning_rate": 9.985941837532979e-06,
-      "loss": 0.4119,
-      "step": 1993
-    },
-    {
-      "epoch": 0.13031827985099012,
-      "grad_norm": 0.47971484065055847,
-      "learning_rate": 9.985915658574982e-06,
-      "loss": 0.3939,
-      "step": 1994
-    },
-    {
-      "epoch": 0.13038363505653225,
-      "grad_norm": 0.5414712429046631,
-      "learning_rate": 9.985889455298948e-06,
-      "loss": 0.4333,
-      "step": 1995
-    },
-    {
-      "epoch": 0.13044899026207438,
-      "grad_norm": 0.46881482005119324,
-      "learning_rate": 9.985863227705002e-06,
-      "loss": 0.3941,
-      "step": 1996
-    },
-    {
-      "epoch": 0.1305143454676165,
-      "grad_norm": 0.5257583856582642,
-      "learning_rate": 9.985836975793272e-06,
-      "loss": 0.5,
-      "step": 1997
-    },
-    {
-      "epoch": 0.1305797006731586,
-      "grad_norm": 0.4880787134170532,
-      "learning_rate": 9.985810699563892e-06,
-      "loss": 0.418,
-      "step": 1998
-    },
-    {
-      "epoch": 0.13064505587870073,
-      "grad_norm": 0.45849093794822693,
-      "learning_rate": 9.985784399016984e-06,
-      "loss": 0.3674,
-      "step": 1999
-    },
-    {
-      "epoch": 0.13071041108424286,
-      "grad_norm": 0.5036157965660095,
-      "learning_rate": 9.985758074152678e-06,
-      "loss": 0.4192,
-      "step": 2000
-    },
-    {
-      "epoch": 0.130775766289785,
-      "grad_norm": 0.5034522414207458,
-      "learning_rate": 9.985731724971103e-06,
-      "loss": 0.4102,
-      "step": 2001
-    },
-    {
-      "epoch": 0.1308411214953271,
-      "grad_norm": 0.5690385103225708,
-      "learning_rate": 9.985705351472388e-06,
-      "loss": 0.4832,
-      "step": 2002
-    },
-    {
-      "epoch": 0.13090647670086922,
-      "grad_norm": 0.4804292619228363,
-      "learning_rate": 9.985678953656658e-06,
-      "loss": 0.4295,
-      "step": 2003
-    },
-    {
-      "epoch": 0.13097183190641135,
-      "grad_norm": 0.5212149620056152,
-      "learning_rate": 9.985652531524049e-06,
-      "loss": 0.4224,
-      "step": 2004
-    },
-    {
-      "epoch": 0.13103718711195347,
-      "grad_norm": 0.5116698145866394,
-      "learning_rate": 9.98562608507468e-06,
-      "loss": 0.4392,
-      "step": 2005
-    },
-    {
-      "epoch": 0.1311025423174956,
-      "grad_norm": 0.5497850775718689,
-      "learning_rate": 9.98559961430869e-06,
-      "loss": 0.4952,
-      "step": 2006
-    },
-    {
-      "epoch": 0.1311678975230377,
-      "grad_norm": 0.49011555314064026,
-      "learning_rate": 9.985573119226202e-06,
-      "loss": 0.429,
-      "step": 2007
-    },
-    {
-      "epoch": 0.13123325272857983,
-      "grad_norm": 0.5441197156906128,
-      "learning_rate": 9.985546599827346e-06,
-      "loss": 0.5136,
-      "step": 2008
-    },
-    {
-      "epoch": 0.13129860793412196,
-      "grad_norm": 0.4889719486236572,
-      "learning_rate": 9.985520056112252e-06,
-      "loss": 0.3924,
-      "step": 2009
-    },
-    {
-      "epoch": 0.13136396313966409,
-      "grad_norm": 0.7058687210083008,
-      "learning_rate": 9.98549348808105e-06,
-      "loss": 0.4088,
-      "step": 2010
-    },
-    {
-      "epoch": 0.1314293183452062,
-      "grad_norm": 0.48820960521698,
-      "learning_rate": 9.98546689573387e-06,
-      "loss": 0.4078,
-      "step": 2011
-    },
-    {
-      "epoch": 0.13149467355074831,
-      "grad_norm": 0.484792023897171,
-      "learning_rate": 9.98544027907084e-06,
-      "loss": 0.3738,
-      "step": 2012
-    },
-    {
-      "epoch": 0.13156002875629044,
-      "grad_norm": 0.44519925117492676,
-      "learning_rate": 9.98541363809209e-06,
-      "loss": 0.3907,
-      "step": 2013
-    },
-    {
-      "epoch": 0.13162538396183257,
-      "grad_norm": 0.4528064727783203,
-      "learning_rate": 9.98538697279775e-06,
-      "loss": 0.3685,
-      "step": 2014
-    },
-    {
-      "epoch": 0.13169073916737467,
-      "grad_norm": 0.5622209310531616,
-      "learning_rate": 9.985360283187954e-06,
-      "loss": 0.4623,
-      "step": 2015
-    },
-    {
-      "epoch": 0.1317560943729168,
-      "grad_norm": 0.5047494769096375,
-      "learning_rate": 9.985333569262827e-06,
-      "loss": 0.4528,
-      "step": 2016
-    },
-    {
-      "epoch": 0.13182144957845893,
-      "grad_norm": 0.4966709017753601,
-      "learning_rate": 9.9853068310225e-06,
-      "loss": 0.4212,
-      "step": 2017
-    },
-    {
-      "epoch": 0.13188680478400105,
-      "grad_norm": 0.46429166197776794,
-      "learning_rate": 9.985280068467104e-06,
-      "loss": 0.3791,
-      "step": 2018
-    },
-    {
-      "epoch": 0.13195215998954316,
-      "grad_norm": 0.4956360161304474,
-      "learning_rate": 9.985253281596773e-06,
-      "loss": 0.415,
-      "step": 2019
-    },
-    {
-      "epoch": 0.13201751519508528,
-      "grad_norm": 0.5155148506164551,
-      "learning_rate": 9.985226470411633e-06,
-      "loss": 0.4122,
-      "step": 2020
-    },
-    {
-      "epoch": 0.1320828704006274,
-      "grad_norm": 0.48214486241340637,
-      "learning_rate": 9.985199634911816e-06,
-      "loss": 0.3825,
-      "step": 2021
-    },
-    {
-      "epoch": 0.13214822560616954,
-      "grad_norm": 0.5534549951553345,
-      "learning_rate": 9.985172775097453e-06,
-      "loss": 0.4773,
-      "step": 2022
-    },
-    {
-      "epoch": 0.13221358081171164,
-      "grad_norm": 0.45360979437828064,
-      "learning_rate": 9.985145890968677e-06,
-      "loss": 0.3465,
-      "step": 2023
-    },
-    {
-      "epoch": 0.13227893601725377,
-      "grad_norm": 0.4827433228492737,
-      "learning_rate": 9.985118982525616e-06,
-      "loss": 0.4082,
-      "step": 2024
-    },
-    {
-      "epoch": 0.1323442912227959,
-      "grad_norm": 0.49237141013145447,
-      "learning_rate": 9.985092049768403e-06,
-      "loss": 0.4253,
-      "step": 2025
-    },
-    {
-      "epoch": 0.13240964642833802,
-      "grad_norm": 0.5407829284667969,
-      "learning_rate": 9.985065092697171e-06,
-      "loss": 0.4795,
-      "step": 2026
-    },
-    {
-      "epoch": 0.13247500163388015,
-      "grad_norm": 0.49884968996047974,
-      "learning_rate": 9.985038111312048e-06,
-      "loss": 0.4081,
-      "step": 2027
-    },
-    {
-      "epoch": 0.13254035683942225,
-      "grad_norm": 0.49501216411590576,
-      "learning_rate": 9.985011105613167e-06,
-      "loss": 0.3897,
-      "step": 2028
-    },
-    {
-      "epoch": 0.13260571204496438,
-      "grad_norm": 0.4731520712375641,
-      "learning_rate": 9.984984075600658e-06,
-      "loss": 0.3976,
-      "step": 2029
-    },
-    {
-      "epoch": 0.1326710672505065,
-      "grad_norm": 0.4776824414730072,
-      "learning_rate": 9.984957021274658e-06,
-      "loss": 0.3961,
-      "step": 2030
-    },
-    {
-      "epoch": 0.13273642245604864,
-      "grad_norm": 0.5426200032234192,
-      "learning_rate": 9.984929942635295e-06,
-      "loss": 0.4394,
-      "step": 2031
-    },
-    {
-      "epoch": 0.13280177766159074,
-      "grad_norm": 0.48932594060897827,
-      "learning_rate": 9.9849028396827e-06,
-      "loss": 0.4227,
-      "step": 2032
-    },
-    {
-      "epoch": 0.13286713286713286,
-      "grad_norm": 0.5132553577423096,
-      "learning_rate": 9.984875712417008e-06,
-      "loss": 0.4344,
-      "step": 2033
-    },
-    {
-      "epoch": 0.132932488072675,
-      "grad_norm": 0.4714146852493286,
-      "learning_rate": 9.984848560838352e-06,
-      "loss": 0.3729,
-      "step": 2034
-    },
-    {
-      "epoch": 0.13299784327821712,
-      "grad_norm": 0.4908793568611145,
-      "learning_rate": 9.98482138494686e-06,
-      "loss": 0.4192,
-      "step": 2035
-    },
-    {
-      "epoch": 0.13306319848375922,
-      "grad_norm": 0.5225798487663269,
-      "learning_rate": 9.984794184742668e-06,
-      "loss": 0.4606,
-      "step": 2036
-    },
-    {
-      "epoch": 0.13312855368930135,
-      "grad_norm": 0.5129513144493103,
-      "learning_rate": 9.984766960225907e-06,
-      "loss": 0.4287,
-      "step": 2037
-    },
-    {
-      "epoch": 0.13319390889484348,
-      "grad_norm": 0.45756110548973083,
-      "learning_rate": 9.984739711396712e-06,
-      "loss": 0.3792,
-      "step": 2038
-    },
-    {
-      "epoch": 0.1332592641003856,
-      "grad_norm": 0.48158565163612366,
-      "learning_rate": 9.984712438255213e-06,
-      "loss": 0.4197,
-      "step": 2039
-    },
-    {
-      "epoch": 0.1333246193059277,
-      "grad_norm": 0.5298853516578674,
-      "learning_rate": 9.984685140801547e-06,
-      "loss": 0.5092,
-      "step": 2040
-    },
-    {
-      "epoch": 0.13338997451146983,
-      "grad_norm": 0.5147441625595093,
-      "learning_rate": 9.984657819035844e-06,
-      "loss": 0.4191,
-      "step": 2041
-    },
-    {
-      "epoch": 0.13345532971701196,
-      "grad_norm": 0.4683409631252289,
-      "learning_rate": 9.984630472958237e-06,
-      "loss": 0.3968,
-      "step": 2042
-    },
-    {
-      "epoch": 0.1335206849225541,
-      "grad_norm": 0.4909045100212097,
-      "learning_rate": 9.98460310256886e-06,
-      "loss": 0.425,
-      "step": 2043
-    },
-    {
-      "epoch": 0.1335860401280962,
-      "grad_norm": 0.500486433506012,
-      "learning_rate": 9.984575707867847e-06,
-      "loss": 0.446,
-      "step": 2044
-    },
-    {
-      "epoch": 0.13365139533363832,
-      "grad_norm": 0.4790321886539459,
-      "learning_rate": 9.984548288855334e-06,
-      "loss": 0.4058,
-      "step": 2045
-    },
-    {
-      "epoch": 0.13371675053918045,
-      "grad_norm": 0.4697878360748291,
-      "learning_rate": 9.98452084553145e-06,
-      "loss": 0.3785,
-      "step": 2046
-    },
-    {
-      "epoch": 0.13378210574472257,
-      "grad_norm": 0.46032053232192993,
-      "learning_rate": 9.984493377896331e-06,
-      "loss": 0.3861,
-      "step": 2047
-    },
-    {
-      "epoch": 0.1338474609502647,
-      "grad_norm": 0.5933749675750732,
-      "learning_rate": 9.98446588595011e-06,
-      "loss": 0.4091,
-      "step": 2048
-    },
-    {
-      "epoch": 0.1339128161558068,
-      "grad_norm": 0.5029727816581726,
-      "learning_rate": 9.984438369692923e-06,
-      "loss": 0.4177,
-      "step": 2049
-    },
-    {
-      "epoch": 0.13397817136134893,
-      "grad_norm": 0.49849143624305725,
-      "learning_rate": 9.984410829124905e-06,
-      "loss": 0.3631,
-      "step": 2050
-    },
-    {
-      "epoch": 0.13404352656689106,
-      "grad_norm": 0.5268356204032898,
-      "learning_rate": 9.984383264246188e-06,
-      "loss": 0.4745,
-      "step": 2051
-    },
-    {
-      "epoch": 0.13410888177243319,
-      "grad_norm": 0.47841522097587585,
-      "learning_rate": 9.984355675056904e-06,
-      "loss": 0.4315,
-      "step": 2052
-    },
-    {
-      "epoch": 0.1341742369779753,
-      "grad_norm": 0.4980556070804596,
-      "learning_rate": 9.984328061557193e-06,
-      "loss": 0.4177,
-      "step": 2053
-    },
-    {
-      "epoch": 0.13423959218351741,
-      "grad_norm": 0.5080947279930115,
-      "learning_rate": 9.984300423747189e-06,
-      "loss": 0.3703,
-      "step": 2054
-    },
-    {
-      "epoch": 0.13430494738905954,
-      "grad_norm": 0.5016592741012573,
-      "learning_rate": 9.984272761627022e-06,
-      "loss": 0.4279,
-      "step": 2055
-    },
-    {
-      "epoch": 0.13437030259460167,
-      "grad_norm": 0.46037498116493225,
-      "learning_rate": 9.984245075196832e-06,
-      "loss": 0.3866,
-      "step": 2056
-    },
-    {
-      "epoch": 0.13443565780014377,
-      "grad_norm": 0.5036309957504272,
-      "learning_rate": 9.98421736445675e-06,
-      "loss": 0.438,
-      "step": 2057
-    },
-    {
-      "epoch": 0.1345010130056859,
-      "grad_norm": 0.485921710729599,
-      "learning_rate": 9.984189629406915e-06,
-      "loss": 0.4015,
-      "step": 2058
-    },
-    {
-      "epoch": 0.13456636821122803,
-      "grad_norm": 0.47186097502708435,
-      "learning_rate": 9.98416187004746e-06,
-      "loss": 0.4148,
-      "step": 2059
-    },
-    {
-      "epoch": 0.13463172341677015,
-      "grad_norm": 0.49796515703201294,
-      "learning_rate": 9.98413408637852e-06,
-      "loss": 0.3819,
-      "step": 2060
-    },
-    {
-      "epoch": 0.13469707862231226,
-      "grad_norm": 0.5647782683372498,
-      "learning_rate": 9.984106278400234e-06,
-      "loss": 0.5065,
-      "step": 2061
-    },
-    {
-      "epoch": 0.13476243382785438,
-      "grad_norm": 0.5064375996589661,
-      "learning_rate": 9.984078446112732e-06,
-      "loss": 0.4517,
-      "step": 2062
-    },
-    {
-      "epoch": 0.1348277890333965,
-      "grad_norm": 0.5200279951095581,
-      "learning_rate": 9.984050589516156e-06,
-      "loss": 0.4384,
-      "step": 2063
-    },
-    {
-      "epoch": 0.13489314423893864,
-      "grad_norm": 0.6002118587493896,
-      "learning_rate": 9.984022708610636e-06,
-      "loss": 0.5024,
-      "step": 2064
-    },
-    {
-      "epoch": 0.13495849944448074,
-      "grad_norm": 0.49059632420539856,
-      "learning_rate": 9.98399480339631e-06,
-      "loss": 0.444,
-      "step": 2065
-    },
-    {
-      "epoch": 0.13502385465002287,
-      "grad_norm": 0.5100719928741455,
-      "learning_rate": 9.983966873873317e-06,
-      "loss": 0.446,
-      "step": 2066
-    },
-    {
-      "epoch": 0.135089209855565,
-      "grad_norm": 0.4898199439048767,
-      "learning_rate": 9.983938920041792e-06,
-      "loss": 0.3841,
-      "step": 2067
-    },
-    {
-      "epoch": 0.13515456506110712,
-      "grad_norm": 0.5595495700836182,
-      "learning_rate": 9.983910941901867e-06,
-      "loss": 0.4472,
-      "step": 2068
-    },
-    {
-      "epoch": 0.13521992026664925,
-      "grad_norm": 0.5707384943962097,
-      "learning_rate": 9.983882939453683e-06,
-      "loss": 0.4348,
-      "step": 2069
-    },
-    {
-      "epoch": 0.13528527547219135,
-      "grad_norm": 0.5026779770851135,
-      "learning_rate": 9.983854912697374e-06,
-      "loss": 0.4351,
-      "step": 2070
-    },
-    {
-      "epoch": 0.13535063067773348,
-      "grad_norm": 0.498727023601532,
-      "learning_rate": 9.983826861633082e-06,
-      "loss": 0.4126,
-      "step": 2071
-    },
-    {
-      "epoch": 0.1354159858832756,
-      "grad_norm": 0.5695613026618958,
-      "learning_rate": 9.983798786260936e-06,
-      "loss": 0.4841,
-      "step": 2072
-    },
-    {
-      "epoch": 0.13548134108881774,
-      "grad_norm": 0.5176199674606323,
-      "learning_rate": 9.983770686581079e-06,
-      "loss": 0.4299,
-      "step": 2073
-    },
-    {
-      "epoch": 0.13554669629435984,
-      "grad_norm": 0.5157866477966309,
-      "learning_rate": 9.983742562593643e-06,
-      "loss": 0.4234,
-      "step": 2074
-    },
-    {
-      "epoch": 0.13561205149990196,
-      "grad_norm": 0.4823054373264313,
-      "learning_rate": 9.98371441429877e-06,
-      "loss": 0.3912,
-      "step": 2075
-    },
-    {
-      "epoch": 0.1356774067054441,
-      "grad_norm": 0.5139957070350647,
-      "learning_rate": 9.983686241696595e-06,
-      "loss": 0.465,
-      "step": 2076
-    },
-    {
-      "epoch": 0.13574276191098622,
-      "grad_norm": 0.5168617367744446,
-      "learning_rate": 9.983658044787257e-06,
-      "loss": 0.4429,
-      "step": 2077
-    },
-    {
-      "epoch": 0.13580811711652832,
-      "grad_norm": 0.5060180425643921,
-      "learning_rate": 9.98362982357089e-06,
-      "loss": 0.4591,
-      "step": 2078
-    },
-    {
-      "epoch": 0.13587347232207045,
-      "grad_norm": 0.5484333634376526,
-      "learning_rate": 9.983601578047636e-06,
-      "loss": 0.5163,
-      "step": 2079
-    },
-    {
-      "epoch": 0.13593882752761258,
-      "grad_norm": 0.5179669857025146,
-      "learning_rate": 9.983573308217629e-06,
-      "loss": 0.452,
-      "step": 2080
-    },
-    {
-      "epoch": 0.1360041827331547,
-      "grad_norm": 0.4726775884628296,
-      "learning_rate": 9.983545014081008e-06,
-      "loss": 0.4041,
-      "step": 2081
-    },
-    {
-      "epoch": 0.1360695379386968,
-      "grad_norm": 0.5156143307685852,
-      "learning_rate": 9.983516695637914e-06,
-      "loss": 0.4764,
-      "step": 2082
-    },
-    {
-      "epoch": 0.13613489314423893,
-      "grad_norm": 0.5028104186058044,
-      "learning_rate": 9.98348835288848e-06,
-      "loss": 0.393,
-      "step": 2083
-    },
-    {
-      "epoch": 0.13620024834978106,
-      "grad_norm": 0.48415282368659973,
-      "learning_rate": 9.983459985832848e-06,
-      "loss": 0.4175,
-      "step": 2084
-    },
-    {
-      "epoch": 0.1362656035553232,
-      "grad_norm": 0.5132015347480774,
-      "learning_rate": 9.983431594471156e-06,
-      "loss": 0.4379,
-      "step": 2085
-    },
-    {
-      "epoch": 0.1363309587608653,
-      "grad_norm": 0.47953587770462036,
-      "learning_rate": 9.98340317880354e-06,
-      "loss": 0.4583,
-      "step": 2086
-    },
-    {
-      "epoch": 0.13639631396640742,
-      "grad_norm": 0.4793354272842407,
-      "learning_rate": 9.983374738830142e-06,
-      "loss": 0.4065,
-      "step": 2087
-    },
-    {
-      "epoch": 0.13646166917194955,
-      "grad_norm": 0.516028106212616,
-      "learning_rate": 9.983346274551097e-06,
-      "loss": 0.4177,
-      "step": 2088
-    },
-    {
-      "epoch": 0.13652702437749167,
-      "grad_norm": 0.5440710783004761,
-      "learning_rate": 9.983317785966549e-06,
-      "loss": 0.5044,
-      "step": 2089
-    },
-    {
-      "epoch": 0.1365923795830338,
-      "grad_norm": 0.4665113389492035,
-      "learning_rate": 9.98328927307663e-06,
-      "loss": 0.3802,
-      "step": 2090
-    },
-    {
-      "epoch": 0.1366577347885759,
-      "grad_norm": 0.48358651995658875,
-      "learning_rate": 9.983260735881486e-06,
-      "loss": 0.3992,
-      "step": 2091
-    },
-    {
-      "epoch": 0.13672308999411803,
-      "grad_norm": 0.48970726132392883,
-      "learning_rate": 9.98323217438125e-06,
-      "loss": 0.3982,
-      "step": 2092
-    },
-    {
-      "epoch": 0.13678844519966016,
-      "grad_norm": 0.4874531924724579,
-      "learning_rate": 9.983203588576067e-06,
-      "loss": 0.4169,
-      "step": 2093
-    },
-    {
-      "epoch": 0.13685380040520229,
-      "grad_norm": 0.5156344771385193,
-      "learning_rate": 9.983174978466072e-06,
-      "loss": 0.3945,
-      "step": 2094
-    },
-    {
-      "epoch": 0.1369191556107444,
-      "grad_norm": 0.49989965558052063,
-      "learning_rate": 9.983146344051407e-06,
-      "loss": 0.42,
-      "step": 2095
-    },
-    {
-      "epoch": 0.13698451081628651,
-      "grad_norm": 0.5101418495178223,
-      "learning_rate": 9.98311768533221e-06,
-      "loss": 0.4517,
-      "step": 2096
-    },
-    {
-      "epoch": 0.13704986602182864,
-      "grad_norm": 0.5618047714233398,
-      "learning_rate": 9.983089002308623e-06,
-      "loss": 0.5002,
-      "step": 2097
-    },
-    {
-      "epoch": 0.13711522122737077,
-      "grad_norm": 0.44366690516471863,
-      "learning_rate": 9.983060294980786e-06,
-      "loss": 0.3677,
-      "step": 2098
-    },
-    {
-      "epoch": 0.13718057643291287,
-      "grad_norm": 0.488775372505188,
-      "learning_rate": 9.983031563348835e-06,
-      "loss": 0.4209,
-      "step": 2099
-    },
-    {
-      "epoch": 0.137245931638455,
-      "grad_norm": 0.4840698838233948,
-      "learning_rate": 9.983002807412913e-06,
-      "loss": 0.3831,
-      "step": 2100
-    },
-    {
-      "epoch": 0.13731128684399713,
-      "grad_norm": 0.5189481377601624,
-      "learning_rate": 9.98297402717316e-06,
-      "loss": 0.445,
-      "step": 2101
-    },
-    {
-      "epoch": 0.13737664204953925,
-      "grad_norm": 0.47477179765701294,
-      "learning_rate": 9.982945222629719e-06,
-      "loss": 0.373,
-      "step": 2102
-    },
-    {
-      "epoch": 0.13744199725508136,
-      "grad_norm": 0.5027723908424377,
-      "learning_rate": 9.982916393782725e-06,
-      "loss": 0.4158,
-      "step": 2103
-    },
-    {
-      "epoch": 0.13750735246062348,
-      "grad_norm": 0.507691502571106,
-      "learning_rate": 9.982887540632323e-06,
-      "loss": 0.3809,
-      "step": 2104
-    },
-    {
-      "epoch": 0.1375727076661656,
-      "grad_norm": 0.5458555221557617,
-      "learning_rate": 9.982858663178651e-06,
-      "loss": 0.4419,
-      "step": 2105
-    },
-    {
-      "epoch": 0.13763806287170774,
-      "grad_norm": 0.5059611201286316,
-      "learning_rate": 9.982829761421853e-06,
-      "loss": 0.4294,
-      "step": 2106
-    },
-    {
-      "epoch": 0.13770341807724984,
-      "grad_norm": 0.5017284750938416,
-      "learning_rate": 9.982800835362067e-06,
-      "loss": 0.3752,
-      "step": 2107
-    },
-    {
-      "epoch": 0.13776877328279197,
-      "grad_norm": 0.5887925624847412,
-      "learning_rate": 9.982771884999433e-06,
-      "loss": 0.5432,
-      "step": 2108
-    },
-    {
-      "epoch": 0.1378341284883341,
-      "grad_norm": 0.48969125747680664,
-      "learning_rate": 9.982742910334098e-06,
-      "loss": 0.4059,
-      "step": 2109
-    },
-    {
-      "epoch": 0.13789948369387622,
-      "grad_norm": 0.47747862339019775,
-      "learning_rate": 9.982713911366198e-06,
-      "loss": 0.4054,
-      "step": 2110
-    },
-    {
-      "epoch": 0.13796483889941835,
-      "grad_norm": 0.536247193813324,
-      "learning_rate": 9.982684888095874e-06,
-      "loss": 0.4363,
-      "step": 2111
-    },
-    {
-      "epoch": 0.13803019410496045,
-      "grad_norm": 0.5041215419769287,
-      "learning_rate": 9.982655840523272e-06,
-      "loss": 0.3947,
-      "step": 2112
-    },
-    {
-      "epoch": 0.13809554931050258,
-      "grad_norm": 0.5441122055053711,
-      "learning_rate": 9.982626768648533e-06,
-      "loss": 0.4576,
-      "step": 2113
-    },
-    {
-      "epoch": 0.1381609045160447,
-      "grad_norm": 0.5348301529884338,
-      "learning_rate": 9.982597672471795e-06,
-      "loss": 0.4644,
-      "step": 2114
-    },
-    {
-      "epoch": 0.13822625972158684,
-      "grad_norm": 0.49890631437301636,
-      "learning_rate": 9.982568551993202e-06,
-      "loss": 0.4001,
-      "step": 2115
-    },
-    {
-      "epoch": 0.13829161492712894,
-      "grad_norm": 0.5681653618812561,
-      "learning_rate": 9.982539407212895e-06,
-      "loss": 0.4812,
-      "step": 2116
-    },
-    {
-      "epoch": 0.13835697013267106,
-      "grad_norm": 0.5000870227813721,
-      "learning_rate": 9.982510238131018e-06,
-      "loss": 0.4108,
-      "step": 2117
-    },
-    {
-      "epoch": 0.1384223253382132,
-      "grad_norm": 0.49189293384552,
-      "learning_rate": 9.982481044747713e-06,
-      "loss": 0.4109,
-      "step": 2118
-    },
-    {
-      "epoch": 0.13848768054375532,
-      "grad_norm": 0.49221333861351013,
-      "learning_rate": 9.982451827063121e-06,
-      "loss": 0.415,
-      "step": 2119
-    },
-    {
-      "epoch": 0.13855303574929742,
-      "grad_norm": 0.5470876693725586,
-      "learning_rate": 9.982422585077386e-06,
-      "loss": 0.4894,
-      "step": 2120
-    },
-    {
-      "epoch": 0.13861839095483955,
-      "grad_norm": 0.4839327335357666,
-      "learning_rate": 9.98239331879065e-06,
-      "loss": 0.4405,
-      "step": 2121
-    },
-    {
-      "epoch": 0.13868374616038168,
-      "grad_norm": 0.5111157894134521,
-      "learning_rate": 9.982364028203056e-06,
-      "loss": 0.4661,
-      "step": 2122
-    },
-    {
-      "epoch": 0.1387491013659238,
-      "grad_norm": 0.5053145885467529,
-      "learning_rate": 9.982334713314748e-06,
-      "loss": 0.4127,
-      "step": 2123
-    },
-    {
-      "epoch": 0.1388144565714659,
-      "grad_norm": 0.47632038593292236,
-      "learning_rate": 9.982305374125865e-06,
-      "loss": 0.4067,
-      "step": 2124
-    },
-    {
-      "epoch": 0.13887981177700803,
-      "grad_norm": 0.48555120825767517,
-      "learning_rate": 9.982276010636554e-06,
-      "loss": 0.3996,
-      "step": 2125
-    },
-    {
-      "epoch": 0.13894516698255016,
-      "grad_norm": 0.4961962103843689,
-      "learning_rate": 9.982246622846955e-06,
-      "loss": 0.4187,
-      "step": 2126
-    },
-    {
-      "epoch": 0.1390105221880923,
-      "grad_norm": 0.48168522119522095,
-      "learning_rate": 9.982217210757214e-06,
-      "loss": 0.4124,
-      "step": 2127
-    },
-    {
-      "epoch": 0.1390758773936344,
-      "grad_norm": 0.48840004205703735,
-      "learning_rate": 9.982187774367475e-06,
-      "loss": 0.4133,
-      "step": 2128
-    },
-    {
-      "epoch": 0.13914123259917652,
-      "grad_norm": 0.45378541946411133,
-      "learning_rate": 9.982158313677879e-06,
-      "loss": 0.3904,
-      "step": 2129
-    },
-    {
-      "epoch": 0.13920658780471865,
-      "grad_norm": 0.608108401298523,
-      "learning_rate": 9.98212882868857e-06,
-      "loss": 0.4546,
-      "step": 2130
-    },
-    {
-      "epoch": 0.13927194301026077,
-      "grad_norm": 0.4957594573497772,
-      "learning_rate": 9.982099319399696e-06,
-      "loss": 0.4246,
-      "step": 2131
-    },
-    {
-      "epoch": 0.1393372982158029,
-      "grad_norm": 0.47023138403892517,
-      "learning_rate": 9.982069785811395e-06,
-      "loss": 0.3833,
-      "step": 2132
-    },
-    {
-      "epoch": 0.139402653421345,
-      "grad_norm": 0.48780372738838196,
-      "learning_rate": 9.982040227923815e-06,
-      "loss": 0.4143,
-      "step": 2133
-    },
-    {
-      "epoch": 0.13946800862688713,
-      "grad_norm": 0.5033027529716492,
-      "learning_rate": 9.982010645737098e-06,
-      "loss": 0.4193,
-      "step": 2134
-    },
-    {
-      "epoch": 0.13953336383242926,
-      "grad_norm": 0.5258437991142273,
-      "learning_rate": 9.98198103925139e-06,
-      "loss": 0.4729,
-      "step": 2135
-    },
-    {
-      "epoch": 0.13959871903797139,
-      "grad_norm": 0.5443706512451172,
-      "learning_rate": 9.981951408466834e-06,
-      "loss": 0.4389,
-      "step": 2136
-    },
-    {
-      "epoch": 0.1396640742435135,
-      "grad_norm": 0.5517335534095764,
-      "learning_rate": 9.981921753383574e-06,
-      "loss": 0.4902,
-      "step": 2137
-    },
-    {
-      "epoch": 0.13972942944905561,
-      "grad_norm": 0.4983333647251129,
-      "learning_rate": 9.981892074001758e-06,
-      "loss": 0.4121,
-      "step": 2138
-    },
-    {
-      "epoch": 0.13979478465459774,
-      "grad_norm": 0.49221929907798767,
-      "learning_rate": 9.981862370321527e-06,
-      "loss": 0.4307,
-      "step": 2139
-    },
-    {
-      "epoch": 0.13986013986013987,
-      "grad_norm": 0.4637519419193268,
-      "learning_rate": 9.981832642343027e-06,
-      "loss": 0.383,
-      "step": 2140
-    },
-    {
-      "epoch": 0.13992549506568197,
-      "grad_norm": 0.5155725479125977,
-      "learning_rate": 9.981802890066404e-06,
-      "loss": 0.4103,
-      "step": 2141
-    },
-    {
-      "epoch": 0.1399908502712241,
-      "grad_norm": 0.4840599000453949,
-      "learning_rate": 9.981773113491801e-06,
-      "loss": 0.388,
-      "step": 2142
-    },
-    {
-      "epoch": 0.14005620547676623,
-      "grad_norm": 0.4303397834300995,
-      "learning_rate": 9.981743312619367e-06,
-      "loss": 0.3225,
-      "step": 2143
-    },
-    {
-      "epoch": 0.14012156068230835,
-      "grad_norm": 0.5074151754379272,
-      "learning_rate": 9.981713487449243e-06,
-      "loss": 0.4638,
-      "step": 2144
-    },
-    {
-      "epoch": 0.14018691588785046,
-      "grad_norm": 0.4615415036678314,
-      "learning_rate": 9.981683637981579e-06,
-      "loss": 0.407,
-      "step": 2145
-    },
-    {
-      "epoch": 0.14025227109339258,
-      "grad_norm": 0.5153157114982605,
-      "learning_rate": 9.981653764216516e-06,
-      "loss": 0.4803,
-      "step": 2146
-    },
-    {
-      "epoch": 0.1403176262989347,
-      "grad_norm": 0.4582500159740448,
-      "learning_rate": 9.981623866154202e-06,
-      "loss": 0.3967,
-      "step": 2147
-    },
-    {
-      "epoch": 0.14038298150447684,
-      "grad_norm": 0.48393577337265015,
-      "learning_rate": 9.981593943794783e-06,
-      "loss": 0.3556,
-      "step": 2148
-    },
-    {
-      "epoch": 0.14044833671001894,
-      "grad_norm": 0.5076748132705688,
-      "learning_rate": 9.981563997138405e-06,
-      "loss": 0.422,
-      "step": 2149
-    },
-    {
-      "epoch": 0.14051369191556107,
-      "grad_norm": 0.5000056028366089,
-      "learning_rate": 9.981534026185213e-06,
-      "loss": 0.428,
-      "step": 2150
-    },
-    {
-      "epoch": 0.1405790471211032,
-      "grad_norm": 0.47516247630119324,
-      "learning_rate": 9.981504030935354e-06,
-      "loss": 0.3851,
-      "step": 2151
-    },
-    {
-      "epoch": 0.14064440232664532,
-      "grad_norm": 0.5118639469146729,
-      "learning_rate": 9.981474011388974e-06,
-      "loss": 0.4027,
-      "step": 2152
-    },
-    {
-      "epoch": 0.14070975753218745,
-      "grad_norm": 0.49031445384025574,
-      "learning_rate": 9.98144396754622e-06,
-      "loss": 0.4167,
-      "step": 2153
-    },
-    {
-      "epoch": 0.14077511273772955,
-      "grad_norm": 0.4741804897785187,
-      "learning_rate": 9.981413899407237e-06,
-      "loss": 0.386,
-      "step": 2154
-    },
-    {
-      "epoch": 0.14084046794327168,
-      "grad_norm": 0.5037586688995361,
-      "learning_rate": 9.981383806972174e-06,
-      "loss": 0.4392,
-      "step": 2155
-    },
-    {
-      "epoch": 0.1409058231488138,
-      "grad_norm": 0.4791088402271271,
-      "learning_rate": 9.981353690241174e-06,
-      "loss": 0.3811,
-      "step": 2156
-    },
-    {
-      "epoch": 0.14097117835435594,
-      "grad_norm": 0.446625292301178,
-      "learning_rate": 9.981323549214388e-06,
-      "loss": 0.3497,
-      "step": 2157
-    },
-    {
-      "epoch": 0.14103653355989804,
-      "grad_norm": 0.5049254298210144,
-      "learning_rate": 9.981293383891962e-06,
-      "loss": 0.4535,
-      "step": 2158
-    },
-    {
-      "epoch": 0.14110188876544016,
-      "grad_norm": 0.4784114360809326,
-      "learning_rate": 9.981263194274043e-06,
-      "loss": 0.423,
-      "step": 2159
-    },
-    {
-      "epoch": 0.1411672439709823,
-      "grad_norm": 0.5221757888793945,
-      "learning_rate": 9.981232980360776e-06,
-      "loss": 0.4802,
-      "step": 2160
-    },
-    {
-      "epoch": 0.14123259917652442,
-      "grad_norm": 0.5037352442741394,
-      "learning_rate": 9.981202742152309e-06,
-      "loss": 0.4063,
-      "step": 2161
-    },
-    {
-      "epoch": 0.14129795438206652,
-      "grad_norm": 0.4714972972869873,
-      "learning_rate": 9.981172479648793e-06,
-      "loss": 0.3518,
-      "step": 2162
-    },
-    {
-      "epoch": 0.14136330958760865,
-      "grad_norm": 0.48040875792503357,
-      "learning_rate": 9.981142192850373e-06,
-      "loss": 0.416,
-      "step": 2163
-    },
-    {
-      "epoch": 0.14142866479315078,
-      "grad_norm": 0.47911337018013,
-      "learning_rate": 9.981111881757195e-06,
-      "loss": 0.3878,
-      "step": 2164
-    },
-    {
-      "epoch": 0.1414940199986929,
-      "grad_norm": 0.5378538966178894,
-      "learning_rate": 9.98108154636941e-06,
-      "loss": 0.4468,
-      "step": 2165
-    },
-    {
-      "epoch": 0.141559375204235,
-      "grad_norm": 0.4630666971206665,
-      "learning_rate": 9.981051186687165e-06,
-      "loss": 0.3939,
-      "step": 2166
-    },
-    {
-      "epoch": 0.14162473040977713,
-      "grad_norm": 0.4693721532821655,
-      "learning_rate": 9.981020802710608e-06,
-      "loss": 0.4077,
-      "step": 2167
-    },
-    {
-      "epoch": 0.14169008561531926,
-      "grad_norm": 0.48182645440101624,
-      "learning_rate": 9.980990394439887e-06,
-      "loss": 0.405,
-      "step": 2168
-    },
-    {
-      "epoch": 0.1417554408208614,
-      "grad_norm": 0.5250624418258667,
-      "learning_rate": 9.980959961875149e-06,
-      "loss": 0.4425,
-      "step": 2169
-    },
-    {
-      "epoch": 0.1418207960264035,
-      "grad_norm": 0.5258338451385498,
-      "learning_rate": 9.980929505016544e-06,
-      "loss": 0.4352,
-      "step": 2170
-    },
-    {
-      "epoch": 0.14188615123194562,
-      "grad_norm": 0.49704474210739136,
-      "learning_rate": 9.980899023864222e-06,
-      "loss": 0.4166,
-      "step": 2171
-    },
-    {
-      "epoch": 0.14195150643748775,
-      "grad_norm": 0.5207509398460388,
-      "learning_rate": 9.980868518418327e-06,
-      "loss": 0.4443,
-      "step": 2172
-    },
-    {
-      "epoch": 0.14201686164302987,
-      "grad_norm": 0.44627878069877625,
-      "learning_rate": 9.980837988679013e-06,
-      "loss": 0.3656,
-      "step": 2173
-    },
-    {
-      "epoch": 0.142082216848572,
-      "grad_norm": 0.5200849771499634,
-      "learning_rate": 9.980807434646426e-06,
-      "loss": 0.4662,
-      "step": 2174
-    },
-    {
-      "epoch": 0.1421475720541141,
-      "grad_norm": 0.5185167193412781,
-      "learning_rate": 9.980776856320715e-06,
-      "loss": 0.4289,
-      "step": 2175
-    },
-    {
-      "epoch": 0.14221292725965623,
-      "grad_norm": 0.5304030179977417,
-      "learning_rate": 9.980746253702031e-06,
-      "loss": 0.4672,
-      "step": 2176
-    },
-    {
-      "epoch": 0.14227828246519836,
-      "grad_norm": 0.5466896295547485,
-      "learning_rate": 9.98071562679052e-06,
-      "loss": 0.477,
-      "step": 2177
-    },
-    {
-      "epoch": 0.14234363767074049,
-      "grad_norm": 0.5005002617835999,
-      "learning_rate": 9.980684975586335e-06,
-      "loss": 0.4571,
-      "step": 2178
-    },
-    {
-      "epoch": 0.1424089928762826,
-      "grad_norm": 0.4743422269821167,
-      "learning_rate": 9.980654300089624e-06,
-      "loss": 0.3921,
-      "step": 2179
-    },
-    {
-      "epoch": 0.14247434808182471,
-      "grad_norm": 0.5106698274612427,
-      "learning_rate": 9.980623600300536e-06,
-      "loss": 0.4562,
-      "step": 2180
-    },
-    {
-      "epoch": 0.14253970328736684,
-      "grad_norm": 0.4898502826690674,
-      "learning_rate": 9.980592876219221e-06,
-      "loss": 0.4064,
-      "step": 2181
-    },
-    {
-      "epoch": 0.14260505849290897,
-      "grad_norm": 0.4879530370235443,
-      "learning_rate": 9.98056212784583e-06,
-      "loss": 0.434,
-      "step": 2182
-    },
-    {
-      "epoch": 0.14267041369845107,
-      "grad_norm": 0.46781447529792786,
-      "learning_rate": 9.980531355180512e-06,
-      "loss": 0.3684,
-      "step": 2183
-    },
-    {
-      "epoch": 0.1427357689039932,
-      "grad_norm": 0.47800523042678833,
-      "learning_rate": 9.980500558223415e-06,
-      "loss": 0.4082,
-      "step": 2184
-    },
-    {
-      "epoch": 0.14280112410953533,
-      "grad_norm": 0.45998021960258484,
-      "learning_rate": 9.980469736974694e-06,
-      "loss": 0.3637,
-      "step": 2185
-    },
-    {
-      "epoch": 0.14286647931507745,
-      "grad_norm": 0.48784536123275757,
-      "learning_rate": 9.980438891434495e-06,
-      "loss": 0.3754,
-      "step": 2186
-    },
-    {
-      "epoch": 0.14293183452061956,
-      "grad_norm": 0.4959687292575836,
-      "learning_rate": 9.980408021602971e-06,
-      "loss": 0.4298,
-      "step": 2187
-    },
-    {
-      "epoch": 0.14299718972616168,
-      "grad_norm": 0.8836976289749146,
-      "learning_rate": 9.980377127480272e-06,
-      "loss": 0.4265,
-      "step": 2188
-    },
-    {
-      "epoch": 0.1430625449317038,
-      "grad_norm": 0.49875491857528687,
-      "learning_rate": 9.980346209066546e-06,
-      "loss": 0.4197,
-      "step": 2189
-    },
-    {
-      "epoch": 0.14312790013724594,
-      "grad_norm": 0.5402434468269348,
-      "learning_rate": 9.980315266361949e-06,
-      "loss": 0.3996,
-      "step": 2190
-    },
-    {
-      "epoch": 0.14319325534278804,
-      "grad_norm": 0.499809205532074,
-      "learning_rate": 9.980284299366629e-06,
-      "loss": 0.441,
-      "step": 2191
-    },
-    {
-      "epoch": 0.14325861054833017,
-      "grad_norm": 0.5115283727645874,
-      "learning_rate": 9.980253308080736e-06,
-      "loss": 0.4306,
-      "step": 2192
-    },
-    {
-      "epoch": 0.1433239657538723,
-      "grad_norm": 0.47517943382263184,
-      "learning_rate": 9.980222292504422e-06,
-      "loss": 0.3831,
-      "step": 2193
-    },
-    {
-      "epoch": 0.14338932095941442,
-      "grad_norm": 0.47964057326316833,
-      "learning_rate": 9.98019125263784e-06,
-      "loss": 0.3985,
-      "step": 2194
-    },
-    {
-      "epoch": 0.14345467616495655,
-      "grad_norm": 0.5241397023200989,
-      "learning_rate": 9.980160188481138e-06,
-      "loss": 0.4569,
-      "step": 2195
-    },
-    {
-      "epoch": 0.14352003137049865,
-      "grad_norm": 0.49302685260772705,
-      "learning_rate": 9.980129100034473e-06,
-      "loss": 0.4065,
-      "step": 2196
-    },
-    {
-      "epoch": 0.14358538657604078,
-      "grad_norm": 0.5190097689628601,
-      "learning_rate": 9.98009798729799e-06,
-      "loss": 0.4696,
-      "step": 2197
-    },
-    {
-      "epoch": 0.1436507417815829,
-      "grad_norm": 0.48809418082237244,
-      "learning_rate": 9.980066850271844e-06,
-      "loss": 0.4064,
-      "step": 2198
-    },
-    {
-      "epoch": 0.14371609698712504,
-      "grad_norm": 0.5275015830993652,
-      "learning_rate": 9.980035688956186e-06,
-      "loss": 0.446,
-      "step": 2199
-    },
-    {
-      "epoch": 0.14378145219266714,
-      "grad_norm": 0.4656153619289398,
-      "learning_rate": 9.98000450335117e-06,
-      "loss": 0.4146,
-      "step": 2200
-    },
-    {
-      "epoch": 0.14384680739820926,
-      "grad_norm": 0.5004518628120422,
-      "learning_rate": 9.979973293456949e-06,
-      "loss": 0.3737,
-      "step": 2201
-    },
-    {
-      "epoch": 0.1439121626037514,
-      "grad_norm": 0.5236954689025879,
-      "learning_rate": 9.979942059273669e-06,
-      "loss": 0.4274,
-      "step": 2202
-    },
-    {
-      "epoch": 0.14397751780929352,
-      "grad_norm": 0.5039278268814087,
-      "learning_rate": 9.979910800801489e-06,
-      "loss": 0.4355,
-      "step": 2203
-    },
-    {
-      "epoch": 0.14404287301483562,
-      "grad_norm": 0.4495544135570526,
-      "learning_rate": 9.979879518040558e-06,
-      "loss": 0.3711,
-      "step": 2204
-    },
-    {
-      "epoch": 0.14410822822037775,
-      "grad_norm": 0.4917999804019928,
-      "learning_rate": 9.979848210991028e-06,
-      "loss": 0.413,
-      "step": 2205
-    },
-    {
-      "epoch": 0.14417358342591988,
-      "grad_norm": 0.5010198354721069,
-      "learning_rate": 9.979816879653054e-06,
-      "loss": 0.4008,
-      "step": 2206
-    },
-    {
-      "epoch": 0.144238938631462,
-      "grad_norm": 0.5320481657981873,
-      "learning_rate": 9.97978552402679e-06,
-      "loss": 0.4524,
-      "step": 2207
-    },
-    {
-      "epoch": 0.1443042938370041,
-      "grad_norm": 0.4922701120376587,
-      "learning_rate": 9.979754144112386e-06,
-      "loss": 0.3814,
-      "step": 2208
-    },
-    {
-      "epoch": 0.14436964904254623,
-      "grad_norm": 0.4877546429634094,
-      "learning_rate": 9.979722739909994e-06,
-      "loss": 0.4526,
-      "step": 2209
-    },
-    {
-      "epoch": 0.14443500424808836,
-      "grad_norm": 0.5020317435264587,
-      "learning_rate": 9.97969131141977e-06,
-      "loss": 0.4018,
-      "step": 2210
-    },
-    {
-      "epoch": 0.1445003594536305,
-      "grad_norm": 0.5045342445373535,
-      "learning_rate": 9.979659858641866e-06,
-      "loss": 0.4447,
-      "step": 2211
-    },
-    {
-      "epoch": 0.1445657146591726,
-      "grad_norm": 0.5243743062019348,
-      "learning_rate": 9.979628381576437e-06,
-      "loss": 0.4698,
-      "step": 2212
-    },
-    {
-      "epoch": 0.14463106986471472,
-      "grad_norm": 0.5178640484809875,
-      "learning_rate": 9.979596880223634e-06,
-      "loss": 0.3934,
-      "step": 2213
-    },
-    {
-      "epoch": 0.14469642507025685,
-      "grad_norm": 0.49646425247192383,
-      "learning_rate": 9.979565354583612e-06,
-      "loss": 0.4379,
-      "step": 2214
-    },
-    {
-      "epoch": 0.14476178027579897,
-      "grad_norm": 0.4796517789363861,
-      "learning_rate": 9.979533804656526e-06,
-      "loss": 0.4287,
-      "step": 2215
-    },
-    {
-      "epoch": 0.1448271354813411,
-      "grad_norm": 0.5299208164215088,
-      "learning_rate": 9.979502230442527e-06,
-      "loss": 0.4678,
-      "step": 2216
-    },
-    {
-      "epoch": 0.1448924906868832,
-      "grad_norm": 0.49109676480293274,
-      "learning_rate": 9.979470631941773e-06,
-      "loss": 0.4413,
-      "step": 2217
-    },
-    {
-      "epoch": 0.14495784589242533,
-      "grad_norm": 0.5628079771995544,
-      "learning_rate": 9.979439009154412e-06,
-      "loss": 0.4514,
-      "step": 2218
-    },
-    {
-      "epoch": 0.14502320109796746,
-      "grad_norm": 0.48290780186653137,
-      "learning_rate": 9.979407362080604e-06,
-      "loss": 0.414,
-      "step": 2219
-    },
-    {
-      "epoch": 0.14508855630350959,
-      "grad_norm": 0.521253228187561,
-      "learning_rate": 9.979375690720501e-06,
-      "loss": 0.4535,
-      "step": 2220
-    },
-    {
-      "epoch": 0.1451539115090517,
-      "grad_norm": 0.48355838656425476,
-      "learning_rate": 9.979343995074258e-06,
-      "loss": 0.4312,
-      "step": 2221
-    },
-    {
-      "epoch": 0.14521926671459381,
-      "grad_norm": 0.4963189363479614,
-      "learning_rate": 9.97931227514203e-06,
-      "loss": 0.4179,
-      "step": 2222
-    },
-    {
-      "epoch": 0.14528462192013594,
-      "grad_norm": 0.44014713168144226,
-      "learning_rate": 9.97928053092397e-06,
-      "loss": 0.3873,
-      "step": 2223
-    },
-    {
-      "epoch": 0.14534997712567807,
-      "grad_norm": 0.5127825140953064,
-      "learning_rate": 9.979248762420233e-06,
-      "loss": 0.479,
-      "step": 2224
-    },
-    {
-      "epoch": 0.14541533233122017,
-      "grad_norm": 0.5252187848091125,
-      "learning_rate": 9.979216969630976e-06,
-      "loss": 0.4282,
-      "step": 2225
-    },
-    {
-      "epoch": 0.1454806875367623,
-      "grad_norm": 0.5021201372146606,
-      "learning_rate": 9.979185152556353e-06,
-      "loss": 0.4307,
-      "step": 2226
-    },
-    {
-      "epoch": 0.14554604274230443,
-      "grad_norm": 0.5003250241279602,
-      "learning_rate": 9.979153311196519e-06,
-      "loss": 0.4427,
-      "step": 2227
-    },
-    {
-      "epoch": 0.14561139794784655,
-      "grad_norm": 0.4868018627166748,
-      "learning_rate": 9.979121445551629e-06,
-      "loss": 0.3816,
-      "step": 2228
-    },
-    {
-      "epoch": 0.14567675315338866,
-      "grad_norm": 0.49736616015434265,
-      "learning_rate": 9.979089555621838e-06,
-      "loss": 0.4107,
-      "step": 2229
-    },
-    {
-      "epoch": 0.14574210835893078,
-      "grad_norm": 0.5106949210166931,
-      "learning_rate": 9.979057641407303e-06,
-      "loss": 0.4051,
-      "step": 2230
-    },
-    {
-      "epoch": 0.1458074635644729,
-      "grad_norm": 0.4665818214416504,
-      "learning_rate": 9.979025702908181e-06,
-      "loss": 0.3796,
-      "step": 2231
-    },
-    {
-      "epoch": 0.14587281877001504,
-      "grad_norm": 0.4829004406929016,
-      "learning_rate": 9.978993740124623e-06,
-      "loss": 0.4271,
-      "step": 2232
-    },
-    {
-      "epoch": 0.14593817397555714,
-      "grad_norm": 0.5315292477607727,
-      "learning_rate": 9.978961753056789e-06,
-      "loss": 0.4555,
-      "step": 2233
-    },
-    {
-      "epoch": 0.14600352918109927,
-      "grad_norm": 0.530169665813446,
-      "learning_rate": 9.978929741704832e-06,
-      "loss": 0.435,
-      "step": 2234
-    },
-    {
-      "epoch": 0.1460688843866414,
-      "grad_norm": 0.5203970670700073,
-      "learning_rate": 9.978897706068911e-06,
-      "loss": 0.4168,
-      "step": 2235
-    },
-    {
-      "epoch": 0.14613423959218352,
-      "grad_norm": 0.46462902426719666,
-      "learning_rate": 9.97886564614918e-06,
-      "loss": 0.3774,
-      "step": 2236
-    },
-    {
-      "epoch": 0.14619959479772565,
-      "grad_norm": 0.5217627286911011,
-      "learning_rate": 9.978833561945798e-06,
-      "loss": 0.4302,
-      "step": 2237
-    },
-    {
-      "epoch": 0.14626495000326775,
-      "grad_norm": 0.4850352704524994,
-      "learning_rate": 9.97880145345892e-06,
-      "loss": 0.3647,
-      "step": 2238
-    },
-    {
-      "epoch": 0.14633030520880988,
-      "grad_norm": 0.501953661441803,
-      "learning_rate": 9.978769320688702e-06,
-      "loss": 0.4255,
-      "step": 2239
-    },
-    {
-      "epoch": 0.146395660414352,
-      "grad_norm": 0.5132482647895813,
-      "learning_rate": 9.978737163635301e-06,
-      "loss": 0.4256,
-      "step": 2240
-    },
-    {
-      "epoch": 0.14646101561989414,
-      "grad_norm": 0.5114203691482544,
-      "learning_rate": 9.978704982298875e-06,
-      "loss": 0.4578,
-      "step": 2241
-    },
-    {
-      "epoch": 0.14652637082543624,
-      "grad_norm": 0.5146781206130981,
-      "learning_rate": 9.978672776679578e-06,
-      "loss": 0.4518,
-      "step": 2242
-    },
-    {
-      "epoch": 0.14659172603097836,
-      "grad_norm": 0.4847266972064972,
-      "learning_rate": 9.97864054677757e-06,
-      "loss": 0.4169,
-      "step": 2243
-    },
-    {
-      "epoch": 0.1466570812365205,
-      "grad_norm": 0.5032507181167603,
-      "learning_rate": 9.978608292593007e-06,
-      "loss": 0.428,
-      "step": 2244
-    },
-    {
-      "epoch": 0.14672243644206262,
-      "grad_norm": 0.481487900018692,
-      "learning_rate": 9.978576014126047e-06,
-      "loss": 0.4075,
-      "step": 2245
-    },
-    {
-      "epoch": 0.14678779164760472,
-      "grad_norm": 0.541334867477417,
-      "learning_rate": 9.978543711376847e-06,
-      "loss": 0.4248,
-      "step": 2246
-    },
-    {
-      "epoch": 0.14685314685314685,
-      "grad_norm": 0.4564792215824127,
-      "learning_rate": 9.978511384345565e-06,
-      "loss": 0.3525,
-      "step": 2247
-    },
-    {
-      "epoch": 0.14691850205868898,
-      "grad_norm": 0.5100907683372498,
-      "learning_rate": 9.978479033032357e-06,
-      "loss": 0.4246,
-      "step": 2248
-    },
-    {
-      "epoch": 0.1469838572642311,
-      "grad_norm": 0.5410794019699097,
-      "learning_rate": 9.978446657437383e-06,
-      "loss": 0.5128,
-      "step": 2249
-    },
-    {
-      "epoch": 0.1470492124697732,
-      "grad_norm": 0.5061261653900146,
-      "learning_rate": 9.9784142575608e-06,
-      "loss": 0.379,
-      "step": 2250
-    },
-    {
-      "epoch": 0.14711456767531533,
-      "grad_norm": 0.5119841694831848,
-      "learning_rate": 9.978381833402766e-06,
-      "loss": 0.4102,
-      "step": 2251
-    },
-    {
-      "epoch": 0.14717992288085746,
-      "grad_norm": 0.4679925739765167,
-      "learning_rate": 9.97834938496344e-06,
-      "loss": 0.3811,
-      "step": 2252
-    },
-    {
-      "epoch": 0.1472452780863996,
-      "grad_norm": 0.48318007588386536,
-      "learning_rate": 9.978316912242978e-06,
-      "loss": 0.4114,
-      "step": 2253
-    },
-    {
-      "epoch": 0.1473106332919417,
-      "grad_norm": 0.4982635974884033,
-      "learning_rate": 9.97828441524154e-06,
-      "loss": 0.4602,
-      "step": 2254
-    },
-    {
-      "epoch": 0.14737598849748382,
-      "grad_norm": 0.495976060628891,
-      "learning_rate": 9.978251893959286e-06,
-      "loss": 0.3616,
-      "step": 2255
-    },
-    {
-      "epoch": 0.14744134370302595,
-      "grad_norm": 0.48419514298439026,
-      "learning_rate": 9.97821934839637e-06,
-      "loss": 0.4388,
-      "step": 2256
-    },
-    {
-      "epoch": 0.14750669890856807,
-      "grad_norm": 0.4629019796848297,
-      "learning_rate": 9.978186778552955e-06,
-      "loss": 0.4112,
-      "step": 2257
-    },
-    {
-      "epoch": 0.1475720541141102,
-      "grad_norm": 0.5135029554367065,
-      "learning_rate": 9.978154184429198e-06,
-      "loss": 0.4388,
-      "step": 2258
-    },
-    {
-      "epoch": 0.1476374093196523,
-      "grad_norm": 0.46955379843711853,
-      "learning_rate": 9.97812156602526e-06,
-      "loss": 0.3721,
-      "step": 2259
-    },
-    {
-      "epoch": 0.14770276452519443,
-      "grad_norm": 0.5143701434135437,
-      "learning_rate": 9.978088923341296e-06,
-      "loss": 0.429,
-      "step": 2260
-    },
-    {
-      "epoch": 0.14776811973073656,
-      "grad_norm": 0.49218401312828064,
-      "learning_rate": 9.97805625637747e-06,
-      "loss": 0.4341,
-      "step": 2261
-    },
-    {
-      "epoch": 0.14783347493627869,
-      "grad_norm": 0.512100338935852,
-      "learning_rate": 9.978023565133938e-06,
-      "loss": 0.4368,
-      "step": 2262
-    },
-    {
-      "epoch": 0.1478988301418208,
-      "grad_norm": 0.5482844710350037,
-      "learning_rate": 9.977990849610861e-06,
-      "loss": 0.474,
-      "step": 2263
-    },
-    {
-      "epoch": 0.14796418534736291,
-      "grad_norm": 0.5389716029167175,
-      "learning_rate": 9.977958109808396e-06,
-      "loss": 0.4738,
-      "step": 2264
-    },
-    {
-      "epoch": 0.14802954055290504,
-      "grad_norm": 0.4722362756729126,
-      "learning_rate": 9.977925345726707e-06,
-      "loss": 0.3908,
-      "step": 2265
-    },
-    {
-      "epoch": 0.14809489575844717,
-      "grad_norm": 0.5035797953605652,
-      "learning_rate": 9.977892557365953e-06,
-      "loss": 0.4336,
-      "step": 2266
-    },
-    {
-      "epoch": 0.14816025096398927,
-      "grad_norm": 0.49283385276794434,
-      "learning_rate": 9.97785974472629e-06,
-      "loss": 0.4481,
-      "step": 2267
-    },
-    {
-      "epoch": 0.1482256061695314,
-      "grad_norm": 0.48502737283706665,
-      "learning_rate": 9.977826907807882e-06,
-      "loss": 0.4387,
-      "step": 2268
-    },
-    {
-      "epoch": 0.14829096137507353,
-      "grad_norm": 0.5310791730880737,
-      "learning_rate": 9.977794046610886e-06,
-      "loss": 0.4555,
-      "step": 2269
-    },
-    {
-      "epoch": 0.14835631658061565,
-      "grad_norm": 0.5248449444770813,
-      "learning_rate": 9.977761161135465e-06,
-      "loss": 0.4808,
-      "step": 2270
-    },
-    {
-      "epoch": 0.14842167178615776,
-      "grad_norm": 0.5024116039276123,
-      "learning_rate": 9.977728251381777e-06,
-      "loss": 0.4468,
-      "step": 2271
-    },
-    {
-      "epoch": 0.14848702699169988,
-      "grad_norm": 0.4562566876411438,
-      "learning_rate": 9.977695317349986e-06,
-      "loss": 0.3714,
-      "step": 2272
-    },
-    {
-      "epoch": 0.148552382197242,
-      "grad_norm": 0.6429773569107056,
-      "learning_rate": 9.97766235904025e-06,
-      "loss": 0.4526,
-      "step": 2273
-    },
-    {
-      "epoch": 0.14861773740278414,
-      "grad_norm": 0.5034480690956116,
-      "learning_rate": 9.97762937645273e-06,
-      "loss": 0.4759,
-      "step": 2274
-    },
-    {
-      "epoch": 0.14868309260832624,
-      "grad_norm": 0.479637086391449,
-      "learning_rate": 9.977596369587585e-06,
-      "loss": 0.3865,
-      "step": 2275
-    },
-    {
-      "epoch": 0.14874844781386837,
-      "grad_norm": 0.516732394695282,
-      "learning_rate": 9.977563338444982e-06,
-      "loss": 0.4363,
-      "step": 2276
-    },
-    {
-      "epoch": 0.1488138030194105,
-      "grad_norm": 0.4789265990257263,
-      "learning_rate": 9.977530283025076e-06,
-      "loss": 0.3908,
-      "step": 2277
-    },
-    {
-      "epoch": 0.14887915822495262,
-      "grad_norm": 0.5140530467033386,
-      "learning_rate": 9.97749720332803e-06,
-      "loss": 0.4376,
-      "step": 2278
-    },
-    {
-      "epoch": 0.14894451343049475,
-      "grad_norm": 0.473240464925766,
-      "learning_rate": 9.977464099354006e-06,
-      "loss": 0.357,
-      "step": 2279
-    },
-    {
-      "epoch": 0.14900986863603685,
-      "grad_norm": 0.5226386785507202,
-      "learning_rate": 9.977430971103166e-06,
-      "loss": 0.4535,
-      "step": 2280
-    },
-    {
-      "epoch": 0.14907522384157898,
-      "grad_norm": 0.4788806736469269,
-      "learning_rate": 9.97739781857567e-06,
-      "loss": 0.4008,
-      "step": 2281
-    },
-    {
-      "epoch": 0.1491405790471211,
-      "grad_norm": 0.49951809644699097,
-      "learning_rate": 9.97736464177168e-06,
-      "loss": 0.4339,
-      "step": 2282
-    },
-    {
-      "epoch": 0.14920593425266324,
-      "grad_norm": 0.5056042075157166,
-      "learning_rate": 9.977331440691361e-06,
-      "loss": 0.4592,
-      "step": 2283
-    },
-    {
-      "epoch": 0.14927128945820534,
-      "grad_norm": 0.5108303427696228,
-      "learning_rate": 9.97729821533487e-06,
-      "loss": 0.4575,
-      "step": 2284
-    },
-    {
-      "epoch": 0.14933664466374746,
-      "grad_norm": 0.4839963912963867,
-      "learning_rate": 9.977264965702372e-06,
-      "loss": 0.3743,
-      "step": 2285
-    },
-    {
-      "epoch": 0.1494019998692896,
-      "grad_norm": 0.463399738073349,
-      "learning_rate": 9.977231691794027e-06,
-      "loss": 0.3404,
-      "step": 2286
-    },
-    {
-      "epoch": 0.14946735507483172,
-      "grad_norm": 0.5017335414886475,
-      "learning_rate": 9.977198393610001e-06,
-      "loss": 0.4269,
-      "step": 2287
-    },
-    {
-      "epoch": 0.14953271028037382,
-      "grad_norm": 0.47880029678344727,
-      "learning_rate": 9.977165071150453e-06,
-      "loss": 0.4256,
-      "step": 2288
-    },
-    {
-      "epoch": 0.14959806548591595,
-      "grad_norm": 0.5520682334899902,
-      "learning_rate": 9.977131724415548e-06,
-      "loss": 0.4682,
-      "step": 2289
-    },
-    {
-      "epoch": 0.14966342069145808,
-      "grad_norm": 0.5078201293945312,
-      "learning_rate": 9.977098353405445e-06,
-      "loss": 0.4772,
-      "step": 2290
-    },
-    {
-      "epoch": 0.1497287758970002,
-      "grad_norm": 0.49950456619262695,
-      "learning_rate": 9.97706495812031e-06,
-      "loss": 0.4307,
-      "step": 2291
-    },
-    {
-      "epoch": 0.1497941311025423,
-      "grad_norm": 0.5226721167564392,
-      "learning_rate": 9.977031538560305e-06,
-      "loss": 0.4842,
-      "step": 2292
-    },
-    {
-      "epoch": 0.14985948630808443,
-      "grad_norm": 0.5184908509254456,
-      "learning_rate": 9.976998094725594e-06,
-      "loss": 0.4977,
-      "step": 2293
-    },
-    {
-      "epoch": 0.14992484151362656,
-      "grad_norm": 0.49716198444366455,
-      "learning_rate": 9.976964626616339e-06,
-      "loss": 0.457,
-      "step": 2294
-    },
-    {
-      "epoch": 0.1499901967191687,
-      "grad_norm": 0.48878955841064453,
-      "learning_rate": 9.976931134232702e-06,
-      "loss": 0.4397,
-      "step": 2295
-    },
-    {
-      "epoch": 0.1500555519247108,
-      "grad_norm": 1.3947206735610962,
-      "learning_rate": 9.976897617574849e-06,
-      "loss": 0.4807,
-      "step": 2296
-    },
-    {
-      "epoch": 0.15012090713025292,
-      "grad_norm": 0.4955795109272003,
-      "learning_rate": 9.97686407664294e-06,
-      "loss": 0.4134,
-      "step": 2297
-    },
-    {
-      "epoch": 0.15018626233579505,
-      "grad_norm": 0.5125658512115479,
-      "learning_rate": 9.976830511437143e-06,
-      "loss": 0.4344,
-      "step": 2298
-    },
-    {
-      "epoch": 0.15025161754133717,
-      "grad_norm": 0.5206575989723206,
-      "learning_rate": 9.976796921957619e-06,
-      "loss": 0.4545,
-      "step": 2299
-    },
-    {
-      "epoch": 0.1503169727468793,
-      "grad_norm": 0.518939197063446,
-      "learning_rate": 9.976763308204532e-06,
-      "loss": 0.4508,
-      "step": 2300
-    },
-    {
-      "epoch": 0.1503823279524214,
-      "grad_norm": 0.5151715278625488,
-      "learning_rate": 9.976729670178046e-06,
-      "loss": 0.4028,
-      "step": 2301
-    },
-    {
-      "epoch": 0.15044768315796353,
-      "grad_norm": 0.5385341048240662,
-      "learning_rate": 9.976696007878326e-06,
-      "loss": 0.4299,
-      "step": 2302
-    },
-    {
-      "epoch": 0.15051303836350566,
-      "grad_norm": 0.531333327293396,
-      "learning_rate": 9.976662321305535e-06,
-      "loss": 0.4172,
-      "step": 2303
-    },
-    {
-      "epoch": 0.15057839356904779,
-      "grad_norm": 0.48247775435447693,
-      "learning_rate": 9.976628610459837e-06,
-      "loss": 0.4387,
-      "step": 2304
-    },
-    {
-      "epoch": 0.1506437487745899,
-      "grad_norm": 0.5362417101860046,
-      "learning_rate": 9.976594875341399e-06,
-      "loss": 0.4274,
-      "step": 2305
-    },
-    {
-      "epoch": 0.15070910398013201,
-      "grad_norm": 0.5025076866149902,
-      "learning_rate": 9.976561115950383e-06,
-      "loss": 0.4143,
-      "step": 2306
-    },
-    {
-      "epoch": 0.15077445918567414,
-      "grad_norm": 0.475013792514801,
-      "learning_rate": 9.976527332286953e-06,
-      "loss": 0.4127,
-      "step": 2307
-    },
-    {
-      "epoch": 0.15083981439121627,
-      "grad_norm": 0.4805130958557129,
-      "learning_rate": 9.976493524351276e-06,
-      "loss": 0.3979,
-      "step": 2308
-    },
-    {
-      "epoch": 0.15090516959675837,
-      "grad_norm": 0.5044189691543579,
-      "learning_rate": 9.976459692143516e-06,
-      "loss": 0.4128,
-      "step": 2309
-    },
-    {
-      "epoch": 0.1509705248023005,
-      "grad_norm": 0.4822445809841156,
-      "learning_rate": 9.976425835663839e-06,
-      "loss": 0.3999,
-      "step": 2310
-    },
-    {
-      "epoch": 0.15103588000784263,
-      "grad_norm": 0.5312120914459229,
-      "learning_rate": 9.97639195491241e-06,
-      "loss": 0.4142,
-      "step": 2311
-    },
-    {
-      "epoch": 0.15110123521338475,
-      "grad_norm": 0.4583929181098938,
-      "learning_rate": 9.976358049889392e-06,
-      "loss": 0.3819,
-      "step": 2312
-    },
-    {
-      "epoch": 0.15116659041892685,
-      "grad_norm": 0.51595538854599,
-      "learning_rate": 9.97632412059495e-06,
-      "loss": 0.3849,
-      "step": 2313
-    },
-    {
-      "epoch": 0.15123194562446898,
-      "grad_norm": 0.5065730214118958,
-      "learning_rate": 9.976290167029255e-06,
-      "loss": 0.4276,
-      "step": 2314
-    },
-    {
-      "epoch": 0.1512973008300111,
-      "grad_norm": 0.482515424489975,
-      "learning_rate": 9.976256189192465e-06,
-      "loss": 0.4047,
-      "step": 2315
-    },
-    {
-      "epoch": 0.15136265603555324,
-      "grad_norm": 0.4882115423679352,
-      "learning_rate": 9.976222187084753e-06,
-      "loss": 0.4058,
-      "step": 2316
-    },
-    {
-      "epoch": 0.15142801124109534,
-      "grad_norm": 0.4734536111354828,
-      "learning_rate": 9.97618816070628e-06,
-      "loss": 0.4117,
-      "step": 2317
-    },
-    {
-      "epoch": 0.15149336644663747,
-      "grad_norm": 0.5062746405601501,
-      "learning_rate": 9.976154110057214e-06,
-      "loss": 0.3693,
-      "step": 2318
-    },
-    {
-      "epoch": 0.1515587216521796,
-      "grad_norm": 0.4745713174343109,
-      "learning_rate": 9.976120035137719e-06,
-      "loss": 0.427,
-      "step": 2319
-    },
-    {
-      "epoch": 0.15162407685772172,
-      "grad_norm": 0.5089232325553894,
-      "learning_rate": 9.976085935947966e-06,
-      "loss": 0.4183,
-      "step": 2320
-    },
-    {
-      "epoch": 0.15168943206326385,
-      "grad_norm": 0.5024188160896301,
-      "learning_rate": 9.976051812488115e-06,
-      "loss": 0.4173,
-      "step": 2321
-    },
-    {
-      "epoch": 0.15175478726880595,
-      "grad_norm": 0.5488912463188171,
-      "learning_rate": 9.976017664758335e-06,
-      "loss": 0.4848,
-      "step": 2322
-    },
-    {
-      "epoch": 0.15182014247434808,
-      "grad_norm": 0.49083489179611206,
-      "learning_rate": 9.975983492758796e-06,
-      "loss": 0.4546,
-      "step": 2323
-    },
-    {
-      "epoch": 0.1518854976798902,
-      "grad_norm": 0.4858939051628113,
-      "learning_rate": 9.97594929648966e-06,
-      "loss": 0.456,
-      "step": 2324
-    },
-    {
-      "epoch": 0.15195085288543234,
-      "grad_norm": 0.5023922920227051,
-      "learning_rate": 9.975915075951095e-06,
-      "loss": 0.4422,
-      "step": 2325
-    },
-    {
-      "epoch": 0.15201620809097444,
-      "grad_norm": 0.46519041061401367,
-      "learning_rate": 9.975880831143267e-06,
-      "loss": 0.3725,
-      "step": 2326
-    },
-    {
-      "epoch": 0.15208156329651656,
-      "grad_norm": 0.4633754789829254,
-      "learning_rate": 9.975846562066347e-06,
-      "loss": 0.3691,
-      "step": 2327
-    },
-    {
-      "epoch": 0.1521469185020587,
-      "grad_norm": 0.5417261719703674,
-      "learning_rate": 9.9758122687205e-06,
-      "loss": 0.4732,
-      "step": 2328
-    },
-    {
-      "epoch": 0.15221227370760082,
-      "grad_norm": 0.509924590587616,
-      "learning_rate": 9.975777951105889e-06,
-      "loss": 0.473,
-      "step": 2329
-    },
-    {
-      "epoch": 0.15227762891314292,
-      "grad_norm": 0.500942051410675,
-      "learning_rate": 9.975743609222689e-06,
-      "loss": 0.4007,
-      "step": 2330
-    },
-    {
-      "epoch": 0.15234298411868505,
-      "grad_norm": 0.49725082516670227,
-      "learning_rate": 9.97570924307106e-06,
-      "loss": 0.4541,
-      "step": 2331
-    },
-    {
-      "epoch": 0.15240833932422718,
-      "grad_norm": 0.48673009872436523,
-      "learning_rate": 9.975674852651177e-06,
-      "loss": 0.4245,
-      "step": 2332
-    },
-    {
-      "epoch": 0.1524736945297693,
-      "grad_norm": 0.48869961500167847,
-      "learning_rate": 9.975640437963203e-06,
-      "loss": 0.3738,
-      "step": 2333
-    },
-    {
-      "epoch": 0.1525390497353114,
-      "grad_norm": 0.508194088935852,
-      "learning_rate": 9.975605999007306e-06,
-      "loss": 0.4185,
-      "step": 2334
-    },
-    {
-      "epoch": 0.15260440494085353,
-      "grad_norm": 0.5041077136993408,
-      "learning_rate": 9.975571535783654e-06,
-      "loss": 0.3882,
-      "step": 2335
-    },
-    {
-      "epoch": 0.15266976014639566,
-      "grad_norm": 0.5077170133590698,
-      "learning_rate": 9.975537048292419e-06,
-      "loss": 0.4299,
-      "step": 2336
-    },
-    {
-      "epoch": 0.1527351153519378,
-      "grad_norm": 0.496986448764801,
-      "learning_rate": 9.975502536533762e-06,
-      "loss": 0.3915,
-      "step": 2337
-    },
-    {
-      "epoch": 0.15280047055747992,
-      "grad_norm": 0.5121133327484131,
-      "learning_rate": 9.97546800050786e-06,
-      "loss": 0.4435,
-      "step": 2338
-    },
-    {
-      "epoch": 0.15286582576302202,
-      "grad_norm": 0.48407822847366333,
-      "learning_rate": 9.975433440214874e-06,
-      "loss": 0.3995,
-      "step": 2339
-    },
-    {
-      "epoch": 0.15293118096856415,
-      "grad_norm": 0.5398458242416382,
-      "learning_rate": 9.975398855654976e-06,
-      "loss": 0.4322,
-      "step": 2340
-    },
-    {
-      "epoch": 0.15299653617410627,
-      "grad_norm": 0.43759775161743164,
-      "learning_rate": 9.975364246828334e-06,
-      "loss": 0.3736,
-      "step": 2341
-    },
-    {
-      "epoch": 0.1530618913796484,
-      "grad_norm": 0.5048732757568359,
-      "learning_rate": 9.975329613735117e-06,
-      "loss": 0.4444,
-      "step": 2342
-    },
-    {
-      "epoch": 0.1531272465851905,
-      "grad_norm": 0.453186571598053,
-      "learning_rate": 9.975294956375495e-06,
-      "loss": 0.3525,
-      "step": 2343
-    },
-    {
-      "epoch": 0.15319260179073263,
-      "grad_norm": 0.5275481939315796,
-      "learning_rate": 9.975260274749632e-06,
-      "loss": 0.4464,
-      "step": 2344
-    },
-    {
-      "epoch": 0.15325795699627476,
-      "grad_norm": 0.5292560458183289,
-      "learning_rate": 9.975225568857704e-06,
-      "loss": 0.4778,
-      "step": 2345
-    },
-    {
-      "epoch": 0.15332331220181689,
-      "grad_norm": 0.4985980689525604,
-      "learning_rate": 9.975190838699878e-06,
-      "loss": 0.4261,
-      "step": 2346
-    },
-    {
-      "epoch": 0.153388667407359,
-      "grad_norm": 0.4822975695133209,
-      "learning_rate": 9.975156084276323e-06,
-      "loss": 0.4166,
-      "step": 2347
-    },
-    {
-      "epoch": 0.15345402261290111,
-      "grad_norm": 0.5165229439735413,
-      "learning_rate": 9.975121305587205e-06,
-      "loss": 0.4462,
-      "step": 2348
-    },
-    {
-      "epoch": 0.15351937781844324,
-      "grad_norm": 0.49926477670669556,
-      "learning_rate": 9.9750865026327e-06,
-      "loss": 0.4324,
-      "step": 2349
-    },
-    {
-      "epoch": 0.15358473302398537,
-      "grad_norm": 0.5210674405097961,
-      "learning_rate": 9.975051675412974e-06,
-      "loss": 0.4622,
-      "step": 2350
-    },
-    {
-      "epoch": 0.15365008822952747,
-      "grad_norm": 0.5540010333061218,
-      "learning_rate": 9.975016823928196e-06,
-      "loss": 0.4629,
-      "step": 2351
-    },
-    {
-      "epoch": 0.1537154434350696,
-      "grad_norm": 0.5138182640075684,
-      "learning_rate": 9.974981948178539e-06,
-      "loss": 0.4563,
-      "step": 2352
-    },
-    {
-      "epoch": 0.15378079864061173,
-      "grad_norm": 0.47402670979499817,
-      "learning_rate": 9.97494704816417e-06,
-      "loss": 0.4038,
-      "step": 2353
-    },
-    {
-      "epoch": 0.15384615384615385,
-      "grad_norm": 0.49794459342956543,
-      "learning_rate": 9.974912123885263e-06,
-      "loss": 0.4349,
-      "step": 2354
-    },
-    {
-      "epoch": 0.15391150905169595,
-      "grad_norm": 0.5229313373565674,
-      "learning_rate": 9.974877175341984e-06,
-      "loss": 0.4185,
-      "step": 2355
-    },
-    {
-      "epoch": 0.15397686425723808,
-      "grad_norm": 0.5404794812202454,
-      "learning_rate": 9.974842202534507e-06,
-      "loss": 0.4551,
-      "step": 2356
-    },
-    {
-      "epoch": 0.1540422194627802,
-      "grad_norm": 0.5233075022697449,
-      "learning_rate": 9.974807205463e-06,
-      "loss": 0.4446,
-      "step": 2357
-    },
-    {
-      "epoch": 0.15410757466832234,
-      "grad_norm": 0.5049179196357727,
-      "learning_rate": 9.974772184127636e-06,
-      "loss": 0.4438,
-      "step": 2358
-    },
-    {
-      "epoch": 0.15417292987386447,
-      "grad_norm": 0.6475977897644043,
-      "learning_rate": 9.974737138528584e-06,
-      "loss": 0.4383,
-      "step": 2359
-    },
-    {
-      "epoch": 0.15423828507940657,
-      "grad_norm": 0.5115492343902588,
-      "learning_rate": 9.974702068666015e-06,
-      "loss": 0.4222,
-      "step": 2360
-    },
-    {
-      "epoch": 0.1543036402849487,
-      "grad_norm": 0.506045937538147,
-      "learning_rate": 9.974666974540101e-06,
-      "loss": 0.4222,
-      "step": 2361
-    },
-    {
-      "epoch": 0.15436899549049082,
-      "grad_norm": 0.5062224268913269,
-      "learning_rate": 9.974631856151014e-06,
-      "loss": 0.4477,
-      "step": 2362
-    },
-    {
-      "epoch": 0.15443435069603295,
-      "grad_norm": 0.5204629898071289,
-      "learning_rate": 9.974596713498921e-06,
-      "loss": 0.4422,
-      "step": 2363
-    },
-    {
-      "epoch": 0.15449970590157505,
-      "grad_norm": 0.4689907133579254,
-      "learning_rate": 9.974561546583999e-06,
-      "loss": 0.3657,
-      "step": 2364
-    },
-    {
-      "epoch": 0.15456506110711718,
-      "grad_norm": 0.5130204558372498,
-      "learning_rate": 9.974526355406417e-06,
-      "loss": 0.3866,
-      "step": 2365
-    },
-    {
-      "epoch": 0.1546304163126593,
-      "grad_norm": 0.5153915286064148,
-      "learning_rate": 9.974491139966346e-06,
-      "loss": 0.4234,
-      "step": 2366
-    },
-    {
-      "epoch": 0.15469577151820144,
-      "grad_norm": 0.49781179428100586,
-      "learning_rate": 9.974455900263958e-06,
-      "loss": 0.4006,
-      "step": 2367
-    },
-    {
-      "epoch": 0.15476112672374354,
-      "grad_norm": 0.5339179635047913,
-      "learning_rate": 9.974420636299427e-06,
-      "loss": 0.4446,
-      "step": 2368
-    },
-    {
-      "epoch": 0.15482648192928566,
-      "grad_norm": 0.5743839144706726,
-      "learning_rate": 9.97438534807292e-06,
-      "loss": 0.4957,
-      "step": 2369
-    },
-    {
-      "epoch": 0.1548918371348278,
-      "grad_norm": 0.5327855348587036,
-      "learning_rate": 9.974350035584615e-06,
-      "loss": 0.4267,
-      "step": 2370
-    },
-    {
-      "epoch": 0.15495719234036992,
-      "grad_norm": 0.5445069670677185,
-      "learning_rate": 9.974314698834682e-06,
-      "loss": 0.4871,
-      "step": 2371
-    },
-    {
-      "epoch": 0.15502254754591202,
-      "grad_norm": 0.5317440032958984,
-      "learning_rate": 9.97427933782329e-06,
-      "loss": 0.4293,
-      "step": 2372
-    },
-    {
-      "epoch": 0.15508790275145415,
-      "grad_norm": 0.514540433883667,
-      "learning_rate": 9.974243952550617e-06,
-      "loss": 0.382,
-      "step": 2373
-    },
-    {
-      "epoch": 0.15515325795699628,
-      "grad_norm": 0.46227580308914185,
-      "learning_rate": 9.974208543016833e-06,
-      "loss": 0.3745,
-      "step": 2374
-    },
-    {
-      "epoch": 0.1552186131625384,
-      "grad_norm": 0.48030710220336914,
-      "learning_rate": 9.97417310922211e-06,
-      "loss": 0.3821,
-      "step": 2375
-    },
-    {
-      "epoch": 0.1552839683680805,
-      "grad_norm": 0.5295577049255371,
-      "learning_rate": 9.97413765116662e-06,
-      "loss": 0.4289,
-      "step": 2376
-    },
-    {
-      "epoch": 0.15534932357362263,
-      "grad_norm": 0.49830979108810425,
-      "learning_rate": 9.97410216885054e-06,
-      "loss": 0.358,
-      "step": 2377
-    },
-    {
-      "epoch": 0.15541467877916476,
-      "grad_norm": 0.54817795753479,
-      "learning_rate": 9.974066662274038e-06,
-      "loss": 0.4759,
-      "step": 2378
-    },
-    {
-      "epoch": 0.1554800339847069,
-      "grad_norm": 0.47363972663879395,
-      "learning_rate": 9.974031131437292e-06,
-      "loss": 0.3983,
-      "step": 2379
-    },
-    {
-      "epoch": 0.15554538919024902,
-      "grad_norm": 0.5195755958557129,
-      "learning_rate": 9.973995576340471e-06,
-      "loss": 0.441,
-      "step": 2380
-    },
-    {
-      "epoch": 0.15561074439579112,
-      "grad_norm": 0.477298378944397,
-      "learning_rate": 9.973959996983753e-06,
-      "loss": 0.4067,
-      "step": 2381
-    },
-    {
-      "epoch": 0.15567609960133325,
-      "grad_norm": 0.5366641879081726,
-      "learning_rate": 9.973924393367307e-06,
-      "loss": 0.4643,
-      "step": 2382
-    },
-    {
-      "epoch": 0.15574145480687537,
-      "grad_norm": 0.5306047201156616,
-      "learning_rate": 9.97388876549131e-06,
-      "loss": 0.446,
-      "step": 2383
-    },
-    {
-      "epoch": 0.1558068100124175,
-      "grad_norm": 0.5169602036476135,
-      "learning_rate": 9.973853113355933e-06,
-      "loss": 0.4324,
-      "step": 2384
-    },
-    {
-      "epoch": 0.1558721652179596,
-      "grad_norm": 0.4701854884624481,
-      "learning_rate": 9.973817436961352e-06,
-      "loss": 0.396,
-      "step": 2385
-    },
-    {
-      "epoch": 0.15593752042350173,
-      "grad_norm": 0.5466340780258179,
-      "learning_rate": 9.973781736307739e-06,
-      "loss": 0.4145,
-      "step": 2386
-    },
-    {
-      "epoch": 0.15600287562904386,
-      "grad_norm": 0.4791712164878845,
-      "learning_rate": 9.973746011395271e-06,
-      "loss": 0.3664,
-      "step": 2387
-    },
-    {
-      "epoch": 0.15606823083458599,
-      "grad_norm": 0.5368568301200867,
-      "learning_rate": 9.97371026222412e-06,
-      "loss": 0.436,
-      "step": 2388
-    },
-    {
-      "epoch": 0.1561335860401281,
-      "grad_norm": 0.5561971664428711,
-      "learning_rate": 9.973674488794462e-06,
-      "loss": 0.4419,
-      "step": 2389
-    },
-    {
-      "epoch": 0.15619894124567021,
-      "grad_norm": 0.45606914162635803,
-      "learning_rate": 9.973638691106468e-06,
-      "loss": 0.3394,
-      "step": 2390
-    },
-    {
-      "epoch": 0.15626429645121234,
-      "grad_norm": 0.4836057126522064,
-      "learning_rate": 9.973602869160317e-06,
-      "loss": 0.4044,
-      "step": 2391
-    },
-    {
-      "epoch": 0.15632965165675447,
-      "grad_norm": 0.525128960609436,
-      "learning_rate": 9.97356702295618e-06,
-      "loss": 0.4277,
-      "step": 2392
-    },
-    {
-      "epoch": 0.15639500686229657,
-      "grad_norm": 0.5319295525550842,
-      "learning_rate": 9.973531152494237e-06,
-      "loss": 0.4373,
-      "step": 2393
-    },
-    {
-      "epoch": 0.1564603620678387,
-      "grad_norm": 0.5011122822761536,
-      "learning_rate": 9.973495257774657e-06,
-      "loss": 0.4233,
-      "step": 2394
-    },
-    {
-      "epoch": 0.15652571727338083,
-      "grad_norm": 0.5131736993789673,
-      "learning_rate": 9.973459338797619e-06,
-      "loss": 0.4591,
-      "step": 2395
-    },
-    {
-      "epoch": 0.15659107247892295,
-      "grad_norm": 0.495587021112442,
-      "learning_rate": 9.973423395563295e-06,
-      "loss": 0.4431,
-      "step": 2396
-    },
-    {
-      "epoch": 0.15665642768446505,
-      "grad_norm": 0.5457156896591187,
-      "learning_rate": 9.973387428071864e-06,
-      "loss": 0.461,
-      "step": 2397
-    },
-    {
-      "epoch": 0.15672178289000718,
-      "grad_norm": 0.49077412486076355,
-      "learning_rate": 9.973351436323498e-06,
-      "loss": 0.4373,
-      "step": 2398
-    },
-    {
-      "epoch": 0.1567871380955493,
-      "grad_norm": 0.4816787838935852,
-      "learning_rate": 9.973315420318375e-06,
-      "loss": 0.4248,
-      "step": 2399
-    },
-    {
-      "epoch": 0.15685249330109144,
-      "grad_norm": 0.468403160572052,
-      "learning_rate": 9.973279380056669e-06,
-      "loss": 0.3655,
-      "step": 2400
-    },
-    {
-      "epoch": 0.15691784850663357,
-      "grad_norm": 0.4903174340724945,
-      "learning_rate": 9.973243315538559e-06,
-      "loss": 0.3717,
-      "step": 2401
-    },
-    {
-      "epoch": 0.15698320371217567,
-      "grad_norm": 0.5325028300285339,
-      "learning_rate": 9.973207226764215e-06,
-      "loss": 0.4761,
-      "step": 2402
-    },
-    {
-      "epoch": 0.1570485589177178,
-      "grad_norm": 0.5138307213783264,
-      "learning_rate": 9.97317111373382e-06,
-      "loss": 0.4406,
-      "step": 2403
-    },
-    {
-      "epoch": 0.15711391412325992,
-      "grad_norm": 0.48092320561408997,
-      "learning_rate": 9.973134976447543e-06,
-      "loss": 0.4219,
-      "step": 2404
-    },
-    {
-      "epoch": 0.15717926932880205,
-      "grad_norm": 0.4826708436012268,
-      "learning_rate": 9.973098814905566e-06,
-      "loss": 0.4325,
-      "step": 2405
-    },
-    {
-      "epoch": 0.15724462453434415,
-      "grad_norm": 0.4795357584953308,
-      "learning_rate": 9.973062629108064e-06,
-      "loss": 0.3773,
-      "step": 2406
-    },
-    {
-      "epoch": 0.15730997973988628,
-      "grad_norm": 0.46927645802497864,
-      "learning_rate": 9.97302641905521e-06,
-      "loss": 0.4059,
-      "step": 2407
-    },
-    {
-      "epoch": 0.1573753349454284,
-      "grad_norm": 0.5348656177520752,
-      "learning_rate": 9.972990184747185e-06,
-      "loss": 0.4843,
-      "step": 2408
-    },
-    {
-      "epoch": 0.15744069015097054,
-      "grad_norm": 0.48582735657691956,
-      "learning_rate": 9.972953926184164e-06,
-      "loss": 0.4215,
-      "step": 2409
-    },
-    {
-      "epoch": 0.15750604535651264,
-      "grad_norm": 0.5237491130828857,
-      "learning_rate": 9.972917643366325e-06,
-      "loss": 0.478,
-      "step": 2410
-    },
-    {
-      "epoch": 0.15757140056205476,
-      "grad_norm": 0.47684186697006226,
-      "learning_rate": 9.97288133629384e-06,
-      "loss": 0.4309,
-      "step": 2411
-    },
-    {
-      "epoch": 0.1576367557675969,
-      "grad_norm": 0.4899986982345581,
-      "learning_rate": 9.972845004966895e-06,
-      "loss": 0.402,
-      "step": 2412
-    },
-    {
-      "epoch": 0.15770211097313902,
-      "grad_norm": 0.48216915130615234,
-      "learning_rate": 9.972808649385658e-06,
-      "loss": 0.4335,
-      "step": 2413
-    },
-    {
-      "epoch": 0.15776746617868112,
-      "grad_norm": 0.49452242255210876,
-      "learning_rate": 9.972772269550313e-06,
-      "loss": 0.4367,
-      "step": 2414
-    },
-    {
-      "epoch": 0.15783282138422325,
-      "grad_norm": 0.515508770942688,
-      "learning_rate": 9.972735865461034e-06,
-      "loss": 0.4591,
-      "step": 2415
-    },
-    {
-      "epoch": 0.15789817658976538,
-      "grad_norm": 0.48209694027900696,
-      "learning_rate": 9.972699437118e-06,
-      "loss": 0.4206,
-      "step": 2416
-    },
-    {
-      "epoch": 0.1579635317953075,
-      "grad_norm": 0.4766177535057068,
-      "learning_rate": 9.972662984521388e-06,
-      "loss": 0.3985,
-      "step": 2417
-    },
-    {
-      "epoch": 0.1580288870008496,
-      "grad_norm": 0.511594831943512,
-      "learning_rate": 9.972626507671375e-06,
-      "loss": 0.4561,
-      "step": 2418
-    },
-    {
-      "epoch": 0.15809424220639173,
-      "grad_norm": 0.4937410354614258,
-      "learning_rate": 9.972590006568142e-06,
-      "loss": 0.392,
-      "step": 2419
-    },
-    {
-      "epoch": 0.15815959741193386,
-      "grad_norm": 0.479414165019989,
-      "learning_rate": 9.972553481211862e-06,
-      "loss": 0.4053,
-      "step": 2420
-    },
-    {
-      "epoch": 0.158224952617476,
-      "grad_norm": 0.5139066576957703,
-      "learning_rate": 9.972516931602718e-06,
-      "loss": 0.4514,
-      "step": 2421
-    },
-    {
-      "epoch": 0.15829030782301812,
-      "grad_norm": 0.5182710886001587,
-      "learning_rate": 9.972480357740886e-06,
-      "loss": 0.4503,
-      "step": 2422
-    },
-    {
-      "epoch": 0.15835566302856022,
-      "grad_norm": 0.5157301425933838,
-      "learning_rate": 9.972443759626544e-06,
-      "loss": 0.4582,
-      "step": 2423
-    },
-    {
-      "epoch": 0.15842101823410235,
-      "grad_norm": 0.48315101861953735,
-      "learning_rate": 9.972407137259872e-06,
-      "loss": 0.3745,
-      "step": 2424
-    },
-    {
-      "epoch": 0.15848637343964447,
-      "grad_norm": 0.5033113956451416,
-      "learning_rate": 9.972370490641047e-06,
-      "loss": 0.3872,
-      "step": 2425
-    },
-    {
-      "epoch": 0.1585517286451866,
-      "grad_norm": 0.4903390407562256,
-      "learning_rate": 9.972333819770248e-06,
-      "loss": 0.394,
-      "step": 2426
-    },
-    {
-      "epoch": 0.1586170838507287,
-      "grad_norm": 0.4640309512615204,
-      "learning_rate": 9.972297124647657e-06,
-      "loss": 0.4091,
-      "step": 2427
-    },
-    {
-      "epoch": 0.15868243905627083,
-      "grad_norm": 0.48285433650016785,
-      "learning_rate": 9.972260405273447e-06,
-      "loss": 0.4245,
-      "step": 2428
-    },
-    {
-      "epoch": 0.15874779426181296,
-      "grad_norm": 0.5143033862113953,
-      "learning_rate": 9.972223661647802e-06,
-      "loss": 0.4609,
-      "step": 2429
-    },
-    {
-      "epoch": 0.15881314946735509,
-      "grad_norm": 0.47964897751808167,
-      "learning_rate": 9.972186893770901e-06,
-      "loss": 0.3741,
-      "step": 2430
-    },
-    {
-      "epoch": 0.1588785046728972,
-      "grad_norm": 0.5177162289619446,
-      "learning_rate": 9.97215010164292e-06,
-      "loss": 0.4548,
-      "step": 2431
-    },
-    {
-      "epoch": 0.15894385987843931,
-      "grad_norm": 0.501973032951355,
-      "learning_rate": 9.972113285264042e-06,
-      "loss": 0.4082,
-      "step": 2432
-    },
-    {
-      "epoch": 0.15900921508398144,
-      "grad_norm": 0.45796096324920654,
-      "learning_rate": 9.972076444634443e-06,
-      "loss": 0.3939,
-      "step": 2433
-    },
-    {
-      "epoch": 0.15907457028952357,
-      "grad_norm": 0.48177218437194824,
-      "learning_rate": 9.972039579754305e-06,
-      "loss": 0.3938,
-      "step": 2434
-    },
-    {
-      "epoch": 0.15913992549506567,
-      "grad_norm": 0.4820548892021179,
-      "learning_rate": 9.97200269062381e-06,
-      "loss": 0.4491,
-      "step": 2435
-    },
-    {
-      "epoch": 0.1592052807006078,
-      "grad_norm": 0.4811043441295624,
-      "learning_rate": 9.971965777243133e-06,
-      "loss": 0.3995,
-      "step": 2436
-    },
-    {
-      "epoch": 0.15927063590614993,
-      "grad_norm": 0.8044670820236206,
-      "learning_rate": 9.971928839612458e-06,
-      "loss": 0.4068,
-      "step": 2437
-    },
-    {
-      "epoch": 0.15933599111169205,
-      "grad_norm": 0.5364180207252502,
-      "learning_rate": 9.971891877731961e-06,
-      "loss": 0.4235,
-      "step": 2438
-    },
-    {
-      "epoch": 0.15940134631723415,
-      "grad_norm": 0.4931538999080658,
-      "learning_rate": 9.971854891601828e-06,
-      "loss": 0.3894,
-      "step": 2439
-    },
-    {
-      "epoch": 0.15946670152277628,
-      "grad_norm": 0.5401843786239624,
-      "learning_rate": 9.971817881222234e-06,
-      "loss": 0.4618,
-      "step": 2440
-    },
-    {
-      "epoch": 0.1595320567283184,
-      "grad_norm": 0.4642857015132904,
-      "learning_rate": 9.971780846593363e-06,
-      "loss": 0.3731,
-      "step": 2441
-    },
-    {
-      "epoch": 0.15959741193386054,
-      "grad_norm": 0.5497761964797974,
-      "learning_rate": 9.971743787715393e-06,
-      "loss": 0.4824,
-      "step": 2442
-    },
-    {
-      "epoch": 0.15966276713940267,
-      "grad_norm": 0.5208721160888672,
-      "learning_rate": 9.971706704588509e-06,
-      "loss": 0.4415,
-      "step": 2443
-    },
-    {
-      "epoch": 0.15972812234494477,
-      "grad_norm": 0.45102331042289734,
-      "learning_rate": 9.971669597212887e-06,
-      "loss": 0.3535,
-      "step": 2444
-    },
-    {
-      "epoch": 0.1597934775504869,
-      "grad_norm": 0.5041860342025757,
-      "learning_rate": 9.971632465588709e-06,
-      "loss": 0.4654,
-      "step": 2445
-    },
-    {
-      "epoch": 0.15985883275602902,
-      "grad_norm": 0.5482191443443298,
-      "learning_rate": 9.97159530971616e-06,
-      "loss": 0.4445,
-      "step": 2446
-    },
-    {
-      "epoch": 0.15992418796157115,
-      "grad_norm": 0.5334155559539795,
-      "learning_rate": 9.971558129595415e-06,
-      "loss": 0.4317,
-      "step": 2447
-    },
-    {
-      "epoch": 0.15998954316711325,
-      "grad_norm": 0.4832179546356201,
-      "learning_rate": 9.971520925226662e-06,
-      "loss": 0.4177,
-      "step": 2448
-    },
-    {
-      "epoch": 0.16005489837265538,
-      "grad_norm": 0.5031760931015015,
-      "learning_rate": 9.971483696610078e-06,
-      "loss": 0.4143,
-      "step": 2449
-    },
-    {
-      "epoch": 0.1601202535781975,
-      "grad_norm": 0.49227967858314514,
-      "learning_rate": 9.971446443745845e-06,
-      "loss": 0.4254,
-      "step": 2450
-    },
-    {
-      "epoch": 0.16018560878373964,
-      "grad_norm": 0.5087870359420776,
-      "learning_rate": 9.971409166634144e-06,
-      "loss": 0.4446,
-      "step": 2451
-    },
-    {
-      "epoch": 0.16025096398928174,
-      "grad_norm": 0.5398125648498535,
-      "learning_rate": 9.971371865275162e-06,
-      "loss": 0.4469,
-      "step": 2452
-    },
-    {
-      "epoch": 0.16031631919482386,
-      "grad_norm": 0.4663192629814148,
-      "learning_rate": 9.971334539669075e-06,
-      "loss": 0.3717,
-      "step": 2453
-    },
-    {
-      "epoch": 0.160381674400366,
-      "grad_norm": 0.4872797429561615,
-      "learning_rate": 9.971297189816068e-06,
-      "loss": 0.3952,
-      "step": 2454
-    },
-    {
-      "epoch": 0.16044702960590812,
-      "grad_norm": 0.4967391788959503,
-      "learning_rate": 9.971259815716322e-06,
-      "loss": 0.388,
-      "step": 2455
-    },
-    {
-      "epoch": 0.16051238481145022,
-      "grad_norm": 0.4867062568664551,
-      "learning_rate": 9.971222417370019e-06,
-      "loss": 0.3991,
-      "step": 2456
-    },
-    {
-      "epoch": 0.16057774001699235,
-      "grad_norm": 0.47665491700172424,
-      "learning_rate": 9.971184994777342e-06,
-      "loss": 0.3841,
-      "step": 2457
-    },
-    {
-      "epoch": 0.16064309522253448,
-      "grad_norm": 0.5055465698242188,
-      "learning_rate": 9.971147547938475e-06,
-      "loss": 0.3978,
-      "step": 2458
-    },
-    {
-      "epoch": 0.1607084504280766,
-      "grad_norm": 0.4906710386276245,
-      "learning_rate": 9.971110076853599e-06,
-      "loss": 0.3905,
-      "step": 2459
-    },
-    {
-      "epoch": 0.1607738056336187,
-      "grad_norm": 0.486117959022522,
-      "learning_rate": 9.971072581522894e-06,
-      "loss": 0.3927,
-      "step": 2460
-    },
-    {
-      "epoch": 0.16083916083916083,
-      "grad_norm": 0.4956723749637604,
-      "learning_rate": 9.971035061946549e-06,
-      "loss": 0.4235,
-      "step": 2461
-    },
-    {
-      "epoch": 0.16090451604470296,
-      "grad_norm": 0.4936509132385254,
-      "learning_rate": 9.970997518124743e-06,
-      "loss": 0.4027,
-      "step": 2462
-    },
-    {
-      "epoch": 0.1609698712502451,
-      "grad_norm": 0.47628098726272583,
-      "learning_rate": 9.97095995005766e-06,
-      "loss": 0.4145,
-      "step": 2463
-    },
-    {
-      "epoch": 0.16103522645578722,
-      "grad_norm": 0.5037341713905334,
-      "learning_rate": 9.970922357745483e-06,
-      "loss": 0.433,
-      "step": 2464
-    },
-    {
-      "epoch": 0.16110058166132932,
-      "grad_norm": 0.521263599395752,
-      "learning_rate": 9.970884741188397e-06,
-      "loss": 0.4394,
-      "step": 2465
-    },
-    {
-      "epoch": 0.16116593686687145,
-      "grad_norm": 0.4817979633808136,
-      "learning_rate": 9.970847100386582e-06,
-      "loss": 0.4414,
-      "step": 2466
-    },
-    {
-      "epoch": 0.16123129207241357,
-      "grad_norm": 0.4828839600086212,
-      "learning_rate": 9.970809435340226e-06,
-      "loss": 0.4241,
-      "step": 2467
-    },
-    {
-      "epoch": 0.1612966472779557,
-      "grad_norm": 0.5036912560462952,
-      "learning_rate": 9.970771746049508e-06,
-      "loss": 0.4186,
-      "step": 2468
-    },
-    {
-      "epoch": 0.1613620024834978,
-      "grad_norm": 0.5398948192596436,
-      "learning_rate": 9.970734032514616e-06,
-      "loss": 0.4558,
-      "step": 2469
-    },
-    {
-      "epoch": 0.16142735768903993,
-      "grad_norm": 0.5140867829322815,
-      "learning_rate": 9.97069629473573e-06,
-      "loss": 0.4256,
-      "step": 2470
-    },
-    {
-      "epoch": 0.16149271289458206,
-      "grad_norm": 0.5130822062492371,
-      "learning_rate": 9.970658532713038e-06,
-      "loss": 0.4516,
-      "step": 2471
-    },
-    {
-      "epoch": 0.16155806810012419,
-      "grad_norm": 0.4797975718975067,
-      "learning_rate": 9.970620746446721e-06,
-      "loss": 0.4297,
-      "step": 2472
-    },
-    {
-      "epoch": 0.16162342330566629,
-      "grad_norm": 0.4829849600791931,
-      "learning_rate": 9.970582935936966e-06,
-      "loss": 0.3745,
-      "step": 2473
-    },
-    {
-      "epoch": 0.16168877851120841,
-      "grad_norm": 0.5111587047576904,
-      "learning_rate": 9.970545101183956e-06,
-      "loss": 0.4004,
-      "step": 2474
-    },
-    {
-      "epoch": 0.16175413371675054,
-      "grad_norm": 0.5087136626243591,
-      "learning_rate": 9.970507242187873e-06,
-      "loss": 0.4819,
-      "step": 2475
-    },
-    {
-      "epoch": 0.16181948892229267,
-      "grad_norm": 0.5141459107398987,
-      "learning_rate": 9.970469358948906e-06,
-      "loss": 0.4173,
-      "step": 2476
-    },
-    {
-      "epoch": 0.16188484412783477,
-      "grad_norm": 0.46680575609207153,
-      "learning_rate": 9.97043145146724e-06,
-      "loss": 0.4384,
-      "step": 2477
-    },
-    {
-      "epoch": 0.1619501993333769,
-      "grad_norm": 0.4982910752296448,
-      "learning_rate": 9.970393519743055e-06,
-      "loss": 0.4524,
-      "step": 2478
-    },
-    {
-      "epoch": 0.16201555453891903,
-      "grad_norm": 0.48790621757507324,
-      "learning_rate": 9.97035556377654e-06,
-      "loss": 0.4261,
-      "step": 2479
-    },
-    {
-      "epoch": 0.16208090974446115,
-      "grad_norm": 0.5400761365890503,
-      "learning_rate": 9.970317583567879e-06,
-      "loss": 0.45,
-      "step": 2480
-    },
-    {
-      "epoch": 0.16214626495000325,
-      "grad_norm": 0.5190091133117676,
-      "learning_rate": 9.970279579117257e-06,
-      "loss": 0.4153,
-      "step": 2481
-    },
-    {
-      "epoch": 0.16221162015554538,
-      "grad_norm": 0.48533931374549866,
-      "learning_rate": 9.970241550424859e-06,
-      "loss": 0.4143,
-      "step": 2482
-    },
-    {
-      "epoch": 0.1622769753610875,
-      "grad_norm": 0.4670906662940979,
-      "learning_rate": 9.970203497490873e-06,
-      "loss": 0.383,
-      "step": 2483
-    },
-    {
-      "epoch": 0.16234233056662964,
-      "grad_norm": 0.49258172512054443,
-      "learning_rate": 9.970165420315481e-06,
-      "loss": 0.4109,
-      "step": 2484
-    },
-    {
-      "epoch": 0.16240768577217177,
-      "grad_norm": 0.5390580892562866,
-      "learning_rate": 9.97012731889887e-06,
-      "loss": 0.4761,
-      "step": 2485
-    },
-    {
-      "epoch": 0.16247304097771387,
-      "grad_norm": 0.46807584166526794,
-      "learning_rate": 9.970089193241229e-06,
-      "loss": 0.397,
-      "step": 2486
-    },
-    {
-      "epoch": 0.162538396183256,
-      "grad_norm": 0.5004175901412964,
-      "learning_rate": 9.97005104334274e-06,
-      "loss": 0.4167,
-      "step": 2487
-    },
-    {
-      "epoch": 0.16260375138879812,
-      "grad_norm": 0.4912553131580353,
-      "learning_rate": 9.97001286920359e-06,
-      "loss": 0.3806,
-      "step": 2488
-    },
-    {
-      "epoch": 0.16266910659434025,
-      "grad_norm": 0.5068600177764893,
-      "learning_rate": 9.969974670823963e-06,
-      "loss": 0.4448,
-      "step": 2489
-    },
-    {
-      "epoch": 0.16273446179988235,
-      "grad_norm": 0.5258041024208069,
-      "learning_rate": 9.969936448204051e-06,
-      "loss": 0.4242,
-      "step": 2490
-    },
-    {
-      "epoch": 0.16279981700542448,
-      "grad_norm": 0.5120575428009033,
-      "learning_rate": 9.969898201344037e-06,
-      "loss": 0.4393,
-      "step": 2491
-    },
-    {
-      "epoch": 0.1628651722109666,
-      "grad_norm": 0.5244059562683105,
-      "learning_rate": 9.969859930244106e-06,
-      "loss": 0.4359,
-      "step": 2492
-    },
-    {
-      "epoch": 0.16293052741650874,
-      "grad_norm": 0.5091240406036377,
-      "learning_rate": 9.969821634904447e-06,
-      "loss": 0.413,
-      "step": 2493
-    },
-    {
-      "epoch": 0.16299588262205084,
-      "grad_norm": 0.5374748706817627,
-      "learning_rate": 9.969783315325246e-06,
-      "loss": 0.4473,
-      "step": 2494
-    },
-    {
-      "epoch": 0.16306123782759296,
-      "grad_norm": 0.4869825839996338,
-      "learning_rate": 9.969744971506691e-06,
-      "loss": 0.3928,
-      "step": 2495
-    },
-    {
-      "epoch": 0.1631265930331351,
-      "grad_norm": 0.49428319931030273,
-      "learning_rate": 9.969706603448967e-06,
-      "loss": 0.402,
-      "step": 2496
-    },
-    {
-      "epoch": 0.16319194823867722,
-      "grad_norm": 0.43656015396118164,
-      "learning_rate": 9.969668211152264e-06,
-      "loss": 0.3381,
-      "step": 2497
-    },
-    {
-      "epoch": 0.16325730344421932,
-      "grad_norm": 0.4965074360370636,
-      "learning_rate": 9.969629794616766e-06,
-      "loss": 0.3864,
-      "step": 2498
-    },
-    {
-      "epoch": 0.16332265864976145,
-      "grad_norm": 0.5049904584884644,
-      "learning_rate": 9.969591353842663e-06,
-      "loss": 0.4281,
-      "step": 2499
-    },
-    {
-      "epoch": 0.16338801385530358,
-      "grad_norm": 0.523197591304779,
-      "learning_rate": 9.96955288883014e-06,
-      "loss": 0.4588,
-      "step": 2500
-    },
-    {
-      "epoch": 0.1634533690608457,
-      "grad_norm": 0.5229784250259399,
-      "learning_rate": 9.969514399579385e-06,
-      "loss": 0.4401,
-      "step": 2501
-    },
-    {
-      "epoch": 0.1635187242663878,
-      "grad_norm": 0.5095672011375427,
-      "learning_rate": 9.96947588609059e-06,
-      "loss": 0.4417,
-      "step": 2502
-    },
-    {
-      "epoch": 0.16358407947192993,
-      "grad_norm": 0.4908083975315094,
-      "learning_rate": 9.969437348363937e-06,
-      "loss": 0.434,
-      "step": 2503
-    },
-    {
-      "epoch": 0.16364943467747206,
-      "grad_norm": 0.5618243217468262,
-      "learning_rate": 9.969398786399616e-06,
-      "loss": 0.48,
-      "step": 2504
-    },
-    {
-      "epoch": 0.1637147898830142,
-      "grad_norm": 0.5383886694908142,
-      "learning_rate": 9.969360200197818e-06,
-      "loss": 0.4726,
-      "step": 2505
-    },
-    {
-      "epoch": 0.16378014508855632,
-      "grad_norm": 0.48987555503845215,
-      "learning_rate": 9.969321589758726e-06,
-      "loss": 0.4145,
-      "step": 2506
-    },
-    {
-      "epoch": 0.16384550029409842,
-      "grad_norm": 0.5051498413085938,
-      "learning_rate": 9.969282955082534e-06,
-      "loss": 0.436,
-      "step": 2507
-    },
-    {
-      "epoch": 0.16391085549964055,
-      "grad_norm": 0.4904925525188446,
-      "learning_rate": 9.969244296169425e-06,
-      "loss": 0.4305,
-      "step": 2508
-    },
-    {
-      "epoch": 0.16397621070518267,
-      "grad_norm": 0.44761115312576294,
-      "learning_rate": 9.969205613019592e-06,
-      "loss": 0.3561,
-      "step": 2509
-    },
-    {
-      "epoch": 0.1640415659107248,
-      "grad_norm": 0.5262337327003479,
-      "learning_rate": 9.969166905633219e-06,
-      "loss": 0.4472,
-      "step": 2510
-    },
-    {
-      "epoch": 0.1641069211162669,
-      "grad_norm": 0.4879244863986969,
-      "learning_rate": 9.9691281740105e-06,
-      "loss": 0.382,
-      "step": 2511
-    },
-    {
-      "epoch": 0.16417227632180903,
-      "grad_norm": 0.4646982252597809,
-      "learning_rate": 9.96908941815162e-06,
-      "loss": 0.3673,
-      "step": 2512
-    },
-    {
-      "epoch": 0.16423763152735116,
-      "grad_norm": 0.45479026436805725,
-      "learning_rate": 9.96905063805677e-06,
-      "loss": 0.3925,
-      "step": 2513
-    },
-    {
-      "epoch": 0.16430298673289329,
-      "grad_norm": 0.4475038945674896,
-      "learning_rate": 9.96901183372614e-06,
-      "loss": 0.3981,
-      "step": 2514
-    },
-    {
-      "epoch": 0.16436834193843539,
-      "grad_norm": 0.5348302721977234,
-      "learning_rate": 9.968973005159916e-06,
-      "loss": 0.4298,
-      "step": 2515
-    },
-    {
-      "epoch": 0.16443369714397751,
-      "grad_norm": 0.4796079695224762,
-      "learning_rate": 9.96893415235829e-06,
-      "loss": 0.3671,
-      "step": 2516
-    },
-    {
-      "epoch": 0.16449905234951964,
-      "grad_norm": 0.5769591927528381,
-      "learning_rate": 9.968895275321451e-06,
-      "loss": 0.5147,
-      "step": 2517
-    },
-    {
-      "epoch": 0.16456440755506177,
-      "grad_norm": 0.4583422839641571,
-      "learning_rate": 9.968856374049587e-06,
-      "loss": 0.3913,
-      "step": 2518
-    },
-    {
-      "epoch": 0.16462976276060387,
-      "grad_norm": 0.48605695366859436,
-      "learning_rate": 9.96881744854289e-06,
-      "loss": 0.3907,
-      "step": 2519
-    },
-    {
-      "epoch": 0.164695117966146,
-      "grad_norm": 0.4812755882740021,
-      "learning_rate": 9.96877849880155e-06,
-      "loss": 0.426,
-      "step": 2520
-    },
-    {
-      "epoch": 0.16476047317168813,
-      "grad_norm": 0.4877561032772064,
-      "learning_rate": 9.968739524825754e-06,
-      "loss": 0.4104,
-      "step": 2521
-    },
-    {
-      "epoch": 0.16482582837723025,
-      "grad_norm": 0.511563241481781,
-      "learning_rate": 9.968700526615696e-06,
-      "loss": 0.4241,
-      "step": 2522
-    },
-    {
-      "epoch": 0.16489118358277235,
-      "grad_norm": 0.4960847496986389,
-      "learning_rate": 9.968661504171562e-06,
-      "loss": 0.4505,
-      "step": 2523
-    },
-    {
-      "epoch": 0.16495653878831448,
-      "grad_norm": 0.5165857076644897,
-      "learning_rate": 9.968622457493547e-06,
-      "loss": 0.4381,
-      "step": 2524
-    },
-    {
-      "epoch": 0.1650218939938566,
-      "grad_norm": 0.4729250967502594,
-      "learning_rate": 9.968583386581836e-06,
-      "loss": 0.3937,
-      "step": 2525
-    },
-    {
-      "epoch": 0.16508724919939874,
-      "grad_norm": 0.4830740690231323,
-      "learning_rate": 9.968544291436625e-06,
-      "loss": 0.4158,
-      "step": 2526
-    },
-    {
-      "epoch": 0.16515260440494087,
-      "grad_norm": 0.5134458541870117,
-      "learning_rate": 9.9685051720581e-06,
-      "loss": 0.4746,
-      "step": 2527
-    },
-    {
-      "epoch": 0.16521795961048297,
-      "grad_norm": 0.48444420099258423,
-      "learning_rate": 9.968466028446456e-06,
-      "loss": 0.4025,
-      "step": 2528
-    },
-    {
-      "epoch": 0.1652833148160251,
-      "grad_norm": 0.4610399603843689,
-      "learning_rate": 9.968426860601882e-06,
-      "loss": 0.3912,
-      "step": 2529
-    },
-    {
-      "epoch": 0.16534867002156722,
-      "grad_norm": 0.4778654873371124,
-      "learning_rate": 9.968387668524569e-06,
-      "loss": 0.4188,
-      "step": 2530
-    },
-    {
-      "epoch": 0.16541402522710935,
-      "grad_norm": 0.49111485481262207,
-      "learning_rate": 9.968348452214708e-06,
-      "loss": 0.4031,
-      "step": 2531
-    },
-    {
-      "epoch": 0.16547938043265145,
-      "grad_norm": 0.501004159450531,
-      "learning_rate": 9.968309211672489e-06,
-      "loss": 0.4069,
-      "step": 2532
-    },
-    {
-      "epoch": 0.16554473563819358,
-      "grad_norm": 0.4897302985191345,
-      "learning_rate": 9.968269946898106e-06,
-      "loss": 0.4174,
-      "step": 2533
-    },
-    {
-      "epoch": 0.1656100908437357,
-      "grad_norm": 0.5068104267120361,
-      "learning_rate": 9.968230657891748e-06,
-      "loss": 0.4256,
-      "step": 2534
-    },
-    {
-      "epoch": 0.16567544604927784,
-      "grad_norm": 0.5204775929450989,
-      "learning_rate": 9.968191344653608e-06,
-      "loss": 0.4388,
-      "step": 2535
-    },
-    {
-      "epoch": 0.16574080125481994,
-      "grad_norm": 0.5054092407226562,
-      "learning_rate": 9.968152007183879e-06,
-      "loss": 0.4322,
-      "step": 2536
-    },
-    {
-      "epoch": 0.16580615646036206,
-      "grad_norm": 0.48208674788475037,
-      "learning_rate": 9.96811264548275e-06,
-      "loss": 0.3779,
-      "step": 2537
-    },
-    {
-      "epoch": 0.1658715116659042,
-      "grad_norm": 0.5330373048782349,
-      "learning_rate": 9.968073259550417e-06,
-      "loss": 0.417,
-      "step": 2538
-    },
-    {
-      "epoch": 0.16593686687144632,
-      "grad_norm": 0.4920862317085266,
-      "learning_rate": 9.968033849387067e-06,
-      "loss": 0.4066,
-      "step": 2539
-    },
-    {
-      "epoch": 0.16600222207698842,
-      "grad_norm": 0.5067939758300781,
-      "learning_rate": 9.967994414992897e-06,
-      "loss": 0.4024,
-      "step": 2540
-    },
-    {
-      "epoch": 0.16606757728253055,
-      "grad_norm": 0.47380581498146057,
-      "learning_rate": 9.967954956368095e-06,
-      "loss": 0.4062,
-      "step": 2541
-    },
-    {
-      "epoch": 0.16613293248807268,
-      "grad_norm": 0.48494410514831543,
-      "learning_rate": 9.967915473512857e-06,
-      "loss": 0.409,
-      "step": 2542
-    },
-    {
-      "epoch": 0.1661982876936148,
-      "grad_norm": 0.4796536862850189,
-      "learning_rate": 9.967875966427374e-06,
-      "loss": 0.4195,
-      "step": 2543
-    },
-    {
-      "epoch": 0.1662636428991569,
-      "grad_norm": 0.5089160203933716,
-      "learning_rate": 9.96783643511184e-06,
-      "loss": 0.4332,
-      "step": 2544
-    },
-    {
-      "epoch": 0.16632899810469903,
-      "grad_norm": 0.4698975384235382,
-      "learning_rate": 9.967796879566445e-06,
-      "loss": 0.3886,
-      "step": 2545
-    },
-    {
-      "epoch": 0.16639435331024116,
-      "grad_norm": 0.4938991665840149,
-      "learning_rate": 9.967757299791383e-06,
-      "loss": 0.4114,
-      "step": 2546
-    },
-    {
-      "epoch": 0.1664597085157833,
-      "grad_norm": 0.42095014452934265,
-      "learning_rate": 9.967717695786849e-06,
-      "loss": 0.3436,
-      "step": 2547
-    },
-    {
-      "epoch": 0.16652506372132542,
-      "grad_norm": 0.4695545732975006,
-      "learning_rate": 9.967678067553035e-06,
-      "loss": 0.4201,
-      "step": 2548
-    },
-    {
-      "epoch": 0.16659041892686752,
-      "grad_norm": 0.4768037796020508,
-      "learning_rate": 9.967638415090132e-06,
-      "loss": 0.3815,
-      "step": 2549
-    },
-    {
-      "epoch": 0.16665577413240965,
-      "grad_norm": 0.45915326476097107,
-      "learning_rate": 9.967598738398338e-06,
-      "loss": 0.3719,
-      "step": 2550
-    },
-    {
-      "epoch": 0.16672112933795177,
-      "grad_norm": 0.47874993085861206,
-      "learning_rate": 9.967559037477842e-06,
-      "loss": 0.3782,
-      "step": 2551
-    },
-    {
-      "epoch": 0.1667864845434939,
-      "grad_norm": 0.4765692949295044,
-      "learning_rate": 9.967519312328842e-06,
-      "loss": 0.4424,
-      "step": 2552
-    },
-    {
-      "epoch": 0.166851839749036,
-      "grad_norm": 0.4355213940143585,
-      "learning_rate": 9.967479562951527e-06,
-      "loss": 0.3788,
-      "step": 2553
-    },
-    {
-      "epoch": 0.16691719495457813,
-      "grad_norm": 0.44575202465057373,
-      "learning_rate": 9.967439789346096e-06,
-      "loss": 0.3618,
-      "step": 2554
-    },
-    {
-      "epoch": 0.16698255016012026,
-      "grad_norm": 0.4374285638332367,
-      "learning_rate": 9.967399991512739e-06,
-      "loss": 0.39,
-      "step": 2555
-    },
-    {
-      "epoch": 0.16704790536566239,
-      "grad_norm": 0.5247700214385986,
-      "learning_rate": 9.96736016945165e-06,
-      "loss": 0.4986,
-      "step": 2556
-    },
-    {
-      "epoch": 0.16711326057120449,
-      "grad_norm": 0.502058207988739,
-      "learning_rate": 9.967320323163025e-06,
-      "loss": 0.4027,
-      "step": 2557
-    },
-    {
-      "epoch": 0.16717861577674661,
-      "grad_norm": 0.45537254214286804,
-      "learning_rate": 9.967280452647059e-06,
-      "loss": 0.3588,
-      "step": 2558
-    },
-    {
-      "epoch": 0.16724397098228874,
-      "grad_norm": 0.4959736168384552,
-      "learning_rate": 9.967240557903946e-06,
-      "loss": 0.38,
-      "step": 2559
-    },
-    {
-      "epoch": 0.16730932618783087,
-      "grad_norm": 0.5058095455169678,
-      "learning_rate": 9.967200638933878e-06,
-      "loss": 0.436,
-      "step": 2560
-    },
-    {
-      "epoch": 0.16737468139337297,
-      "grad_norm": 0.47594496607780457,
-      "learning_rate": 9.967160695737053e-06,
-      "loss": 0.4214,
-      "step": 2561
-    },
-    {
-      "epoch": 0.1674400365989151,
-      "grad_norm": 0.5219940543174744,
-      "learning_rate": 9.967120728313664e-06,
-      "loss": 0.3942,
-      "step": 2562
-    },
-    {
-      "epoch": 0.16750539180445723,
-      "grad_norm": 0.4856316149234772,
-      "learning_rate": 9.967080736663907e-06,
-      "loss": 0.4215,
-      "step": 2563
-    },
-    {
-      "epoch": 0.16757074700999935,
-      "grad_norm": 0.49029284715652466,
-      "learning_rate": 9.967040720787976e-06,
-      "loss": 0.3627,
-      "step": 2564
-    },
-    {
-      "epoch": 0.16763610221554145,
-      "grad_norm": 0.43458911776542664,
-      "learning_rate": 9.967000680686067e-06,
-      "loss": 0.3616,
-      "step": 2565
-    },
-    {
-      "epoch": 0.16770145742108358,
-      "grad_norm": 0.43978604674339294,
-      "learning_rate": 9.966960616358374e-06,
-      "loss": 0.3626,
-      "step": 2566
-    },
-    {
-      "epoch": 0.1677668126266257,
-      "grad_norm": 0.4900888502597809,
-      "learning_rate": 9.966920527805095e-06,
-      "loss": 0.3947,
-      "step": 2567
-    },
-    {
-      "epoch": 0.16783216783216784,
-      "grad_norm": 0.5364375114440918,
-      "learning_rate": 9.966880415026423e-06,
-      "loss": 0.4354,
-      "step": 2568
-    },
-    {
-      "epoch": 0.16789752303770997,
-      "grad_norm": 0.508897066116333,
-      "learning_rate": 9.966840278022553e-06,
-      "loss": 0.4934,
-      "step": 2569
-    },
-    {
-      "epoch": 0.16796287824325207,
-      "grad_norm": 0.48830941319465637,
-      "learning_rate": 9.966800116793684e-06,
-      "loss": 0.4073,
-      "step": 2570
-    },
-    {
-      "epoch": 0.1680282334487942,
-      "grad_norm": 0.45657533407211304,
-      "learning_rate": 9.96675993134001e-06,
-      "loss": 0.4061,
-      "step": 2571
-    },
-    {
-      "epoch": 0.16809358865433632,
-      "grad_norm": 0.4951988160610199,
-      "learning_rate": 9.966719721661728e-06,
-      "loss": 0.4522,
-      "step": 2572
-    },
-    {
-      "epoch": 0.16815894385987845,
-      "grad_norm": 0.47685354948043823,
-      "learning_rate": 9.966679487759032e-06,
-      "loss": 0.4018,
-      "step": 2573
-    },
-    {
-      "epoch": 0.16822429906542055,
-      "grad_norm": 0.5142900943756104,
-      "learning_rate": 9.96663922963212e-06,
-      "loss": 0.4395,
-      "step": 2574
-    },
-    {
-      "epoch": 0.16828965427096268,
-      "grad_norm": 0.5248950719833374,
-      "learning_rate": 9.966598947281187e-06,
-      "loss": 0.4512,
-      "step": 2575
-    },
-    {
-      "epoch": 0.1683550094765048,
-      "grad_norm": 0.5207715034484863,
-      "learning_rate": 9.966558640706432e-06,
-      "loss": 0.4485,
-      "step": 2576
-    },
-    {
-      "epoch": 0.16842036468204694,
-      "grad_norm": 0.4778415560722351,
-      "learning_rate": 9.966518309908049e-06,
-      "loss": 0.3832,
-      "step": 2577
-    },
-    {
-      "epoch": 0.16848571988758904,
-      "grad_norm": 0.47981250286102295,
-      "learning_rate": 9.966477954886236e-06,
-      "loss": 0.4027,
-      "step": 2578
-    },
-    {
-      "epoch": 0.16855107509313116,
-      "grad_norm": 0.5018543004989624,
-      "learning_rate": 9.966437575641188e-06,
-      "loss": 0.4194,
-      "step": 2579
-    },
-    {
-      "epoch": 0.1686164302986733,
-      "grad_norm": 0.5558579564094543,
-      "learning_rate": 9.966397172173106e-06,
-      "loss": 0.4751,
-      "step": 2580
-    },
-    {
-      "epoch": 0.16868178550421542,
-      "grad_norm": 0.5145378112792969,
-      "learning_rate": 9.966356744482182e-06,
-      "loss": 0.4732,
-      "step": 2581
-    },
-    {
-      "epoch": 0.16874714070975752,
-      "grad_norm": 0.44169601798057556,
-      "learning_rate": 9.966316292568616e-06,
-      "loss": 0.3465,
-      "step": 2582
-    },
-    {
-      "epoch": 0.16881249591529965,
-      "grad_norm": 0.4938541054725647,
-      "learning_rate": 9.966275816432606e-06,
-      "loss": 0.4023,
-      "step": 2583
-    },
-    {
-      "epoch": 0.16887785112084178,
-      "grad_norm": 0.5080822706222534,
-      "learning_rate": 9.96623531607435e-06,
-      "loss": 0.4284,
-      "step": 2584
-    },
-    {
-      "epoch": 0.1689432063263839,
-      "grad_norm": 0.4997018575668335,
-      "learning_rate": 9.966194791494039e-06,
-      "loss": 0.3967,
-      "step": 2585
-    },
-    {
-      "epoch": 0.169008561531926,
-      "grad_norm": 0.45045390725135803,
-      "learning_rate": 9.966154242691879e-06,
-      "loss": 0.4077,
-      "step": 2586
-    },
-    {
-      "epoch": 0.16907391673746813,
-      "grad_norm": 0.4637119472026825,
-      "learning_rate": 9.966113669668063e-06,
-      "loss": 0.3846,
-      "step": 2587
-    },
-    {
-      "epoch": 0.16913927194301026,
-      "grad_norm": 0.5090743899345398,
-      "learning_rate": 9.966073072422791e-06,
-      "loss": 0.3832,
-      "step": 2588
-    },
-    {
-      "epoch": 0.1692046271485524,
-      "grad_norm": 0.5185267925262451,
-      "learning_rate": 9.96603245095626e-06,
-      "loss": 0.4207,
-      "step": 2589
-    },
-    {
-      "epoch": 0.16926998235409452,
-      "grad_norm": 0.4877210855484009,
-      "learning_rate": 9.96599180526867e-06,
-      "loss": 0.4343,
-      "step": 2590
-    },
-    {
-      "epoch": 0.16933533755963662,
-      "grad_norm": 0.4766295254230499,
-      "learning_rate": 9.965951135360214e-06,
-      "loss": 0.4267,
-      "step": 2591
-    },
-    {
-      "epoch": 0.16940069276517875,
-      "grad_norm": 0.5057688355445862,
-      "learning_rate": 9.965910441231097e-06,
-      "loss": 0.418,
-      "step": 2592
-    },
-    {
-      "epoch": 0.16946604797072087,
-      "grad_norm": 0.4822060763835907,
-      "learning_rate": 9.965869722881515e-06,
-      "loss": 0.3854,
-      "step": 2593
-    },
-    {
-      "epoch": 0.169531403176263,
-      "grad_norm": 0.49696168303489685,
-      "learning_rate": 9.965828980311662e-06,
-      "loss": 0.372,
-      "step": 2594
-    },
-    {
-      "epoch": 0.1695967583818051,
-      "grad_norm": 0.5186379551887512,
-      "learning_rate": 9.965788213521743e-06,
-      "loss": 0.4242,
-      "step": 2595
-    },
-    {
-      "epoch": 0.16966211358734723,
-      "grad_norm": 0.48256298899650574,
-      "learning_rate": 9.965747422511956e-06,
-      "loss": 0.4027,
-      "step": 2596
-    },
-    {
-      "epoch": 0.16972746879288936,
-      "grad_norm": 0.45609068870544434,
-      "learning_rate": 9.965706607282497e-06,
-      "loss": 0.3708,
-      "step": 2597
-    },
-    {
-      "epoch": 0.16979282399843149,
-      "grad_norm": 0.47587496042251587,
-      "learning_rate": 9.965665767833567e-06,
-      "loss": 0.4422,
-      "step": 2598
-    },
-    {
-      "epoch": 0.16985817920397359,
-      "grad_norm": 0.5188522934913635,
-      "learning_rate": 9.965624904165364e-06,
-      "loss": 0.4643,
-      "step": 2599
-    },
-    {
-      "epoch": 0.16992353440951571,
-      "grad_norm": 0.5248739719390869,
-      "learning_rate": 9.965584016278089e-06,
-      "loss": 0.4377,
-      "step": 2600
-    },
-    {
-      "epoch": 0.16998888961505784,
-      "grad_norm": 0.4977901577949524,
-      "learning_rate": 9.96554310417194e-06,
-      "loss": 0.4275,
-      "step": 2601
-    },
-    {
-      "epoch": 0.17005424482059997,
-      "grad_norm": 0.48617416620254517,
-      "learning_rate": 9.965502167847117e-06,
-      "loss": 0.402,
-      "step": 2602
-    },
-    {
-      "epoch": 0.17011960002614207,
-      "grad_norm": 0.50570148229599,
-      "learning_rate": 9.96546120730382e-06,
-      "loss": 0.4729,
-      "step": 2603
-    },
-    {
-      "epoch": 0.1701849552316842,
-      "grad_norm": 0.5073838233947754,
-      "learning_rate": 9.965420222542248e-06,
-      "loss": 0.4264,
-      "step": 2604
-    },
-    {
-      "epoch": 0.17025031043722633,
-      "grad_norm": 0.4870482385158539,
-      "learning_rate": 9.965379213562602e-06,
-      "loss": 0.4014,
-      "step": 2605
-    },
-    {
-      "epoch": 0.17031566564276845,
-      "grad_norm": 0.5134543180465698,
-      "learning_rate": 9.965338180365081e-06,
-      "loss": 0.4449,
-      "step": 2606
-    },
-    {
-      "epoch": 0.17038102084831055,
-      "grad_norm": 0.5562323927879333,
-      "learning_rate": 9.965297122949886e-06,
-      "loss": 0.4601,
-      "step": 2607
-    },
-    {
-      "epoch": 0.17044637605385268,
-      "grad_norm": 0.7858109474182129,
-      "learning_rate": 9.965256041317217e-06,
-      "loss": 0.4765,
-      "step": 2608
-    },
-    {
-      "epoch": 0.1705117312593948,
-      "grad_norm": 0.4887913465499878,
-      "learning_rate": 9.965214935467274e-06,
-      "loss": 0.4496,
-      "step": 2609
-    },
-    {
-      "epoch": 0.17057708646493694,
-      "grad_norm": 0.5138968825340271,
-      "learning_rate": 9.965173805400257e-06,
-      "loss": 0.4569,
-      "step": 2610
-    },
-    {
-      "epoch": 0.17064244167047907,
-      "grad_norm": 0.47897759079933167,
-      "learning_rate": 9.965132651116368e-06,
-      "loss": 0.4029,
-      "step": 2611
-    },
-    {
-      "epoch": 0.17070779687602117,
-      "grad_norm": 0.4786286950111389,
-      "learning_rate": 9.965091472615807e-06,
-      "loss": 0.4086,
-      "step": 2612
-    },
-    {
-      "epoch": 0.1707731520815633,
-      "grad_norm": 0.4858819544315338,
-      "learning_rate": 9.965050269898774e-06,
-      "loss": 0.3894,
-      "step": 2613
-    },
-    {
-      "epoch": 0.17083850728710542,
-      "grad_norm": 0.4524725079536438,
-      "learning_rate": 9.96500904296547e-06,
-      "loss": 0.3649,
-      "step": 2614
-    },
-    {
-      "epoch": 0.17090386249264755,
-      "grad_norm": 0.5363240838050842,
-      "learning_rate": 9.964967791816098e-06,
-      "loss": 0.4893,
-      "step": 2615
-    },
-    {
-      "epoch": 0.17096921769818965,
-      "grad_norm": 0.4693722128868103,
-      "learning_rate": 9.96492651645086e-06,
-      "loss": 0.3968,
-      "step": 2616
-    },
-    {
-      "epoch": 0.17103457290373178,
-      "grad_norm": 0.5352193713188171,
-      "learning_rate": 9.964885216869953e-06,
-      "loss": 0.4488,
-      "step": 2617
-    },
-    {
-      "epoch": 0.1710999281092739,
-      "grad_norm": 0.45310312509536743,
-      "learning_rate": 9.96484389307358e-06,
-      "loss": 0.3974,
-      "step": 2618
-    },
-    {
-      "epoch": 0.17116528331481604,
-      "grad_norm": 0.49187517166137695,
-      "learning_rate": 9.964802545061944e-06,
-      "loss": 0.4147,
-      "step": 2619
-    },
-    {
-      "epoch": 0.17123063852035814,
-      "grad_norm": 0.5644070506095886,
-      "learning_rate": 9.964761172835247e-06,
-      "loss": 0.4787,
-      "step": 2620
-    },
-    {
-      "epoch": 0.17129599372590026,
-      "grad_norm": 0.4508492648601532,
-      "learning_rate": 9.96471977639369e-06,
-      "loss": 0.3417,
-      "step": 2621
-    },
-    {
-      "epoch": 0.1713613489314424,
-      "grad_norm": 0.4854797124862671,
-      "learning_rate": 9.964678355737474e-06,
-      "loss": 0.3945,
-      "step": 2622
-    },
-    {
-      "epoch": 0.17142670413698452,
-      "grad_norm": 0.523059606552124,
-      "learning_rate": 9.964636910866802e-06,
-      "loss": 0.4428,
-      "step": 2623
-    },
-    {
-      "epoch": 0.17149205934252662,
-      "grad_norm": 0.5245983600616455,
-      "learning_rate": 9.964595441781874e-06,
-      "loss": 0.4341,
-      "step": 2624
-    },
-    {
-      "epoch": 0.17155741454806875,
-      "grad_norm": 0.47294411063194275,
-      "learning_rate": 9.964553948482897e-06,
-      "loss": 0.3847,
-      "step": 2625
-    },
-    {
-      "epoch": 0.17162276975361088,
-      "grad_norm": 0.48770037293434143,
-      "learning_rate": 9.964512430970071e-06,
-      "loss": 0.4038,
-      "step": 2626
-    },
-    {
-      "epoch": 0.171688124959153,
-      "grad_norm": 0.45670706033706665,
-      "learning_rate": 9.964470889243595e-06,
-      "loss": 0.3577,
-      "step": 2627
-    },
-    {
-      "epoch": 0.1717534801646951,
-      "grad_norm": 0.5225144028663635,
-      "learning_rate": 9.964429323303675e-06,
-      "loss": 0.442,
-      "step": 2628
-    },
-    {
-      "epoch": 0.17181883537023723,
-      "grad_norm": 0.5164922475814819,
-      "learning_rate": 9.964387733150516e-06,
-      "loss": 0.4187,
-      "step": 2629
-    },
-    {
-      "epoch": 0.17188419057577936,
-      "grad_norm": 0.4995775520801544,
-      "learning_rate": 9.964346118784315e-06,
-      "loss": 0.3911,
-      "step": 2630
-    },
-    {
-      "epoch": 0.1719495457813215,
-      "grad_norm": 0.48990604281425476,
-      "learning_rate": 9.96430448020528e-06,
-      "loss": 0.4182,
-      "step": 2631
-    },
-    {
-      "epoch": 0.17201490098686362,
-      "grad_norm": 0.48489493131637573,
-      "learning_rate": 9.964262817413613e-06,
-      "loss": 0.3983,
-      "step": 2632
-    },
-    {
-      "epoch": 0.17208025619240572,
-      "grad_norm": 0.5082417726516724,
-      "learning_rate": 9.964221130409515e-06,
-      "loss": 0.3983,
-      "step": 2633
-    },
-    {
-      "epoch": 0.17214561139794785,
-      "grad_norm": 0.503032386302948,
-      "learning_rate": 9.96417941919319e-06,
-      "loss": 0.4297,
-      "step": 2634
-    },
-    {
-      "epoch": 0.17221096660348997,
-      "grad_norm": 0.5166121125221252,
-      "learning_rate": 9.964137683764844e-06,
-      "loss": 0.4047,
-      "step": 2635
-    },
-    {
-      "epoch": 0.1722763218090321,
-      "grad_norm": 0.4831024706363678,
-      "learning_rate": 9.964095924124678e-06,
-      "loss": 0.4258,
-      "step": 2636
-    },
-    {
-      "epoch": 0.1723416770145742,
-      "grad_norm": 0.48391157388687134,
-      "learning_rate": 9.964054140272895e-06,
-      "loss": 0.4064,
-      "step": 2637
-    },
-    {
-      "epoch": 0.17240703222011633,
-      "grad_norm": 0.6155975461006165,
-      "learning_rate": 9.964012332209703e-06,
-      "loss": 0.4597,
-      "step": 2638
-    },
-    {
-      "epoch": 0.17247238742565846,
-      "grad_norm": 0.5119083523750305,
-      "learning_rate": 9.963970499935303e-06,
-      "loss": 0.4272,
-      "step": 2639
-    },
-    {
-      "epoch": 0.17253774263120059,
-      "grad_norm": 0.5019258260726929,
-      "learning_rate": 9.963928643449898e-06,
-      "loss": 0.4597,
-      "step": 2640
-    },
-    {
-      "epoch": 0.17260309783674269,
-      "grad_norm": 0.4831063151359558,
-      "learning_rate": 9.963886762753692e-06,
-      "loss": 0.3992,
-      "step": 2641
-    },
-    {
-      "epoch": 0.17266845304228481,
-      "grad_norm": 0.5370878577232361,
-      "learning_rate": 9.963844857846895e-06,
-      "loss": 0.4319,
-      "step": 2642
-    },
-    {
-      "epoch": 0.17273380824782694,
-      "grad_norm": 0.5012040138244629,
-      "learning_rate": 9.963802928729702e-06,
-      "loss": 0.4204,
-      "step": 2643
-    },
-    {
-      "epoch": 0.17279916345336907,
-      "grad_norm": 0.523209810256958,
-      "learning_rate": 9.963760975402325e-06,
-      "loss": 0.4792,
-      "step": 2644
-    },
-    {
-      "epoch": 0.17286451865891117,
-      "grad_norm": 0.45911481976509094,
-      "learning_rate": 9.963718997864967e-06,
-      "loss": 0.3701,
-      "step": 2645
-    },
-    {
-      "epoch": 0.1729298738644533,
-      "grad_norm": 0.557472288608551,
-      "learning_rate": 9.963676996117827e-06,
-      "loss": 0.4708,
-      "step": 2646
-    },
-    {
-      "epoch": 0.17299522906999543,
-      "grad_norm": 0.49956074357032776,
-      "learning_rate": 9.96363497016112e-06,
-      "loss": 0.4201,
-      "step": 2647
-    },
-    {
-      "epoch": 0.17306058427553755,
-      "grad_norm": 0.4768489897251129,
-      "learning_rate": 9.963592919995044e-06,
-      "loss": 0.3931,
-      "step": 2648
-    },
-    {
-      "epoch": 0.17312593948107965,
-      "grad_norm": 0.4986908733844757,
-      "learning_rate": 9.963550845619805e-06,
-      "loss": 0.4258,
-      "step": 2649
-    },
-    {
-      "epoch": 0.17319129468662178,
-      "grad_norm": 0.4949187636375427,
-      "learning_rate": 9.963508747035611e-06,
-      "loss": 0.4068,
-      "step": 2650
-    },
-    {
-      "epoch": 0.1732566498921639,
-      "grad_norm": 0.494733601808548,
-      "learning_rate": 9.963466624242664e-06,
-      "loss": 0.4559,
-      "step": 2651
-    },
-    {
-      "epoch": 0.17332200509770604,
-      "grad_norm": 0.49882426857948303,
-      "learning_rate": 9.96342447724117e-06,
-      "loss": 0.3474,
-      "step": 2652
-    },
-    {
-      "epoch": 0.17338736030324817,
-      "grad_norm": 0.5101995468139648,
-      "learning_rate": 9.963382306031336e-06,
-      "loss": 0.3964,
-      "step": 2653
-    },
-    {
-      "epoch": 0.17345271550879027,
-      "grad_norm": 0.4387817084789276,
-      "learning_rate": 9.963340110613368e-06,
-      "loss": 0.3374,
-      "step": 2654
-    },
-    {
-      "epoch": 0.1735180707143324,
-      "grad_norm": 0.4991458058357239,
-      "learning_rate": 9.963297890987469e-06,
-      "loss": 0.3834,
-      "step": 2655
-    },
-    {
-      "epoch": 0.17358342591987452,
-      "grad_norm": 0.53550124168396,
-      "learning_rate": 9.963255647153848e-06,
-      "loss": 0.4925,
-      "step": 2656
-    },
-    {
-      "epoch": 0.17364878112541665,
-      "grad_norm": 0.48368826508522034,
-      "learning_rate": 9.96321337911271e-06,
-      "loss": 0.422,
-      "step": 2657
-    },
-    {
-      "epoch": 0.17371413633095875,
-      "grad_norm": 0.5215728878974915,
-      "learning_rate": 9.963171086864261e-06,
-      "loss": 0.4427,
-      "step": 2658
-    },
-    {
-      "epoch": 0.17377949153650088,
-      "grad_norm": 0.5231907963752747,
-      "learning_rate": 9.963128770408705e-06,
-      "loss": 0.4695,
-      "step": 2659
-    },
-    {
-      "epoch": 0.173844846742043,
-      "grad_norm": 0.5197647213935852,
-      "learning_rate": 9.963086429746253e-06,
-      "loss": 0.4168,
-      "step": 2660
-    },
-    {
-      "epoch": 0.17391020194758514,
-      "grad_norm": 0.5657692551612854,
-      "learning_rate": 9.963044064877108e-06,
-      "loss": 0.5545,
-      "step": 2661
-    },
-    {
-      "epoch": 0.17397555715312724,
-      "grad_norm": 0.47776147723197937,
-      "learning_rate": 9.963001675801478e-06,
-      "loss": 0.4143,
-      "step": 2662
-    },
-    {
-      "epoch": 0.17404091235866936,
-      "grad_norm": 0.47372832894325256,
-      "learning_rate": 9.96295926251957e-06,
-      "loss": 0.4079,
-      "step": 2663
-    },
-    {
-      "epoch": 0.1741062675642115,
-      "grad_norm": 0.5402714610099792,
-      "learning_rate": 9.96291682503159e-06,
-      "loss": 0.4726,
-      "step": 2664
-    },
-    {
-      "epoch": 0.17417162276975362,
-      "grad_norm": 0.48774057626724243,
-      "learning_rate": 9.962874363337745e-06,
-      "loss": 0.4021,
-      "step": 2665
-    },
-    {
-      "epoch": 0.17423697797529572,
-      "grad_norm": 0.4426060914993286,
-      "learning_rate": 9.962831877438242e-06,
-      "loss": 0.3216,
-      "step": 2666
-    },
-    {
-      "epoch": 0.17430233318083785,
-      "grad_norm": 0.49408379197120667,
-      "learning_rate": 9.962789367333289e-06,
-      "loss": 0.4005,
-      "step": 2667
-    },
-    {
-      "epoch": 0.17436768838637998,
-      "grad_norm": 0.4574377238750458,
-      "learning_rate": 9.962746833023093e-06,
-      "loss": 0.3721,
-      "step": 2668
-    },
-    {
-      "epoch": 0.1744330435919221,
-      "grad_norm": 0.46665847301483154,
-      "learning_rate": 9.962704274507862e-06,
-      "loss": 0.3392,
-      "step": 2669
-    },
-    {
-      "epoch": 0.1744983987974642,
-      "grad_norm": 0.50373774766922,
-      "learning_rate": 9.962661691787802e-06,
-      "loss": 0.4304,
-      "step": 2670
-    },
-    {
-      "epoch": 0.17456375400300633,
-      "grad_norm": 0.5277829766273499,
-      "learning_rate": 9.962619084863124e-06,
-      "loss": 0.44,
-      "step": 2671
-    },
-    {
-      "epoch": 0.17462910920854846,
-      "grad_norm": 0.4681248366832733,
-      "learning_rate": 9.96257645373403e-06,
-      "loss": 0.3894,
-      "step": 2672
-    },
-    {
-      "epoch": 0.1746944644140906,
-      "grad_norm": 0.4964708983898163,
-      "learning_rate": 9.962533798400734e-06,
-      "loss": 0.3904,
-      "step": 2673
-    },
-    {
-      "epoch": 0.17475981961963272,
-      "grad_norm": 0.5258198976516724,
-      "learning_rate": 9.96249111886344e-06,
-      "loss": 0.486,
-      "step": 2674
-    },
-    {
-      "epoch": 0.17482517482517482,
-      "grad_norm": 0.45436540246009827,
-      "learning_rate": 9.96244841512236e-06,
-      "loss": 0.4057,
-      "step": 2675
-    },
-    {
-      "epoch": 0.17489053003071695,
-      "grad_norm": 0.4477421045303345,
-      "learning_rate": 9.962405687177696e-06,
-      "loss": 0.3382,
-      "step": 2676
-    },
-    {
-      "epoch": 0.17495588523625907,
-      "grad_norm": 0.5042634606361389,
-      "learning_rate": 9.962362935029664e-06,
-      "loss": 0.4325,
-      "step": 2677
-    },
-    {
-      "epoch": 0.1750212404418012,
-      "grad_norm": 0.46567997336387634,
-      "learning_rate": 9.962320158678467e-06,
-      "loss": 0.3978,
-      "step": 2678
-    },
-    {
-      "epoch": 0.1750865956473433,
-      "grad_norm": 0.49018993973731995,
-      "learning_rate": 9.962277358124315e-06,
-      "loss": 0.4545,
-      "step": 2679
-    },
-    {
-      "epoch": 0.17515195085288543,
-      "grad_norm": 0.4478982388973236,
-      "learning_rate": 9.962234533367418e-06,
-      "loss": 0.3594,
-      "step": 2680
-    },
-    {
-      "epoch": 0.17521730605842756,
-      "grad_norm": 0.4821394681930542,
-      "learning_rate": 9.962191684407986e-06,
-      "loss": 0.3919,
-      "step": 2681
-    },
-    {
-      "epoch": 0.17528266126396969,
-      "grad_norm": 0.48290926218032837,
-      "learning_rate": 9.962148811246224e-06,
-      "loss": 0.401,
-      "step": 2682
-    },
-    {
-      "epoch": 0.17534801646951179,
-      "grad_norm": 0.45694294571876526,
-      "learning_rate": 9.962105913882342e-06,
-      "loss": 0.378,
-      "step": 2683
-    },
-    {
-      "epoch": 0.17541337167505391,
-      "grad_norm": 0.4834173321723938,
-      "learning_rate": 9.962062992316553e-06,
-      "loss": 0.3825,
-      "step": 2684
-    },
-    {
-      "epoch": 0.17547872688059604,
-      "grad_norm": 0.47401756048202515,
-      "learning_rate": 9.962020046549063e-06,
-      "loss": 0.394,
-      "step": 2685
-    },
-    {
-      "epoch": 0.17554408208613817,
-      "grad_norm": 0.5378835201263428,
-      "learning_rate": 9.961977076580082e-06,
-      "loss": 0.4804,
-      "step": 2686
-    },
-    {
-      "epoch": 0.17560943729168027,
-      "grad_norm": 0.5090337991714478,
-      "learning_rate": 9.961934082409819e-06,
-      "loss": 0.4134,
-      "step": 2687
-    },
-    {
-      "epoch": 0.1756747924972224,
-      "grad_norm": 0.4982738196849823,
-      "learning_rate": 9.961891064038487e-06,
-      "loss": 0.4509,
-      "step": 2688
-    },
-    {
-      "epoch": 0.17574014770276453,
-      "grad_norm": 0.46464022994041443,
-      "learning_rate": 9.961848021466291e-06,
-      "loss": 0.362,
-      "step": 2689
-    },
-    {
-      "epoch": 0.17580550290830665,
-      "grad_norm": 0.5181958675384521,
-      "learning_rate": 9.961804954693443e-06,
-      "loss": 0.4662,
-      "step": 2690
-    },
-    {
-      "epoch": 0.17587085811384875,
-      "grad_norm": 0.4751102030277252,
-      "learning_rate": 9.961761863720154e-06,
-      "loss": 0.395,
-      "step": 2691
-    },
-    {
-      "epoch": 0.17593621331939088,
-      "grad_norm": 0.49805060029029846,
-      "learning_rate": 9.961718748546633e-06,
-      "loss": 0.4122,
-      "step": 2692
-    },
-    {
-      "epoch": 0.176001568524933,
-      "grad_norm": 0.4755553901195526,
-      "learning_rate": 9.961675609173092e-06,
-      "loss": 0.4316,
-      "step": 2693
-    },
-    {
-      "epoch": 0.17606692373047514,
-      "grad_norm": 0.4815382659435272,
-      "learning_rate": 9.961632445599741e-06,
-      "loss": 0.3892,
-      "step": 2694
-    },
-    {
-      "epoch": 0.17613227893601727,
-      "grad_norm": 0.5086824297904968,
-      "learning_rate": 9.961589257826786e-06,
-      "loss": 0.3946,
-      "step": 2695
-    },
-    {
-      "epoch": 0.17619763414155937,
-      "grad_norm": 0.5206469297409058,
-      "learning_rate": 9.961546045854444e-06,
-      "loss": 0.4368,
-      "step": 2696
-    },
-    {
-      "epoch": 0.1762629893471015,
-      "grad_norm": 0.4665102958679199,
-      "learning_rate": 9.961502809682924e-06,
-      "loss": 0.3791,
-      "step": 2697
-    },
-    {
-      "epoch": 0.17632834455264362,
-      "grad_norm": 0.43253853917121887,
-      "learning_rate": 9.961459549312434e-06,
-      "loss": 0.3437,
-      "step": 2698
-    },
-    {
-      "epoch": 0.17639369975818575,
-      "grad_norm": 0.4989438056945801,
-      "learning_rate": 9.961416264743188e-06,
-      "loss": 0.4552,
-      "step": 2699
-    },
-    {
-      "epoch": 0.17645905496372785,
-      "grad_norm": 0.5197848081588745,
-      "learning_rate": 9.961372955975395e-06,
-      "loss": 0.4618,
-      "step": 2700
-    },
-    {
-      "epoch": 0.17652441016926998,
-      "grad_norm": 0.44897815585136414,
-      "learning_rate": 9.961329623009268e-06,
-      "loss": 0.3801,
-      "step": 2701
-    },
-    {
-      "epoch": 0.1765897653748121,
-      "grad_norm": 0.49561241269111633,
-      "learning_rate": 9.961286265845018e-06,
-      "loss": 0.3879,
-      "step": 2702
-    },
-    {
-      "epoch": 0.17665512058035424,
-      "grad_norm": 0.5102766752243042,
-      "learning_rate": 9.961242884482856e-06,
-      "loss": 0.4175,
-      "step": 2703
-    },
-    {
-      "epoch": 0.17672047578589634,
-      "grad_norm": 0.5269936919212341,
-      "learning_rate": 9.961199478922994e-06,
-      "loss": 0.4487,
-      "step": 2704
-    },
-    {
-      "epoch": 0.17678583099143846,
-      "grad_norm": 0.44980356097221375,
-      "learning_rate": 9.961156049165641e-06,
-      "loss": 0.357,
-      "step": 2705
-    },
-    {
-      "epoch": 0.1768511861969806,
-      "grad_norm": 0.4681536853313446,
-      "learning_rate": 9.961112595211014e-06,
-      "loss": 0.3701,
-      "step": 2706
-    },
-    {
-      "epoch": 0.17691654140252272,
-      "grad_norm": 0.48139142990112305,
-      "learning_rate": 9.96106911705932e-06,
-      "loss": 0.3979,
-      "step": 2707
-    },
-    {
-      "epoch": 0.17698189660806482,
-      "grad_norm": 0.5184427499771118,
-      "learning_rate": 9.961025614710775e-06,
-      "loss": 0.4142,
-      "step": 2708
-    },
-    {
-      "epoch": 0.17704725181360695,
-      "grad_norm": 0.4597087800502777,
-      "learning_rate": 9.96098208816559e-06,
-      "loss": 0.3845,
-      "step": 2709
-    },
-    {
-      "epoch": 0.17711260701914908,
-      "grad_norm": 0.4684363901615143,
-      "learning_rate": 9.960938537423976e-06,
-      "loss": 0.416,
-      "step": 2710
-    },
-    {
-      "epoch": 0.1771779622246912,
-      "grad_norm": 0.48280373215675354,
-      "learning_rate": 9.960894962486145e-06,
-      "loss": 0.412,
-      "step": 2711
-    },
-    {
-      "epoch": 0.1772433174302333,
-      "grad_norm": 0.47760751843452454,
-      "learning_rate": 9.96085136335231e-06,
-      "loss": 0.3837,
-      "step": 2712
-    },
-    {
-      "epoch": 0.17730867263577543,
-      "grad_norm": 0.4962843656539917,
-      "learning_rate": 9.960807740022686e-06,
-      "loss": 0.4204,
-      "step": 2713
-    },
-    {
-      "epoch": 0.17737402784131756,
-      "grad_norm": 0.4643685817718506,
-      "learning_rate": 9.960764092497485e-06,
-      "loss": 0.4018,
-      "step": 2714
-    },
-    {
-      "epoch": 0.1774393830468597,
-      "grad_norm": 0.4813143014907837,
-      "learning_rate": 9.960720420776917e-06,
-      "loss": 0.4005,
-      "step": 2715
-    },
-    {
-      "epoch": 0.17750473825240182,
-      "grad_norm": 0.4844858944416046,
-      "learning_rate": 9.960676724861198e-06,
-      "loss": 0.4202,
-      "step": 2716
-    },
-    {
-      "epoch": 0.17757009345794392,
-      "grad_norm": 0.506568968296051,
-      "learning_rate": 9.96063300475054e-06,
-      "loss": 0.4362,
-      "step": 2717
-    },
-    {
-      "epoch": 0.17763544866348605,
-      "grad_norm": 0.48755761981010437,
-      "learning_rate": 9.960589260445156e-06,
-      "loss": 0.396,
-      "step": 2718
-    },
-    {
-      "epoch": 0.17770080386902817,
-      "grad_norm": 0.486409991979599,
-      "learning_rate": 9.960545491945259e-06,
-      "loss": 0.4419,
-      "step": 2719
-    },
-    {
-      "epoch": 0.1777661590745703,
-      "grad_norm": 0.4849031865596771,
-      "learning_rate": 9.960501699251063e-06,
-      "loss": 0.3919,
-      "step": 2720
-    },
-    {
-      "epoch": 0.1778315142801124,
-      "grad_norm": 0.49966585636138916,
-      "learning_rate": 9.960457882362784e-06,
-      "loss": 0.408,
-      "step": 2721
-    },
-    {
-      "epoch": 0.17789686948565453,
-      "grad_norm": 0.4932485520839691,
-      "learning_rate": 9.960414041280632e-06,
-      "loss": 0.4624,
-      "step": 2722
-    },
-    {
-      "epoch": 0.17796222469119666,
-      "grad_norm": 0.46550408005714417,
-      "learning_rate": 9.960370176004822e-06,
-      "loss": 0.3948,
-      "step": 2723
-    },
-    {
-      "epoch": 0.17802757989673879,
-      "grad_norm": 0.5008115768432617,
-      "learning_rate": 9.960326286535569e-06,
-      "loss": 0.4448,
-      "step": 2724
-    },
-    {
-      "epoch": 0.17809293510228089,
-      "grad_norm": 0.5356562733650208,
-      "learning_rate": 9.960282372873086e-06,
-      "loss": 0.516,
-      "step": 2725
-    },
-    {
-      "epoch": 0.17815829030782301,
-      "grad_norm": 0.5430569648742676,
-      "learning_rate": 9.960238435017586e-06,
-      "loss": 0.4586,
-      "step": 2726
-    },
-    {
-      "epoch": 0.17822364551336514,
-      "grad_norm": 0.48317262530326843,
-      "learning_rate": 9.960194472969286e-06,
-      "loss": 0.4023,
-      "step": 2727
-    },
-    {
-      "epoch": 0.17828900071890727,
-      "grad_norm": 0.5187653303146362,
-      "learning_rate": 9.9601504867284e-06,
-      "loss": 0.4474,
-      "step": 2728
-    },
-    {
-      "epoch": 0.17835435592444937,
-      "grad_norm": 0.4397761821746826,
-      "learning_rate": 9.960106476295142e-06,
-      "loss": 0.3827,
-      "step": 2729
-    },
-    {
-      "epoch": 0.1784197111299915,
-      "grad_norm": 0.44425103068351746,
-      "learning_rate": 9.960062441669724e-06,
-      "loss": 0.3933,
-      "step": 2730
-    },
-    {
-      "epoch": 0.17848506633553363,
-      "grad_norm": 0.47415691614151,
-      "learning_rate": 9.960018382852364e-06,
-      "loss": 0.3889,
-      "step": 2731
-    },
-    {
-      "epoch": 0.17855042154107575,
-      "grad_norm": 0.489446759223938,
-      "learning_rate": 9.959974299843277e-06,
-      "loss": 0.4232,
-      "step": 2732
-    },
-    {
-      "epoch": 0.17861577674661785,
-      "grad_norm": 0.4992932379245758,
-      "learning_rate": 9.959930192642676e-06,
-      "loss": 0.448,
-      "step": 2733
-    },
-    {
-      "epoch": 0.17868113195215998,
-      "grad_norm": 0.47179922461509705,
-      "learning_rate": 9.959886061250778e-06,
-      "loss": 0.3851,
-      "step": 2734
-    },
-    {
-      "epoch": 0.1787464871577021,
-      "grad_norm": 0.47768720984458923,
-      "learning_rate": 9.959841905667798e-06,
-      "loss": 0.4111,
-      "step": 2735
-    },
-    {
-      "epoch": 0.17881184236324424,
-      "grad_norm": 0.4497457444667816,
-      "learning_rate": 9.95979772589395e-06,
-      "loss": 0.3718,
-      "step": 2736
-    },
-    {
-      "epoch": 0.17887719756878637,
-      "grad_norm": 0.4446553885936737,
-      "learning_rate": 9.95975352192945e-06,
-      "loss": 0.3936,
-      "step": 2737
-    },
-    {
-      "epoch": 0.17894255277432847,
-      "grad_norm": 0.541711151599884,
-      "learning_rate": 9.959709293774513e-06,
-      "loss": 0.4346,
-      "step": 2738
-    },
-    {
-      "epoch": 0.1790079079798706,
-      "grad_norm": 0.4848293364048004,
-      "learning_rate": 9.959665041429355e-06,
-      "loss": 0.39,
-      "step": 2739
-    },
-    {
-      "epoch": 0.17907326318541272,
-      "grad_norm": 0.4830087721347809,
-      "learning_rate": 9.959620764894196e-06,
-      "loss": 0.4342,
-      "step": 2740
-    },
-    {
-      "epoch": 0.17913861839095485,
-      "grad_norm": 0.48588570952415466,
-      "learning_rate": 9.959576464169245e-06,
-      "loss": 0.3978,
-      "step": 2741
-    },
-    {
-      "epoch": 0.17920397359649695,
-      "grad_norm": 0.5507774353027344,
-      "learning_rate": 9.959532139254723e-06,
-      "loss": 0.4369,
-      "step": 2742
-    },
-    {
-      "epoch": 0.17926932880203908,
-      "grad_norm": 0.468826562166214,
-      "learning_rate": 9.959487790150844e-06,
-      "loss": 0.4161,
-      "step": 2743
-    },
-    {
-      "epoch": 0.1793346840075812,
-      "grad_norm": 0.45922598242759705,
-      "learning_rate": 9.959443416857827e-06,
-      "loss": 0.4035,
-      "step": 2744
-    },
-    {
-      "epoch": 0.17940003921312334,
-      "grad_norm": 0.5572922825813293,
-      "learning_rate": 9.959399019375884e-06,
-      "loss": 0.4935,
-      "step": 2745
-    },
-    {
-      "epoch": 0.17946539441866544,
-      "grad_norm": 0.5046947002410889,
-      "learning_rate": 9.959354597705232e-06,
-      "loss": 0.4202,
-      "step": 2746
-    },
-    {
-      "epoch": 0.17953074962420756,
-      "grad_norm": 0.5219874978065491,
-      "learning_rate": 9.959310151846092e-06,
-      "loss": 0.441,
-      "step": 2747
-    },
-    {
-      "epoch": 0.1795961048297497,
-      "grad_norm": 0.5299882888793945,
-      "learning_rate": 9.95926568179868e-06,
-      "loss": 0.4974,
-      "step": 2748
-    },
-    {
-      "epoch": 0.17966146003529182,
-      "grad_norm": 0.4867027997970581,
-      "learning_rate": 9.959221187563208e-06,
-      "loss": 0.4215,
-      "step": 2749
-    },
-    {
-      "epoch": 0.17972681524083392,
-      "grad_norm": 0.4667873978614807,
-      "learning_rate": 9.959176669139898e-06,
-      "loss": 0.4003,
-      "step": 2750
-    },
-    {
-      "epoch": 0.17979217044637605,
-      "grad_norm": 0.5644049048423767,
-      "learning_rate": 9.959132126528965e-06,
-      "loss": 0.4377,
-      "step": 2751
-    },
-    {
-      "epoch": 0.17985752565191818,
-      "grad_norm": 0.4796614944934845,
-      "learning_rate": 9.959087559730627e-06,
-      "loss": 0.4016,
-      "step": 2752
-    },
-    {
-      "epoch": 0.1799228808574603,
-      "grad_norm": 0.49181193113327026,
-      "learning_rate": 9.959042968745101e-06,
-      "loss": 0.3768,
-      "step": 2753
-    },
-    {
-      "epoch": 0.1799882360630024,
-      "grad_norm": 0.5339409112930298,
-      "learning_rate": 9.958998353572605e-06,
-      "loss": 0.4718,
-      "step": 2754
-    },
-    {
-      "epoch": 0.18005359126854453,
-      "grad_norm": 0.4908137619495392,
-      "learning_rate": 9.958953714213355e-06,
-      "loss": 0.4309,
-      "step": 2755
-    },
-    {
-      "epoch": 0.18011894647408666,
-      "grad_norm": 0.44965025782585144,
-      "learning_rate": 9.95890905066757e-06,
-      "loss": 0.3547,
-      "step": 2756
-    },
-    {
-      "epoch": 0.1801843016796288,
-      "grad_norm": 0.4824472963809967,
-      "learning_rate": 9.958864362935468e-06,
-      "loss": 0.438,
-      "step": 2757
-    },
-    {
-      "epoch": 0.18024965688517092,
-      "grad_norm": 0.5025871992111206,
-      "learning_rate": 9.958819651017266e-06,
-      "loss": 0.3762,
-      "step": 2758
-    },
-    {
-      "epoch": 0.18031501209071302,
-      "grad_norm": 0.5372518301010132,
-      "learning_rate": 9.958774914913183e-06,
-      "loss": 0.4312,
-      "step": 2759
-    },
-    {
-      "epoch": 0.18038036729625515,
-      "grad_norm": 0.46665894985198975,
-      "learning_rate": 9.958730154623436e-06,
-      "loss": 0.4116,
-      "step": 2760
-    },
-    {
-      "epoch": 0.18044572250179727,
-      "grad_norm": 0.4289593994617462,
-      "learning_rate": 9.958685370148244e-06,
-      "loss": 0.3363,
-      "step": 2761
-    },
-    {
-      "epoch": 0.1805110777073394,
-      "grad_norm": 0.5162213444709778,
-      "learning_rate": 9.958640561487826e-06,
-      "loss": 0.4618,
-      "step": 2762
-    },
-    {
-      "epoch": 0.1805764329128815,
-      "grad_norm": 0.5332930088043213,
-      "learning_rate": 9.958595728642401e-06,
-      "loss": 0.4222,
-      "step": 2763
-    },
-    {
-      "epoch": 0.18064178811842363,
-      "grad_norm": 0.4913944900035858,
-      "learning_rate": 9.958550871612186e-06,
-      "loss": 0.3846,
-      "step": 2764
-    },
-    {
-      "epoch": 0.18070714332396576,
-      "grad_norm": 0.451699823141098,
-      "learning_rate": 9.958505990397402e-06,
-      "loss": 0.3482,
-      "step": 2765
-    },
-    {
-      "epoch": 0.18077249852950789,
-      "grad_norm": 0.46061721444129944,
-      "learning_rate": 9.958461084998265e-06,
-      "loss": 0.357,
-      "step": 2766
-    },
-    {
-      "epoch": 0.18083785373504999,
-      "grad_norm": 0.54320228099823,
-      "learning_rate": 9.958416155414996e-06,
-      "loss": 0.5188,
-      "step": 2767
-    },
-    {
-      "epoch": 0.18090320894059211,
-      "grad_norm": 0.49299147725105286,
-      "learning_rate": 9.958371201647814e-06,
-      "loss": 0.3738,
-      "step": 2768
-    },
-    {
-      "epoch": 0.18096856414613424,
-      "grad_norm": 0.4782959818840027,
-      "learning_rate": 9.958326223696938e-06,
-      "loss": 0.4039,
-      "step": 2769
-    },
-    {
-      "epoch": 0.18103391935167637,
-      "grad_norm": 0.4619569480419159,
-      "learning_rate": 9.958281221562586e-06,
-      "loss": 0.397,
-      "step": 2770
-    },
-    {
-      "epoch": 0.18109927455721847,
-      "grad_norm": 0.501964271068573,
-      "learning_rate": 9.958236195244981e-06,
-      "loss": 0.4531,
-      "step": 2771
-    },
-    {
-      "epoch": 0.1811646297627606,
-      "grad_norm": 0.5149459838867188,
-      "learning_rate": 9.958191144744339e-06,
-      "loss": 0.4573,
-      "step": 2772
-    },
-    {
-      "epoch": 0.18122998496830273,
-      "grad_norm": 0.5695905089378357,
-      "learning_rate": 9.958146070060881e-06,
-      "loss": 0.486,
-      "step": 2773
-    },
-    {
-      "epoch": 0.18129534017384485,
-      "grad_norm": 0.5138512849807739,
-      "learning_rate": 9.958100971194827e-06,
-      "loss": 0.4179,
-      "step": 2774
-    },
-    {
-      "epoch": 0.18136069537938695,
-      "grad_norm": 0.533556342124939,
-      "learning_rate": 9.958055848146398e-06,
-      "loss": 0.4039,
-      "step": 2775
-    },
-    {
-      "epoch": 0.18142605058492908,
-      "grad_norm": 0.48465925455093384,
-      "learning_rate": 9.95801070091581e-06,
-      "loss": 0.3961,
-      "step": 2776
-    },
-    {
-      "epoch": 0.1814914057904712,
-      "grad_norm": 0.4893660843372345,
-      "learning_rate": 9.957965529503288e-06,
-      "loss": 0.4316,
-      "step": 2777
-    },
-    {
-      "epoch": 0.18155676099601334,
-      "grad_norm": 0.49492719769477844,
-      "learning_rate": 9.957920333909051e-06,
-      "loss": 0.4537,
-      "step": 2778
-    },
-    {
-      "epoch": 0.18162211620155547,
-      "grad_norm": 0.5087707042694092,
-      "learning_rate": 9.957875114133319e-06,
-      "loss": 0.438,
-      "step": 2779
-    },
-    {
-      "epoch": 0.18168747140709757,
-      "grad_norm": 0.43353673815727234,
-      "learning_rate": 9.957829870176312e-06,
-      "loss": 0.3576,
-      "step": 2780
-    },
-    {
-      "epoch": 0.1817528266126397,
-      "grad_norm": 0.4844864010810852,
-      "learning_rate": 9.957784602038252e-06,
-      "loss": 0.3669,
-      "step": 2781
-    },
-    {
-      "epoch": 0.18181818181818182,
-      "grad_norm": 0.5208877921104431,
-      "learning_rate": 9.957739309719357e-06,
-      "loss": 0.4753,
-      "step": 2782
-    },
-    {
-      "epoch": 0.18188353702372395,
-      "grad_norm": 0.4982951581478119,
-      "learning_rate": 9.957693993219852e-06,
-      "loss": 0.4126,
-      "step": 2783
-    },
-    {
-      "epoch": 0.18194889222926605,
-      "grad_norm": 0.48803335428237915,
-      "learning_rate": 9.957648652539955e-06,
-      "loss": 0.4386,
-      "step": 2784
-    },
-    {
-      "epoch": 0.18201424743480818,
-      "grad_norm": 0.44746482372283936,
-      "learning_rate": 9.957603287679888e-06,
-      "loss": 0.3802,
-      "step": 2785
-    },
-    {
-      "epoch": 0.1820796026403503,
-      "grad_norm": 0.5041863322257996,
-      "learning_rate": 9.957557898639871e-06,
-      "loss": 0.4748,
-      "step": 2786
-    },
-    {
-      "epoch": 0.18214495784589244,
-      "grad_norm": 0.49817386269569397,
-      "learning_rate": 9.957512485420129e-06,
-      "loss": 0.372,
-      "step": 2787
-    },
-    {
-      "epoch": 0.18221031305143454,
-      "grad_norm": 0.6099628210067749,
-      "learning_rate": 9.95746704802088e-06,
-      "loss": 0.4915,
-      "step": 2788
-    },
-    {
-      "epoch": 0.18227566825697666,
-      "grad_norm": 0.4703923463821411,
-      "learning_rate": 9.957421586442346e-06,
-      "loss": 0.3882,
-      "step": 2789
-    },
-    {
-      "epoch": 0.1823410234625188,
-      "grad_norm": 0.5085967779159546,
-      "learning_rate": 9.957376100684751e-06,
-      "loss": 0.4065,
-      "step": 2790
-    },
-    {
-      "epoch": 0.18240637866806092,
-      "grad_norm": 0.4918684959411621,
-      "learning_rate": 9.957330590748315e-06,
-      "loss": 0.3979,
-      "step": 2791
-    },
-    {
-      "epoch": 0.18247173387360302,
-      "grad_norm": 0.5107892751693726,
-      "learning_rate": 9.95728505663326e-06,
-      "loss": 0.4268,
-      "step": 2792
-    },
-    {
-      "epoch": 0.18253708907914515,
-      "grad_norm": 0.46673840284347534,
-      "learning_rate": 9.957239498339809e-06,
-      "loss": 0.3606,
-      "step": 2793
-    },
-    {
-      "epoch": 0.18260244428468728,
-      "grad_norm": 0.45932820439338684,
-      "learning_rate": 9.957193915868184e-06,
-      "loss": 0.3767,
-      "step": 2794
-    },
-    {
-      "epoch": 0.1826677994902294,
-      "grad_norm": 0.562187910079956,
-      "learning_rate": 9.957148309218605e-06,
-      "loss": 0.4199,
-      "step": 2795
-    },
-    {
-      "epoch": 0.1827331546957715,
-      "grad_norm": 0.4595358371734619,
-      "learning_rate": 9.957102678391297e-06,
-      "loss": 0.3737,
-      "step": 2796
-    },
-    {
-      "epoch": 0.18279850990131363,
-      "grad_norm": 0.5276442766189575,
-      "learning_rate": 9.957057023386482e-06,
-      "loss": 0.4423,
-      "step": 2797
-    },
-    {
-      "epoch": 0.18286386510685576,
-      "grad_norm": 0.49165117740631104,
-      "learning_rate": 9.957011344204384e-06,
-      "loss": 0.4345,
-      "step": 2798
-    },
-    {
-      "epoch": 0.1829292203123979,
-      "grad_norm": 0.4598459303379059,
-      "learning_rate": 9.956965640845223e-06,
-      "loss": 0.3825,
-      "step": 2799
-    },
-    {
-      "epoch": 0.18299457551794002,
-      "grad_norm": 0.4706825315952301,
-      "learning_rate": 9.956919913309225e-06,
-      "loss": 0.3574,
-      "step": 2800
-    },
-    {
-      "epoch": 0.18305993072348212,
-      "grad_norm": 0.4787357449531555,
-      "learning_rate": 9.956874161596609e-06,
-      "loss": 0.3828,
-      "step": 2801
-    },
-    {
-      "epoch": 0.18312528592902425,
-      "grad_norm": 0.5086193680763245,
-      "learning_rate": 9.9568283857076e-06,
-      "loss": 0.4072,
-      "step": 2802
-    },
-    {
-      "epoch": 0.18319064113456637,
-      "grad_norm": 0.5089016556739807,
-      "learning_rate": 9.956782585642424e-06,
-      "loss": 0.4356,
-      "step": 2803
-    },
-    {
-      "epoch": 0.1832559963401085,
-      "grad_norm": 0.4908629059791565,
-      "learning_rate": 9.956736761401302e-06,
-      "loss": 0.3937,
-      "step": 2804
-    },
-    {
-      "epoch": 0.1833213515456506,
-      "grad_norm": 0.5212690234184265,
-      "learning_rate": 9.956690912984457e-06,
-      "loss": 0.4683,
-      "step": 2805
-    },
-    {
-      "epoch": 0.18338670675119273,
-      "grad_norm": 0.5048547983169556,
-      "learning_rate": 9.956645040392111e-06,
-      "loss": 0.4163,
-      "step": 2806
-    },
-    {
-      "epoch": 0.18345206195673486,
-      "grad_norm": 0.5160130858421326,
-      "learning_rate": 9.956599143624495e-06,
-      "loss": 0.4377,
-      "step": 2807
-    },
-    {
-      "epoch": 0.18351741716227699,
-      "grad_norm": 0.47323620319366455,
-      "learning_rate": 9.956553222681825e-06,
-      "loss": 0.4034,
-      "step": 2808
-    },
-    {
-      "epoch": 0.18358277236781909,
-      "grad_norm": 0.4883844256401062,
-      "learning_rate": 9.956507277564328e-06,
-      "loss": 0.396,
-      "step": 2809
-    },
-    {
-      "epoch": 0.18364812757336121,
-      "grad_norm": 0.49567705392837524,
-      "learning_rate": 9.956461308272227e-06,
-      "loss": 0.3822,
-      "step": 2810
-    },
-    {
-      "epoch": 0.18371348277890334,
-      "grad_norm": 0.5009114146232605,
-      "learning_rate": 9.956415314805747e-06,
-      "loss": 0.429,
-      "step": 2811
-    },
-    {
-      "epoch": 0.18377883798444547,
-      "grad_norm": 0.4522407352924347,
-      "learning_rate": 9.956369297165113e-06,
-      "loss": 0.3981,
-      "step": 2812
-    },
-    {
-      "epoch": 0.18384419318998757,
-      "grad_norm": 0.46104896068573,
-      "learning_rate": 9.956323255350551e-06,
-      "loss": 0.3663,
-      "step": 2813
-    },
-    {
-      "epoch": 0.1839095483955297,
-      "grad_norm": 0.6915690898895264,
-      "learning_rate": 9.956277189362281e-06,
-      "loss": 0.4644,
-      "step": 2814
-    },
-    {
-      "epoch": 0.18397490360107183,
-      "grad_norm": 0.47293227910995483,
-      "learning_rate": 9.95623109920053e-06,
-      "loss": 0.4012,
-      "step": 2815
-    },
-    {
-      "epoch": 0.18404025880661395,
-      "grad_norm": 0.49006471037864685,
-      "learning_rate": 9.956184984865524e-06,
-      "loss": 0.4488,
-      "step": 2816
-    },
-    {
-      "epoch": 0.18410561401215605,
-      "grad_norm": 0.47831088304519653,
-      "learning_rate": 9.956138846357488e-06,
-      "loss": 0.4071,
-      "step": 2817
-    },
-    {
-      "epoch": 0.18417096921769818,
-      "grad_norm": 0.5056982040405273,
-      "learning_rate": 9.956092683676644e-06,
-      "loss": 0.4303,
-      "step": 2818
-    },
-    {
-      "epoch": 0.1842363244232403,
-      "grad_norm": 0.5214632153511047,
-      "learning_rate": 9.95604649682322e-06,
-      "loss": 0.3883,
-      "step": 2819
-    },
-    {
-      "epoch": 0.18430167962878244,
-      "grad_norm": 0.5154610276222229,
-      "learning_rate": 9.95600028579744e-06,
-      "loss": 0.4622,
-      "step": 2820
-    },
-    {
-      "epoch": 0.18436703483432457,
-      "grad_norm": 0.5146011710166931,
-      "learning_rate": 9.955954050599529e-06,
-      "loss": 0.477,
-      "step": 2821
-    },
-    {
-      "epoch": 0.18443239003986667,
-      "grad_norm": 0.5352776050567627,
-      "learning_rate": 9.955907791229713e-06,
-      "loss": 0.444,
-      "step": 2822
-    },
-    {
-      "epoch": 0.1844977452454088,
-      "grad_norm": 0.5003596544265747,
-      "learning_rate": 9.955861507688219e-06,
-      "loss": 0.422,
-      "step": 2823
-    },
-    {
-      "epoch": 0.18456310045095092,
-      "grad_norm": 0.5347896814346313,
-      "learning_rate": 9.955815199975271e-06,
-      "loss": 0.507,
-      "step": 2824
-    },
-    {
-      "epoch": 0.18462845565649305,
-      "grad_norm": 0.5119844675064087,
-      "learning_rate": 9.955768868091096e-06,
-      "loss": 0.4488,
-      "step": 2825
-    },
-    {
-      "epoch": 0.18469381086203515,
-      "grad_norm": 0.5100293159484863,
-      "learning_rate": 9.95572251203592e-06,
-      "loss": 0.4442,
-      "step": 2826
-    },
-    {
-      "epoch": 0.18475916606757728,
-      "grad_norm": 0.4475594460964203,
-      "learning_rate": 9.955676131809966e-06,
-      "loss": 0.3668,
-      "step": 2827
-    },
-    {
-      "epoch": 0.1848245212731194,
-      "grad_norm": 0.4683065116405487,
-      "learning_rate": 9.955629727413465e-06,
-      "loss": 0.4293,
-      "step": 2828
-    },
-    {
-      "epoch": 0.18488987647866154,
-      "grad_norm": 0.4584593176841736,
-      "learning_rate": 9.955583298846638e-06,
-      "loss": 0.3729,
-      "step": 2829
-    },
-    {
-      "epoch": 0.18495523168420364,
-      "grad_norm": 0.4597283899784088,
-      "learning_rate": 9.955536846109717e-06,
-      "loss": 0.3884,
-      "step": 2830
-    },
-    {
-      "epoch": 0.18502058688974576,
-      "grad_norm": 0.4883746802806854,
-      "learning_rate": 9.955490369202926e-06,
-      "loss": 0.425,
-      "step": 2831
-    },
-    {
-      "epoch": 0.1850859420952879,
-      "grad_norm": 0.4885231554508209,
-      "learning_rate": 9.95544386812649e-06,
-      "loss": 0.419,
-      "step": 2832
-    },
-    {
-      "epoch": 0.18515129730083002,
-      "grad_norm": 0.4709199368953705,
-      "learning_rate": 9.95539734288064e-06,
-      "loss": 0.4117,
-      "step": 2833
-    },
-    {
-      "epoch": 0.18521665250637212,
-      "grad_norm": 0.48245787620544434,
-      "learning_rate": 9.955350793465599e-06,
-      "loss": 0.4371,
-      "step": 2834
-    },
-    {
-      "epoch": 0.18528200771191425,
-      "grad_norm": 0.42119458317756653,
-      "learning_rate": 9.955304219881597e-06,
-      "loss": 0.3272,
-      "step": 2835
-    },
-    {
-      "epoch": 0.18534736291745638,
-      "grad_norm": 0.476041316986084,
-      "learning_rate": 9.955257622128857e-06,
-      "loss": 0.3919,
-      "step": 2836
-    },
-    {
-      "epoch": 0.1854127181229985,
-      "grad_norm": 0.48731857538223267,
-      "learning_rate": 9.95521100020761e-06,
-      "loss": 0.4009,
-      "step": 2837
-    },
-    {
-      "epoch": 0.1854780733285406,
-      "grad_norm": 0.5108567476272583,
-      "learning_rate": 9.955164354118082e-06,
-      "loss": 0.4774,
-      "step": 2838
-    },
-    {
-      "epoch": 0.18554342853408273,
-      "grad_norm": 0.47319793701171875,
-      "learning_rate": 9.955117683860502e-06,
-      "loss": 0.4241,
-      "step": 2839
-    },
-    {
-      "epoch": 0.18560878373962486,
-      "grad_norm": 0.4684969186782837,
-      "learning_rate": 9.955070989435097e-06,
-      "loss": 0.3701,
-      "step": 2840
-    },
-    {
-      "epoch": 0.185674138945167,
-      "grad_norm": 0.5024217367172241,
-      "learning_rate": 9.955024270842093e-06,
-      "loss": 0.4255,
-      "step": 2841
-    },
-    {
-      "epoch": 0.18573949415070912,
-      "grad_norm": 0.485853374004364,
-      "learning_rate": 9.954977528081718e-06,
-      "loss": 0.3623,
-      "step": 2842
-    },
-    {
-      "epoch": 0.18580484935625122,
-      "grad_norm": 0.49599698185920715,
-      "learning_rate": 9.954930761154202e-06,
-      "loss": 0.4565,
-      "step": 2843
-    },
-    {
-      "epoch": 0.18587020456179335,
-      "grad_norm": 0.49961188435554504,
-      "learning_rate": 9.954883970059774e-06,
-      "loss": 0.4354,
-      "step": 2844
-    },
-    {
-      "epoch": 0.18593555976733547,
-      "grad_norm": 0.4934704601764679,
-      "learning_rate": 9.954837154798657e-06,
-      "loss": 0.4318,
-      "step": 2845
-    },
-    {
-      "epoch": 0.1860009149728776,
-      "grad_norm": 0.47270822525024414,
-      "learning_rate": 9.954790315371084e-06,
-      "loss": 0.3625,
-      "step": 2846
-    },
-    {
-      "epoch": 0.1860662701784197,
-      "grad_norm": 0.4581604301929474,
-      "learning_rate": 9.954743451777283e-06,
-      "loss": 0.362,
-      "step": 2847
-    },
-    {
-      "epoch": 0.18613162538396183,
-      "grad_norm": 0.5095042586326599,
-      "learning_rate": 9.954696564017481e-06,
-      "loss": 0.4033,
-      "step": 2848
-    },
-    {
-      "epoch": 0.18619698058950396,
-      "grad_norm": 0.5206222534179688,
-      "learning_rate": 9.954649652091908e-06,
-      "loss": 0.4295,
-      "step": 2849
-    },
-    {
-      "epoch": 0.18626233579504609,
-      "grad_norm": 0.4605466425418854,
-      "learning_rate": 9.954602716000792e-06,
-      "loss": 0.3965,
-      "step": 2850
-    },
-    {
-      "epoch": 0.18632769100058819,
-      "grad_norm": 0.4901019334793091,
-      "learning_rate": 9.954555755744362e-06,
-      "loss": 0.4067,
-      "step": 2851
-    },
-    {
-      "epoch": 0.18639304620613031,
-      "grad_norm": 0.5255517959594727,
-      "learning_rate": 9.954508771322846e-06,
-      "loss": 0.4333,
-      "step": 2852
-    },
-    {
-      "epoch": 0.18645840141167244,
-      "grad_norm": 0.4537615180015564,
-      "learning_rate": 9.954461762736477e-06,
-      "loss": 0.369,
-      "step": 2853
-    },
-    {
-      "epoch": 0.18652375661721457,
-      "grad_norm": 0.4792485237121582,
-      "learning_rate": 9.954414729985478e-06,
-      "loss": 0.3713,
-      "step": 2854
-    },
-    {
-      "epoch": 0.18658911182275667,
-      "grad_norm": 0.4976639449596405,
-      "learning_rate": 9.954367673070085e-06,
-      "loss": 0.4186,
-      "step": 2855
-    },
-    {
-      "epoch": 0.1866544670282988,
-      "grad_norm": 0.49446138739585876,
-      "learning_rate": 9.954320591990522e-06,
-      "loss": 0.4391,
-      "step": 2856
-    },
-    {
-      "epoch": 0.18671982223384093,
-      "grad_norm": 0.5113913416862488,
-      "learning_rate": 9.954273486747022e-06,
-      "loss": 0.4553,
-      "step": 2857
-    },
-    {
-      "epoch": 0.18678517743938305,
-      "grad_norm": 0.480999231338501,
-      "learning_rate": 9.954226357339815e-06,
-      "loss": 0.4054,
-      "step": 2858
-    },
-    {
-      "epoch": 0.18685053264492515,
-      "grad_norm": 0.6251979470252991,
-      "learning_rate": 9.954179203769128e-06,
-      "loss": 0.3825,
-      "step": 2859
-    },
-    {
-      "epoch": 0.18691588785046728,
-      "grad_norm": 0.48471885919570923,
-      "learning_rate": 9.954132026035195e-06,
-      "loss": 0.4072,
-      "step": 2860
-    },
-    {
-      "epoch": 0.1869812430560094,
-      "grad_norm": 0.4829366207122803,
-      "learning_rate": 9.954084824138243e-06,
-      "loss": 0.3581,
-      "step": 2861
-    },
-    {
-      "epoch": 0.18704659826155154,
-      "grad_norm": 0.5017015933990479,
-      "learning_rate": 9.954037598078501e-06,
-      "loss": 0.4154,
-      "step": 2862
-    },
-    {
-      "epoch": 0.18711195346709367,
-      "grad_norm": 0.5041373372077942,
-      "learning_rate": 9.953990347856203e-06,
-      "loss": 0.417,
-      "step": 2863
-    },
-    {
-      "epoch": 0.18717730867263577,
-      "grad_norm": 0.5277778506278992,
-      "learning_rate": 9.95394307347158e-06,
-      "loss": 0.403,
-      "step": 2864
-    },
-    {
-      "epoch": 0.1872426638781779,
-      "grad_norm": 0.5407474637031555,
-      "learning_rate": 9.953895774924857e-06,
-      "loss": 0.4373,
-      "step": 2865
-    },
-    {
-      "epoch": 0.18730801908372002,
-      "grad_norm": 0.5035947561264038,
-      "learning_rate": 9.95384845221627e-06,
-      "loss": 0.3907,
-      "step": 2866
-    },
-    {
-      "epoch": 0.18737337428926215,
-      "grad_norm": 0.5016087889671326,
-      "learning_rate": 9.953801105346047e-06,
-      "loss": 0.4095,
-      "step": 2867
-    },
-    {
-      "epoch": 0.18743872949480425,
-      "grad_norm": 0.4977370798587799,
-      "learning_rate": 9.95375373431442e-06,
-      "loss": 0.3988,
-      "step": 2868
-    },
-    {
-      "epoch": 0.18750408470034638,
-      "grad_norm": 0.5564749836921692,
-      "learning_rate": 9.95370633912162e-06,
-      "loss": 0.4316,
-      "step": 2869
-    },
-    {
-      "epoch": 0.1875694399058885,
-      "grad_norm": 0.5515015125274658,
-      "learning_rate": 9.95365891976788e-06,
-      "loss": 0.438,
-      "step": 2870
-    },
-    {
-      "epoch": 0.18763479511143064,
-      "grad_norm": 0.5201045274734497,
-      "learning_rate": 9.953611476253427e-06,
-      "loss": 0.4856,
-      "step": 2871
-    },
-    {
-      "epoch": 0.18770015031697274,
-      "grad_norm": 0.5049932599067688,
-      "learning_rate": 9.953564008578494e-06,
-      "loss": 0.4252,
-      "step": 2872
-    },
-    {
-      "epoch": 0.18776550552251486,
-      "grad_norm": 0.5389835238456726,
-      "learning_rate": 9.953516516743316e-06,
-      "loss": 0.407,
-      "step": 2873
-    },
-    {
-      "epoch": 0.187830860728057,
-      "grad_norm": 0.5269778966903687,
-      "learning_rate": 9.953469000748119e-06,
-      "loss": 0.4253,
-      "step": 2874
-    },
-    {
-      "epoch": 0.18789621593359912,
-      "grad_norm": 0.5148095488548279,
-      "learning_rate": 9.95342146059314e-06,
-      "loss": 0.3965,
-      "step": 2875
-    },
-    {
-      "epoch": 0.18796157113914122,
-      "grad_norm": 0.49052777886390686,
-      "learning_rate": 9.953373896278606e-06,
-      "loss": 0.4362,
-      "step": 2876
-    },
-    {
-      "epoch": 0.18802692634468335,
-      "grad_norm": 0.5031803250312805,
-      "learning_rate": 9.953326307804754e-06,
-      "loss": 0.464,
-      "step": 2877
-    },
-    {
-      "epoch": 0.18809228155022548,
-      "grad_norm": 0.4811420142650604,
-      "learning_rate": 9.953278695171813e-06,
-      "loss": 0.3791,
-      "step": 2878
-    },
-    {
-      "epoch": 0.1881576367557676,
-      "grad_norm": 0.5202829241752625,
-      "learning_rate": 9.953231058380016e-06,
-      "loss": 0.4402,
-      "step": 2879
-    },
-    {
-      "epoch": 0.1882229919613097,
-      "grad_norm": 0.4753965437412262,
-      "learning_rate": 9.953183397429594e-06,
-      "loss": 0.4359,
-      "step": 2880
-    },
-    {
-      "epoch": 0.18828834716685183,
-      "grad_norm": 0.5079650282859802,
-      "learning_rate": 9.95313571232078e-06,
-      "loss": 0.4241,
-      "step": 2881
-    },
-    {
-      "epoch": 0.18835370237239396,
-      "grad_norm": 0.5138829946517944,
-      "learning_rate": 9.95308800305381e-06,
-      "loss": 0.4323,
-      "step": 2882
-    },
-    {
-      "epoch": 0.1884190575779361,
-      "grad_norm": 0.47650250792503357,
-      "learning_rate": 9.95304026962891e-06,
-      "loss": 0.3822,
-      "step": 2883
-    },
-    {
-      "epoch": 0.18848441278347822,
-      "grad_norm": 0.49669456481933594,
-      "learning_rate": 9.95299251204632e-06,
-      "loss": 0.4221,
-      "step": 2884
-    },
-    {
-      "epoch": 0.18854976798902032,
-      "grad_norm": 0.47730210423469543,
-      "learning_rate": 9.952944730306269e-06,
-      "loss": 0.4177,
-      "step": 2885
-    },
-    {
-      "epoch": 0.18861512319456245,
-      "grad_norm": 0.5034216046333313,
-      "learning_rate": 9.95289692440899e-06,
-      "loss": 0.4301,
-      "step": 2886
-    },
-    {
-      "epoch": 0.18868047840010457,
-      "grad_norm": 0.47410061955451965,
-      "learning_rate": 9.952849094354718e-06,
-      "loss": 0.4136,
-      "step": 2887
-    },
-    {
-      "epoch": 0.1887458336056467,
-      "grad_norm": 0.502212643623352,
-      "learning_rate": 9.952801240143683e-06,
-      "loss": 0.4182,
-      "step": 2888
-    },
-    {
-      "epoch": 0.1888111888111888,
-      "grad_norm": 0.47118884325027466,
-      "learning_rate": 9.952753361776122e-06,
-      "loss": 0.4227,
-      "step": 2889
-    },
-    {
-      "epoch": 0.18887654401673093,
-      "grad_norm": 0.49516400694847107,
-      "learning_rate": 9.952705459252268e-06,
-      "loss": 0.4109,
-      "step": 2890
-    },
-    {
-      "epoch": 0.18894189922227306,
-      "grad_norm": 0.4778333902359009,
-      "learning_rate": 9.952657532572351e-06,
-      "loss": 0.3907,
-      "step": 2891
-    },
-    {
-      "epoch": 0.18900725442781519,
-      "grad_norm": 0.5070626139640808,
-      "learning_rate": 9.95260958173661e-06,
-      "loss": 0.4215,
-      "step": 2892
-    },
-    {
-      "epoch": 0.18907260963335729,
-      "grad_norm": 0.4762381613254547,
-      "learning_rate": 9.952561606745276e-06,
-      "loss": 0.3843,
-      "step": 2893
-    },
-    {
-      "epoch": 0.18913796483889941,
-      "grad_norm": 0.5275992751121521,
-      "learning_rate": 9.952513607598582e-06,
-      "loss": 0.4819,
-      "step": 2894
-    },
-    {
-      "epoch": 0.18920332004444154,
-      "grad_norm": 0.47402840852737427,
-      "learning_rate": 9.952465584296764e-06,
-      "loss": 0.3812,
-      "step": 2895
-    },
-    {
-      "epoch": 0.18926867524998367,
-      "grad_norm": 0.4844459295272827,
-      "learning_rate": 9.952417536840056e-06,
-      "loss": 0.4263,
-      "step": 2896
-    },
-    {
-      "epoch": 0.18933403045552577,
-      "grad_norm": 0.4950529634952545,
-      "learning_rate": 9.952369465228692e-06,
-      "loss": 0.4427,
-      "step": 2897
-    },
-    {
-      "epoch": 0.1893993856610679,
-      "grad_norm": 0.4626738727092743,
-      "learning_rate": 9.952321369462906e-06,
-      "loss": 0.377,
-      "step": 2898
-    },
-    {
-      "epoch": 0.18946474086661003,
-      "grad_norm": 0.49446913599967957,
-      "learning_rate": 9.952273249542934e-06,
-      "loss": 0.4158,
-      "step": 2899
-    },
-    {
-      "epoch": 0.18953009607215215,
-      "grad_norm": 0.47827136516571045,
-      "learning_rate": 9.952225105469008e-06,
-      "loss": 0.4088,
-      "step": 2900
-    },
-    {
-      "epoch": 0.18959545127769425,
-      "grad_norm": 0.47055137157440186,
-      "learning_rate": 9.952176937241367e-06,
-      "loss": 0.3736,
-      "step": 2901
-    },
-    {
-      "epoch": 0.18966080648323638,
-      "grad_norm": 0.5122933983802795,
-      "learning_rate": 9.95212874486024e-06,
-      "loss": 0.4083,
-      "step": 2902
-    },
-    {
-      "epoch": 0.1897261616887785,
-      "grad_norm": 0.49091583490371704,
-      "learning_rate": 9.952080528325868e-06,
-      "loss": 0.4331,
-      "step": 2903
-    },
-    {
-      "epoch": 0.18979151689432064,
-      "grad_norm": 0.5090427994728088,
-      "learning_rate": 9.952032287638484e-06,
-      "loss": 0.3838,
-      "step": 2904
-    },
-    {
-      "epoch": 0.18985687209986277,
-      "grad_norm": 0.5443885326385498,
-      "learning_rate": 9.951984022798322e-06,
-      "loss": 0.47,
-      "step": 2905
-    },
-    {
-      "epoch": 0.18992222730540487,
-      "grad_norm": 0.4834836423397064,
-      "learning_rate": 9.95193573380562e-06,
-      "loss": 0.4003,
-      "step": 2906
-    },
-    {
-      "epoch": 0.189987582510947,
-      "grad_norm": 0.49716243147850037,
-      "learning_rate": 9.951887420660609e-06,
-      "loss": 0.3943,
-      "step": 2907
-    },
-    {
-      "epoch": 0.19005293771648912,
-      "grad_norm": 0.4701566994190216,
-      "learning_rate": 9.95183908336353e-06,
-      "loss": 0.3968,
-      "step": 2908
-    },
-    {
-      "epoch": 0.19011829292203125,
-      "grad_norm": 0.5198107957839966,
-      "learning_rate": 9.951790721914615e-06,
-      "loss": 0.484,
-      "step": 2909
-    },
-    {
-      "epoch": 0.19018364812757335,
-      "grad_norm": 0.5403050780296326,
-      "learning_rate": 9.951742336314101e-06,
-      "loss": 0.4833,
-      "step": 2910
-    },
-    {
-      "epoch": 0.19024900333311548,
-      "grad_norm": 0.9144891500473022,
-      "learning_rate": 9.951693926562225e-06,
-      "loss": 0.3682,
-      "step": 2911
-    },
-    {
-      "epoch": 0.1903143585386576,
-      "grad_norm": 0.5206704139709473,
-      "learning_rate": 9.951645492659222e-06,
-      "loss": 0.4376,
-      "step": 2912
-    },
-    {
-      "epoch": 0.19037971374419974,
-      "grad_norm": 0.48614439368247986,
-      "learning_rate": 9.95159703460533e-06,
-      "loss": 0.4016,
-      "step": 2913
-    },
-    {
-      "epoch": 0.19044506894974184,
-      "grad_norm": 0.526782214641571,
-      "learning_rate": 9.95154855240078e-06,
-      "loss": 0.4798,
-      "step": 2914
-    },
-    {
-      "epoch": 0.19051042415528396,
-      "grad_norm": 0.47169598937034607,
-      "learning_rate": 9.951500046045815e-06,
-      "loss": 0.3736,
-      "step": 2915
-    },
-    {
-      "epoch": 0.1905757793608261,
-      "grad_norm": 0.4730372428894043,
-      "learning_rate": 9.95145151554067e-06,
-      "loss": 0.4016,
-      "step": 2916
-    },
-    {
-      "epoch": 0.19064113456636822,
-      "grad_norm": 0.5132974982261658,
-      "learning_rate": 9.95140296088558e-06,
-      "loss": 0.4528,
-      "step": 2917
-    },
-    {
-      "epoch": 0.19070648977191032,
-      "grad_norm": 0.511885404586792,
-      "learning_rate": 9.95135438208078e-06,
-      "loss": 0.431,
-      "step": 2918
-    },
-    {
-      "epoch": 0.19077184497745245,
-      "grad_norm": 0.5020025968551636,
-      "learning_rate": 9.951305779126512e-06,
-      "loss": 0.421,
-      "step": 2919
-    },
-    {
-      "epoch": 0.19083720018299458,
-      "grad_norm": 0.4954475164413452,
-      "learning_rate": 9.951257152023008e-06,
-      "loss": 0.4234,
-      "step": 2920
-    },
-    {
-      "epoch": 0.1909025553885367,
-      "grad_norm": 0.4862518906593323,
-      "learning_rate": 9.951208500770509e-06,
-      "loss": 0.3815,
-      "step": 2921
-    },
-    {
-      "epoch": 0.1909679105940788,
-      "grad_norm": 0.4983312487602234,
-      "learning_rate": 9.951159825369251e-06,
-      "loss": 0.4354,
-      "step": 2922
-    },
-    {
-      "epoch": 0.19103326579962093,
-      "grad_norm": 0.5410768985748291,
-      "learning_rate": 9.951111125819472e-06,
-      "loss": 0.4781,
-      "step": 2923
-    },
-    {
-      "epoch": 0.19109862100516306,
-      "grad_norm": 0.46137362718582153,
-      "learning_rate": 9.951062402121407e-06,
-      "loss": 0.3955,
-      "step": 2924
-    },
-    {
-      "epoch": 0.1911639762107052,
-      "grad_norm": 0.4605845510959625,
-      "learning_rate": 9.951013654275297e-06,
-      "loss": 0.3869,
-      "step": 2925
-    },
-    {
-      "epoch": 0.19122933141624732,
-      "grad_norm": 0.4990957975387573,
-      "learning_rate": 9.950964882281378e-06,
-      "loss": 0.4599,
-      "step": 2926
-    },
-    {
-      "epoch": 0.19129468662178942,
-      "grad_norm": 0.46196988224983215,
-      "learning_rate": 9.950916086139888e-06,
-      "loss": 0.3958,
-      "step": 2927
-    },
-    {
-      "epoch": 0.19136004182733155,
-      "grad_norm": 0.5079265832901001,
-      "learning_rate": 9.950867265851065e-06,
-      "loss": 0.4342,
-      "step": 2928
-    },
-    {
-      "epoch": 0.19142539703287367,
-      "grad_norm": 0.49464908242225647,
-      "learning_rate": 9.950818421415146e-06,
-      "loss": 0.3949,
-      "step": 2929
-    },
-    {
-      "epoch": 0.1914907522384158,
-      "grad_norm": 0.491862028837204,
-      "learning_rate": 9.950769552832372e-06,
-      "loss": 0.429,
-      "step": 2930
-    },
-    {
-      "epoch": 0.1915561074439579,
-      "grad_norm": 0.4822671115398407,
-      "learning_rate": 9.95072066010298e-06,
-      "loss": 0.4328,
-      "step": 2931
-    },
-    {
-      "epoch": 0.19162146264950003,
-      "grad_norm": 0.49650049209594727,
-      "learning_rate": 9.950671743227206e-06,
-      "loss": 0.4325,
-      "step": 2932
-    },
-    {
-      "epoch": 0.19168681785504216,
-      "grad_norm": 0.4699820280075073,
-      "learning_rate": 9.950622802205295e-06,
-      "loss": 0.3996,
-      "step": 2933
-    },
-    {
-      "epoch": 0.19175217306058429,
-      "grad_norm": 0.48909181356430054,
-      "learning_rate": 9.950573837037478e-06,
-      "loss": 0.3859,
-      "step": 2934
-    },
-    {
-      "epoch": 0.19181752826612639,
-      "grad_norm": 0.4794492721557617,
-      "learning_rate": 9.950524847723997e-06,
-      "loss": 0.3664,
-      "step": 2935
-    },
-    {
-      "epoch": 0.19188288347166851,
-      "grad_norm": 0.5136559009552002,
-      "learning_rate": 9.950475834265093e-06,
-      "loss": 0.4798,
-      "step": 2936
-    },
-    {
-      "epoch": 0.19194823867721064,
-      "grad_norm": 0.4615060091018677,
-      "learning_rate": 9.950426796661004e-06,
-      "loss": 0.3884,
-      "step": 2937
-    },
-    {
-      "epoch": 0.19201359388275277,
-      "grad_norm": 0.47666165232658386,
-      "learning_rate": 9.950377734911966e-06,
-      "loss": 0.4247,
-      "step": 2938
-    },
-    {
-      "epoch": 0.19207894908829487,
-      "grad_norm": 0.4553219676017761,
-      "learning_rate": 9.950328649018223e-06,
-      "loss": 0.3678,
-      "step": 2939
-    },
-    {
-      "epoch": 0.192144304293837,
-      "grad_norm": 0.49629679322242737,
-      "learning_rate": 9.950279538980012e-06,
-      "loss": 0.4022,
-      "step": 2940
-    },
-    {
-      "epoch": 0.19220965949937913,
-      "grad_norm": 0.44633710384368896,
-      "learning_rate": 9.950230404797571e-06,
-      "loss": 0.3578,
-      "step": 2941
-    },
-    {
-      "epoch": 0.19227501470492125,
-      "grad_norm": 0.4926506280899048,
-      "learning_rate": 9.950181246471143e-06,
-      "loss": 0.4071,
-      "step": 2942
-    },
-    {
-      "epoch": 0.19234036991046335,
-      "grad_norm": 0.48583802580833435,
-      "learning_rate": 9.950132064000967e-06,
-      "loss": 0.4112,
-      "step": 2943
-    },
-    {
-      "epoch": 0.19240572511600548,
-      "grad_norm": 0.4745608866214752,
-      "learning_rate": 9.95008285738728e-06,
-      "loss": 0.3989,
-      "step": 2944
-    },
-    {
-      "epoch": 0.1924710803215476,
-      "grad_norm": 0.4737425744533539,
-      "learning_rate": 9.950033626630324e-06,
-      "loss": 0.4038,
-      "step": 2945
-    },
-    {
-      "epoch": 0.19253643552708974,
-      "grad_norm": 0.4979565143585205,
-      "learning_rate": 9.94998437173034e-06,
-      "loss": 0.3971,
-      "step": 2946
-    },
-    {
-      "epoch": 0.19260179073263187,
-      "grad_norm": 0.49709323048591614,
-      "learning_rate": 9.949935092687566e-06,
-      "loss": 0.4702,
-      "step": 2947
-    },
-    {
-      "epoch": 0.19266714593817397,
-      "grad_norm": 0.45512285828590393,
-      "learning_rate": 9.949885789502246e-06,
-      "loss": 0.3935,
-      "step": 2948
-    },
-    {
-      "epoch": 0.1927325011437161,
-      "grad_norm": 0.5323231220245361,
-      "learning_rate": 9.949836462174618e-06,
-      "loss": 0.4397,
-      "step": 2949
-    },
-    {
-      "epoch": 0.19279785634925822,
-      "grad_norm": 0.5058055520057678,
-      "learning_rate": 9.949787110704921e-06,
-      "loss": 0.4414,
-      "step": 2950
-    },
-    {
-      "epoch": 0.19286321155480035,
-      "grad_norm": 0.4774121344089508,
-      "learning_rate": 9.949737735093398e-06,
-      "loss": 0.3904,
-      "step": 2951
-    },
-    {
-      "epoch": 0.19292856676034245,
-      "grad_norm": 0.4506870210170746,
-      "learning_rate": 9.94968833534029e-06,
-      "loss": 0.3784,
-      "step": 2952
-    },
-    {
-      "epoch": 0.19299392196588458,
-      "grad_norm": 0.4744364321231842,
-      "learning_rate": 9.949638911445838e-06,
-      "loss": 0.3831,
-      "step": 2953
-    },
-    {
-      "epoch": 0.1930592771714267,
-      "grad_norm": 0.5002384781837463,
-      "learning_rate": 9.94958946341028e-06,
-      "loss": 0.4185,
-      "step": 2954
-    },
-    {
-      "epoch": 0.19312463237696884,
-      "grad_norm": 0.49564677476882935,
-      "learning_rate": 9.949539991233863e-06,
-      "loss": 0.4085,
-      "step": 2955
-    },
-    {
-      "epoch": 0.19318998758251094,
-      "grad_norm": 0.5562955737113953,
-      "learning_rate": 9.949490494916822e-06,
-      "loss": 0.4773,
-      "step": 2956
-    },
-    {
-      "epoch": 0.19325534278805306,
-      "grad_norm": 0.4735763967037201,
-      "learning_rate": 9.9494409744594e-06,
-      "loss": 0.3838,
-      "step": 2957
-    },
-    {
-      "epoch": 0.1933206979935952,
-      "grad_norm": 0.4815181791782379,
-      "learning_rate": 9.949391429861843e-06,
-      "loss": 0.4107,
-      "step": 2958
-    },
-    {
-      "epoch": 0.19338605319913732,
-      "grad_norm": 0.4927268624305725,
-      "learning_rate": 9.94934186112439e-06,
-      "loss": 0.4567,
-      "step": 2959
-    },
-    {
-      "epoch": 0.19345140840467942,
-      "grad_norm": 0.4868094325065613,
-      "learning_rate": 9.949292268247279e-06,
-      "loss": 0.4535,
-      "step": 2960
-    },
-    {
-      "epoch": 0.19351676361022155,
-      "grad_norm": 0.4772026836872101,
-      "learning_rate": 9.949242651230756e-06,
-      "loss": 0.4221,
-      "step": 2961
-    },
-    {
-      "epoch": 0.19358211881576368,
-      "grad_norm": 0.49434521794319153,
-      "learning_rate": 9.949193010075063e-06,
-      "loss": 0.439,
-      "step": 2962
-    },
-    {
-      "epoch": 0.1936474740213058,
-      "grad_norm": 0.4620160758495331,
-      "learning_rate": 9.949143344780438e-06,
-      "loss": 0.4054,
-      "step": 2963
-    },
-    {
-      "epoch": 0.1937128292268479,
-      "grad_norm": 0.46887150406837463,
-      "learning_rate": 9.949093655347128e-06,
-      "loss": 0.4319,
-      "step": 2964
-    },
-    {
-      "epoch": 0.19377818443239003,
-      "grad_norm": 0.5082137584686279,
-      "learning_rate": 9.949043941775376e-06,
-      "loss": 0.4468,
-      "step": 2965
-    },
-    {
-      "epoch": 0.19384353963793216,
-      "grad_norm": 0.5001013278961182,
-      "learning_rate": 9.94899420406542e-06,
-      "loss": 0.4206,
-      "step": 2966
-    },
-    {
-      "epoch": 0.1939088948434743,
-      "grad_norm": 0.4783688187599182,
-      "learning_rate": 9.948944442217505e-06,
-      "loss": 0.4112,
-      "step": 2967
-    },
-    {
-      "epoch": 0.19397425004901642,
-      "grad_norm": 0.4642834961414337,
-      "learning_rate": 9.948894656231873e-06,
-      "loss": 0.4158,
-      "step": 2968
-    },
-    {
-      "epoch": 0.19403960525455852,
-      "grad_norm": 0.49629834294319153,
-      "learning_rate": 9.948844846108769e-06,
-      "loss": 0.4633,
-      "step": 2969
-    },
-    {
-      "epoch": 0.19410496046010065,
-      "grad_norm": 0.5198444724082947,
-      "learning_rate": 9.948795011848434e-06,
-      "loss": 0.4758,
-      "step": 2970
-    },
-    {
-      "epoch": 0.19417031566564277,
-      "grad_norm": 0.4808283746242523,
-      "learning_rate": 9.94874515345111e-06,
-      "loss": 0.4345,
-      "step": 2971
-    },
-    {
-      "epoch": 0.1942356708711849,
-      "grad_norm": 0.4733369052410126,
-      "learning_rate": 9.948695270917042e-06,
-      "loss": 0.4091,
-      "step": 2972
-    },
-    {
-      "epoch": 0.194301026076727,
-      "grad_norm": 0.49000704288482666,
-      "learning_rate": 9.948645364246473e-06,
-      "loss": 0.4366,
-      "step": 2973
-    },
-    {
-      "epoch": 0.19436638128226913,
-      "grad_norm": 0.49197790026664734,
-      "learning_rate": 9.948595433439645e-06,
-      "loss": 0.445,
-      "step": 2974
-    },
-    {
-      "epoch": 0.19443173648781126,
-      "grad_norm": 0.5117666125297546,
-      "learning_rate": 9.948545478496804e-06,
-      "loss": 0.4724,
-      "step": 2975
-    },
-    {
-      "epoch": 0.19449709169335339,
-      "grad_norm": 0.5310428142547607,
-      "learning_rate": 9.94849549941819e-06,
-      "loss": 0.4818,
-      "step": 2976
-    },
-    {
-      "epoch": 0.19456244689889549,
-      "grad_norm": 0.5212787389755249,
-      "learning_rate": 9.948445496204053e-06,
-      "loss": 0.4772,
-      "step": 2977
-    },
-    {
-      "epoch": 0.1946278021044376,
-      "grad_norm": 0.48359397053718567,
-      "learning_rate": 9.948395468854631e-06,
-      "loss": 0.4264,
-      "step": 2978
-    },
-    {
-      "epoch": 0.19469315730997974,
-      "grad_norm": 0.45742395520210266,
-      "learning_rate": 9.948345417370171e-06,
-      "loss": 0.3659,
-      "step": 2979
-    },
-    {
-      "epoch": 0.19475851251552187,
-      "grad_norm": 0.48544666171073914,
-      "learning_rate": 9.948295341750915e-06,
-      "loss": 0.407,
-      "step": 2980
-    },
-    {
-      "epoch": 0.19482386772106397,
-      "grad_norm": 0.4681014120578766,
-      "learning_rate": 9.948245241997109e-06,
-      "loss": 0.3803,
-      "step": 2981
-    },
-    {
-      "epoch": 0.1948892229266061,
-      "grad_norm": 0.4730170965194702,
-      "learning_rate": 9.948195118108997e-06,
-      "loss": 0.4058,
-      "step": 2982
-    },
-    {
-      "epoch": 0.19495457813214823,
-      "grad_norm": 0.5104221105575562,
-      "learning_rate": 9.948144970086822e-06,
-      "loss": 0.4847,
-      "step": 2983
-    },
-    {
-      "epoch": 0.19501993333769035,
-      "grad_norm": 0.47281453013420105,
-      "learning_rate": 9.94809479793083e-06,
-      "loss": 0.4039,
-      "step": 2984
-    },
-    {
-      "epoch": 0.19508528854323245,
-      "grad_norm": 0.4699348509311676,
-      "learning_rate": 9.948044601641266e-06,
-      "loss": 0.3886,
-      "step": 2985
-    },
-    {
-      "epoch": 0.19515064374877458,
-      "grad_norm": 0.488786518573761,
-      "learning_rate": 9.947994381218373e-06,
-      "loss": 0.4694,
-      "step": 2986
-    },
-    {
-      "epoch": 0.1952159989543167,
-      "grad_norm": 0.5252128839492798,
-      "learning_rate": 9.9479441366624e-06,
-      "loss": 0.416,
-      "step": 2987
-    },
-    {
-      "epoch": 0.19528135415985884,
-      "grad_norm": 0.5052991509437561,
-      "learning_rate": 9.947893867973586e-06,
-      "loss": 0.4482,
-      "step": 2988
-    },
-    {
-      "epoch": 0.19534670936540097,
-      "grad_norm": 0.47287124395370483,
-      "learning_rate": 9.947843575152182e-06,
-      "loss": 0.3973,
-      "step": 2989
-    },
-    {
-      "epoch": 0.19541206457094307,
-      "grad_norm": 0.4809347987174988,
-      "learning_rate": 9.94779325819843e-06,
-      "loss": 0.3785,
-      "step": 2990
-    },
-    {
-      "epoch": 0.1954774197764852,
-      "grad_norm": 0.4892086982727051,
-      "learning_rate": 9.947742917112577e-06,
-      "loss": 0.4301,
-      "step": 2991
-    },
-    {
-      "epoch": 0.19554277498202732,
-      "grad_norm": 0.49457159638404846,
-      "learning_rate": 9.947692551894867e-06,
-      "loss": 0.4404,
-      "step": 2992
-    },
-    {
-      "epoch": 0.19560813018756945,
-      "grad_norm": 0.4926825761795044,
-      "learning_rate": 9.947642162545546e-06,
-      "loss": 0.3892,
-      "step": 2993
-    },
-    {
-      "epoch": 0.19567348539311155,
-      "grad_norm": 0.49925148487091064,
-      "learning_rate": 9.94759174906486e-06,
-      "loss": 0.4404,
-      "step": 2994
-    },
-    {
-      "epoch": 0.19573884059865368,
-      "grad_norm": 0.5001039505004883,
-      "learning_rate": 9.947541311453056e-06,
-      "loss": 0.4217,
-      "step": 2995
-    },
-    {
-      "epoch": 0.1958041958041958,
-      "grad_norm": 0.4720327854156494,
-      "learning_rate": 9.947490849710378e-06,
-      "loss": 0.4309,
-      "step": 2996
-    },
-    {
-      "epoch": 0.19586955100973794,
-      "grad_norm": 0.4850936532020569,
-      "learning_rate": 9.947440363837073e-06,
-      "loss": 0.3798,
-      "step": 2997
-    },
-    {
-      "epoch": 0.19593490621528004,
-      "grad_norm": 0.5073444247245789,
-      "learning_rate": 9.947389853833389e-06,
-      "loss": 0.4547,
-      "step": 2998
-    },
-    {
-      "epoch": 0.19600026142082216,
-      "grad_norm": 0.46684056520462036,
-      "learning_rate": 9.94733931969957e-06,
-      "loss": 0.402,
-      "step": 2999
-    },
-    {
-      "epoch": 0.1960656166263643,
-      "grad_norm": 0.5287966132164001,
-      "learning_rate": 9.947288761435863e-06,
-      "loss": 0.4397,
-      "step": 3000
-    },
-    {
-      "epoch": 0.19613097183190642,
-      "grad_norm": 0.47407302260398865,
-      "learning_rate": 9.947238179042515e-06,
-      "loss": 0.4266,
-      "step": 3001
-    },
-    {
-      "epoch": 0.19619632703744852,
-      "grad_norm": 0.5004463195800781,
-      "learning_rate": 9.947187572519772e-06,
-      "loss": 0.3904,
-      "step": 3002
-    },
-    {
-      "epoch": 0.19626168224299065,
-      "grad_norm": 0.4812561273574829,
-      "learning_rate": 9.947136941867881e-06,
-      "loss": 0.4051,
-      "step": 3003
-    },
-    {
-      "epoch": 0.19632703744853278,
-      "grad_norm": 0.47599342465400696,
-      "learning_rate": 9.947086287087091e-06,
-      "loss": 0.3976,
-      "step": 3004
-    },
-    {
-      "epoch": 0.1963923926540749,
-      "grad_norm": 0.4551607370376587,
-      "learning_rate": 9.947035608177646e-06,
-      "loss": 0.3557,
-      "step": 3005
-    },
-    {
-      "epoch": 0.19645774785961703,
-      "grad_norm": 0.4981020390987396,
-      "learning_rate": 9.946984905139793e-06,
-      "loss": 0.4282,
-      "step": 3006
-    },
-    {
-      "epoch": 0.19652310306515913,
-      "grad_norm": 0.48331186175346375,
-      "learning_rate": 9.946934177973783e-06,
-      "loss": 0.4099,
-      "step": 3007
-    },
-    {
-      "epoch": 0.19658845827070126,
-      "grad_norm": 0.47015613317489624,
-      "learning_rate": 9.946883426679862e-06,
-      "loss": 0.4361,
-      "step": 3008
-    },
-    {
-      "epoch": 0.1966538134762434,
-      "grad_norm": 0.45380699634552,
-      "learning_rate": 9.946832651258277e-06,
-      "loss": 0.4162,
-      "step": 3009
-    },
-    {
-      "epoch": 0.19671916868178552,
-      "grad_norm": 0.4749274253845215,
-      "learning_rate": 9.946781851709273e-06,
-      "loss": 0.3918,
-      "step": 3010
-    },
-    {
-      "epoch": 0.19678452388732762,
-      "grad_norm": 0.4943445920944214,
-      "learning_rate": 9.946731028033102e-06,
-      "loss": 0.4397,
-      "step": 3011
-    },
-    {
-      "epoch": 0.19684987909286975,
-      "grad_norm": 0.5177479982376099,
-      "learning_rate": 9.946680180230008e-06,
-      "loss": 0.441,
-      "step": 3012
-    },
-    {
-      "epoch": 0.19691523429841187,
-      "grad_norm": 0.43935540318489075,
-      "learning_rate": 9.946629308300242e-06,
-      "loss": 0.3378,
-      "step": 3013
-    },
-    {
-      "epoch": 0.196980589503954,
-      "grad_norm": 0.47517162561416626,
-      "learning_rate": 9.946578412244053e-06,
-      "loss": 0.4125,
-      "step": 3014
-    },
-    {
-      "epoch": 0.1970459447094961,
-      "grad_norm": 0.4808422327041626,
-      "learning_rate": 9.946527492061686e-06,
-      "loss": 0.4199,
-      "step": 3015
-    },
-    {
-      "epoch": 0.19711129991503823,
-      "grad_norm": 0.48982810974121094,
-      "learning_rate": 9.94647654775339e-06,
-      "loss": 0.4366,
-      "step": 3016
-    },
-    {
-      "epoch": 0.19717665512058036,
-      "grad_norm": 0.45494890213012695,
-      "learning_rate": 9.946425579319414e-06,
-      "loss": 0.377,
-      "step": 3017
-    },
-    {
-      "epoch": 0.19724201032612249,
-      "grad_norm": 0.46580860018730164,
-      "learning_rate": 9.946374586760008e-06,
-      "loss": 0.3914,
-      "step": 3018
-    },
-    {
-      "epoch": 0.19730736553166459,
-      "grad_norm": 0.48293864727020264,
-      "learning_rate": 9.94632357007542e-06,
-      "loss": 0.4335,
-      "step": 3019
-    },
-    {
-      "epoch": 0.1973727207372067,
-      "grad_norm": 0.48486408591270447,
-      "learning_rate": 9.946272529265898e-06,
-      "loss": 0.4141,
-      "step": 3020
-    },
-    {
-      "epoch": 0.19743807594274884,
-      "grad_norm": 0.4494650363922119,
-      "learning_rate": 9.946221464331692e-06,
-      "loss": 0.3404,
-      "step": 3021
-    },
-    {
-      "epoch": 0.19750343114829097,
-      "grad_norm": 0.48047712445259094,
-      "learning_rate": 9.946170375273047e-06,
-      "loss": 0.4371,
-      "step": 3022
-    },
-    {
-      "epoch": 0.19756878635383307,
-      "grad_norm": 0.5063717365264893,
-      "learning_rate": 9.946119262090218e-06,
-      "loss": 0.4458,
-      "step": 3023
-    },
-    {
-      "epoch": 0.1976341415593752,
-      "grad_norm": 0.46469295024871826,
-      "learning_rate": 9.946068124783454e-06,
-      "loss": 0.4195,
-      "step": 3024
-    },
-    {
-      "epoch": 0.19769949676491733,
-      "grad_norm": 0.49715572595596313,
-      "learning_rate": 9.946016963353e-06,
-      "loss": 0.4232,
-      "step": 3025
-    },
-    {
-      "epoch": 0.19776485197045945,
-      "grad_norm": 0.467678427696228,
-      "learning_rate": 9.945965777799107e-06,
-      "loss": 0.4288,
-      "step": 3026
-    },
-    {
-      "epoch": 0.19783020717600158,
-      "grad_norm": 0.5184245705604553,
-      "learning_rate": 9.945914568122026e-06,
-      "loss": 0.4764,
-      "step": 3027
-    },
-    {
-      "epoch": 0.19789556238154368,
-      "grad_norm": 0.48791640996932983,
-      "learning_rate": 9.945863334322008e-06,
-      "loss": 0.3664,
-      "step": 3028
-    },
-    {
-      "epoch": 0.1979609175870858,
-      "grad_norm": 0.4951387941837311,
-      "learning_rate": 9.945812076399299e-06,
-      "loss": 0.4006,
-      "step": 3029
-    },
-    {
-      "epoch": 0.19802627279262794,
-      "grad_norm": 0.49412912130355835,
-      "learning_rate": 9.945760794354152e-06,
-      "loss": 0.4226,
-      "step": 3030
-    },
-    {
-      "epoch": 0.19809162799817007,
-      "grad_norm": 0.43828338384628296,
-      "learning_rate": 9.945709488186817e-06,
-      "loss": 0.3588,
-      "step": 3031
-    },
-    {
-      "epoch": 0.19815698320371217,
-      "grad_norm": 0.46683600544929504,
-      "learning_rate": 9.945658157897542e-06,
-      "loss": 0.4145,
-      "step": 3032
-    },
-    {
-      "epoch": 0.1982223384092543,
-      "grad_norm": 0.47453755140304565,
-      "learning_rate": 9.945606803486578e-06,
-      "loss": 0.4173,
-      "step": 3033
-    },
-    {
-      "epoch": 0.19828769361479642,
-      "grad_norm": 0.4920898377895355,
-      "learning_rate": 9.945555424954179e-06,
-      "loss": 0.4405,
-      "step": 3034
-    },
-    {
-      "epoch": 0.19835304882033855,
-      "grad_norm": 0.4704682230949402,
-      "learning_rate": 9.945504022300591e-06,
-      "loss": 0.3836,
-      "step": 3035
-    },
-    {
-      "epoch": 0.19841840402588065,
-      "grad_norm": 0.48477762937545776,
-      "learning_rate": 9.945452595526067e-06,
-      "loss": 0.3873,
-      "step": 3036
-    },
-    {
-      "epoch": 0.19848375923142278,
-      "grad_norm": 0.48065242171287537,
-      "learning_rate": 9.945401144630858e-06,
-      "loss": 0.3985,
-      "step": 3037
-    },
-    {
-      "epoch": 0.1985491144369649,
-      "grad_norm": 0.49803870916366577,
-      "learning_rate": 9.945349669615214e-06,
-      "loss": 0.45,
-      "step": 3038
-    },
-    {
-      "epoch": 0.19861446964250704,
-      "grad_norm": 0.5130840539932251,
-      "learning_rate": 9.945298170479388e-06,
-      "loss": 0.4503,
-      "step": 3039
-    },
-    {
-      "epoch": 0.19867982484804914,
-      "grad_norm": 0.49603888392448425,
-      "learning_rate": 9.945246647223626e-06,
-      "loss": 0.4267,
-      "step": 3040
-    },
-    {
-      "epoch": 0.19874518005359126,
-      "grad_norm": 0.4903644323348999,
-      "learning_rate": 9.945195099848185e-06,
-      "loss": 0.4383,
-      "step": 3041
-    },
-    {
-      "epoch": 0.1988105352591334,
-      "grad_norm": 0.5150305032730103,
-      "learning_rate": 9.945143528353315e-06,
-      "loss": 0.4184,
-      "step": 3042
-    },
-    {
-      "epoch": 0.19887589046467552,
-      "grad_norm": 0.5166090130805969,
-      "learning_rate": 9.945091932739266e-06,
-      "loss": 0.3575,
-      "step": 3043
-    },
-    {
-      "epoch": 0.19894124567021762,
-      "grad_norm": 0.5110989809036255,
-      "learning_rate": 9.94504031300629e-06,
-      "loss": 0.4477,
-      "step": 3044
-    },
-    {
-      "epoch": 0.19900660087575975,
-      "grad_norm": 0.5148541927337646,
-      "learning_rate": 9.94498866915464e-06,
-      "loss": 0.4442,
-      "step": 3045
-    },
-    {
-      "epoch": 0.19907195608130188,
-      "grad_norm": 0.5030977725982666,
-      "learning_rate": 9.944937001184567e-06,
-      "loss": 0.4418,
-      "step": 3046
-    },
-    {
-      "epoch": 0.199137311286844,
-      "grad_norm": 0.5268440246582031,
-      "learning_rate": 9.944885309096323e-06,
-      "loss": 0.4133,
-      "step": 3047
-    },
-    {
-      "epoch": 0.19920266649238613,
-      "grad_norm": 0.49540185928344727,
-      "learning_rate": 9.94483359289016e-06,
-      "loss": 0.451,
-      "step": 3048
-    },
-    {
-      "epoch": 0.19926802169792823,
-      "grad_norm": 0.4891604483127594,
-      "learning_rate": 9.944781852566332e-06,
-      "loss": 0.4171,
-      "step": 3049
-    },
-    {
-      "epoch": 0.19933337690347036,
-      "grad_norm": 0.48614418506622314,
-      "learning_rate": 9.944730088125088e-06,
-      "loss": 0.3818,
-      "step": 3050
-    },
-    {
-      "epoch": 0.1993987321090125,
-      "grad_norm": 0.4653158485889435,
-      "learning_rate": 9.944678299566683e-06,
-      "loss": 0.4341,
-      "step": 3051
-    },
-    {
-      "epoch": 0.19946408731455462,
-      "grad_norm": 0.5147649645805359,
-      "learning_rate": 9.94462648689137e-06,
-      "loss": 0.4217,
-      "step": 3052
-    },
-    {
-      "epoch": 0.19952944252009672,
-      "grad_norm": 0.4894815981388092,
-      "learning_rate": 9.944574650099398e-06,
-      "loss": 0.4613,
-      "step": 3053
-    },
-    {
-      "epoch": 0.19959479772563885,
-      "grad_norm": 0.4685996472835541,
-      "learning_rate": 9.944522789191025e-06,
-      "loss": 0.4172,
-      "step": 3054
-    },
-    {
-      "epoch": 0.19966015293118097,
-      "grad_norm": 0.4641586244106293,
-      "learning_rate": 9.944470904166501e-06,
-      "loss": 0.3646,
-      "step": 3055
-    },
-    {
-      "epoch": 0.1997255081367231,
-      "grad_norm": 0.7271279096603394,
-      "learning_rate": 9.94441899502608e-06,
-      "loss": 0.4043,
-      "step": 3056
-    },
-    {
-      "epoch": 0.1997908633422652,
-      "grad_norm": 0.4952391982078552,
-      "learning_rate": 9.944367061770012e-06,
-      "loss": 0.4222,
-      "step": 3057
-    },
-    {
-      "epoch": 0.19985621854780733,
-      "grad_norm": 0.4307084381580353,
-      "learning_rate": 9.944315104398556e-06,
-      "loss": 0.3664,
-      "step": 3058
-    },
-    {
-      "epoch": 0.19992157375334946,
-      "grad_norm": 0.5306459665298462,
-      "learning_rate": 9.94426312291196e-06,
-      "loss": 0.4592,
-      "step": 3059
-    },
-    {
-      "epoch": 0.19998692895889159,
-      "grad_norm": 0.4990803301334381,
-      "learning_rate": 9.94421111731048e-06,
-      "loss": 0.4043,
-      "step": 3060
-    },
-    {
-      "epoch": 0.20005228416443369,
-      "grad_norm": 0.45972147583961487,
-      "learning_rate": 9.944159087594372e-06,
-      "loss": 0.3694,
-      "step": 3061
-    },
-    {
-      "epoch": 0.2001176393699758,
-      "grad_norm": 0.4808419942855835,
-      "learning_rate": 9.944107033763886e-06,
-      "loss": 0.4052,
-      "step": 3062
-    },
-    {
-      "epoch": 0.20018299457551794,
-      "grad_norm": 0.5073785781860352,
-      "learning_rate": 9.944054955819275e-06,
-      "loss": 0.4372,
-      "step": 3063
-    },
-    {
-      "epoch": 0.20024834978106007,
-      "grad_norm": 0.5173817873001099,
-      "learning_rate": 9.944002853760798e-06,
-      "loss": 0.4257,
-      "step": 3064
-    },
-    {
-      "epoch": 0.20031370498660217,
-      "grad_norm": 0.5190584063529968,
-      "learning_rate": 9.943950727588706e-06,
-      "loss": 0.4242,
-      "step": 3065
-    },
-    {
-      "epoch": 0.2003790601921443,
-      "grad_norm": 0.4993712306022644,
-      "learning_rate": 9.943898577303253e-06,
-      "loss": 0.3869,
-      "step": 3066
-    },
-    {
-      "epoch": 0.20044441539768643,
-      "grad_norm": 0.5062596201896667,
-      "learning_rate": 9.943846402904693e-06,
-      "loss": 0.4487,
-      "step": 3067
-    },
-    {
-      "epoch": 0.20050977060322855,
-      "grad_norm": 0.47708597779273987,
-      "learning_rate": 9.943794204393282e-06,
-      "loss": 0.424,
-      "step": 3068
-    },
-    {
-      "epoch": 0.20057512580877068,
-      "grad_norm": 0.470600426197052,
-      "learning_rate": 9.943741981769275e-06,
-      "loss": 0.4098,
-      "step": 3069
-    },
-    {
-      "epoch": 0.20064048101431278,
-      "grad_norm": 0.47591736912727356,
-      "learning_rate": 9.943689735032926e-06,
-      "loss": 0.3667,
-      "step": 3070
-    },
-    {
-      "epoch": 0.2007058362198549,
-      "grad_norm": 0.48058611154556274,
-      "learning_rate": 9.943637464184488e-06,
-      "loss": 0.3966,
-      "step": 3071
-    },
-    {
-      "epoch": 0.20077119142539704,
-      "grad_norm": 0.4728534519672394,
-      "learning_rate": 9.943585169224216e-06,
-      "loss": 0.4437,
-      "step": 3072
-    },
-    {
-      "epoch": 0.20083654663093917,
-      "grad_norm": 0.4847133457660675,
-      "learning_rate": 9.94353285015237e-06,
-      "loss": 0.4125,
-      "step": 3073
-    },
-    {
-      "epoch": 0.20090190183648127,
-      "grad_norm": 0.5271087288856506,
-      "learning_rate": 9.9434805069692e-06,
-      "loss": 0.4308,
-      "step": 3074
-    },
-    {
-      "epoch": 0.2009672570420234,
-      "grad_norm": 0.4755455553531647,
-      "learning_rate": 9.943428139674963e-06,
-      "loss": 0.3931,
-      "step": 3075
-    },
-    {
-      "epoch": 0.20103261224756552,
-      "grad_norm": 0.466139018535614,
-      "learning_rate": 9.943375748269914e-06,
-      "loss": 0.378,
-      "step": 3076
-    },
-    {
-      "epoch": 0.20109796745310765,
-      "grad_norm": 0.47393783926963806,
-      "learning_rate": 9.94332333275431e-06,
-      "loss": 0.4291,
-      "step": 3077
-    },
-    {
-      "epoch": 0.20116332265864975,
-      "grad_norm": 0.5035730004310608,
-      "learning_rate": 9.943270893128405e-06,
-      "loss": 0.4132,
-      "step": 3078
-    },
-    {
-      "epoch": 0.20122867786419188,
-      "grad_norm": 0.49975332617759705,
-      "learning_rate": 9.943218429392456e-06,
-      "loss": 0.3801,
-      "step": 3079
-    },
-    {
-      "epoch": 0.201294033069734,
-      "grad_norm": 0.523624837398529,
-      "learning_rate": 9.943165941546717e-06,
-      "loss": 0.4561,
-      "step": 3080
-    },
-    {
-      "epoch": 0.20135938827527614,
-      "grad_norm": 0.5602368712425232,
-      "learning_rate": 9.943113429591445e-06,
-      "loss": 0.4861,
-      "step": 3081
-    },
-    {
-      "epoch": 0.20142474348081824,
-      "grad_norm": 0.4771738052368164,
-      "learning_rate": 9.943060893526896e-06,
-      "loss": 0.4053,
-      "step": 3082
-    },
-    {
-      "epoch": 0.20149009868636036,
-      "grad_norm": 0.4846934676170349,
-      "learning_rate": 9.94300833335333e-06,
-      "loss": 0.4158,
-      "step": 3083
-    },
-    {
-      "epoch": 0.2015554538919025,
-      "grad_norm": 0.47020387649536133,
-      "learning_rate": 9.942955749070996e-06,
-      "loss": 0.3534,
-      "step": 3084
-    },
-    {
-      "epoch": 0.20162080909744462,
-      "grad_norm": 0.4568374454975128,
-      "learning_rate": 9.942903140680157e-06,
-      "loss": 0.3812,
-      "step": 3085
-    },
-    {
-      "epoch": 0.20168616430298672,
-      "grad_norm": 0.45790353417396545,
-      "learning_rate": 9.942850508181065e-06,
-      "loss": 0.3674,
-      "step": 3086
-    },
-    {
-      "epoch": 0.20175151950852885,
-      "grad_norm": 0.6400083303451538,
-      "learning_rate": 9.942797851573982e-06,
-      "loss": 0.3992,
-      "step": 3087
-    },
-    {
-      "epoch": 0.20181687471407098,
-      "grad_norm": 0.44926542043685913,
-      "learning_rate": 9.942745170859158e-06,
-      "loss": 0.4104,
-      "step": 3088
-    },
-    {
-      "epoch": 0.2018822299196131,
-      "grad_norm": 0.5154117345809937,
-      "learning_rate": 9.942692466036854e-06,
-      "loss": 0.4709,
-      "step": 3089
-    },
-    {
-      "epoch": 0.20194758512515523,
-      "grad_norm": 0.5078647136688232,
-      "learning_rate": 9.942639737107327e-06,
-      "loss": 0.4232,
-      "step": 3090
-    },
-    {
-      "epoch": 0.20201294033069733,
-      "grad_norm": 0.4890657961368561,
-      "learning_rate": 9.942586984070834e-06,
-      "loss": 0.4083,
-      "step": 3091
-    },
-    {
-      "epoch": 0.20207829553623946,
-      "grad_norm": 0.4696938991546631,
-      "learning_rate": 9.942534206927631e-06,
-      "loss": 0.3894,
-      "step": 3092
-    },
-    {
-      "epoch": 0.2021436507417816,
-      "grad_norm": 0.48734769225120544,
-      "learning_rate": 9.94248140567798e-06,
-      "loss": 0.4047,
-      "step": 3093
-    },
-    {
-      "epoch": 0.20220900594732372,
-      "grad_norm": 0.4499852955341339,
-      "learning_rate": 9.942428580322132e-06,
-      "loss": 0.3395,
-      "step": 3094
-    },
-    {
-      "epoch": 0.20227436115286582,
-      "grad_norm": 0.48252198100090027,
-      "learning_rate": 9.942375730860347e-06,
-      "loss": 0.4034,
-      "step": 3095
-    },
-    {
-      "epoch": 0.20233971635840795,
-      "grad_norm": 0.5231744050979614,
-      "learning_rate": 9.942322857292886e-06,
-      "loss": 0.4517,
-      "step": 3096
-    },
-    {
-      "epoch": 0.20240507156395007,
-      "grad_norm": 0.4870540201663971,
-      "learning_rate": 9.94226995962e-06,
-      "loss": 0.45,
-      "step": 3097
-    },
-    {
-      "epoch": 0.2024704267694922,
-      "grad_norm": 0.5164012908935547,
-      "learning_rate": 9.942217037841955e-06,
-      "loss": 0.4722,
-      "step": 3098
-    },
-    {
-      "epoch": 0.2025357819750343,
-      "grad_norm": 0.4839988350868225,
-      "learning_rate": 9.942164091959004e-06,
-      "loss": 0.4341,
-      "step": 3099
-    },
-    {
-      "epoch": 0.20260113718057643,
-      "grad_norm": 0.5070463418960571,
-      "learning_rate": 9.942111121971407e-06,
-      "loss": 0.4233,
-      "step": 3100
-    },
-    {
-      "epoch": 0.20266649238611856,
-      "grad_norm": 0.44807514548301697,
-      "learning_rate": 9.942058127879421e-06,
-      "loss": 0.3667,
-      "step": 3101
-    },
-    {
-      "epoch": 0.20273184759166069,
-      "grad_norm": 0.4492264688014984,
-      "learning_rate": 9.942005109683305e-06,
-      "loss": 0.3564,
-      "step": 3102
-    },
-    {
-      "epoch": 0.20279720279720279,
-      "grad_norm": 0.5319750308990479,
-      "learning_rate": 9.94195206738332e-06,
-      "loss": 0.4389,
-      "step": 3103
-    },
-    {
-      "epoch": 0.2028625580027449,
-      "grad_norm": 0.5499125719070435,
-      "learning_rate": 9.941899000979722e-06,
-      "loss": 0.4612,
-      "step": 3104
-    },
-    {
-      "epoch": 0.20292791320828704,
-      "grad_norm": 0.5554568767547607,
-      "learning_rate": 9.94184591047277e-06,
-      "loss": 0.5242,
-      "step": 3105
-    },
-    {
-      "epoch": 0.20299326841382917,
-      "grad_norm": 0.5147949457168579,
-      "learning_rate": 9.941792795862723e-06,
-      "loss": 0.4334,
-      "step": 3106
-    },
-    {
-      "epoch": 0.20305862361937127,
-      "grad_norm": 0.4976714551448822,
-      "learning_rate": 9.941739657149843e-06,
-      "loss": 0.4439,
-      "step": 3107
-    },
-    {
-      "epoch": 0.2031239788249134,
-      "grad_norm": 0.4922333359718323,
-      "learning_rate": 9.941686494334384e-06,
-      "loss": 0.4226,
-      "step": 3108
-    },
-    {
-      "epoch": 0.20318933403045553,
-      "grad_norm": 0.5006389021873474,
-      "learning_rate": 9.941633307416609e-06,
-      "loss": 0.4031,
-      "step": 3109
-    },
-    {
-      "epoch": 0.20325468923599765,
-      "grad_norm": 0.4451974630355835,
-      "learning_rate": 9.941580096396776e-06,
-      "loss": 0.3712,
-      "step": 3110
-    },
-    {
-      "epoch": 0.20332004444153978,
-      "grad_norm": 0.4745657444000244,
-      "learning_rate": 9.941526861275146e-06,
-      "loss": 0.4287,
-      "step": 3111
-    },
-    {
-      "epoch": 0.20338539964708188,
-      "grad_norm": 0.5245086550712585,
-      "learning_rate": 9.941473602051978e-06,
-      "loss": 0.4101,
-      "step": 3112
-    },
-    {
-      "epoch": 0.203450754852624,
-      "grad_norm": 0.44496220350265503,
-      "learning_rate": 9.94142031872753e-06,
-      "loss": 0.3706,
-      "step": 3113
-    },
-    {
-      "epoch": 0.20351611005816614,
-      "grad_norm": 0.505380392074585,
-      "learning_rate": 9.941367011302063e-06,
-      "loss": 0.3862,
-      "step": 3114
-    },
-    {
-      "epoch": 0.20358146526370827,
-      "grad_norm": 0.4921588599681854,
-      "learning_rate": 9.94131367977584e-06,
-      "loss": 0.4421,
-      "step": 3115
-    },
-    {
-      "epoch": 0.20364682046925037,
-      "grad_norm": 0.46233102679252625,
-      "learning_rate": 9.941260324149114e-06,
-      "loss": 0.3797,
-      "step": 3116
-    },
-    {
-      "epoch": 0.2037121756747925,
-      "grad_norm": 0.528011679649353,
-      "learning_rate": 9.941206944422153e-06,
-      "loss": 0.488,
-      "step": 3117
-    },
-    {
-      "epoch": 0.20377753088033462,
-      "grad_norm": 0.5147208571434021,
-      "learning_rate": 9.941153540595211e-06,
-      "loss": 0.4665,
-      "step": 3118
-    },
-    {
-      "epoch": 0.20384288608587675,
-      "grad_norm": 0.554547131061554,
-      "learning_rate": 9.941100112668554e-06,
-      "loss": 0.45,
-      "step": 3119
-    },
-    {
-      "epoch": 0.20390824129141885,
-      "grad_norm": 0.4568560719490051,
-      "learning_rate": 9.94104666064244e-06,
-      "loss": 0.3713,
-      "step": 3120
-    },
-    {
-      "epoch": 0.20397359649696098,
-      "grad_norm": 0.46647655963897705,
-      "learning_rate": 9.940993184517126e-06,
-      "loss": 0.4206,
-      "step": 3121
-    },
-    {
-      "epoch": 0.2040389517025031,
-      "grad_norm": 0.454797625541687,
-      "learning_rate": 9.940939684292881e-06,
-      "loss": 0.38,
-      "step": 3122
-    },
-    {
-      "epoch": 0.20410430690804524,
-      "grad_norm": 0.5423403978347778,
-      "learning_rate": 9.94088615996996e-06,
-      "loss": 0.4573,
-      "step": 3123
-    },
-    {
-      "epoch": 0.20416966211358734,
-      "grad_norm": 0.48902904987335205,
-      "learning_rate": 9.940832611548625e-06,
-      "loss": 0.4239,
-      "step": 3124
-    },
-    {
-      "epoch": 0.20423501731912946,
-      "grad_norm": 0.4641076326370239,
-      "learning_rate": 9.940779039029138e-06,
-      "loss": 0.3552,
-      "step": 3125
-    },
-    {
-      "epoch": 0.2043003725246716,
-      "grad_norm": 0.47573062777519226,
-      "learning_rate": 9.94072544241176e-06,
-      "loss": 0.4293,
-      "step": 3126
-    },
-    {
-      "epoch": 0.20436572773021372,
-      "grad_norm": 0.5493038892745972,
-      "learning_rate": 9.940671821696752e-06,
-      "loss": 0.4896,
-      "step": 3127
-    },
-    {
-      "epoch": 0.20443108293575582,
-      "grad_norm": 0.46159788966178894,
-      "learning_rate": 9.940618176884376e-06,
-      "loss": 0.3674,
-      "step": 3128
-    },
-    {
-      "epoch": 0.20449643814129795,
-      "grad_norm": 0.5037823915481567,
-      "learning_rate": 9.940564507974895e-06,
-      "loss": 0.4155,
-      "step": 3129
-    },
-    {
-      "epoch": 0.20456179334684008,
-      "grad_norm": 0.47430261969566345,
-      "learning_rate": 9.940510814968567e-06,
-      "loss": 0.4106,
-      "step": 3130
-    },
-    {
-      "epoch": 0.2046271485523822,
-      "grad_norm": 0.44317811727523804,
-      "learning_rate": 9.940457097865656e-06,
-      "loss": 0.3631,
-      "step": 3131
-    },
-    {
-      "epoch": 0.20469250375792433,
-      "grad_norm": 0.44111067056655884,
-      "learning_rate": 9.940403356666427e-06,
-      "loss": 0.3688,
-      "step": 3132
-    },
-    {
-      "epoch": 0.20475785896346643,
-      "grad_norm": 0.5026799440383911,
-      "learning_rate": 9.940349591371137e-06,
-      "loss": 0.5078,
-      "step": 3133
-    },
-    {
-      "epoch": 0.20482321416900856,
-      "grad_norm": 0.47906067967414856,
-      "learning_rate": 9.94029580198005e-06,
-      "loss": 0.4116,
-      "step": 3134
-    },
-    {
-      "epoch": 0.2048885693745507,
-      "grad_norm": 0.5190334916114807,
-      "learning_rate": 9.940241988493432e-06,
-      "loss": 0.4791,
-      "step": 3135
-    },
-    {
-      "epoch": 0.20495392458009282,
-      "grad_norm": 0.45592185854911804,
-      "learning_rate": 9.940188150911541e-06,
-      "loss": 0.3882,
-      "step": 3136
-    },
-    {
-      "epoch": 0.20501927978563492,
-      "grad_norm": 0.4982894957065582,
-      "learning_rate": 9.94013428923464e-06,
-      "loss": 0.4061,
-      "step": 3137
-    },
-    {
-      "epoch": 0.20508463499117704,
-      "grad_norm": 0.430267333984375,
-      "learning_rate": 9.940080403462993e-06,
-      "loss": 0.3177,
-      "step": 3138
-    },
-    {
-      "epoch": 0.20514999019671917,
-      "grad_norm": 0.48851391673088074,
-      "learning_rate": 9.940026493596863e-06,
-      "loss": 0.4162,
-      "step": 3139
-    },
-    {
-      "epoch": 0.2052153454022613,
-      "grad_norm": 0.4630681276321411,
-      "learning_rate": 9.939972559636511e-06,
-      "loss": 0.4115,
-      "step": 3140
-    },
-    {
-      "epoch": 0.2052807006078034,
-      "grad_norm": 0.44982555508613586,
-      "learning_rate": 9.939918601582203e-06,
-      "loss": 0.3826,
-      "step": 3141
-    },
-    {
-      "epoch": 0.20534605581334553,
-      "grad_norm": 0.4807620644569397,
-      "learning_rate": 9.939864619434201e-06,
-      "loss": 0.3713,
-      "step": 3142
-    },
-    {
-      "epoch": 0.20541141101888766,
-      "grad_norm": 0.4503217041492462,
-      "learning_rate": 9.939810613192766e-06,
-      "loss": 0.3403,
-      "step": 3143
-    },
-    {
-      "epoch": 0.20547676622442979,
-      "grad_norm": 0.5286787748336792,
-      "learning_rate": 9.939756582858164e-06,
-      "loss": 0.4696,
-      "step": 3144
-    },
-    {
-      "epoch": 0.20554212142997189,
-      "grad_norm": 0.5015026330947876,
-      "learning_rate": 9.939702528430658e-06,
-      "loss": 0.4455,
-      "step": 3145
-    },
-    {
-      "epoch": 0.205607476635514,
-      "grad_norm": 0.5203262567520142,
-      "learning_rate": 9.939648449910513e-06,
-      "loss": 0.4345,
-      "step": 3146
-    },
-    {
-      "epoch": 0.20567283184105614,
-      "grad_norm": 0.48993492126464844,
-      "learning_rate": 9.93959434729799e-06,
-      "loss": 0.3685,
-      "step": 3147
-    },
-    {
-      "epoch": 0.20573818704659827,
-      "grad_norm": 0.5506002902984619,
-      "learning_rate": 9.939540220593353e-06,
-      "loss": 0.4938,
-      "step": 3148
-    },
-    {
-      "epoch": 0.20580354225214037,
-      "grad_norm": 0.4672248959541321,
-      "learning_rate": 9.939486069796869e-06,
-      "loss": 0.3733,
-      "step": 3149
-    },
-    {
-      "epoch": 0.2058688974576825,
-      "grad_norm": 0.5117464661598206,
-      "learning_rate": 9.9394318949088e-06,
-      "loss": 0.4284,
-      "step": 3150
-    },
-    {
-      "epoch": 0.20593425266322463,
-      "grad_norm": 0.5215551853179932,
-      "learning_rate": 9.939377695929409e-06,
-      "loss": 0.471,
-      "step": 3151
-    },
-    {
-      "epoch": 0.20599960786876675,
-      "grad_norm": 0.5296312570571899,
-      "learning_rate": 9.939323472858963e-06,
-      "loss": 0.4314,
-      "step": 3152
-    },
-    {
-      "epoch": 0.20606496307430888,
-      "grad_norm": 0.5131133794784546,
-      "learning_rate": 9.939269225697726e-06,
-      "loss": 0.4579,
-      "step": 3153
-    },
-    {
-      "epoch": 0.20613031827985098,
-      "grad_norm": 0.5022668838500977,
-      "learning_rate": 9.939214954445959e-06,
-      "loss": 0.4262,
-      "step": 3154
-    },
-    {
-      "epoch": 0.2061956734853931,
-      "grad_norm": 0.49072396755218506,
-      "learning_rate": 9.939160659103933e-06,
-      "loss": 0.4047,
-      "step": 3155
-    },
-    {
-      "epoch": 0.20626102869093524,
-      "grad_norm": 0.5091725587844849,
-      "learning_rate": 9.939106339671907e-06,
-      "loss": 0.422,
-      "step": 3156
-    },
-    {
-      "epoch": 0.20632638389647737,
-      "grad_norm": 0.4800470173358917,
-      "learning_rate": 9.939051996150149e-06,
-      "loss": 0.4108,
-      "step": 3157
-    },
-    {
-      "epoch": 0.20639173910201947,
-      "grad_norm": 0.514090359210968,
-      "learning_rate": 9.938997628538924e-06,
-      "loss": 0.4693,
-      "step": 3158
-    },
-    {
-      "epoch": 0.2064570943075616,
-      "grad_norm": 0.48027849197387695,
-      "learning_rate": 9.938943236838496e-06,
-      "loss": 0.4376,
-      "step": 3159
-    },
-    {
-      "epoch": 0.20652244951310372,
-      "grad_norm": 0.4613378345966339,
-      "learning_rate": 9.93888882104913e-06,
-      "loss": 0.4015,
-      "step": 3160
-    },
-    {
-      "epoch": 0.20658780471864585,
-      "grad_norm": 0.44781532883644104,
-      "learning_rate": 9.938834381171093e-06,
-      "loss": 0.3736,
-      "step": 3161
-    },
-    {
-      "epoch": 0.20665315992418795,
-      "grad_norm": 0.4846932291984558,
-      "learning_rate": 9.93877991720465e-06,
-      "loss": 0.362,
-      "step": 3162
-    },
-    {
-      "epoch": 0.20671851512973008,
-      "grad_norm": 0.49717938899993896,
-      "learning_rate": 9.938725429150066e-06,
-      "loss": 0.443,
-      "step": 3163
-    },
-    {
-      "epoch": 0.2067838703352722,
-      "grad_norm": 0.5367235541343689,
-      "learning_rate": 9.938670917007606e-06,
-      "loss": 0.4483,
-      "step": 3164
-    },
-    {
-      "epoch": 0.20684922554081434,
-      "grad_norm": 0.4786202013492584,
-      "learning_rate": 9.93861638077754e-06,
-      "loss": 0.4262,
-      "step": 3165
-    },
-    {
-      "epoch": 0.20691458074635644,
-      "grad_norm": 0.5516684651374817,
-      "learning_rate": 9.938561820460128e-06,
-      "loss": 0.5142,
-      "step": 3166
-    },
-    {
-      "epoch": 0.20697993595189856,
-      "grad_norm": 0.4452001452445984,
-      "learning_rate": 9.938507236055642e-06,
-      "loss": 0.3254,
-      "step": 3167
-    },
-    {
-      "epoch": 0.2070452911574407,
-      "grad_norm": 0.5052228569984436,
-      "learning_rate": 9.938452627564344e-06,
-      "loss": 0.4298,
-      "step": 3168
-    },
-    {
-      "epoch": 0.20711064636298282,
-      "grad_norm": 0.5003474354743958,
-      "learning_rate": 9.938397994986501e-06,
-      "loss": 0.4518,
-      "step": 3169
-    },
-    {
-      "epoch": 0.20717600156852492,
-      "grad_norm": 0.48262444138526917,
-      "learning_rate": 9.938343338322381e-06,
-      "loss": 0.4085,
-      "step": 3170
-    },
-    {
-      "epoch": 0.20724135677406705,
-      "grad_norm": 0.5018842816352844,
-      "learning_rate": 9.938288657572248e-06,
-      "loss": 0.4547,
-      "step": 3171
-    },
-    {
-      "epoch": 0.20730671197960918,
-      "grad_norm": 0.47284314036369324,
-      "learning_rate": 9.938233952736372e-06,
-      "loss": 0.3755,
-      "step": 3172
-    },
-    {
-      "epoch": 0.2073720671851513,
-      "grad_norm": 0.49318069219589233,
-      "learning_rate": 9.938179223815019e-06,
-      "loss": 0.4385,
-      "step": 3173
-    },
-    {
-      "epoch": 0.20743742239069343,
-      "grad_norm": 0.4700956642627716,
-      "learning_rate": 9.938124470808454e-06,
-      "loss": 0.3693,
-      "step": 3174
-    },
-    {
-      "epoch": 0.20750277759623553,
-      "grad_norm": 0.5164834260940552,
-      "learning_rate": 9.938069693716945e-06,
-      "loss": 0.4665,
-      "step": 3175
-    },
-    {
-      "epoch": 0.20756813280177766,
-      "grad_norm": 0.4650100767612457,
-      "learning_rate": 9.93801489254076e-06,
-      "loss": 0.4092,
-      "step": 3176
-    },
-    {
-      "epoch": 0.2076334880073198,
-      "grad_norm": 0.43761709332466125,
-      "learning_rate": 9.937960067280165e-06,
-      "loss": 0.3421,
-      "step": 3177
-    },
-    {
-      "epoch": 0.20769884321286192,
-      "grad_norm": 0.48198753595352173,
-      "learning_rate": 9.937905217935428e-06,
-      "loss": 0.4009,
-      "step": 3178
-    },
-    {
-      "epoch": 0.20776419841840402,
-      "grad_norm": 0.45960405468940735,
-      "learning_rate": 9.93785034450682e-06,
-      "loss": 0.3799,
-      "step": 3179
-    },
-    {
-      "epoch": 0.20782955362394614,
-      "grad_norm": 0.49930331110954285,
-      "learning_rate": 9.9377954469946e-06,
-      "loss": 0.3879,
-      "step": 3180
-    },
-    {
-      "epoch": 0.20789490882948827,
-      "grad_norm": 0.4426723122596741,
-      "learning_rate": 9.937740525399044e-06,
-      "loss": 0.3593,
-      "step": 3181
-    },
-    {
-      "epoch": 0.2079602640350304,
-      "grad_norm": 0.4870409667491913,
-      "learning_rate": 9.937685579720415e-06,
-      "loss": 0.4495,
-      "step": 3182
-    },
-    {
-      "epoch": 0.2080256192405725,
-      "grad_norm": 0.4347936809062958,
-      "learning_rate": 9.937630609958986e-06,
-      "loss": 0.3739,
-      "step": 3183
-    },
-    {
-      "epoch": 0.20809097444611463,
-      "grad_norm": 0.5160869359970093,
-      "learning_rate": 9.93757561611502e-06,
-      "loss": 0.4429,
-      "step": 3184
-    },
-    {
-      "epoch": 0.20815632965165676,
-      "grad_norm": 0.4967484772205353,
-      "learning_rate": 9.937520598188786e-06,
-      "loss": 0.4306,
-      "step": 3185
-    },
-    {
-      "epoch": 0.20822168485719889,
-      "grad_norm": 0.4874371290206909,
-      "learning_rate": 9.937465556180555e-06,
-      "loss": 0.4307,
-      "step": 3186
-    },
-    {
-      "epoch": 0.20828704006274099,
-      "grad_norm": 0.49783626198768616,
-      "learning_rate": 9.937410490090593e-06,
-      "loss": 0.4579,
-      "step": 3187
-    },
-    {
-      "epoch": 0.2083523952682831,
-      "grad_norm": 0.5056105852127075,
-      "learning_rate": 9.93735539991917e-06,
-      "loss": 0.4494,
-      "step": 3188
-    },
-    {
-      "epoch": 0.20841775047382524,
-      "grad_norm": 0.49673497676849365,
-      "learning_rate": 9.937300285666556e-06,
-      "loss": 0.4389,
-      "step": 3189
-    },
-    {
-      "epoch": 0.20848310567936737,
-      "grad_norm": 0.4490938186645508,
-      "learning_rate": 9.937245147333016e-06,
-      "loss": 0.3822,
-      "step": 3190
-    },
-    {
-      "epoch": 0.20854846088490947,
-      "grad_norm": 0.5379023551940918,
-      "learning_rate": 9.937189984918822e-06,
-      "loss": 0.3911,
-      "step": 3191
-    },
-    {
-      "epoch": 0.2086138160904516,
-      "grad_norm": 0.5370519161224365,
-      "learning_rate": 9.937134798424242e-06,
-      "loss": 0.454,
-      "step": 3192
-    },
-    {
-      "epoch": 0.20867917129599373,
-      "grad_norm": 0.4845854938030243,
-      "learning_rate": 9.937079587849545e-06,
-      "loss": 0.3959,
-      "step": 3193
-    },
-    {
-      "epoch": 0.20874452650153585,
-      "grad_norm": 0.4893931448459625,
-      "learning_rate": 9.937024353195e-06,
-      "loss": 0.4306,
-      "step": 3194
-    },
-    {
-      "epoch": 0.20880988170707798,
-      "grad_norm": 0.5111810564994812,
-      "learning_rate": 9.936969094460877e-06,
-      "loss": 0.4809,
-      "step": 3195
-    },
-    {
-      "epoch": 0.20887523691262008,
-      "grad_norm": 0.4792070686817169,
-      "learning_rate": 9.936913811647446e-06,
-      "loss": 0.4234,
-      "step": 3196
-    },
-    {
-      "epoch": 0.2089405921181622,
-      "grad_norm": 0.4581071436405182,
-      "learning_rate": 9.936858504754976e-06,
-      "loss": 0.3631,
-      "step": 3197
-    },
-    {
-      "epoch": 0.20900594732370434,
-      "grad_norm": 0.508477509021759,
-      "learning_rate": 9.936803173783735e-06,
-      "loss": 0.445,
-      "step": 3198
-    },
-    {
-      "epoch": 0.20907130252924647,
-      "grad_norm": 0.48508119583129883,
-      "learning_rate": 9.936747818733996e-06,
-      "loss": 0.4139,
-      "step": 3199
-    },
-    {
-      "epoch": 0.20913665773478857,
-      "grad_norm": 0.5168250203132629,
-      "learning_rate": 9.936692439606028e-06,
-      "loss": 0.427,
-      "step": 3200
-    },
-    {
-      "epoch": 0.2092020129403307,
-      "grad_norm": 0.4804687201976776,
-      "learning_rate": 9.936637036400101e-06,
-      "loss": 0.3888,
-      "step": 3201
-    },
-    {
-      "epoch": 0.20926736814587282,
-      "grad_norm": 0.48574692010879517,
-      "learning_rate": 9.936581609116485e-06,
-      "loss": 0.4211,
-      "step": 3202
-    },
-    {
-      "epoch": 0.20933272335141495,
-      "grad_norm": 0.5198602080345154,
-      "learning_rate": 9.936526157755448e-06,
-      "loss": 0.4395,
-      "step": 3203
-    },
-    {
-      "epoch": 0.20939807855695705,
-      "grad_norm": 0.5564453601837158,
-      "learning_rate": 9.936470682317265e-06,
-      "loss": 0.5146,
-      "step": 3204
-    },
-    {
-      "epoch": 0.20946343376249918,
-      "grad_norm": 0.4566188454627991,
-      "learning_rate": 9.936415182802203e-06,
-      "loss": 0.3731,
-      "step": 3205
-    },
-    {
-      "epoch": 0.2095287889680413,
-      "grad_norm": 0.5720825791358948,
-      "learning_rate": 9.936359659210537e-06,
-      "loss": 0.4609,
-      "step": 3206
-    },
-    {
-      "epoch": 0.20959414417358344,
-      "grad_norm": 0.4590819776058197,
-      "learning_rate": 9.936304111542532e-06,
-      "loss": 0.4009,
-      "step": 3207
-    },
-    {
-      "epoch": 0.20965949937912554,
-      "grad_norm": 0.5057934522628784,
-      "learning_rate": 9.936248539798462e-06,
-      "loss": 0.4047,
-      "step": 3208
-    },
-    {
-      "epoch": 0.20972485458466766,
-      "grad_norm": 0.49229344725608826,
-      "learning_rate": 9.936192943978597e-06,
-      "loss": 0.394,
-      "step": 3209
-    },
-    {
-      "epoch": 0.2097902097902098,
-      "grad_norm": 0.50483638048172,
-      "learning_rate": 9.93613732408321e-06,
-      "loss": 0.3974,
-      "step": 3210
-    },
-    {
-      "epoch": 0.20985556499575192,
-      "grad_norm": 0.492837131023407,
-      "learning_rate": 9.936081680112573e-06,
-      "loss": 0.3986,
-      "step": 3211
-    },
-    {
-      "epoch": 0.20992092020129402,
-      "grad_norm": 0.457077294588089,
-      "learning_rate": 9.936026012066952e-06,
-      "loss": 0.3967,
-      "step": 3212
-    },
-    {
-      "epoch": 0.20998627540683615,
-      "grad_norm": 0.4983299970626831,
-      "learning_rate": 9.935970319946627e-06,
-      "loss": 0.4083,
-      "step": 3213
-    },
-    {
-      "epoch": 0.21005163061237828,
-      "grad_norm": 0.49250268936157227,
-      "learning_rate": 9.93591460375186e-06,
-      "loss": 0.3713,
-      "step": 3214
-    },
-    {
-      "epoch": 0.2101169858179204,
-      "grad_norm": 0.5435276031494141,
-      "learning_rate": 9.93585886348293e-06,
-      "loss": 0.4754,
-      "step": 3215
-    },
-    {
-      "epoch": 0.21018234102346253,
-      "grad_norm": 0.49676281213760376,
-      "learning_rate": 9.935803099140106e-06,
-      "loss": 0.4596,
-      "step": 3216
-    },
-    {
-      "epoch": 0.21024769622900463,
-      "grad_norm": 0.4617408812046051,
-      "learning_rate": 9.93574731072366e-06,
-      "loss": 0.3884,
-      "step": 3217
-    },
-    {
-      "epoch": 0.21031305143454676,
-      "grad_norm": 0.5948168039321899,
-      "learning_rate": 9.935691498233864e-06,
-      "loss": 0.508,
-      "step": 3218
-    },
-    {
-      "epoch": 0.2103784066400889,
-      "grad_norm": 0.49031031131744385,
-      "learning_rate": 9.935635661670992e-06,
-      "loss": 0.4222,
-      "step": 3219
-    },
-    {
-      "epoch": 0.21044376184563102,
-      "grad_norm": 0.482632040977478,
-      "learning_rate": 9.935579801035314e-06,
-      "loss": 0.413,
-      "step": 3220
-    },
-    {
-      "epoch": 0.21050911705117312,
-      "grad_norm": 0.5207369327545166,
-      "learning_rate": 9.935523916327103e-06,
-      "loss": 0.4704,
-      "step": 3221
-    },
-    {
-      "epoch": 0.21057447225671524,
-      "grad_norm": 0.5311384797096252,
-      "learning_rate": 9.935468007546634e-06,
-      "loss": 0.5367,
-      "step": 3222
-    },
-    {
-      "epoch": 0.21063982746225737,
-      "grad_norm": 0.5066277980804443,
-      "learning_rate": 9.935412074694175e-06,
-      "loss": 0.4196,
-      "step": 3223
-    },
-    {
-      "epoch": 0.2107051826677995,
-      "grad_norm": 0.4708802402019501,
-      "learning_rate": 9.935356117770003e-06,
-      "loss": 0.3943,
-      "step": 3224
-    },
-    {
-      "epoch": 0.2107705378733416,
-      "grad_norm": 0.48777833580970764,
-      "learning_rate": 9.935300136774389e-06,
-      "loss": 0.4278,
-      "step": 3225
-    },
-    {
-      "epoch": 0.21083589307888373,
-      "grad_norm": 0.4551773965358734,
-      "learning_rate": 9.935244131707607e-06,
-      "loss": 0.379,
-      "step": 3226
-    },
-    {
-      "epoch": 0.21090124828442586,
-      "grad_norm": 0.4466029405593872,
-      "learning_rate": 9.93518810256993e-06,
-      "loss": 0.3787,
-      "step": 3227
-    },
-    {
-      "epoch": 0.21096660348996799,
-      "grad_norm": 0.46648839116096497,
-      "learning_rate": 9.935132049361631e-06,
-      "loss": 0.402,
-      "step": 3228
-    },
-    {
-      "epoch": 0.21103195869551009,
-      "grad_norm": 0.5277845859527588,
-      "learning_rate": 9.935075972082982e-06,
-      "loss": 0.4619,
-      "step": 3229
-    },
-    {
-      "epoch": 0.2110973139010522,
-      "grad_norm": 0.476776659488678,
-      "learning_rate": 9.93501987073426e-06,
-      "loss": 0.3646,
-      "step": 3230
-    },
-    {
-      "epoch": 0.21116266910659434,
-      "grad_norm": 0.4704614579677582,
-      "learning_rate": 9.934963745315733e-06,
-      "loss": 0.391,
-      "step": 3231
-    },
-    {
-      "epoch": 0.21122802431213647,
-      "grad_norm": 0.47129425406455994,
-      "learning_rate": 9.934907595827681e-06,
-      "loss": 0.3842,
-      "step": 3232
-    },
-    {
-      "epoch": 0.21129337951767857,
-      "grad_norm": 0.4966982305049896,
-      "learning_rate": 9.934851422270374e-06,
-      "loss": 0.4296,
-      "step": 3233
-    },
-    {
-      "epoch": 0.2113587347232207,
-      "grad_norm": 0.464455246925354,
-      "learning_rate": 9.934795224644088e-06,
-      "loss": 0.3967,
-      "step": 3234
-    },
-    {
-      "epoch": 0.21142408992876283,
-      "grad_norm": 0.5066690444946289,
-      "learning_rate": 9.934739002949094e-06,
-      "loss": 0.4719,
-      "step": 3235
-    },
-    {
-      "epoch": 0.21148944513430495,
-      "grad_norm": 0.4529123604297638,
-      "learning_rate": 9.93468275718567e-06,
-      "loss": 0.3636,
-      "step": 3236
-    },
-    {
-      "epoch": 0.21155480033984708,
-      "grad_norm": 0.49601155519485474,
-      "learning_rate": 9.934626487354088e-06,
-      "loss": 0.3912,
-      "step": 3237
-    },
-    {
-      "epoch": 0.21162015554538918,
-      "grad_norm": 0.4670489728450775,
-      "learning_rate": 9.934570193454625e-06,
-      "loss": 0.3863,
-      "step": 3238
-    },
-    {
-      "epoch": 0.2116855107509313,
-      "grad_norm": 0.5014293193817139,
-      "learning_rate": 9.934513875487552e-06,
-      "loss": 0.4399,
-      "step": 3239
-    },
-    {
-      "epoch": 0.21175086595647344,
-      "grad_norm": 0.49930065870285034,
-      "learning_rate": 9.934457533453144e-06,
-      "loss": 0.4593,
-      "step": 3240
-    },
-    {
-      "epoch": 0.21181622116201557,
-      "grad_norm": 0.48156407475471497,
-      "learning_rate": 9.93440116735168e-06,
-      "loss": 0.3905,
-      "step": 3241
-    },
-    {
-      "epoch": 0.21188157636755767,
-      "grad_norm": 0.5224270820617676,
-      "learning_rate": 9.93434477718343e-06,
-      "loss": 0.4378,
-      "step": 3242
-    },
-    {
-      "epoch": 0.2119469315730998,
-      "grad_norm": 0.4587031900882721,
-      "learning_rate": 9.934288362948672e-06,
-      "loss": 0.3758,
-      "step": 3243
-    },
-    {
-      "epoch": 0.21201228677864192,
-      "grad_norm": 0.46535447239875793,
-      "learning_rate": 9.93423192464768e-06,
-      "loss": 0.3866,
-      "step": 3244
-    },
-    {
-      "epoch": 0.21207764198418405,
-      "grad_norm": 0.4809655547142029,
-      "learning_rate": 9.934175462280729e-06,
-      "loss": 0.4016,
-      "step": 3245
-    },
-    {
-      "epoch": 0.21214299718972615,
-      "grad_norm": 0.4861283302307129,
-      "learning_rate": 9.934118975848094e-06,
-      "loss": 0.3678,
-      "step": 3246
-    },
-    {
-      "epoch": 0.21220835239526828,
-      "grad_norm": 0.47646233439445496,
-      "learning_rate": 9.934062465350053e-06,
-      "loss": 0.4075,
-      "step": 3247
-    },
-    {
-      "epoch": 0.2122737076008104,
-      "grad_norm": 0.5074921250343323,
-      "learning_rate": 9.934005930786881e-06,
-      "loss": 0.441,
-      "step": 3248
-    },
-    {
-      "epoch": 0.21233906280635254,
-      "grad_norm": 0.4906167984008789,
-      "learning_rate": 9.933949372158852e-06,
-      "loss": 0.411,
-      "step": 3249
-    },
-    {
-      "epoch": 0.21240441801189464,
-      "grad_norm": 0.48985132575035095,
-      "learning_rate": 9.933892789466242e-06,
-      "loss": 0.4294,
-      "step": 3250
-    },
-    {
-      "epoch": 0.21246977321743676,
-      "grad_norm": 0.509037435054779,
-      "learning_rate": 9.933836182709328e-06,
-      "loss": 0.4287,
-      "step": 3251
-    },
-    {
-      "epoch": 0.2125351284229789,
-      "grad_norm": 0.48478612303733826,
-      "learning_rate": 9.933779551888385e-06,
-      "loss": 0.426,
-      "step": 3252
-    },
-    {
-      "epoch": 0.21260048362852102,
-      "grad_norm": 0.46264371275901794,
-      "learning_rate": 9.93372289700369e-06,
-      "loss": 0.3676,
-      "step": 3253
-    },
-    {
-      "epoch": 0.21266583883406312,
-      "grad_norm": 0.45310333371162415,
-      "learning_rate": 9.933666218055522e-06,
-      "loss": 0.3736,
-      "step": 3254
-    },
-    {
-      "epoch": 0.21273119403960525,
-      "grad_norm": 0.48330405354499817,
-      "learning_rate": 9.93360951504415e-06,
-      "loss": 0.4298,
-      "step": 3255
-    },
-    {
-      "epoch": 0.21279654924514738,
-      "grad_norm": 0.49773460626602173,
-      "learning_rate": 9.93355278796986e-06,
-      "loss": 0.4021,
-      "step": 3256
-    },
-    {
-      "epoch": 0.2128619044506895,
-      "grad_norm": 0.4893500506877899,
-      "learning_rate": 9.93349603683292e-06,
-      "loss": 0.4424,
-      "step": 3257
-    },
-    {
-      "epoch": 0.21292725965623163,
-      "grad_norm": 0.49103978276252747,
-      "learning_rate": 9.933439261633612e-06,
-      "loss": 0.4454,
-      "step": 3258
-    },
-    {
-      "epoch": 0.21299261486177373,
-      "grad_norm": 0.5382106900215149,
-      "learning_rate": 9.933382462372212e-06,
-      "loss": 0.4142,
-      "step": 3259
-    },
-    {
-      "epoch": 0.21305797006731586,
-      "grad_norm": 0.5229787826538086,
-      "learning_rate": 9.933325639048996e-06,
-      "loss": 0.4539,
-      "step": 3260
-    },
-    {
-      "epoch": 0.213123325272858,
-      "grad_norm": 0.4758698344230652,
-      "learning_rate": 9.933268791664243e-06,
-      "loss": 0.4277,
-      "step": 3261
-    },
-    {
-      "epoch": 0.21318868047840012,
-      "grad_norm": 0.5039137601852417,
-      "learning_rate": 9.933211920218229e-06,
-      "loss": 0.4301,
-      "step": 3262
-    },
-    {
-      "epoch": 0.21325403568394222,
-      "grad_norm": 0.49445581436157227,
-      "learning_rate": 9.93315502471123e-06,
-      "loss": 0.4329,
-      "step": 3263
-    },
-    {
-      "epoch": 0.21331939088948434,
-      "grad_norm": 0.4823456108570099,
-      "learning_rate": 9.933098105143526e-06,
-      "loss": 0.3474,
-      "step": 3264
-    },
-    {
-      "epoch": 0.21338474609502647,
-      "grad_norm": 0.47261980175971985,
-      "learning_rate": 9.933041161515393e-06,
-      "loss": 0.41,
-      "step": 3265
-    },
-    {
-      "epoch": 0.2134501013005686,
-      "grad_norm": 0.48997393250465393,
-      "learning_rate": 9.93298419382711e-06,
-      "loss": 0.4375,
-      "step": 3266
-    },
-    {
-      "epoch": 0.2135154565061107,
-      "grad_norm": 0.47008273005485535,
-      "learning_rate": 9.932927202078952e-06,
-      "loss": 0.3836,
-      "step": 3267
-    },
-    {
-      "epoch": 0.21358081171165283,
-      "grad_norm": 0.5084006786346436,
-      "learning_rate": 9.932870186271202e-06,
-      "loss": 0.3936,
-      "step": 3268
-    },
-    {
-      "epoch": 0.21364616691719496,
-      "grad_norm": 0.47581514716148376,
-      "learning_rate": 9.932813146404134e-06,
-      "loss": 0.3835,
-      "step": 3269
-    },
-    {
-      "epoch": 0.21371152212273709,
-      "grad_norm": 0.4586402177810669,
-      "learning_rate": 9.932756082478028e-06,
-      "loss": 0.3853,
-      "step": 3270
-    },
-    {
-      "epoch": 0.21377687732827919,
-      "grad_norm": 0.45872631669044495,
-      "learning_rate": 9.932698994493159e-06,
-      "loss": 0.3721,
-      "step": 3271
-    },
-    {
-      "epoch": 0.2138422325338213,
-      "grad_norm": 0.5588338375091553,
-      "learning_rate": 9.932641882449811e-06,
-      "loss": 0.4109,
-      "step": 3272
-    },
-    {
-      "epoch": 0.21390758773936344,
-      "grad_norm": 0.5490834712982178,
-      "learning_rate": 9.93258474634826e-06,
-      "loss": 0.3999,
-      "step": 3273
-    },
-    {
-      "epoch": 0.21397294294490557,
-      "grad_norm": 0.4640160799026489,
-      "learning_rate": 9.932527586188783e-06,
-      "loss": 0.403,
-      "step": 3274
-    },
-    {
-      "epoch": 0.21403829815044767,
-      "grad_norm": 0.458261102437973,
-      "learning_rate": 9.93247040197166e-06,
-      "loss": 0.3818,
-      "step": 3275
-    },
-    {
-      "epoch": 0.2141036533559898,
-      "grad_norm": 0.4928989112377167,
-      "learning_rate": 9.93241319369717e-06,
-      "loss": 0.4057,
-      "step": 3276
-    },
-    {
-      "epoch": 0.21416900856153193,
-      "grad_norm": 0.49751630425453186,
-      "learning_rate": 9.932355961365594e-06,
-      "loss": 0.4303,
-      "step": 3277
-    },
-    {
-      "epoch": 0.21423436376707405,
-      "grad_norm": 0.47694024443626404,
-      "learning_rate": 9.932298704977207e-06,
-      "loss": 0.3878,
-      "step": 3278
-    },
-    {
-      "epoch": 0.21429971897261618,
-      "grad_norm": 0.5090917348861694,
-      "learning_rate": 9.932241424532291e-06,
-      "loss": 0.4125,
-      "step": 3279
-    },
-    {
-      "epoch": 0.21436507417815828,
-      "grad_norm": 0.4575689435005188,
-      "learning_rate": 9.932184120031125e-06,
-      "loss": 0.4127,
-      "step": 3280
-    },
-    {
-      "epoch": 0.2144304293837004,
-      "grad_norm": 0.5061025619506836,
-      "learning_rate": 9.93212679147399e-06,
-      "loss": 0.4034,
-      "step": 3281
-    },
-    {
-      "epoch": 0.21449578458924254,
-      "grad_norm": 0.4857165813446045,
-      "learning_rate": 9.932069438861163e-06,
-      "loss": 0.4119,
-      "step": 3282
-    },
-    {
-      "epoch": 0.21456113979478467,
-      "grad_norm": 0.44739657640457153,
-      "learning_rate": 9.932012062192923e-06,
-      "loss": 0.3425,
-      "step": 3283
-    },
-    {
-      "epoch": 0.21462649500032677,
-      "grad_norm": 0.4993947148323059,
-      "learning_rate": 9.931954661469553e-06,
-      "loss": 0.4619,
-      "step": 3284
-    },
-    {
-      "epoch": 0.2146918502058689,
-      "grad_norm": 0.473056823015213,
-      "learning_rate": 9.931897236691332e-06,
-      "loss": 0.4243,
-      "step": 3285
-    },
-    {
-      "epoch": 0.21475720541141102,
-      "grad_norm": 0.5073739290237427,
-      "learning_rate": 9.93183978785854e-06,
-      "loss": 0.456,
-      "step": 3286
-    },
-    {
-      "epoch": 0.21482256061695315,
-      "grad_norm": 0.486345112323761,
-      "learning_rate": 9.931782314971454e-06,
-      "loss": 0.394,
-      "step": 3287
-    },
-    {
-      "epoch": 0.21488791582249525,
-      "grad_norm": 0.4326786696910858,
-      "learning_rate": 9.931724818030359e-06,
-      "loss": 0.3208,
-      "step": 3288
-    },
-    {
-      "epoch": 0.21495327102803738,
-      "grad_norm": 0.501753032207489,
-      "learning_rate": 9.931667297035535e-06,
-      "loss": 0.4218,
-      "step": 3289
-    },
-    {
-      "epoch": 0.2150186262335795,
-      "grad_norm": 0.5065110921859741,
-      "learning_rate": 9.93160975198726e-06,
-      "loss": 0.4332,
-      "step": 3290
-    },
-    {
-      "epoch": 0.21508398143912164,
-      "grad_norm": 0.4691832661628723,
-      "learning_rate": 9.931552182885815e-06,
-      "loss": 0.3712,
-      "step": 3291
-    },
-    {
-      "epoch": 0.21514933664466374,
-      "grad_norm": 0.47420334815979004,
-      "learning_rate": 9.93149458973148e-06,
-      "loss": 0.384,
-      "step": 3292
-    },
-    {
-      "epoch": 0.21521469185020586,
-      "grad_norm": 0.538435161113739,
-      "learning_rate": 9.931436972524539e-06,
-      "loss": 0.4659,
-      "step": 3293
-    },
-    {
-      "epoch": 0.215280047055748,
-      "grad_norm": 0.48547956347465515,
-      "learning_rate": 9.931379331265272e-06,
-      "loss": 0.4189,
-      "step": 3294
-    },
-    {
-      "epoch": 0.21534540226129012,
-      "grad_norm": 0.5160083174705505,
-      "learning_rate": 9.931321665953961e-06,
-      "loss": 0.4416,
-      "step": 3295
-    },
-    {
-      "epoch": 0.21541075746683222,
-      "grad_norm": 0.5340427160263062,
-      "learning_rate": 9.931263976590883e-06,
-      "loss": 0.4075,
-      "step": 3296
-    },
-    {
-      "epoch": 0.21547611267237435,
-      "grad_norm": 0.4734879434108734,
-      "learning_rate": 9.931206263176325e-06,
-      "loss": 0.4118,
-      "step": 3297
-    },
-    {
-      "epoch": 0.21554146787791648,
-      "grad_norm": 0.46744245290756226,
-      "learning_rate": 9.931148525710563e-06,
-      "loss": 0.405,
-      "step": 3298
-    },
-    {
-      "epoch": 0.2156068230834586,
-      "grad_norm": 0.46901634335517883,
-      "learning_rate": 9.931090764193882e-06,
-      "loss": 0.3863,
-      "step": 3299
-    },
-    {
-      "epoch": 0.21567217828900073,
-      "grad_norm": 0.4528055787086487,
-      "learning_rate": 9.931032978626564e-06,
-      "loss": 0.3505,
-      "step": 3300
-    },
-    {
-      "epoch": 0.21573753349454283,
-      "grad_norm": 0.46030697226524353,
-      "learning_rate": 9.930975169008889e-06,
-      "loss": 0.3723,
-      "step": 3301
-    },
-    {
-      "epoch": 0.21580288870008496,
-      "grad_norm": 0.5202280879020691,
-      "learning_rate": 9.930917335341142e-06,
-      "loss": 0.4326,
-      "step": 3302
-    },
-    {
-      "epoch": 0.2158682439056271,
-      "grad_norm": 0.49749305844306946,
-      "learning_rate": 9.930859477623601e-06,
-      "loss": 0.4446,
-      "step": 3303
-    },
-    {
-      "epoch": 0.21593359911116922,
-      "grad_norm": 0.5041854977607727,
-      "learning_rate": 9.93080159585655e-06,
-      "loss": 0.4553,
-      "step": 3304
-    },
-    {
-      "epoch": 0.21599895431671132,
-      "grad_norm": 0.45173487067222595,
-      "learning_rate": 9.930743690040272e-06,
-      "loss": 0.3483,
-      "step": 3305
-    },
-    {
-      "epoch": 0.21606430952225344,
-      "grad_norm": 0.4890424311161041,
-      "learning_rate": 9.93068576017505e-06,
-      "loss": 0.3785,
-      "step": 3306
-    },
-    {
-      "epoch": 0.21612966472779557,
-      "grad_norm": 0.5173264741897583,
-      "learning_rate": 9.930627806261162e-06,
-      "loss": 0.4447,
-      "step": 3307
-    },
-    {
-      "epoch": 0.2161950199333377,
-      "grad_norm": 0.46183329820632935,
-      "learning_rate": 9.930569828298897e-06,
-      "loss": 0.4042,
-      "step": 3308
-    },
-    {
-      "epoch": 0.2162603751388798,
-      "grad_norm": 0.47356879711151123,
-      "learning_rate": 9.930511826288534e-06,
-      "loss": 0.417,
-      "step": 3309
-    },
-    {
-      "epoch": 0.21632573034442193,
-      "grad_norm": 0.4511333405971527,
-      "learning_rate": 9.930453800230358e-06,
-      "loss": 0.4107,
-      "step": 3310
-    },
-    {
-      "epoch": 0.21639108554996406,
-      "grad_norm": 0.4873167872428894,
-      "learning_rate": 9.930395750124648e-06,
-      "loss": 0.4136,
-      "step": 3311
-    },
-    {
-      "epoch": 0.21645644075550619,
-      "grad_norm": 0.4364887773990631,
-      "learning_rate": 9.93033767597169e-06,
-      "loss": 0.3903,
-      "step": 3312
-    },
-    {
-      "epoch": 0.21652179596104829,
-      "grad_norm": 0.49707648158073425,
-      "learning_rate": 9.93027957777177e-06,
-      "loss": 0.4272,
-      "step": 3313
-    },
-    {
-      "epoch": 0.2165871511665904,
-      "grad_norm": 0.5045496821403503,
-      "learning_rate": 9.930221455525167e-06,
-      "loss": 0.4627,
-      "step": 3314
-    },
-    {
-      "epoch": 0.21665250637213254,
-      "grad_norm": 0.4444250762462616,
-      "learning_rate": 9.930163309232167e-06,
-      "loss": 0.399,
-      "step": 3315
-    },
-    {
-      "epoch": 0.21671786157767467,
-      "grad_norm": 0.497793585062027,
-      "learning_rate": 9.93010513889305e-06,
-      "loss": 0.4363,
-      "step": 3316
-    },
-    {
-      "epoch": 0.21678321678321677,
-      "grad_norm": 0.4658908247947693,
-      "learning_rate": 9.930046944508104e-06,
-      "loss": 0.4037,
-      "step": 3317
-    },
-    {
-      "epoch": 0.2168485719887589,
-      "grad_norm": 0.4523892402648926,
-      "learning_rate": 9.92998872607761e-06,
-      "loss": 0.3937,
-      "step": 3318
-    },
-    {
-      "epoch": 0.21691392719430103,
-      "grad_norm": 0.47562769055366516,
-      "learning_rate": 9.929930483601855e-06,
-      "loss": 0.4209,
-      "step": 3319
-    },
-    {
-      "epoch": 0.21697928239984315,
-      "grad_norm": 0.48166099190711975,
-      "learning_rate": 9.92987221708112e-06,
-      "loss": 0.4187,
-      "step": 3320
-    },
-    {
-      "epoch": 0.21704463760538528,
-      "grad_norm": 0.4934748709201813,
-      "learning_rate": 9.92981392651569e-06,
-      "loss": 0.4222,
-      "step": 3321
-    },
-    {
-      "epoch": 0.21710999281092738,
-      "grad_norm": 0.5159883499145508,
-      "learning_rate": 9.92975561190585e-06,
-      "loss": 0.4605,
-      "step": 3322
-    },
-    {
-      "epoch": 0.2171753480164695,
-      "grad_norm": 0.5119197964668274,
-      "learning_rate": 9.929697273251884e-06,
-      "loss": 0.4384,
-      "step": 3323
-    },
-    {
-      "epoch": 0.21724070322201164,
-      "grad_norm": 0.501603364944458,
-      "learning_rate": 9.929638910554077e-06,
-      "loss": 0.4107,
-      "step": 3324
-    },
-    {
-      "epoch": 0.21730605842755377,
-      "grad_norm": 0.48651987314224243,
-      "learning_rate": 9.929580523812712e-06,
-      "loss": 0.4142,
-      "step": 3325
-    },
-    {
-      "epoch": 0.21737141363309587,
-      "grad_norm": 0.4588828980922699,
-      "learning_rate": 9.929522113028075e-06,
-      "loss": 0.3878,
-      "step": 3326
-    },
-    {
-      "epoch": 0.217436768838638,
-      "grad_norm": 0.4741341173648834,
-      "learning_rate": 9.929463678200452e-06,
-      "loss": 0.3995,
-      "step": 3327
-    },
-    {
-      "epoch": 0.21750212404418012,
-      "grad_norm": 0.49042367935180664,
-      "learning_rate": 9.929405219330127e-06,
-      "loss": 0.4059,
-      "step": 3328
-    },
-    {
-      "epoch": 0.21756747924972225,
-      "grad_norm": 0.48561954498291016,
-      "learning_rate": 9.929346736417387e-06,
-      "loss": 0.4061,
-      "step": 3329
-    },
-    {
-      "epoch": 0.21763283445526435,
-      "grad_norm": 0.46217429637908936,
-      "learning_rate": 9.929288229462513e-06,
-      "loss": 0.368,
-      "step": 3330
-    },
-    {
-      "epoch": 0.21769818966080648,
-      "grad_norm": 0.5045156478881836,
-      "learning_rate": 9.929229698465792e-06,
-      "loss": 0.4699,
-      "step": 3331
-    },
-    {
-      "epoch": 0.2177635448663486,
-      "grad_norm": 0.4598250985145569,
-      "learning_rate": 9.92917114342751e-06,
-      "loss": 0.4179,
-      "step": 3332
-    },
-    {
-      "epoch": 0.21782890007189074,
-      "grad_norm": 0.4896177053451538,
-      "learning_rate": 9.929112564347953e-06,
-      "loss": 0.4419,
-      "step": 3333
-    },
-    {
-      "epoch": 0.21789425527743284,
-      "grad_norm": 0.4439111649990082,
-      "learning_rate": 9.929053961227408e-06,
-      "loss": 0.3774,
-      "step": 3334
-    },
-    {
-      "epoch": 0.21795961048297496,
-      "grad_norm": 0.482452392578125,
-      "learning_rate": 9.928995334066158e-06,
-      "loss": 0.4369,
-      "step": 3335
-    },
-    {
-      "epoch": 0.2180249656885171,
-      "grad_norm": 0.43713170289993286,
-      "learning_rate": 9.92893668286449e-06,
-      "loss": 0.3457,
-      "step": 3336
-    },
-    {
-      "epoch": 0.21809032089405922,
-      "grad_norm": 0.4810910224914551,
-      "learning_rate": 9.92887800762269e-06,
-      "loss": 0.4441,
-      "step": 3337
-    },
-    {
-      "epoch": 0.21815567609960132,
-      "grad_norm": 0.44690775871276855,
-      "learning_rate": 9.928819308341048e-06,
-      "loss": 0.3682,
-      "step": 3338
-    },
-    {
-      "epoch": 0.21822103130514345,
-      "grad_norm": 0.4413868486881256,
-      "learning_rate": 9.92876058501984e-06,
-      "loss": 0.3701,
-      "step": 3339
-    },
-    {
-      "epoch": 0.21828638651068558,
-      "grad_norm": 0.5194119811058044,
-      "learning_rate": 9.928701837659365e-06,
-      "loss": 0.4488,
-      "step": 3340
-    },
-    {
-      "epoch": 0.2183517417162277,
-      "grad_norm": 0.4844612777233124,
-      "learning_rate": 9.9286430662599e-06,
-      "loss": 0.4242,
-      "step": 3341
-    },
-    {
-      "epoch": 0.21841709692176983,
-      "grad_norm": 0.5143688321113586,
-      "learning_rate": 9.928584270821737e-06,
-      "loss": 0.4585,
-      "step": 3342
-    },
-    {
-      "epoch": 0.21848245212731193,
-      "grad_norm": 0.4815845787525177,
-      "learning_rate": 9.928525451345162e-06,
-      "loss": 0.4028,
-      "step": 3343
-    },
-    {
-      "epoch": 0.21854780733285406,
-      "grad_norm": 0.5016213059425354,
-      "learning_rate": 9.92846660783046e-06,
-      "loss": 0.4078,
-      "step": 3344
-    },
-    {
-      "epoch": 0.2186131625383962,
-      "grad_norm": 0.45264023542404175,
-      "learning_rate": 9.92840774027792e-06,
-      "loss": 0.3467,
-      "step": 3345
-    },
-    {
-      "epoch": 0.21867851774393832,
-      "grad_norm": 0.5195870995521545,
-      "learning_rate": 9.928348848687825e-06,
-      "loss": 0.4393,
-      "step": 3346
-    },
-    {
-      "epoch": 0.21874387294948042,
-      "grad_norm": 0.5074870586395264,
-      "learning_rate": 9.928289933060469e-06,
-      "loss": 0.4728,
-      "step": 3347
-    },
-    {
-      "epoch": 0.21880922815502254,
-      "grad_norm": 0.48438355326652527,
-      "learning_rate": 9.928230993396134e-06,
-      "loss": 0.3779,
-      "step": 3348
-    },
-    {
-      "epoch": 0.21887458336056467,
-      "grad_norm": 0.4741329252719879,
-      "learning_rate": 9.92817202969511e-06,
-      "loss": 0.3937,
-      "step": 3349
-    },
-    {
-      "epoch": 0.2189399385661068,
-      "grad_norm": 0.5010062456130981,
-      "learning_rate": 9.928113041957682e-06,
-      "loss": 0.4131,
-      "step": 3350
-    },
-    {
-      "epoch": 0.2190052937716489,
-      "grad_norm": 0.53528892993927,
-      "learning_rate": 9.92805403018414e-06,
-      "loss": 0.4795,
-      "step": 3351
-    },
-    {
-      "epoch": 0.21907064897719103,
-      "grad_norm": 0.46200308203697205,
-      "learning_rate": 9.927994994374771e-06,
-      "loss": 0.4124,
-      "step": 3352
-    },
-    {
-      "epoch": 0.21913600418273316,
-      "grad_norm": 0.4657052457332611,
-      "learning_rate": 9.927935934529864e-06,
-      "loss": 0.382,
-      "step": 3353
-    },
-    {
-      "epoch": 0.21920135938827529,
-      "grad_norm": 0.5091944336891174,
-      "learning_rate": 9.927876850649706e-06,
-      "loss": 0.4547,
-      "step": 3354
-    },
-    {
-      "epoch": 0.21926671459381739,
-      "grad_norm": 0.4988209307193756,
-      "learning_rate": 9.927817742734585e-06,
-      "loss": 0.4231,
-      "step": 3355
-    },
-    {
-      "epoch": 0.2193320697993595,
-      "grad_norm": 0.4642221927642822,
-      "learning_rate": 9.927758610784791e-06,
-      "loss": 0.3752,
-      "step": 3356
-    },
-    {
-      "epoch": 0.21939742500490164,
-      "grad_norm": 0.47392427921295166,
-      "learning_rate": 9.92769945480061e-06,
-      "loss": 0.3678,
-      "step": 3357
-    },
-    {
-      "epoch": 0.21946278021044377,
-      "grad_norm": 0.4886663854122162,
-      "learning_rate": 9.92764027478233e-06,
-      "loss": 0.4318,
-      "step": 3358
-    },
-    {
-      "epoch": 0.21952813541598587,
-      "grad_norm": 0.4901321232318878,
-      "learning_rate": 9.927581070730244e-06,
-      "loss": 0.3917,
-      "step": 3359
-    },
-    {
-      "epoch": 0.219593490621528,
-      "grad_norm": 0.4995364546775818,
-      "learning_rate": 9.927521842644637e-06,
-      "loss": 0.4202,
-      "step": 3360
-    },
-    {
-      "epoch": 0.21965884582707013,
-      "grad_norm": 0.5176507830619812,
-      "learning_rate": 9.927462590525801e-06,
-      "loss": 0.4486,
-      "step": 3361
-    },
-    {
-      "epoch": 0.21972420103261225,
-      "grad_norm": 0.4921971261501312,
-      "learning_rate": 9.92740331437402e-06,
-      "loss": 0.4425,
-      "step": 3362
-    },
-    {
-      "epoch": 0.21978955623815438,
-      "grad_norm": 0.45165786147117615,
-      "learning_rate": 9.927344014189587e-06,
-      "loss": 0.3702,
-      "step": 3363
-    },
-    {
-      "epoch": 0.21985491144369648,
-      "grad_norm": 0.520483136177063,
-      "learning_rate": 9.92728468997279e-06,
-      "loss": 0.4732,
-      "step": 3364
-    },
-    {
-      "epoch": 0.2199202666492386,
-      "grad_norm": 0.5097583532333374,
-      "learning_rate": 9.927225341723918e-06,
-      "loss": 0.4181,
-      "step": 3365
-    },
-    {
-      "epoch": 0.21998562185478074,
-      "grad_norm": 0.4923904240131378,
-      "learning_rate": 9.927165969443262e-06,
-      "loss": 0.4158,
-      "step": 3366
-    },
-    {
-      "epoch": 0.22005097706032287,
-      "grad_norm": 0.4719454050064087,
-      "learning_rate": 9.927106573131112e-06,
-      "loss": 0.4026,
-      "step": 3367
-    },
-    {
-      "epoch": 0.22011633226586497,
-      "grad_norm": 0.4841327965259552,
-      "learning_rate": 9.927047152787754e-06,
-      "loss": 0.4223,
-      "step": 3368
-    },
-    {
-      "epoch": 0.2201816874714071,
-      "grad_norm": 0.488750696182251,
-      "learning_rate": 9.92698770841348e-06,
-      "loss": 0.4718,
-      "step": 3369
-    },
-    {
-      "epoch": 0.22024704267694922,
-      "grad_norm": 0.501806378364563,
-      "learning_rate": 9.926928240008583e-06,
-      "loss": 0.4108,
-      "step": 3370
-    },
-    {
-      "epoch": 0.22031239788249135,
-      "grad_norm": 0.4870474636554718,
-      "learning_rate": 9.926868747573348e-06,
-      "loss": 0.4144,
-      "step": 3371
-    },
-    {
-      "epoch": 0.22037775308803345,
-      "grad_norm": 0.5012169480323792,
-      "learning_rate": 9.926809231108068e-06,
-      "loss": 0.4378,
-      "step": 3372
-    },
-    {
-      "epoch": 0.22044310829357558,
-      "grad_norm": 0.5123572945594788,
-      "learning_rate": 9.926749690613031e-06,
-      "loss": 0.4185,
-      "step": 3373
-    },
-    {
-      "epoch": 0.2205084634991177,
-      "grad_norm": 0.47672030329704285,
-      "learning_rate": 9.92669012608853e-06,
-      "loss": 0.3625,
-      "step": 3374
-    },
-    {
-      "epoch": 0.22057381870465984,
-      "grad_norm": 0.5566274523735046,
-      "learning_rate": 9.926630537534855e-06,
-      "loss": 0.4724,
-      "step": 3375
-    },
-    {
-      "epoch": 0.22063917391020194,
-      "grad_norm": 0.46616724133491516,
-      "learning_rate": 9.926570924952295e-06,
-      "loss": 0.3786,
-      "step": 3376
-    },
-    {
-      "epoch": 0.22070452911574406,
-      "grad_norm": 0.48081544041633606,
-      "learning_rate": 9.926511288341143e-06,
-      "loss": 0.3832,
-      "step": 3377
-    },
-    {
-      "epoch": 0.2207698843212862,
-      "grad_norm": 0.5178582072257996,
-      "learning_rate": 9.926451627701687e-06,
-      "loss": 0.4417,
-      "step": 3378
-    },
-    {
-      "epoch": 0.22083523952682832,
-      "grad_norm": 0.4961230456829071,
-      "learning_rate": 9.926391943034222e-06,
-      "loss": 0.4278,
-      "step": 3379
-    },
-    {
-      "epoch": 0.22090059473237042,
-      "grad_norm": 0.4684832692146301,
-      "learning_rate": 9.926332234339035e-06,
-      "loss": 0.3777,
-      "step": 3380
-    },
-    {
-      "epoch": 0.22096594993791255,
-      "grad_norm": 0.5126791000366211,
-      "learning_rate": 9.92627250161642e-06,
-      "loss": 0.476,
-      "step": 3381
-    },
-    {
-      "epoch": 0.22103130514345468,
-      "grad_norm": 0.453173965215683,
-      "learning_rate": 9.926212744866668e-06,
-      "loss": 0.3912,
-      "step": 3382
-    },
-    {
-      "epoch": 0.2210966603489968,
-      "grad_norm": 0.5246379971504211,
-      "learning_rate": 9.926152964090068e-06,
-      "loss": 0.4641,
-      "step": 3383
-    },
-    {
-      "epoch": 0.22116201555453893,
-      "grad_norm": 0.4921092987060547,
-      "learning_rate": 9.926093159286916e-06,
-      "loss": 0.4173,
-      "step": 3384
-    },
-    {
-      "epoch": 0.22122737076008103,
-      "grad_norm": 0.4708023965358734,
-      "learning_rate": 9.926033330457498e-06,
-      "loss": 0.4066,
-      "step": 3385
-    },
-    {
-      "epoch": 0.22129272596562316,
-      "grad_norm": 0.49583831429481506,
-      "learning_rate": 9.925973477602111e-06,
-      "loss": 0.4209,
-      "step": 3386
-    },
-    {
-      "epoch": 0.2213580811711653,
-      "grad_norm": 0.46903926134109497,
-      "learning_rate": 9.925913600721045e-06,
-      "loss": 0.3758,
-      "step": 3387
-    },
-    {
-      "epoch": 0.22142343637670742,
-      "grad_norm": 0.4628562331199646,
-      "learning_rate": 9.925853699814592e-06,
-      "loss": 0.3919,
-      "step": 3388
-    },
-    {
-      "epoch": 0.22148879158224952,
-      "grad_norm": 0.4656374454498291,
-      "learning_rate": 9.925793774883042e-06,
-      "loss": 0.3627,
-      "step": 3389
-    },
-    {
-      "epoch": 0.22155414678779164,
-      "grad_norm": 0.49630263447761536,
-      "learning_rate": 9.925733825926691e-06,
-      "loss": 0.4053,
-      "step": 3390
-    },
-    {
-      "epoch": 0.22161950199333377,
-      "grad_norm": 0.5168403387069702,
-      "learning_rate": 9.92567385294583e-06,
-      "loss": 0.3891,
-      "step": 3391
-    },
-    {
-      "epoch": 0.2216848571988759,
-      "grad_norm": 0.4836116135120392,
-      "learning_rate": 9.92561385594075e-06,
-      "loss": 0.4171,
-      "step": 3392
-    },
-    {
-      "epoch": 0.221750212404418,
-      "grad_norm": 0.4696779251098633,
-      "learning_rate": 9.925553834911745e-06,
-      "loss": 0.3903,
-      "step": 3393
-    },
-    {
-      "epoch": 0.22181556760996013,
-      "grad_norm": 0.46139299869537354,
-      "learning_rate": 9.925493789859107e-06,
-      "loss": 0.3961,
-      "step": 3394
-    },
-    {
-      "epoch": 0.22188092281550226,
-      "grad_norm": 0.4747457206249237,
-      "learning_rate": 9.92543372078313e-06,
-      "loss": 0.4106,
-      "step": 3395
-    },
-    {
-      "epoch": 0.22194627802104439,
-      "grad_norm": 0.4985063374042511,
-      "learning_rate": 9.925373627684107e-06,
-      "loss": 0.4419,
-      "step": 3396
-    },
-    {
-      "epoch": 0.22201163322658649,
-      "grad_norm": 0.5045239925384521,
-      "learning_rate": 9.925313510562331e-06,
-      "loss": 0.4206,
-      "step": 3397
-    },
-    {
-      "epoch": 0.2220769884321286,
-      "grad_norm": 0.5201502442359924,
-      "learning_rate": 9.925253369418093e-06,
-      "loss": 0.4159,
-      "step": 3398
-    },
-    {
-      "epoch": 0.22214234363767074,
-      "grad_norm": 0.5161673426628113,
-      "learning_rate": 9.925193204251689e-06,
-      "loss": 0.4384,
-      "step": 3399
-    },
-    {
-      "epoch": 0.22220769884321287,
-      "grad_norm": 0.48066210746765137,
-      "learning_rate": 9.925133015063412e-06,
-      "loss": 0.4421,
-      "step": 3400
-    },
-    {
-      "epoch": 0.22227305404875497,
-      "grad_norm": 0.4755731523036957,
-      "learning_rate": 9.925072801853554e-06,
-      "loss": 0.3963,
-      "step": 3401
-    },
-    {
-      "epoch": 0.2223384092542971,
-      "grad_norm": 0.48620015382766724,
-      "learning_rate": 9.92501256462241e-06,
-      "loss": 0.4184,
-      "step": 3402
-    },
-    {
-      "epoch": 0.22240376445983923,
-      "grad_norm": 0.49302372336387634,
-      "learning_rate": 9.924952303370274e-06,
-      "loss": 0.4506,
-      "step": 3403
-    },
-    {
-      "epoch": 0.22246911966538135,
-      "grad_norm": 0.4507593810558319,
-      "learning_rate": 9.92489201809744e-06,
-      "loss": 0.3732,
-      "step": 3404
-    },
-    {
-      "epoch": 0.22253447487092348,
-      "grad_norm": 0.4765528738498688,
-      "learning_rate": 9.9248317088042e-06,
-      "loss": 0.4214,
-      "step": 3405
-    },
-    {
-      "epoch": 0.22259983007646558,
-      "grad_norm": 0.5671145915985107,
-      "learning_rate": 9.92477137549085e-06,
-      "loss": 0.4977,
-      "step": 3406
-    },
-    {
-      "epoch": 0.2226651852820077,
-      "grad_norm": 0.48747357726097107,
-      "learning_rate": 9.924711018157684e-06,
-      "loss": 0.4504,
-      "step": 3407
-    },
-    {
-      "epoch": 0.22273054048754984,
-      "grad_norm": 0.48180848360061646,
-      "learning_rate": 9.924650636804997e-06,
-      "loss": 0.3956,
-      "step": 3408
-    },
-    {
-      "epoch": 0.22279589569309197,
-      "grad_norm": 0.4759974181652069,
-      "learning_rate": 9.924590231433082e-06,
-      "loss": 0.4175,
-      "step": 3409
-    },
-    {
-      "epoch": 0.22286125089863407,
-      "grad_norm": 0.4900985360145569,
-      "learning_rate": 9.924529802042236e-06,
-      "loss": 0.451,
-      "step": 3410
-    },
-    {
-      "epoch": 0.2229266061041762,
-      "grad_norm": 0.4859611690044403,
-      "learning_rate": 9.92446934863275e-06,
-      "loss": 0.4109,
-      "step": 3411
-    },
-    {
-      "epoch": 0.22299196130971832,
-      "grad_norm": 0.46664854884147644,
-      "learning_rate": 9.924408871204923e-06,
-      "loss": 0.4346,
-      "step": 3412
-    },
-    {
-      "epoch": 0.22305731651526045,
-      "grad_norm": 0.4617181420326233,
-      "learning_rate": 9.924348369759045e-06,
-      "loss": 0.3751,
-      "step": 3413
-    },
-    {
-      "epoch": 0.22312267172080255,
-      "grad_norm": 0.49750587344169617,
-      "learning_rate": 9.924287844295417e-06,
-      "loss": 0.4764,
-      "step": 3414
-    },
-    {
-      "epoch": 0.22318802692634468,
-      "grad_norm": 0.47653597593307495,
-      "learning_rate": 9.92422729481433e-06,
-      "loss": 0.4216,
-      "step": 3415
-    },
-    {
-      "epoch": 0.2232533821318868,
-      "grad_norm": 0.4832088053226471,
-      "learning_rate": 9.92416672131608e-06,
-      "loss": 0.4099,
-      "step": 3416
-    },
-    {
-      "epoch": 0.22331873733742894,
-      "grad_norm": 0.4652642607688904,
-      "learning_rate": 9.924106123800964e-06,
-      "loss": 0.3525,
-      "step": 3417
-    },
-    {
-      "epoch": 0.22338409254297104,
-      "grad_norm": 0.5141023993492126,
-      "learning_rate": 9.924045502269275e-06,
-      "loss": 0.463,
-      "step": 3418
-    },
-    {
-      "epoch": 0.22344944774851316,
-      "grad_norm": 0.4620446562767029,
-      "learning_rate": 9.923984856721312e-06,
-      "loss": 0.381,
-      "step": 3419
-    },
-    {
-      "epoch": 0.2235148029540553,
-      "grad_norm": 0.47818121314048767,
-      "learning_rate": 9.923924187157368e-06,
-      "loss": 0.42,
-      "step": 3420
-    },
-    {
-      "epoch": 0.22358015815959742,
-      "grad_norm": 0.4844329059123993,
-      "learning_rate": 9.92386349357774e-06,
-      "loss": 0.42,
-      "step": 3421
-    },
-    {
-      "epoch": 0.22364551336513952,
-      "grad_norm": 0.5048141479492188,
-      "learning_rate": 9.923802775982724e-06,
-      "loss": 0.4725,
-      "step": 3422
-    },
-    {
-      "epoch": 0.22371086857068165,
-      "grad_norm": 0.47144418954849243,
-      "learning_rate": 9.923742034372618e-06,
-      "loss": 0.4094,
-      "step": 3423
-    },
-    {
-      "epoch": 0.22377622377622378,
-      "grad_norm": 0.49419599771499634,
-      "learning_rate": 9.923681268747714e-06,
-      "loss": 0.4514,
-      "step": 3424
-    },
-    {
-      "epoch": 0.2238415789817659,
-      "grad_norm": 0.49246811866760254,
-      "learning_rate": 9.92362047910831e-06,
-      "loss": 0.4264,
-      "step": 3425
-    },
-    {
-      "epoch": 0.22390693418730803,
-      "grad_norm": 0.46921661496162415,
-      "learning_rate": 9.923559665454707e-06,
-      "loss": 0.4393,
-      "step": 3426
-    },
-    {
-      "epoch": 0.22397228939285013,
-      "grad_norm": 0.4811283051967621,
-      "learning_rate": 9.923498827787195e-06,
-      "loss": 0.3721,
-      "step": 3427
-    },
-    {
-      "epoch": 0.22403764459839226,
-      "grad_norm": 0.5040667653083801,
-      "learning_rate": 9.923437966106074e-06,
-      "loss": 0.4154,
-      "step": 3428
-    },
-    {
-      "epoch": 0.2241029998039344,
-      "grad_norm": 0.4838813543319702,
-      "learning_rate": 9.92337708041164e-06,
-      "loss": 0.4096,
-      "step": 3429
-    },
-    {
-      "epoch": 0.22416835500947652,
-      "grad_norm": 0.5002124309539795,
-      "learning_rate": 9.923316170704192e-06,
-      "loss": 0.4395,
-      "step": 3430
-    },
-    {
-      "epoch": 0.22423371021501862,
-      "grad_norm": 0.4668019413948059,
-      "learning_rate": 9.923255236984024e-06,
-      "loss": 0.3919,
-      "step": 3431
-    },
-    {
-      "epoch": 0.22429906542056074,
-      "grad_norm": 0.4683651328086853,
-      "learning_rate": 9.923194279251435e-06,
-      "loss": 0.3844,
-      "step": 3432
-    },
-    {
-      "epoch": 0.22436442062610287,
-      "grad_norm": 0.46975427865982056,
-      "learning_rate": 9.923133297506721e-06,
-      "loss": 0.4201,
-      "step": 3433
-    },
-    {
-      "epoch": 0.224429775831645,
-      "grad_norm": 0.48653876781463623,
-      "learning_rate": 9.923072291750182e-06,
-      "loss": 0.3965,
-      "step": 3434
-    },
-    {
-      "epoch": 0.2244951310371871,
-      "grad_norm": 0.5011444091796875,
-      "learning_rate": 9.923011261982113e-06,
-      "loss": 0.4005,
-      "step": 3435
-    },
-    {
-      "epoch": 0.22456048624272923,
-      "grad_norm": 0.4734165370464325,
-      "learning_rate": 9.922950208202812e-06,
-      "loss": 0.414,
-      "step": 3436
-    },
-    {
-      "epoch": 0.22462584144827136,
-      "grad_norm": 0.4409369230270386,
-      "learning_rate": 9.922889130412578e-06,
-      "loss": 0.3678,
-      "step": 3437
-    },
-    {
-      "epoch": 0.22469119665381349,
-      "grad_norm": 0.4692237973213196,
-      "learning_rate": 9.922828028611708e-06,
-      "loss": 0.3892,
-      "step": 3438
-    },
-    {
-      "epoch": 0.22475655185935559,
-      "grad_norm": 0.5060802102088928,
-      "learning_rate": 9.922766902800502e-06,
-      "loss": 0.3997,
-      "step": 3439
-    },
-    {
-      "epoch": 0.2248219070648977,
-      "grad_norm": 0.5373620390892029,
-      "learning_rate": 9.922705752979254e-06,
-      "loss": 0.4634,
-      "step": 3440
-    },
-    {
-      "epoch": 0.22488726227043984,
-      "grad_norm": 0.4730021357536316,
-      "learning_rate": 9.922644579148267e-06,
-      "loss": 0.4026,
-      "step": 3441
-    },
-    {
-      "epoch": 0.22495261747598197,
-      "grad_norm": 0.4861553907394409,
-      "learning_rate": 9.922583381307835e-06,
-      "loss": 0.4224,
-      "step": 3442
-    },
-    {
-      "epoch": 0.22501797268152407,
-      "grad_norm": 0.507655918598175,
-      "learning_rate": 9.922522159458259e-06,
-      "loss": 0.4462,
-      "step": 3443
-    },
-    {
-      "epoch": 0.2250833278870662,
-      "grad_norm": 0.47881460189819336,
-      "learning_rate": 9.922460913599838e-06,
-      "loss": 0.3974,
-      "step": 3444
-    },
-    {
-      "epoch": 0.22514868309260833,
-      "grad_norm": 0.49684351682662964,
-      "learning_rate": 9.922399643732867e-06,
-      "loss": 0.4459,
-      "step": 3445
-    },
-    {
-      "epoch": 0.22521403829815045,
-      "grad_norm": 0.4507707953453064,
-      "learning_rate": 9.92233834985765e-06,
-      "loss": 0.3473,
-      "step": 3446
-    },
-    {
-      "epoch": 0.22527939350369258,
-      "grad_norm": 0.4546440839767456,
-      "learning_rate": 9.922277031974484e-06,
-      "loss": 0.4452,
-      "step": 3447
-    },
-    {
-      "epoch": 0.22534474870923468,
-      "grad_norm": 0.47355177998542786,
-      "learning_rate": 9.922215690083667e-06,
-      "loss": 0.3942,
-      "step": 3448
-    },
-    {
-      "epoch": 0.2254101039147768,
-      "grad_norm": 0.46548977494239807,
-      "learning_rate": 9.9221543241855e-06,
-      "loss": 0.4144,
-      "step": 3449
-    },
-    {
-      "epoch": 0.22547545912031894,
-      "grad_norm": 0.4996720850467682,
-      "learning_rate": 9.92209293428028e-06,
-      "loss": 0.4264,
-      "step": 3450
-    },
-    {
-      "epoch": 0.22554081432586107,
-      "grad_norm": 0.44034871459007263,
-      "learning_rate": 9.922031520368307e-06,
-      "loss": 0.3552,
-      "step": 3451
-    },
-    {
-      "epoch": 0.22560616953140317,
-      "grad_norm": 0.43639031052589417,
-      "learning_rate": 9.921970082449881e-06,
-      "loss": 0.3225,
-      "step": 3452
-    },
-    {
-      "epoch": 0.2256715247369453,
-      "grad_norm": 0.502032458782196,
-      "learning_rate": 9.921908620525303e-06,
-      "loss": 0.4183,
-      "step": 3453
-    },
-    {
-      "epoch": 0.22573687994248742,
-      "grad_norm": 0.4731839895248413,
-      "learning_rate": 9.921847134594871e-06,
-      "loss": 0.4144,
-      "step": 3454
-    },
-    {
-      "epoch": 0.22580223514802955,
-      "grad_norm": 0.512500524520874,
-      "learning_rate": 9.921785624658887e-06,
-      "loss": 0.4391,
-      "step": 3455
-    },
-    {
-      "epoch": 0.22586759035357165,
-      "grad_norm": 0.5110539197921753,
-      "learning_rate": 9.921724090717646e-06,
-      "loss": 0.4066,
-      "step": 3456
-    },
-    {
-      "epoch": 0.22593294555911378,
-      "grad_norm": 0.478849858045578,
-      "learning_rate": 9.921662532771455e-06,
-      "loss": 0.419,
-      "step": 3457
-    },
-    {
-      "epoch": 0.2259983007646559,
-      "grad_norm": 0.5071877837181091,
-      "learning_rate": 9.92160095082061e-06,
-      "loss": 0.4284,
-      "step": 3458
-    },
-    {
-      "epoch": 0.22606365597019804,
-      "grad_norm": 0.5038055777549744,
-      "learning_rate": 9.92153934486541e-06,
-      "loss": 0.4439,
-      "step": 3459
-    },
-    {
-      "epoch": 0.22612901117574014,
-      "grad_norm": 0.5345181226730347,
-      "learning_rate": 9.921477714906158e-06,
-      "loss": 0.4447,
-      "step": 3460
-    },
-    {
-      "epoch": 0.22619436638128226,
-      "grad_norm": 0.4881491959095001,
-      "learning_rate": 9.921416060943157e-06,
-      "loss": 0.3998,
-      "step": 3461
-    },
-    {
-      "epoch": 0.2262597215868244,
-      "grad_norm": 0.4545268416404724,
-      "learning_rate": 9.921354382976703e-06,
-      "loss": 0.3934,
-      "step": 3462
-    },
-    {
-      "epoch": 0.22632507679236652,
-      "grad_norm": 0.47923189401626587,
-      "learning_rate": 9.9212926810071e-06,
-      "loss": 0.4024,
-      "step": 3463
-    },
-    {
-      "epoch": 0.22639043199790862,
-      "grad_norm": 0.5075094699859619,
-      "learning_rate": 9.921230955034645e-06,
-      "loss": 0.469,
-      "step": 3464
-    },
-    {
-      "epoch": 0.22645578720345075,
-      "grad_norm": 0.46883073449134827,
-      "learning_rate": 9.921169205059644e-06,
-      "loss": 0.4028,
-      "step": 3465
-    },
-    {
-      "epoch": 0.22652114240899288,
-      "grad_norm": 0.46435844898223877,
-      "learning_rate": 9.921107431082395e-06,
-      "loss": 0.4042,
-      "step": 3466
-    },
-    {
-      "epoch": 0.226586497614535,
-      "grad_norm": 0.44952312111854553,
-      "learning_rate": 9.921045633103201e-06,
-      "loss": 0.3556,
-      "step": 3467
-    },
-    {
-      "epoch": 0.22665185282007713,
-      "grad_norm": 0.44821831583976746,
-      "learning_rate": 9.920983811122363e-06,
-      "loss": 0.4041,
-      "step": 3468
-    },
-    {
-      "epoch": 0.22671720802561923,
-      "grad_norm": 0.46383920311927795,
-      "learning_rate": 9.92092196514018e-06,
-      "loss": 0.3687,
-      "step": 3469
-    },
-    {
-      "epoch": 0.22678256323116136,
-      "grad_norm": 0.4864318072795868,
-      "learning_rate": 9.920860095156956e-06,
-      "loss": 0.4704,
-      "step": 3470
-    },
-    {
-      "epoch": 0.2268479184367035,
-      "grad_norm": 0.46455636620521545,
-      "learning_rate": 9.920798201172996e-06,
-      "loss": 0.3975,
-      "step": 3471
-    },
-    {
-      "epoch": 0.22691327364224562,
-      "grad_norm": 0.47325679659843445,
-      "learning_rate": 9.920736283188596e-06,
-      "loss": 0.4323,
-      "step": 3472
-    },
-    {
-      "epoch": 0.22697862884778772,
-      "grad_norm": 0.49008190631866455,
-      "learning_rate": 9.92067434120406e-06,
-      "loss": 0.4475,
-      "step": 3473
-    },
-    {
-      "epoch": 0.22704398405332984,
-      "grad_norm": 0.45270171761512756,
-      "learning_rate": 9.92061237521969e-06,
-      "loss": 0.3751,
-      "step": 3474
-    },
-    {
-      "epoch": 0.22710933925887197,
-      "grad_norm": 0.4598439931869507,
-      "learning_rate": 9.920550385235791e-06,
-      "loss": 0.3957,
-      "step": 3475
-    },
-    {
-      "epoch": 0.2271746944644141,
-      "grad_norm": 0.5912983417510986,
-      "learning_rate": 9.920488371252662e-06,
-      "loss": 0.4098,
-      "step": 3476
-    },
-    {
-      "epoch": 0.2272400496699562,
-      "grad_norm": 0.49598702788352966,
-      "learning_rate": 9.920426333270607e-06,
-      "loss": 0.4417,
-      "step": 3477
-    },
-    {
-      "epoch": 0.22730540487549833,
-      "grad_norm": 0.4662719666957855,
-      "learning_rate": 9.920364271289929e-06,
-      "loss": 0.3967,
-      "step": 3478
-    },
-    {
-      "epoch": 0.22737076008104046,
-      "grad_norm": 0.4649987518787384,
-      "learning_rate": 9.920302185310928e-06,
-      "loss": 0.4069,
-      "step": 3479
-    },
-    {
-      "epoch": 0.22743611528658259,
-      "grad_norm": 0.4543205201625824,
-      "learning_rate": 9.920240075333909e-06,
-      "loss": 0.3966,
-      "step": 3480
-    },
-    {
-      "epoch": 0.22750147049212469,
-      "grad_norm": 0.47058290243148804,
-      "learning_rate": 9.920177941359174e-06,
-      "loss": 0.4021,
-      "step": 3481
-    },
-    {
-      "epoch": 0.2275668256976668,
-      "grad_norm": 0.5269588828086853,
-      "learning_rate": 9.920115783387028e-06,
-      "loss": 0.4509,
-      "step": 3482
-    },
-    {
-      "epoch": 0.22763218090320894,
-      "grad_norm": 0.48189884424209595,
-      "learning_rate": 9.920053601417773e-06,
-      "loss": 0.3597,
-      "step": 3483
-    },
-    {
-      "epoch": 0.22769753610875107,
-      "grad_norm": 0.49318572878837585,
-      "learning_rate": 9.919991395451713e-06,
-      "loss": 0.4451,
-      "step": 3484
-    },
-    {
-      "epoch": 0.22776289131429317,
-      "grad_norm": 0.44078201055526733,
-      "learning_rate": 9.919929165489149e-06,
-      "loss": 0.3647,
-      "step": 3485
-    },
-    {
-      "epoch": 0.2278282465198353,
-      "grad_norm": 0.4564656615257263,
-      "learning_rate": 9.919866911530386e-06,
-      "loss": 0.3636,
-      "step": 3486
-    },
-    {
-      "epoch": 0.22789360172537743,
-      "grad_norm": 0.5045684576034546,
-      "learning_rate": 9.919804633575727e-06,
-      "loss": 0.4093,
-      "step": 3487
-    },
-    {
-      "epoch": 0.22795895693091955,
-      "grad_norm": 0.48582541942596436,
-      "learning_rate": 9.919742331625477e-06,
-      "loss": 0.4026,
-      "step": 3488
-    },
-    {
-      "epoch": 0.22802431213646168,
-      "grad_norm": 0.49130311608314514,
-      "learning_rate": 9.91968000567994e-06,
-      "loss": 0.4286,
-      "step": 3489
-    },
-    {
-      "epoch": 0.22808966734200378,
-      "grad_norm": 0.4653776288032532,
-      "learning_rate": 9.91961765573942e-06,
-      "loss": 0.3151,
-      "step": 3490
-    },
-    {
-      "epoch": 0.2281550225475459,
-      "grad_norm": 0.49312624335289,
-      "learning_rate": 9.919555281804219e-06,
-      "loss": 0.4269,
-      "step": 3491
-    },
-    {
-      "epoch": 0.22822037775308804,
-      "grad_norm": 0.5089777708053589,
-      "learning_rate": 9.919492883874642e-06,
-      "loss": 0.4033,
-      "step": 3492
-    },
-    {
-      "epoch": 0.22828573295863017,
-      "grad_norm": 0.4697995185852051,
-      "learning_rate": 9.919430461950996e-06,
-      "loss": 0.357,
-      "step": 3493
-    },
-    {
-      "epoch": 0.22835108816417227,
-      "grad_norm": 0.4682181179523468,
-      "learning_rate": 9.919368016033581e-06,
-      "loss": 0.4047,
-      "step": 3494
-    },
-    {
-      "epoch": 0.2284164433697144,
-      "grad_norm": 0.48286086320877075,
-      "learning_rate": 9.919305546122704e-06,
-      "loss": 0.4092,
-      "step": 3495
-    },
-    {
-      "epoch": 0.22848179857525652,
-      "grad_norm": 0.4905316233634949,
-      "learning_rate": 9.919243052218672e-06,
-      "loss": 0.4078,
-      "step": 3496
-    },
-    {
-      "epoch": 0.22854715378079865,
-      "grad_norm": 0.5208161473274231,
-      "learning_rate": 9.919180534321787e-06,
-      "loss": 0.4205,
-      "step": 3497
-    },
-    {
-      "epoch": 0.22861250898634075,
-      "grad_norm": 0.4441189467906952,
-      "learning_rate": 9.919117992432352e-06,
-      "loss": 0.3484,
-      "step": 3498
-    },
-    {
-      "epoch": 0.22867786419188288,
-      "grad_norm": 0.4950271546840668,
-      "learning_rate": 9.919055426550676e-06,
-      "loss": 0.4509,
-      "step": 3499
-    },
-    {
-      "epoch": 0.228743219397425,
-      "grad_norm": 0.4683346748352051,
-      "learning_rate": 9.918992836677064e-06,
-      "loss": 0.4291,
-      "step": 3500
-    },
-    {
-      "epoch": 0.22880857460296714,
-      "grad_norm": 0.49860504269599915,
-      "learning_rate": 9.918930222811818e-06,
-      "loss": 0.4008,
-      "step": 3501
-    },
-    {
-      "epoch": 0.22887392980850924,
-      "grad_norm": 0.48979246616363525,
-      "learning_rate": 9.918867584955245e-06,
-      "loss": 0.4318,
-      "step": 3502
-    },
-    {
-      "epoch": 0.22893928501405136,
-      "grad_norm": 0.4749269485473633,
-      "learning_rate": 9.918804923107651e-06,
-      "loss": 0.3796,
-      "step": 3503
-    },
-    {
-      "epoch": 0.2290046402195935,
-      "grad_norm": 0.47051766514778137,
-      "learning_rate": 9.918742237269341e-06,
-      "loss": 0.374,
-      "step": 3504
-    },
-    {
-      "epoch": 0.22906999542513562,
-      "grad_norm": 0.4721180200576782,
-      "learning_rate": 9.918679527440623e-06,
-      "loss": 0.4035,
-      "step": 3505
-    },
-    {
-      "epoch": 0.22913535063067772,
-      "grad_norm": 0.5044035315513611,
-      "learning_rate": 9.9186167936218e-06,
-      "loss": 0.4212,
-      "step": 3506
-    },
-    {
-      "epoch": 0.22920070583621985,
-      "grad_norm": 0.5842257738113403,
-      "learning_rate": 9.918554035813177e-06,
-      "loss": 0.4856,
-      "step": 3507
-    },
-    {
-      "epoch": 0.22926606104176198,
-      "grad_norm": 0.5033165812492371,
-      "learning_rate": 9.918491254015064e-06,
-      "loss": 0.4001,
-      "step": 3508
-    },
-    {
-      "epoch": 0.2293314162473041,
-      "grad_norm": 0.5334236025810242,
-      "learning_rate": 9.918428448227767e-06,
-      "loss": 0.4575,
-      "step": 3509
-    },
-    {
-      "epoch": 0.22939677145284623,
-      "grad_norm": 0.48921439051628113,
-      "learning_rate": 9.918365618451586e-06,
-      "loss": 0.4554,
-      "step": 3510
-    },
-    {
-      "epoch": 0.22946212665838833,
-      "grad_norm": 0.4366236925125122,
-      "learning_rate": 9.918302764686835e-06,
-      "loss": 0.3555,
-      "step": 3511
-    },
-    {
-      "epoch": 0.22952748186393046,
-      "grad_norm": 0.4496869444847107,
-      "learning_rate": 9.918239886933818e-06,
-      "loss": 0.3394,
-      "step": 3512
-    },
-    {
-      "epoch": 0.2295928370694726,
-      "grad_norm": 0.46605080366134644,
-      "learning_rate": 9.91817698519284e-06,
-      "loss": 0.3875,
-      "step": 3513
-    },
-    {
-      "epoch": 0.22965819227501472,
-      "grad_norm": 0.48124197125434875,
-      "learning_rate": 9.918114059464209e-06,
-      "loss": 0.4025,
-      "step": 3514
-    },
-    {
-      "epoch": 0.22972354748055682,
-      "grad_norm": 0.47093167901039124,
-      "learning_rate": 9.918051109748233e-06,
-      "loss": 0.3923,
-      "step": 3515
-    },
-    {
-      "epoch": 0.22978890268609894,
-      "grad_norm": 0.4437168538570404,
-      "learning_rate": 9.917988136045215e-06,
-      "loss": 0.3532,
-      "step": 3516
-    },
-    {
-      "epoch": 0.22985425789164107,
-      "grad_norm": 0.5177018642425537,
-      "learning_rate": 9.917925138355468e-06,
-      "loss": 0.4273,
-      "step": 3517
-    },
-    {
-      "epoch": 0.2299196130971832,
-      "grad_norm": 0.48225924372673035,
-      "learning_rate": 9.917862116679295e-06,
-      "loss": 0.4301,
-      "step": 3518
-    },
-    {
-      "epoch": 0.2299849683027253,
-      "grad_norm": 0.4468449056148529,
-      "learning_rate": 9.917799071017007e-06,
-      "loss": 0.3567,
-      "step": 3519
-    },
-    {
-      "epoch": 0.23005032350826743,
-      "grad_norm": 0.48492759466171265,
-      "learning_rate": 9.917736001368907e-06,
-      "loss": 0.4383,
-      "step": 3520
-    },
-    {
-      "epoch": 0.23011567871380956,
-      "grad_norm": 0.489656001329422,
-      "learning_rate": 9.917672907735306e-06,
-      "loss": 0.3883,
-      "step": 3521
-    },
-    {
-      "epoch": 0.23018103391935169,
-      "grad_norm": 0.490484356880188,
-      "learning_rate": 9.917609790116508e-06,
-      "loss": 0.4052,
-      "step": 3522
-    },
-    {
-      "epoch": 0.23024638912489379,
-      "grad_norm": 0.48557600378990173,
-      "learning_rate": 9.917546648512826e-06,
-      "loss": 0.3701,
-      "step": 3523
-    },
-    {
-      "epoch": 0.2303117443304359,
-      "grad_norm": 0.44677576422691345,
-      "learning_rate": 9.917483482924566e-06,
-      "loss": 0.3621,
-      "step": 3524
-    },
-    {
-      "epoch": 0.23037709953597804,
-      "grad_norm": 0.45892393589019775,
-      "learning_rate": 9.917420293352034e-06,
-      "loss": 0.3976,
-      "step": 3525
-    },
-    {
-      "epoch": 0.23044245474152017,
-      "grad_norm": 0.440434992313385,
-      "learning_rate": 9.91735707979554e-06,
-      "loss": 0.3849,
-      "step": 3526
-    },
-    {
-      "epoch": 0.23050780994706227,
-      "grad_norm": 0.4849710762500763,
-      "learning_rate": 9.917293842255392e-06,
-      "loss": 0.4536,
-      "step": 3527
-    },
-    {
-      "epoch": 0.2305731651526044,
-      "grad_norm": 0.5004308819770813,
-      "learning_rate": 9.917230580731898e-06,
-      "loss": 0.4063,
-      "step": 3528
-    },
-    {
-      "epoch": 0.23063852035814653,
-      "grad_norm": 0.5044596195220947,
-      "learning_rate": 9.917167295225367e-06,
-      "loss": 0.4073,
-      "step": 3529
-    },
-    {
-      "epoch": 0.23070387556368865,
-      "grad_norm": 0.5104182958602905,
-      "learning_rate": 9.917103985736107e-06,
-      "loss": 0.4824,
-      "step": 3530
-    },
-    {
-      "epoch": 0.23076923076923078,
-      "grad_norm": 0.4673587381839752,
-      "learning_rate": 9.917040652264429e-06,
-      "loss": 0.3776,
-      "step": 3531
-    },
-    {
-      "epoch": 0.23083458597477288,
-      "grad_norm": 0.46703916788101196,
-      "learning_rate": 9.91697729481064e-06,
-      "loss": 0.399,
-      "step": 3532
-    },
-    {
-      "epoch": 0.230899941180315,
-      "grad_norm": 0.48946160078048706,
-      "learning_rate": 9.91691391337505e-06,
-      "loss": 0.4151,
-      "step": 3533
-    },
-    {
-      "epoch": 0.23096529638585714,
-      "grad_norm": 0.5334754586219788,
-      "learning_rate": 9.916850507957965e-06,
-      "loss": 0.4416,
-      "step": 3534
-    },
-    {
-      "epoch": 0.23103065159139927,
-      "grad_norm": 0.5059509873390198,
-      "learning_rate": 9.9167870785597e-06,
-      "loss": 0.4324,
-      "step": 3535
-    },
-    {
-      "epoch": 0.23109600679694137,
-      "grad_norm": 0.48675912618637085,
-      "learning_rate": 9.916723625180557e-06,
-      "loss": 0.3674,
-      "step": 3536
-    },
-    {
-      "epoch": 0.2311613620024835,
-      "grad_norm": 0.4450955092906952,
-      "learning_rate": 9.916660147820853e-06,
-      "loss": 0.3888,
-      "step": 3537
-    },
-    {
-      "epoch": 0.23122671720802562,
-      "grad_norm": 0.4775831699371338,
-      "learning_rate": 9.916596646480894e-06,
-      "loss": 0.4156,
-      "step": 3538
-    },
-    {
-      "epoch": 0.23129207241356775,
-      "grad_norm": 0.5021882653236389,
-      "learning_rate": 9.916533121160988e-06,
-      "loss": 0.4414,
-      "step": 3539
-    },
-    {
-      "epoch": 0.23135742761910985,
-      "grad_norm": 0.4575294852256775,
-      "learning_rate": 9.916469571861447e-06,
-      "loss": 0.4026,
-      "step": 3540
-    },
-    {
-      "epoch": 0.23142278282465198,
-      "grad_norm": 0.5298008322715759,
-      "learning_rate": 9.91640599858258e-06,
-      "loss": 0.4382,
-      "step": 3541
-    },
-    {
-      "epoch": 0.2314881380301941,
-      "grad_norm": 0.5370910167694092,
-      "learning_rate": 9.916342401324702e-06,
-      "loss": 0.4831,
-      "step": 3542
-    },
-    {
-      "epoch": 0.23155349323573624,
-      "grad_norm": 0.4930441975593567,
-      "learning_rate": 9.916278780088115e-06,
-      "loss": 0.4447,
-      "step": 3543
-    },
-    {
-      "epoch": 0.23161884844127834,
-      "grad_norm": 0.47559595108032227,
-      "learning_rate": 9.916215134873134e-06,
-      "loss": 0.4133,
-      "step": 3544
-    },
-    {
-      "epoch": 0.23168420364682046,
-      "grad_norm": 0.5217772126197815,
-      "learning_rate": 9.916151465680069e-06,
-      "loss": 0.407,
-      "step": 3545
-    },
-    {
-      "epoch": 0.2317495588523626,
-      "grad_norm": 0.5612869262695312,
-      "learning_rate": 9.91608777250923e-06,
-      "loss": 0.4943,
-      "step": 3546
-    },
-    {
-      "epoch": 0.23181491405790472,
-      "grad_norm": 0.4543122947216034,
-      "learning_rate": 9.916024055360928e-06,
-      "loss": 0.3515,
-      "step": 3547
-    },
-    {
-      "epoch": 0.23188026926344682,
-      "grad_norm": 0.44925716519355774,
-      "learning_rate": 9.915960314235473e-06,
-      "loss": 0.39,
-      "step": 3548
-    },
-    {
-      "epoch": 0.23194562446898895,
-      "grad_norm": 0.5313082933425903,
-      "learning_rate": 9.915896549133178e-06,
-      "loss": 0.4397,
-      "step": 3549
-    },
-    {
-      "epoch": 0.23201097967453108,
-      "grad_norm": 0.5094080567359924,
-      "learning_rate": 9.915832760054351e-06,
-      "loss": 0.4458,
-      "step": 3550
-    },
-    {
-      "epoch": 0.2320763348800732,
-      "grad_norm": 0.4623429477214813,
-      "learning_rate": 9.915768946999305e-06,
-      "loss": 0.4179,
-      "step": 3551
-    },
-    {
-      "epoch": 0.23214169008561533,
-      "grad_norm": 0.4832753539085388,
-      "learning_rate": 9.91570510996835e-06,
-      "loss": 0.3926,
-      "step": 3552
-    },
-    {
-      "epoch": 0.23220704529115743,
-      "grad_norm": 0.48376351594924927,
-      "learning_rate": 9.9156412489618e-06,
-      "loss": 0.4169,
-      "step": 3553
-    },
-    {
-      "epoch": 0.23227240049669956,
-      "grad_norm": 0.4905616343021393,
-      "learning_rate": 9.915577363979963e-06,
-      "loss": 0.4202,
-      "step": 3554
-    },
-    {
-      "epoch": 0.2323377557022417,
-      "grad_norm": 0.48949626088142395,
-      "learning_rate": 9.915513455023154e-06,
-      "loss": 0.4279,
-      "step": 3555
-    },
-    {
-      "epoch": 0.23240311090778382,
-      "grad_norm": 0.4484732449054718,
-      "learning_rate": 9.915449522091682e-06,
-      "loss": 0.3685,
-      "step": 3556
-    },
-    {
-      "epoch": 0.23246846611332592,
-      "grad_norm": 0.44788533449172974,
-      "learning_rate": 9.91538556518586e-06,
-      "loss": 0.3432,
-      "step": 3557
-    },
-    {
-      "epoch": 0.23253382131886804,
-      "grad_norm": 0.44627711176872253,
-      "learning_rate": 9.915321584306e-06,
-      "loss": 0.3753,
-      "step": 3558
-    },
-    {
-      "epoch": 0.23259917652441017,
-      "grad_norm": 0.4644292891025543,
-      "learning_rate": 9.915257579452412e-06,
-      "loss": 0.3836,
-      "step": 3559
-    },
-    {
-      "epoch": 0.2326645317299523,
-      "grad_norm": 0.5058528780937195,
-      "learning_rate": 9.915193550625411e-06,
-      "loss": 0.4646,
-      "step": 3560
-    },
-    {
-      "epoch": 0.2327298869354944,
-      "grad_norm": 0.4695664048194885,
-      "learning_rate": 9.915129497825309e-06,
-      "loss": 0.3794,
-      "step": 3561
-    },
-    {
-      "epoch": 0.23279524214103653,
-      "grad_norm": 0.45301932096481323,
-      "learning_rate": 9.915065421052418e-06,
-      "loss": 0.3863,
-      "step": 3562
-    },
-    {
-      "epoch": 0.23286059734657866,
-      "grad_norm": 0.5022915005683899,
-      "learning_rate": 9.915001320307049e-06,
-      "loss": 0.4001,
-      "step": 3563
-    },
-    {
-      "epoch": 0.23292595255212079,
-      "grad_norm": 0.468749076128006,
-      "learning_rate": 9.914937195589516e-06,
-      "loss": 0.3831,
-      "step": 3564
-    },
-    {
-      "epoch": 0.23299130775766289,
-      "grad_norm": 0.44168904423713684,
-      "learning_rate": 9.914873046900133e-06,
-      "loss": 0.3438,
-      "step": 3565
-    },
-    {
-      "epoch": 0.233056662963205,
-      "grad_norm": 0.4480777084827423,
-      "learning_rate": 9.91480887423921e-06,
-      "loss": 0.3792,
-      "step": 3566
-    },
-    {
-      "epoch": 0.23312201816874714,
-      "grad_norm": 0.4802215099334717,
-      "learning_rate": 9.914744677607063e-06,
-      "loss": 0.3791,
-      "step": 3567
-    },
-    {
-      "epoch": 0.23318737337428927,
-      "grad_norm": 0.523776650428772,
-      "learning_rate": 9.914680457004003e-06,
-      "loss": 0.4683,
-      "step": 3568
-    },
-    {
-      "epoch": 0.23325272857983137,
-      "grad_norm": 0.5157068967819214,
-      "learning_rate": 9.914616212430341e-06,
-      "loss": 0.3736,
-      "step": 3569
-    },
-    {
-      "epoch": 0.2333180837853735,
-      "grad_norm": 0.5125598311424255,
-      "learning_rate": 9.914551943886397e-06,
-      "loss": 0.4395,
-      "step": 3570
-    },
-    {
-      "epoch": 0.23338343899091563,
-      "grad_norm": 0.5070778131484985,
-      "learning_rate": 9.91448765137248e-06,
-      "loss": 0.4615,
-      "step": 3571
-    },
-    {
-      "epoch": 0.23344879419645775,
-      "grad_norm": 0.47507959604263306,
-      "learning_rate": 9.914423334888901e-06,
-      "loss": 0.4407,
-      "step": 3572
-    },
-    {
-      "epoch": 0.23351414940199988,
-      "grad_norm": 0.5215258002281189,
-      "learning_rate": 9.91435899443598e-06,
-      "loss": 0.4708,
-      "step": 3573
-    },
-    {
-      "epoch": 0.23357950460754198,
-      "grad_norm": 0.5039072036743164,
-      "learning_rate": 9.914294630014027e-06,
-      "loss": 0.4433,
-      "step": 3574
-    },
-    {
-      "epoch": 0.2336448598130841,
-      "grad_norm": 0.4702485203742981,
-      "learning_rate": 9.914230241623356e-06,
-      "loss": 0.3941,
-      "step": 3575
-    },
-    {
-      "epoch": 0.23371021501862624,
-      "grad_norm": 0.47556859254837036,
-      "learning_rate": 9.914165829264281e-06,
-      "loss": 0.4122,
-      "step": 3576
-    },
-    {
-      "epoch": 0.23377557022416837,
-      "grad_norm": 0.7105624675750732,
-      "learning_rate": 9.914101392937119e-06,
-      "loss": 0.4393,
-      "step": 3577
-    },
-    {
-      "epoch": 0.23384092542971047,
-      "grad_norm": 0.4980992078781128,
-      "learning_rate": 9.914036932642181e-06,
-      "loss": 0.3901,
-      "step": 3578
-    },
-    {
-      "epoch": 0.2339062806352526,
-      "grad_norm": 0.4689491391181946,
-      "learning_rate": 9.913972448379783e-06,
-      "loss": 0.388,
-      "step": 3579
-    },
-    {
-      "epoch": 0.23397163584079472,
-      "grad_norm": 0.5059848427772522,
-      "learning_rate": 9.913907940150238e-06,
-      "loss": 0.4531,
-      "step": 3580
-    },
-    {
-      "epoch": 0.23403699104633685,
-      "grad_norm": 0.47146686911582947,
-      "learning_rate": 9.913843407953862e-06,
-      "loss": 0.3656,
-      "step": 3581
-    },
-    {
-      "epoch": 0.23410234625187895,
-      "grad_norm": 0.5070616602897644,
-      "learning_rate": 9.91377885179097e-06,
-      "loss": 0.4375,
-      "step": 3582
-    },
-    {
-      "epoch": 0.23416770145742108,
-      "grad_norm": 0.5187751054763794,
-      "learning_rate": 9.913714271661875e-06,
-      "loss": 0.4443,
-      "step": 3583
-    },
-    {
-      "epoch": 0.2342330566629632,
-      "grad_norm": 0.4969579875469208,
-      "learning_rate": 9.913649667566893e-06,
-      "loss": 0.4393,
-      "step": 3584
-    },
-    {
-      "epoch": 0.23429841186850534,
-      "grad_norm": 0.45504826307296753,
-      "learning_rate": 9.913585039506342e-06,
-      "loss": 0.39,
-      "step": 3585
-    },
-    {
-      "epoch": 0.23436376707404744,
-      "grad_norm": 0.5259515643119812,
-      "learning_rate": 9.913520387480533e-06,
-      "loss": 0.4972,
-      "step": 3586
-    },
-    {
-      "epoch": 0.23442912227958956,
-      "grad_norm": 0.49079811573028564,
-      "learning_rate": 9.913455711489782e-06,
-      "loss": 0.407,
-      "step": 3587
-    },
-    {
-      "epoch": 0.2344944774851317,
-      "grad_norm": 0.4935847818851471,
-      "learning_rate": 9.913391011534406e-06,
-      "loss": 0.4423,
-      "step": 3588
-    },
-    {
-      "epoch": 0.23455983269067382,
-      "grad_norm": 0.45255184173583984,
-      "learning_rate": 9.91332628761472e-06,
-      "loss": 0.4047,
-      "step": 3589
-    },
-    {
-      "epoch": 0.23462518789621592,
-      "grad_norm": 0.5370975732803345,
-      "learning_rate": 9.91326153973104e-06,
-      "loss": 0.3959,
-      "step": 3590
-    },
-    {
-      "epoch": 0.23469054310175805,
-      "grad_norm": 0.534020721912384,
-      "learning_rate": 9.91319676788368e-06,
-      "loss": 0.4743,
-      "step": 3591
-    },
-    {
-      "epoch": 0.23475589830730018,
-      "grad_norm": 0.5184808969497681,
-      "learning_rate": 9.913131972072959e-06,
-      "loss": 0.4356,
-      "step": 3592
-    },
-    {
-      "epoch": 0.2348212535128423,
-      "grad_norm": 0.45330944657325745,
-      "learning_rate": 9.91306715229919e-06,
-      "loss": 0.3782,
-      "step": 3593
-    },
-    {
-      "epoch": 0.23488660871838443,
-      "grad_norm": 0.5136331915855408,
-      "learning_rate": 9.91300230856269e-06,
-      "loss": 0.4294,
-      "step": 3594
-    },
-    {
-      "epoch": 0.23495196392392653,
-      "grad_norm": 0.4611961543560028,
-      "learning_rate": 9.912937440863777e-06,
-      "loss": 0.3651,
-      "step": 3595
-    },
-    {
-      "epoch": 0.23501731912946866,
-      "grad_norm": 1.2540053129196167,
-      "learning_rate": 9.912872549202766e-06,
-      "loss": 0.4077,
-      "step": 3596
-    },
-    {
-      "epoch": 0.2350826743350108,
-      "grad_norm": 0.4653492867946625,
-      "learning_rate": 9.912807633579972e-06,
-      "loss": 0.3764,
-      "step": 3597
-    },
-    {
-      "epoch": 0.23514802954055292,
-      "grad_norm": 0.5431433320045471,
-      "learning_rate": 9.912742693995716e-06,
-      "loss": 0.4674,
-      "step": 3598
-    },
-    {
-      "epoch": 0.23521338474609502,
-      "grad_norm": 0.4963741898536682,
-      "learning_rate": 9.91267773045031e-06,
-      "loss": 0.4226,
-      "step": 3599
-    },
-    {
-      "epoch": 0.23527873995163714,
-      "grad_norm": 0.5268748998641968,
-      "learning_rate": 9.912612742944072e-06,
-      "loss": 0.4736,
-      "step": 3600
-    },
-    {
-      "epoch": 0.23534409515717927,
-      "grad_norm": 0.4762163758277893,
-      "learning_rate": 9.912547731477322e-06,
-      "loss": 0.4385,
-      "step": 3601
-    },
-    {
-      "epoch": 0.2354094503627214,
-      "grad_norm": 0.471513569355011,
-      "learning_rate": 9.912482696050374e-06,
-      "loss": 0.3856,
-      "step": 3602
-    },
-    {
-      "epoch": 0.2354748055682635,
-      "grad_norm": 0.5295483469963074,
-      "learning_rate": 9.912417636663545e-06,
-      "loss": 0.4437,
-      "step": 3603
-    },
-    {
-      "epoch": 0.23554016077380563,
-      "grad_norm": 0.5125187039375305,
-      "learning_rate": 9.912352553317155e-06,
-      "loss": 0.392,
-      "step": 3604
-    },
-    {
-      "epoch": 0.23560551597934776,
-      "grad_norm": 0.5781072378158569,
-      "learning_rate": 9.912287446011518e-06,
-      "loss": 0.3485,
-      "step": 3605
-    },
-    {
-      "epoch": 0.23567087118488989,
-      "grad_norm": 0.49361076951026917,
-      "learning_rate": 9.912222314746955e-06,
-      "loss": 0.4151,
-      "step": 3606
-    },
-    {
-      "epoch": 0.23573622639043199,
-      "grad_norm": 0.5029470324516296,
-      "learning_rate": 9.91215715952378e-06,
-      "loss": 0.4423,
-      "step": 3607
-    },
-    {
-      "epoch": 0.2358015815959741,
-      "grad_norm": 0.4984515309333801,
-      "learning_rate": 9.912091980342316e-06,
-      "loss": 0.4242,
-      "step": 3608
-    },
-    {
-      "epoch": 0.23586693680151624,
-      "grad_norm": 0.4875701069831848,
-      "learning_rate": 9.912026777202874e-06,
-      "loss": 0.3597,
-      "step": 3609
-    },
-    {
-      "epoch": 0.23593229200705837,
-      "grad_norm": 0.5226539969444275,
-      "learning_rate": 9.911961550105779e-06,
-      "loss": 0.4043,
-      "step": 3610
-    },
-    {
-      "epoch": 0.23599764721260047,
-      "grad_norm": 0.4634535014629364,
-      "learning_rate": 9.911896299051345e-06,
-      "loss": 0.3585,
-      "step": 3611
-    },
-    {
-      "epoch": 0.2360630024181426,
-      "grad_norm": 0.5316253900527954,
-      "learning_rate": 9.911831024039888e-06,
-      "loss": 0.4959,
-      "step": 3612
-    },
-    {
-      "epoch": 0.23612835762368473,
-      "grad_norm": 0.4336017966270447,
-      "learning_rate": 9.911765725071734e-06,
-      "loss": 0.3652,
-      "step": 3613
-    },
-    {
-      "epoch": 0.23619371282922685,
-      "grad_norm": 0.45477989315986633,
-      "learning_rate": 9.911700402147195e-06,
-      "loss": 0.4172,
-      "step": 3614
-    },
-    {
-      "epoch": 0.23625906803476898,
-      "grad_norm": 0.5167801976203918,
-      "learning_rate": 9.91163505526659e-06,
-      "loss": 0.4651,
-      "step": 3615
-    },
-    {
-      "epoch": 0.23632442324031108,
-      "grad_norm": 0.5102893710136414,
-      "learning_rate": 9.911569684430242e-06,
-      "loss": 0.4382,
-      "step": 3616
-    },
-    {
-      "epoch": 0.2363897784458532,
-      "grad_norm": 0.4945951998233795,
-      "learning_rate": 9.911504289638465e-06,
-      "loss": 0.4434,
-      "step": 3617
-    },
-    {
-      "epoch": 0.23645513365139534,
-      "grad_norm": 0.498888224363327,
-      "learning_rate": 9.91143887089158e-06,
-      "loss": 0.39,
-      "step": 3618
-    },
-    {
-      "epoch": 0.23652048885693747,
-      "grad_norm": 0.4623052477836609,
-      "learning_rate": 9.911373428189908e-06,
-      "loss": 0.4011,
-      "step": 3619
-    },
-    {
-      "epoch": 0.23658584406247957,
-      "grad_norm": 0.5997217297554016,
-      "learning_rate": 9.911307961533765e-06,
-      "loss": 0.447,
-      "step": 3620
-    },
-    {
-      "epoch": 0.2366511992680217,
-      "grad_norm": 0.49396345019340515,
-      "learning_rate": 9.911242470923472e-06,
-      "loss": 0.4589,
-      "step": 3621
-    },
-    {
-      "epoch": 0.23671655447356382,
-      "grad_norm": 0.5173277258872986,
-      "learning_rate": 9.91117695635935e-06,
-      "loss": 0.4422,
-      "step": 3622
-    },
-    {
-      "epoch": 0.23678190967910595,
-      "grad_norm": 0.502960205078125,
-      "learning_rate": 9.911111417841715e-06,
-      "loss": 0.4131,
-      "step": 3623
-    },
-    {
-      "epoch": 0.23684726488464805,
-      "grad_norm": 0.4695163071155548,
-      "learning_rate": 9.911045855370887e-06,
-      "loss": 0.3577,
-      "step": 3624
-    },
-    {
-      "epoch": 0.23691262009019018,
-      "grad_norm": 0.47594931721687317,
-      "learning_rate": 9.910980268947188e-06,
-      "loss": 0.3931,
-      "step": 3625
-    },
-    {
-      "epoch": 0.2369779752957323,
-      "grad_norm": 0.5421062707901001,
-      "learning_rate": 9.910914658570936e-06,
-      "loss": 0.4604,
-      "step": 3626
-    },
-    {
-      "epoch": 0.23704333050127444,
-      "grad_norm": 0.5341295003890991,
-      "learning_rate": 9.910849024242453e-06,
-      "loss": 0.4852,
-      "step": 3627
-    },
-    {
-      "epoch": 0.23710868570681654,
-      "grad_norm": 0.5298961400985718,
-      "learning_rate": 9.910783365962057e-06,
-      "loss": 0.4514,
-      "step": 3628
-    },
-    {
-      "epoch": 0.23717404091235866,
-      "grad_norm": 0.47505030035972595,
-      "learning_rate": 9.910717683730072e-06,
-      "loss": 0.4018,
-      "step": 3629
-    },
-    {
-      "epoch": 0.2372393961179008,
-      "grad_norm": 0.5131982564926147,
-      "learning_rate": 9.910651977546812e-06,
-      "loss": 0.4161,
-      "step": 3630
-    },
-    {
-      "epoch": 0.23730475132344292,
-      "grad_norm": 0.48219045996665955,
-      "learning_rate": 9.910586247412604e-06,
-      "loss": 0.3945,
-      "step": 3631
-    },
-    {
-      "epoch": 0.23737010652898502,
-      "grad_norm": 0.5161522626876831,
-      "learning_rate": 9.910520493327764e-06,
-      "loss": 0.3998,
-      "step": 3632
-    },
-    {
-      "epoch": 0.23743546173452715,
-      "grad_norm": 0.5477389097213745,
-      "learning_rate": 9.910454715292614e-06,
-      "loss": 0.4687,
-      "step": 3633
-    },
-    {
-      "epoch": 0.23750081694006928,
-      "grad_norm": 0.48694872856140137,
-      "learning_rate": 9.910388913307476e-06,
-      "loss": 0.4175,
-      "step": 3634
-    },
-    {
-      "epoch": 0.2375661721456114,
-      "grad_norm": 0.4663483798503876,
-      "learning_rate": 9.91032308737267e-06,
-      "loss": 0.3902,
-      "step": 3635
-    },
-    {
-      "epoch": 0.23763152735115353,
-      "grad_norm": 0.5143205523490906,
-      "learning_rate": 9.910257237488519e-06,
-      "loss": 0.377,
-      "step": 3636
-    },
-    {
-      "epoch": 0.23769688255669563,
-      "grad_norm": 0.5119413137435913,
-      "learning_rate": 9.91019136365534e-06,
-      "loss": 0.4236,
-      "step": 3637
-    },
-    {
-      "epoch": 0.23776223776223776,
-      "grad_norm": 0.43303024768829346,
-      "learning_rate": 9.910125465873458e-06,
-      "loss": 0.3411,
-      "step": 3638
-    },
-    {
-      "epoch": 0.2378275929677799,
-      "grad_norm": 0.48212507367134094,
-      "learning_rate": 9.910059544143193e-06,
-      "loss": 0.4249,
-      "step": 3639
-    },
-    {
-      "epoch": 0.23789294817332202,
-      "grad_norm": 0.5053635835647583,
-      "learning_rate": 9.909993598464865e-06,
-      "loss": 0.4061,
-      "step": 3640
-    },
-    {
-      "epoch": 0.23795830337886412,
-      "grad_norm": 0.48906877636909485,
-      "learning_rate": 9.909927628838799e-06,
-      "loss": 0.4569,
-      "step": 3641
-    },
-    {
-      "epoch": 0.23802365858440624,
-      "grad_norm": 0.48252391815185547,
-      "learning_rate": 9.909861635265315e-06,
-      "loss": 0.3862,
-      "step": 3642
-    },
-    {
-      "epoch": 0.23808901378994837,
-      "grad_norm": 0.5164976119995117,
-      "learning_rate": 9.909795617744733e-06,
-      "loss": 0.4697,
-      "step": 3643
-    },
-    {
-      "epoch": 0.2381543689954905,
-      "grad_norm": 0.4901650547981262,
-      "learning_rate": 9.909729576277379e-06,
-      "loss": 0.4316,
-      "step": 3644
-    },
-    {
-      "epoch": 0.2382197242010326,
-      "grad_norm": 0.4517502188682556,
-      "learning_rate": 9.909663510863571e-06,
-      "loss": 0.378,
-      "step": 3645
-    },
-    {
-      "epoch": 0.23828507940657473,
-      "grad_norm": 0.4997141361236572,
-      "learning_rate": 9.909597421503635e-06,
-      "loss": 0.432,
-      "step": 3646
-    },
-    {
-      "epoch": 0.23835043461211686,
-      "grad_norm": 0.5163469910621643,
-      "learning_rate": 9.90953130819789e-06,
-      "loss": 0.3907,
-      "step": 3647
-    },
-    {
-      "epoch": 0.23841578981765899,
-      "grad_norm": 0.5234763026237488,
-      "learning_rate": 9.909465170946661e-06,
-      "loss": 0.424,
-      "step": 3648
-    },
-    {
-      "epoch": 0.23848114502320109,
-      "grad_norm": 0.5328369736671448,
-      "learning_rate": 9.909399009750268e-06,
-      "loss": 0.4601,
-      "step": 3649
-    },
-    {
-      "epoch": 0.2385465002287432,
-      "grad_norm": 0.45241212844848633,
-      "learning_rate": 9.909332824609037e-06,
-      "loss": 0.3821,
-      "step": 3650
-    },
-    {
-      "epoch": 0.23861185543428534,
-      "grad_norm": 0.5118778944015503,
-      "learning_rate": 9.90926661552329e-06,
-      "loss": 0.4394,
-      "step": 3651
-    },
-    {
-      "epoch": 0.23867721063982747,
-      "grad_norm": 0.5193794369697571,
-      "learning_rate": 9.909200382493347e-06,
-      "loss": 0.4087,
-      "step": 3652
-    },
-    {
-      "epoch": 0.23874256584536957,
-      "grad_norm": 0.4448736310005188,
-      "learning_rate": 9.909134125519533e-06,
-      "loss": 0.351,
-      "step": 3653
-    },
-    {
-      "epoch": 0.2388079210509117,
-      "grad_norm": 0.4771903455257416,
-      "learning_rate": 9.909067844602172e-06,
-      "loss": 0.4167,
-      "step": 3654
-    },
-    {
-      "epoch": 0.23887327625645383,
-      "grad_norm": 0.5226938128471375,
-      "learning_rate": 9.909001539741587e-06,
-      "loss": 0.4411,
-      "step": 3655
-    },
-    {
-      "epoch": 0.23893863146199595,
-      "grad_norm": 0.48490509390830994,
-      "learning_rate": 9.9089352109381e-06,
-      "loss": 0.3758,
-      "step": 3656
-    },
-    {
-      "epoch": 0.23900398666753808,
-      "grad_norm": 0.47723546624183655,
-      "learning_rate": 9.908868858192036e-06,
-      "loss": 0.435,
-      "step": 3657
-    },
-    {
-      "epoch": 0.23906934187308018,
-      "grad_norm": 0.4633159041404724,
-      "learning_rate": 9.908802481503717e-06,
-      "loss": 0.3985,
-      "step": 3658
-    },
-    {
-      "epoch": 0.2391346970786223,
-      "grad_norm": 0.5509350895881653,
-      "learning_rate": 9.908736080873468e-06,
-      "loss": 0.4614,
-      "step": 3659
-    },
-    {
-      "epoch": 0.23920005228416444,
-      "grad_norm": 0.46527114510536194,
-      "learning_rate": 9.908669656301613e-06,
-      "loss": 0.3911,
-      "step": 3660
-    },
-    {
-      "epoch": 0.23926540748970657,
-      "grad_norm": 0.512252688407898,
-      "learning_rate": 9.908603207788475e-06,
-      "loss": 0.4333,
-      "step": 3661
-    },
-    {
-      "epoch": 0.23933076269524867,
-      "grad_norm": 0.4877110421657562,
-      "learning_rate": 9.908536735334379e-06,
-      "loss": 0.3963,
-      "step": 3662
-    },
-    {
-      "epoch": 0.2393961179007908,
-      "grad_norm": 0.4963582754135132,
-      "learning_rate": 9.908470238939649e-06,
-      "loss": 0.4148,
-      "step": 3663
-    },
-    {
-      "epoch": 0.23946147310633292,
-      "grad_norm": 0.4605066180229187,
-      "learning_rate": 9.908403718604609e-06,
-      "loss": 0.373,
-      "step": 3664
-    },
-    {
-      "epoch": 0.23952682831187505,
-      "grad_norm": 0.468755841255188,
-      "learning_rate": 9.908337174329583e-06,
-      "loss": 0.375,
-      "step": 3665
-    },
-    {
-      "epoch": 0.23959218351741715,
-      "grad_norm": 0.5060827732086182,
-      "learning_rate": 9.908270606114897e-06,
-      "loss": 0.4497,
-      "step": 3666
-    },
-    {
-      "epoch": 0.23965753872295928,
-      "grad_norm": 0.44706472754478455,
-      "learning_rate": 9.908204013960875e-06,
-      "loss": 0.3538,
-      "step": 3667
-    },
-    {
-      "epoch": 0.2397228939285014,
-      "grad_norm": 0.4768923819065094,
-      "learning_rate": 9.90813739786784e-06,
-      "loss": 0.39,
-      "step": 3668
-    },
-    {
-      "epoch": 0.23978824913404354,
-      "grad_norm": 0.4842081069946289,
-      "learning_rate": 9.908070757836121e-06,
-      "loss": 0.4624,
-      "step": 3669
-    },
-    {
-      "epoch": 0.23985360433958564,
-      "grad_norm": 0.47158530354499817,
-      "learning_rate": 9.90800409386604e-06,
-      "loss": 0.4236,
-      "step": 3670
-    },
-    {
-      "epoch": 0.23991895954512776,
-      "grad_norm": 0.5272809267044067,
-      "learning_rate": 9.907937405957921e-06,
-      "loss": 0.3945,
-      "step": 3671
-    },
-    {
-      "epoch": 0.2399843147506699,
-      "grad_norm": 0.4997832775115967,
-      "learning_rate": 9.907870694112092e-06,
-      "loss": 0.4343,
-      "step": 3672
-    },
-    {
-      "epoch": 0.24004966995621202,
-      "grad_norm": 0.4826876223087311,
-      "learning_rate": 9.907803958328879e-06,
-      "loss": 0.4162,
-      "step": 3673
-    },
-    {
-      "epoch": 0.24011502516175415,
-      "grad_norm": 0.5297667384147644,
-      "learning_rate": 9.907737198608604e-06,
-      "loss": 0.495,
-      "step": 3674
-    },
-    {
-      "epoch": 0.24018038036729625,
-      "grad_norm": 0.48894059658050537,
-      "learning_rate": 9.907670414951596e-06,
-      "loss": 0.4015,
-      "step": 3675
-    },
-    {
-      "epoch": 0.24024573557283838,
-      "grad_norm": 0.49508002400398254,
-      "learning_rate": 9.907603607358178e-06,
-      "loss": 0.4239,
-      "step": 3676
-    },
-    {
-      "epoch": 0.2403110907783805,
-      "grad_norm": 0.4823409914970398,
-      "learning_rate": 9.907536775828677e-06,
-      "loss": 0.4233,
-      "step": 3677
-    },
-    {
-      "epoch": 0.24037644598392263,
-      "grad_norm": 0.5320729613304138,
-      "learning_rate": 9.907469920363418e-06,
-      "loss": 0.4583,
-      "step": 3678
-    },
-    {
-      "epoch": 0.24044180118946473,
-      "grad_norm": 0.564666211605072,
-      "learning_rate": 9.90740304096273e-06,
-      "loss": 0.4155,
-      "step": 3679
-    },
-    {
-      "epoch": 0.24050715639500686,
-      "grad_norm": 0.46652790904045105,
-      "learning_rate": 9.907336137626937e-06,
-      "loss": 0.3988,
-      "step": 3680
-    },
-    {
-      "epoch": 0.240572511600549,
-      "grad_norm": 0.47065576910972595,
-      "learning_rate": 9.907269210356364e-06,
-      "loss": 0.3879,
-      "step": 3681
-    },
-    {
-      "epoch": 0.24063786680609112,
-      "grad_norm": 0.4547605514526367,
-      "learning_rate": 9.90720225915134e-06,
-      "loss": 0.3587,
-      "step": 3682
-    },
-    {
-      "epoch": 0.24070322201163322,
-      "grad_norm": 0.523067831993103,
-      "learning_rate": 9.907135284012191e-06,
-      "loss": 0.4263,
-      "step": 3683
-    },
-    {
-      "epoch": 0.24076857721717534,
-      "grad_norm": 0.4699363708496094,
-      "learning_rate": 9.907068284939244e-06,
-      "loss": 0.38,
-      "step": 3684
-    },
-    {
-      "epoch": 0.24083393242271747,
-      "grad_norm": 0.44570061564445496,
-      "learning_rate": 9.907001261932824e-06,
-      "loss": 0.3761,
-      "step": 3685
-    },
-    {
-      "epoch": 0.2408992876282596,
-      "grad_norm": 0.4833367168903351,
-      "learning_rate": 9.906934214993259e-06,
-      "loss": 0.3602,
-      "step": 3686
-    },
-    {
-      "epoch": 0.2409646428338017,
-      "grad_norm": 0.4648684561252594,
-      "learning_rate": 9.906867144120875e-06,
-      "loss": 0.3739,
-      "step": 3687
-    },
-    {
-      "epoch": 0.24102999803934383,
-      "grad_norm": 0.6127973198890686,
-      "learning_rate": 9.906800049316001e-06,
-      "loss": 0.396,
-      "step": 3688
-    },
-    {
-      "epoch": 0.24109535324488596,
-      "grad_norm": 0.4691463112831116,
-      "learning_rate": 9.906732930578963e-06,
-      "loss": 0.4036,
-      "step": 3689
-    },
-    {
-      "epoch": 0.24116070845042809,
-      "grad_norm": 0.4602457582950592,
-      "learning_rate": 9.906665787910089e-06,
-      "loss": 0.3991,
-      "step": 3690
-    },
-    {
-      "epoch": 0.24122606365597019,
-      "grad_norm": 0.486935555934906,
-      "learning_rate": 9.906598621309706e-06,
-      "loss": 0.4189,
-      "step": 3691
-    },
-    {
-      "epoch": 0.2412914188615123,
-      "grad_norm": 0.4824185073375702,
-      "learning_rate": 9.906531430778142e-06,
-      "loss": 0.4131,
-      "step": 3692
-    },
-    {
-      "epoch": 0.24135677406705444,
-      "grad_norm": 0.4648727774620056,
-      "learning_rate": 9.906464216315724e-06,
-      "loss": 0.3911,
-      "step": 3693
-    },
-    {
-      "epoch": 0.24142212927259657,
-      "grad_norm": 0.4880336821079254,
-      "learning_rate": 9.90639697792278e-06,
-      "loss": 0.4357,
-      "step": 3694
-    },
-    {
-      "epoch": 0.2414874844781387,
-      "grad_norm": 0.4606556296348572,
-      "learning_rate": 9.906329715599639e-06,
-      "loss": 0.3877,
-      "step": 3695
-    },
-    {
-      "epoch": 0.2415528396836808,
-      "grad_norm": 0.4391932189464569,
-      "learning_rate": 9.906262429346627e-06,
-      "loss": 0.3512,
-      "step": 3696
-    },
-    {
-      "epoch": 0.24161819488922293,
-      "grad_norm": 0.4812285602092743,
-      "learning_rate": 9.906195119164074e-06,
-      "loss": 0.4152,
-      "step": 3697
-    },
-    {
-      "epoch": 0.24168355009476505,
-      "grad_norm": 0.4891197979450226,
-      "learning_rate": 9.906127785052308e-06,
-      "loss": 0.4649,
-      "step": 3698
-    },
-    {
-      "epoch": 0.24174890530030718,
-      "grad_norm": 0.48923251032829285,
-      "learning_rate": 9.906060427011657e-06,
-      "loss": 0.4232,
-      "step": 3699
-    },
-    {
-      "epoch": 0.24181426050584928,
-      "grad_norm": 0.44078633189201355,
-      "learning_rate": 9.90599304504245e-06,
-      "loss": 0.3612,
-      "step": 3700
-    },
-    {
-      "epoch": 0.2418796157113914,
-      "grad_norm": 0.48194047808647156,
-      "learning_rate": 9.905925639145015e-06,
-      "loss": 0.3981,
-      "step": 3701
-    },
-    {
-      "epoch": 0.24194497091693354,
-      "grad_norm": 0.53200364112854,
-      "learning_rate": 9.905858209319681e-06,
-      "loss": 0.4444,
-      "step": 3702
-    },
-    {
-      "epoch": 0.24201032612247567,
-      "grad_norm": 0.5198573470115662,
-      "learning_rate": 9.905790755566777e-06,
-      "loss": 0.485,
-      "step": 3703
-    },
-    {
-      "epoch": 0.24207568132801777,
-      "grad_norm": 0.4784989655017853,
-      "learning_rate": 9.905723277886631e-06,
-      "loss": 0.4177,
-      "step": 3704
-    },
-    {
-      "epoch": 0.2421410365335599,
-      "grad_norm": 0.49317121505737305,
-      "learning_rate": 9.905655776279576e-06,
-      "loss": 0.4136,
-      "step": 3705
-    },
-    {
-      "epoch": 0.24220639173910202,
-      "grad_norm": 0.5199515223503113,
-      "learning_rate": 9.905588250745936e-06,
-      "loss": 0.4396,
-      "step": 3706
-    },
-    {
-      "epoch": 0.24227174694464415,
-      "grad_norm": 0.5406073927879333,
-      "learning_rate": 9.905520701286043e-06,
-      "loss": 0.429,
-      "step": 3707
-    },
-    {
-      "epoch": 0.24233710215018625,
-      "grad_norm": 0.4321300685405731,
-      "learning_rate": 9.905453127900227e-06,
-      "loss": 0.355,
-      "step": 3708
-    },
-    {
-      "epoch": 0.24240245735572838,
-      "grad_norm": 0.4865424335002899,
-      "learning_rate": 9.905385530588817e-06,
-      "loss": 0.3876,
-      "step": 3709
-    },
-    {
-      "epoch": 0.2424678125612705,
-      "grad_norm": 0.4845200479030609,
-      "learning_rate": 9.905317909352139e-06,
-      "loss": 0.4025,
-      "step": 3710
-    },
-    {
-      "epoch": 0.24253316776681264,
-      "grad_norm": 0.47324949502944946,
-      "learning_rate": 9.90525026419053e-06,
-      "loss": 0.3774,
-      "step": 3711
-    },
-    {
-      "epoch": 0.24259852297235474,
-      "grad_norm": 0.47342386841773987,
-      "learning_rate": 9.905182595104314e-06,
-      "loss": 0.4033,
-      "step": 3712
-    },
-    {
-      "epoch": 0.24266387817789686,
-      "grad_norm": 0.49976396560668945,
-      "learning_rate": 9.905114902093824e-06,
-      "loss": 0.4306,
-      "step": 3713
-    },
-    {
-      "epoch": 0.242729233383439,
-      "grad_norm": 0.4608675241470337,
-      "learning_rate": 9.905047185159389e-06,
-      "loss": 0.4018,
-      "step": 3714
-    },
-    {
-      "epoch": 0.24279458858898112,
-      "grad_norm": 0.4639568626880646,
-      "learning_rate": 9.90497944430134e-06,
-      "loss": 0.4036,
-      "step": 3715
-    },
-    {
-      "epoch": 0.24285994379452325,
-      "grad_norm": 0.4871489107608795,
-      "learning_rate": 9.904911679520006e-06,
-      "loss": 0.4131,
-      "step": 3716
-    },
-    {
-      "epoch": 0.24292529900006535,
-      "grad_norm": 0.4942607581615448,
-      "learning_rate": 9.90484389081572e-06,
-      "loss": 0.4076,
-      "step": 3717
-    },
-    {
-      "epoch": 0.24299065420560748,
-      "grad_norm": 0.5495970845222473,
-      "learning_rate": 9.90477607818881e-06,
-      "loss": 0.4217,
-      "step": 3718
-    },
-    {
-      "epoch": 0.2430560094111496,
-      "grad_norm": 0.4705306887626648,
-      "learning_rate": 9.904708241639606e-06,
-      "loss": 0.4024,
-      "step": 3719
-    },
-    {
-      "epoch": 0.24312136461669173,
-      "grad_norm": 0.44730666279792786,
-      "learning_rate": 9.904640381168444e-06,
-      "loss": 0.3443,
-      "step": 3720
-    },
-    {
-      "epoch": 0.24318671982223383,
-      "grad_norm": 0.47078898549079895,
-      "learning_rate": 9.90457249677565e-06,
-      "loss": 0.3724,
-      "step": 3721
-    },
-    {
-      "epoch": 0.24325207502777596,
-      "grad_norm": 0.4704471230506897,
-      "learning_rate": 9.904504588461558e-06,
-      "loss": 0.3869,
-      "step": 3722
-    },
-    {
-      "epoch": 0.2433174302333181,
-      "grad_norm": 0.46855428814888,
-      "learning_rate": 9.904436656226497e-06,
-      "loss": 0.4121,
-      "step": 3723
-    },
-    {
-      "epoch": 0.24338278543886022,
-      "grad_norm": 0.49998342990875244,
-      "learning_rate": 9.904368700070802e-06,
-      "loss": 0.4077,
-      "step": 3724
-    },
-    {
-      "epoch": 0.24344814064440232,
-      "grad_norm": 0.45472219586372375,
-      "learning_rate": 9.904300719994798e-06,
-      "loss": 0.3806,
-      "step": 3725
-    },
-    {
-      "epoch": 0.24351349584994444,
-      "grad_norm": 0.4763962924480438,
-      "learning_rate": 9.904232715998822e-06,
-      "loss": 0.4262,
-      "step": 3726
-    },
-    {
-      "epoch": 0.24357885105548657,
-      "grad_norm": 0.5250136852264404,
-      "learning_rate": 9.904164688083204e-06,
-      "loss": 0.4254,
-      "step": 3727
-    },
-    {
-      "epoch": 0.2436442062610287,
-      "grad_norm": 0.513821542263031,
-      "learning_rate": 9.904096636248278e-06,
-      "loss": 0.4359,
-      "step": 3728
-    },
-    {
-      "epoch": 0.2437095614665708,
-      "grad_norm": 0.47730183601379395,
-      "learning_rate": 9.90402856049437e-06,
-      "loss": 0.3834,
-      "step": 3729
-    },
-    {
-      "epoch": 0.24377491667211293,
-      "grad_norm": 0.4817773401737213,
-      "learning_rate": 9.903960460821818e-06,
-      "loss": 0.4112,
-      "step": 3730
-    },
-    {
-      "epoch": 0.24384027187765506,
-      "grad_norm": 0.4721023440361023,
-      "learning_rate": 9.903892337230952e-06,
-      "loss": 0.3987,
-      "step": 3731
-    },
-    {
-      "epoch": 0.24390562708319719,
-      "grad_norm": 0.4319049119949341,
-      "learning_rate": 9.903824189722103e-06,
-      "loss": 0.3409,
-      "step": 3732
-    },
-    {
-      "epoch": 0.24397098228873929,
-      "grad_norm": 0.5072712898254395,
-      "learning_rate": 9.903756018295605e-06,
-      "loss": 0.386,
-      "step": 3733
-    },
-    {
-      "epoch": 0.2440363374942814,
-      "grad_norm": 0.45842063426971436,
-      "learning_rate": 9.903687822951791e-06,
-      "loss": 0.3715,
-      "step": 3734
-    },
-    {
-      "epoch": 0.24410169269982354,
-      "grad_norm": 0.5203177332878113,
-      "learning_rate": 9.903619603690991e-06,
-      "loss": 0.4744,
-      "step": 3735
-    },
-    {
-      "epoch": 0.24416704790536567,
-      "grad_norm": 0.46919986605644226,
-      "learning_rate": 9.903551360513542e-06,
-      "loss": 0.4347,
-      "step": 3736
-    },
-    {
-      "epoch": 0.2442324031109078,
-      "grad_norm": 0.4473360776901245,
-      "learning_rate": 9.903483093419773e-06,
-      "loss": 0.358,
-      "step": 3737
-    },
-    {
-      "epoch": 0.2442977583164499,
-      "grad_norm": 0.4448186755180359,
-      "learning_rate": 9.903414802410016e-06,
-      "loss": 0.3769,
-      "step": 3738
-    },
-    {
-      "epoch": 0.24436311352199203,
-      "grad_norm": 0.4455445110797882,
-      "learning_rate": 9.903346487484608e-06,
-      "loss": 0.351,
-      "step": 3739
-    },
-    {
-      "epoch": 0.24442846872753415,
-      "grad_norm": 0.49022209644317627,
-      "learning_rate": 9.903278148643883e-06,
-      "loss": 0.4276,
-      "step": 3740
-    },
-    {
-      "epoch": 0.24449382393307628,
-      "grad_norm": 0.49154970049858093,
-      "learning_rate": 9.903209785888168e-06,
-      "loss": 0.4033,
-      "step": 3741
-    },
-    {
-      "epoch": 0.24455917913861838,
-      "grad_norm": 0.5160120129585266,
-      "learning_rate": 9.903141399217801e-06,
-      "loss": 0.4124,
-      "step": 3742
-    },
-    {
-      "epoch": 0.2446245343441605,
-      "grad_norm": 0.49152106046676636,
-      "learning_rate": 9.903072988633117e-06,
-      "loss": 0.3852,
-      "step": 3743
-    },
-    {
-      "epoch": 0.24468988954970264,
-      "grad_norm": 0.5074083209037781,
-      "learning_rate": 9.903004554134445e-06,
-      "loss": 0.431,
-      "step": 3744
-    },
-    {
-      "epoch": 0.24475524475524477,
-      "grad_norm": 0.4891955554485321,
-      "learning_rate": 9.902936095722123e-06,
-      "loss": 0.3834,
-      "step": 3745
-    },
-    {
-      "epoch": 0.24482059996078687,
-      "grad_norm": 0.4742303788661957,
-      "learning_rate": 9.902867613396482e-06,
-      "loss": 0.3664,
-      "step": 3746
-    },
-    {
-      "epoch": 0.244885955166329,
-      "grad_norm": 0.5135862827301025,
-      "learning_rate": 9.902799107157857e-06,
-      "loss": 0.4371,
-      "step": 3747
-    },
-    {
-      "epoch": 0.24495131037187112,
-      "grad_norm": 0.462126225233078,
-      "learning_rate": 9.902730577006583e-06,
-      "loss": 0.3749,
-      "step": 3748
-    },
-    {
-      "epoch": 0.24501666557741325,
-      "grad_norm": 0.5256122946739197,
-      "learning_rate": 9.902662022942994e-06,
-      "loss": 0.4601,
-      "step": 3749
-    },
-    {
-      "epoch": 0.24508202078295535,
-      "grad_norm": 0.4707671105861664,
-      "learning_rate": 9.902593444967424e-06,
-      "loss": 0.3879,
-      "step": 3750
-    },
-    {
-      "epoch": 0.24514737598849748,
-      "grad_norm": 0.4596821665763855,
-      "learning_rate": 9.902524843080206e-06,
-      "loss": 0.4088,
-      "step": 3751
-    },
-    {
-      "epoch": 0.2452127311940396,
-      "grad_norm": 0.4927043616771698,
-      "learning_rate": 9.902456217281674e-06,
-      "loss": 0.4366,
-      "step": 3752
-    },
-    {
-      "epoch": 0.24527808639958174,
-      "grad_norm": 0.48119106888771057,
-      "learning_rate": 9.90238756757217e-06,
-      "loss": 0.4108,
-      "step": 3753
-    },
-    {
-      "epoch": 0.24534344160512384,
-      "grad_norm": 0.48688697814941406,
-      "learning_rate": 9.90231889395202e-06,
-      "loss": 0.4673,
-      "step": 3754
-    },
-    {
-      "epoch": 0.24540879681066596,
-      "grad_norm": 0.4701494872570038,
-      "learning_rate": 9.902250196421562e-06,
-      "loss": 0.3807,
-      "step": 3755
-    },
-    {
-      "epoch": 0.2454741520162081,
-      "grad_norm": 0.44716522097587585,
-      "learning_rate": 9.902181474981133e-06,
-      "loss": 0.3892,
-      "step": 3756
-    },
-    {
-      "epoch": 0.24553950722175022,
-      "grad_norm": 0.4895716607570648,
-      "learning_rate": 9.902112729631066e-06,
-      "loss": 0.4221,
-      "step": 3757
-    },
-    {
-      "epoch": 0.24560486242729235,
-      "grad_norm": 0.49176111817359924,
-      "learning_rate": 9.902043960371697e-06,
-      "loss": 0.4293,
-      "step": 3758
-    },
-    {
-      "epoch": 0.24567021763283445,
-      "grad_norm": 0.4756181240081787,
-      "learning_rate": 9.901975167203361e-06,
-      "loss": 0.3844,
-      "step": 3759
-    },
-    {
-      "epoch": 0.24573557283837658,
-      "grad_norm": 0.48180314898490906,
-      "learning_rate": 9.901906350126395e-06,
-      "loss": 0.4009,
-      "step": 3760
-    },
-    {
-      "epoch": 0.2458009280439187,
-      "grad_norm": 0.48072636127471924,
-      "learning_rate": 9.901837509141132e-06,
-      "loss": 0.394,
-      "step": 3761
-    },
-    {
-      "epoch": 0.24586628324946083,
-      "grad_norm": 0.4648549258708954,
-      "learning_rate": 9.901768644247911e-06,
-      "loss": 0.3976,
-      "step": 3762
-    },
-    {
-      "epoch": 0.24593163845500293,
-      "grad_norm": 0.418720006942749,
-      "learning_rate": 9.901699755447065e-06,
-      "loss": 0.333,
-      "step": 3763
-    },
-    {
-      "epoch": 0.24599699366054506,
-      "grad_norm": 0.4856666624546051,
-      "learning_rate": 9.901630842738931e-06,
-      "loss": 0.4291,
-      "step": 3764
-    },
-    {
-      "epoch": 0.2460623488660872,
-      "grad_norm": 0.47633805871009827,
-      "learning_rate": 9.901561906123846e-06,
-      "loss": 0.4316,
-      "step": 3765
-    },
-    {
-      "epoch": 0.24612770407162932,
-      "grad_norm": 0.5254993438720703,
-      "learning_rate": 9.901492945602147e-06,
-      "loss": 0.4471,
-      "step": 3766
-    },
-    {
-      "epoch": 0.24619305927717142,
-      "grad_norm": 0.48775529861450195,
-      "learning_rate": 9.901423961174167e-06,
-      "loss": 0.3877,
-      "step": 3767
-    },
-    {
-      "epoch": 0.24625841448271354,
-      "grad_norm": 0.5022530555725098,
-      "learning_rate": 9.901354952840245e-06,
-      "loss": 0.4226,
-      "step": 3768
-    },
-    {
-      "epoch": 0.24632376968825567,
-      "grad_norm": 0.4446774125099182,
-      "learning_rate": 9.901285920600717e-06,
-      "loss": 0.3532,
-      "step": 3769
-    },
-    {
-      "epoch": 0.2463891248937978,
-      "grad_norm": 0.4788687527179718,
-      "learning_rate": 9.901216864455918e-06,
-      "loss": 0.3943,
-      "step": 3770
-    },
-    {
-      "epoch": 0.2464544800993399,
-      "grad_norm": 0.5055045485496521,
-      "learning_rate": 9.901147784406188e-06,
-      "loss": 0.4492,
-      "step": 3771
-    },
-    {
-      "epoch": 0.24651983530488203,
-      "grad_norm": 0.4874025881290436,
-      "learning_rate": 9.90107868045186e-06,
-      "loss": 0.409,
-      "step": 3772
-    },
-    {
-      "epoch": 0.24658519051042416,
-      "grad_norm": 0.5308887958526611,
-      "learning_rate": 9.901009552593277e-06,
-      "loss": 0.4231,
-      "step": 3773
-    },
-    {
-      "epoch": 0.24665054571596629,
-      "grad_norm": 0.48197951912879944,
-      "learning_rate": 9.900940400830771e-06,
-      "loss": 0.4134,
-      "step": 3774
-    },
-    {
-      "epoch": 0.24671590092150839,
-      "grad_norm": 0.5125496983528137,
-      "learning_rate": 9.90087122516468e-06,
-      "loss": 0.3976,
-      "step": 3775
-    },
-    {
-      "epoch": 0.2467812561270505,
-      "grad_norm": 0.4560488164424896,
-      "learning_rate": 9.900802025595342e-06,
-      "loss": 0.3593,
-      "step": 3776
-    },
-    {
-      "epoch": 0.24684661133259264,
-      "grad_norm": 0.4720732867717743,
-      "learning_rate": 9.900732802123097e-06,
-      "loss": 0.4183,
-      "step": 3777
-    },
-    {
-      "epoch": 0.24691196653813477,
-      "grad_norm": 0.5029764771461487,
-      "learning_rate": 9.900663554748278e-06,
-      "loss": 0.4845,
-      "step": 3778
-    },
-    {
-      "epoch": 0.2469773217436769,
-      "grad_norm": 0.46738043427467346,
-      "learning_rate": 9.900594283471226e-06,
-      "loss": 0.4029,
-      "step": 3779
-    },
-    {
-      "epoch": 0.247042676949219,
-      "grad_norm": 0.47529447078704834,
-      "learning_rate": 9.900524988292278e-06,
-      "loss": 0.395,
-      "step": 3780
-    },
-    {
-      "epoch": 0.24710803215476113,
-      "grad_norm": 0.49335628747940063,
-      "learning_rate": 9.900455669211773e-06,
-      "loss": 0.4791,
-      "step": 3781
-    },
-    {
-      "epoch": 0.24717338736030325,
-      "grad_norm": 0.4847749173641205,
-      "learning_rate": 9.900386326230046e-06,
-      "loss": 0.377,
-      "step": 3782
-    },
-    {
-      "epoch": 0.24723874256584538,
-      "grad_norm": 0.48539313673973083,
-      "learning_rate": 9.900316959347439e-06,
-      "loss": 0.4176,
-      "step": 3783
-    },
-    {
-      "epoch": 0.24730409777138748,
-      "grad_norm": 0.47560250759124756,
-      "learning_rate": 9.900247568564287e-06,
-      "loss": 0.4224,
-      "step": 3784
-    },
-    {
-      "epoch": 0.2473694529769296,
-      "grad_norm": 0.5065321922302246,
-      "learning_rate": 9.90017815388093e-06,
-      "loss": 0.4549,
-      "step": 3785
-    },
-    {
-      "epoch": 0.24743480818247174,
-      "grad_norm": 0.41895002126693726,
-      "learning_rate": 9.900108715297707e-06,
-      "loss": 0.3222,
-      "step": 3786
-    },
-    {
-      "epoch": 0.24750016338801387,
-      "grad_norm": 0.4707399308681488,
-      "learning_rate": 9.900039252814957e-06,
-      "loss": 0.4061,
-      "step": 3787
-    },
-    {
-      "epoch": 0.24756551859355597,
-      "grad_norm": 0.4642290771007538,
-      "learning_rate": 9.899969766433018e-06,
-      "loss": 0.3804,
-      "step": 3788
-    },
-    {
-      "epoch": 0.2476308737990981,
-      "grad_norm": 0.4792925715446472,
-      "learning_rate": 9.899900256152228e-06,
-      "loss": 0.36,
-      "step": 3789
-    },
-    {
-      "epoch": 0.24769622900464022,
-      "grad_norm": 0.48355647921562195,
-      "learning_rate": 9.899830721972927e-06,
-      "loss": 0.4061,
-      "step": 3790
-    },
-    {
-      "epoch": 0.24776158421018235,
-      "grad_norm": 0.479915976524353,
-      "learning_rate": 9.899761163895454e-06,
-      "loss": 0.3863,
-      "step": 3791
-    },
-    {
-      "epoch": 0.24782693941572445,
-      "grad_norm": 0.4555598199367523,
-      "learning_rate": 9.89969158192015e-06,
-      "loss": 0.4038,
-      "step": 3792
-    },
-    {
-      "epoch": 0.24789229462126658,
-      "grad_norm": 0.5053451061248779,
-      "learning_rate": 9.899621976047351e-06,
-      "loss": 0.423,
-      "step": 3793
-    },
-    {
-      "epoch": 0.2479576498268087,
-      "grad_norm": 0.503421425819397,
-      "learning_rate": 9.899552346277399e-06,
-      "loss": 0.4331,
-      "step": 3794
-    },
-    {
-      "epoch": 0.24802300503235084,
-      "grad_norm": 0.48323407769203186,
-      "learning_rate": 9.899482692610633e-06,
-      "loss": 0.4149,
-      "step": 3795
-    },
-    {
-      "epoch": 0.24808836023789294,
-      "grad_norm": 0.43359190225601196,
-      "learning_rate": 9.899413015047392e-06,
-      "loss": 0.3766,
-      "step": 3796
-    },
-    {
-      "epoch": 0.24815371544343506,
-      "grad_norm": 0.4996505081653595,
-      "learning_rate": 9.899343313588017e-06,
-      "loss": 0.3819,
-      "step": 3797
-    },
-    {
-      "epoch": 0.2482190706489772,
-      "grad_norm": 0.533610463142395,
-      "learning_rate": 9.899273588232847e-06,
-      "loss": 0.4787,
-      "step": 3798
-    },
-    {
-      "epoch": 0.24828442585451932,
-      "grad_norm": 0.5073145031929016,
-      "learning_rate": 9.899203838982221e-06,
-      "loss": 0.4339,
-      "step": 3799
-    },
-    {
-      "epoch": 0.24834978106006145,
-      "grad_norm": 0.4526595175266266,
-      "learning_rate": 9.899134065836482e-06,
-      "loss": 0.4048,
-      "step": 3800
-    },
-    {
-      "epoch": 0.24841513626560355,
-      "grad_norm": 0.4555380344390869,
-      "learning_rate": 9.89906426879597e-06,
-      "loss": 0.3982,
-      "step": 3801
-    },
-    {
-      "epoch": 0.24848049147114568,
-      "grad_norm": 0.47561123967170715,
-      "learning_rate": 9.898994447861024e-06,
-      "loss": 0.4033,
-      "step": 3802
-    },
-    {
-      "epoch": 0.2485458466766878,
-      "grad_norm": 0.49485746026039124,
-      "learning_rate": 9.898924603031983e-06,
-      "loss": 0.4475,
-      "step": 3803
-    },
-    {
-      "epoch": 0.24861120188222993,
-      "grad_norm": 0.44991278648376465,
-      "learning_rate": 9.898854734309191e-06,
-      "loss": 0.3678,
-      "step": 3804
-    },
-    {
-      "epoch": 0.24867655708777203,
-      "grad_norm": 0.45418450236320496,
-      "learning_rate": 9.898784841692988e-06,
-      "loss": 0.3994,
-      "step": 3805
-    },
-    {
-      "epoch": 0.24874191229331416,
-      "grad_norm": 0.50419682264328,
-      "learning_rate": 9.898714925183713e-06,
-      "loss": 0.4467,
-      "step": 3806
-    },
-    {
-      "epoch": 0.2488072674988563,
-      "grad_norm": 0.4924890995025635,
-      "learning_rate": 9.898644984781708e-06,
-      "loss": 0.4426,
-      "step": 3807
-    },
-    {
-      "epoch": 0.24887262270439842,
-      "grad_norm": 0.47436606884002686,
-      "learning_rate": 9.898575020487315e-06,
-      "loss": 0.4162,
-      "step": 3808
-    },
-    {
-      "epoch": 0.24893797790994052,
-      "grad_norm": 0.46907737851142883,
-      "learning_rate": 9.898505032300875e-06,
-      "loss": 0.4353,
-      "step": 3809
-    },
-    {
-      "epoch": 0.24900333311548264,
-      "grad_norm": 0.47675079107284546,
-      "learning_rate": 9.898435020222728e-06,
-      "loss": 0.3905,
-      "step": 3810
-    },
-    {
-      "epoch": 0.24906868832102477,
-      "grad_norm": 0.44763273000717163,
-      "learning_rate": 9.898364984253216e-06,
-      "loss": 0.378,
-      "step": 3811
-    },
-    {
-      "epoch": 0.2491340435265669,
-      "grad_norm": 0.447813481092453,
-      "learning_rate": 9.898294924392683e-06,
-      "loss": 0.366,
-      "step": 3812
-    },
-    {
-      "epoch": 0.249199398732109,
-      "grad_norm": 0.49202960729599,
-      "learning_rate": 9.898224840641469e-06,
-      "loss": 0.4146,
-      "step": 3813
-    },
-    {
-      "epoch": 0.24926475393765113,
-      "grad_norm": 0.4937524199485779,
-      "learning_rate": 9.898154732999912e-06,
-      "loss": 0.4352,
-      "step": 3814
-    },
-    {
-      "epoch": 0.24933010914319326,
-      "grad_norm": 0.4539417028427124,
-      "learning_rate": 9.89808460146836e-06,
-      "loss": 0.3948,
-      "step": 3815
-    },
-    {
-      "epoch": 0.24939546434873539,
-      "grad_norm": 0.48760756850242615,
-      "learning_rate": 9.898014446047153e-06,
-      "loss": 0.3868,
-      "step": 3816
-    },
-    {
-      "epoch": 0.24946081955427749,
-      "grad_norm": 0.5057722926139832,
-      "learning_rate": 9.897944266736632e-06,
-      "loss": 0.4052,
-      "step": 3817
-    },
-    {
-      "epoch": 0.2495261747598196,
-      "grad_norm": 0.4397360384464264,
-      "learning_rate": 9.89787406353714e-06,
-      "loss": 0.3392,
-      "step": 3818
-    },
-    {
-      "epoch": 0.24959152996536174,
-      "grad_norm": 0.4437639117240906,
-      "learning_rate": 9.897803836449018e-06,
-      "loss": 0.3629,
-      "step": 3819
-    },
-    {
-      "epoch": 0.24965688517090387,
-      "grad_norm": 0.4564015567302704,
-      "learning_rate": 9.897733585472612e-06,
-      "loss": 0.4098,
-      "step": 3820
-    },
-    {
-      "epoch": 0.249722240376446,
-      "grad_norm": 0.48790571093559265,
-      "learning_rate": 9.897663310608261e-06,
-      "loss": 0.3884,
-      "step": 3821
-    },
-    {
-      "epoch": 0.2497875955819881,
-      "grad_norm": 0.5272866487503052,
-      "learning_rate": 9.89759301185631e-06,
-      "loss": 0.4879,
-      "step": 3822
-    },
-    {
-      "epoch": 0.24985295078753023,
-      "grad_norm": 0.48300930857658386,
-      "learning_rate": 9.897522689217102e-06,
-      "loss": 0.4706,
-      "step": 3823
-    },
-    {
-      "epoch": 0.24991830599307235,
-      "grad_norm": 0.48800382018089294,
-      "learning_rate": 9.897452342690979e-06,
-      "loss": 0.4633,
-      "step": 3824
-    },
-    {
-      "epoch": 0.24998366119861448,
-      "grad_norm": 0.48852139711380005,
-      "learning_rate": 9.897381972278284e-06,
-      "loss": 0.4146,
-      "step": 3825
-    },
-    {
-      "epoch": 0.2500490164041566,
-      "grad_norm": 0.5131561160087585,
-      "learning_rate": 9.89731157797936e-06,
-      "loss": 0.4364,
-      "step": 3826
-    },
-    {
-      "epoch": 0.25011437160969874,
-      "grad_norm": 0.45051053166389465,
-      "learning_rate": 9.897241159794552e-06,
-      "loss": 0.3748,
-      "step": 3827
-    },
-    {
-      "epoch": 0.2501797268152408,
-      "grad_norm": 0.5114407539367676,
-      "learning_rate": 9.8971707177242e-06,
-      "loss": 0.4334,
-      "step": 3828
-    },
-    {
-      "epoch": 0.25024508202078294,
-      "grad_norm": 0.44549310207366943,
-      "learning_rate": 9.897100251768652e-06,
-      "loss": 0.3661,
-      "step": 3829
-    },
-    {
-      "epoch": 0.25031043722632507,
-      "grad_norm": 0.49115490913391113,
-      "learning_rate": 9.89702976192825e-06,
-      "loss": 0.3887,
-      "step": 3830
-    },
-    {
-      "epoch": 0.2503757924318672,
-      "grad_norm": 0.491936057806015,
-      "learning_rate": 9.896959248203335e-06,
-      "loss": 0.4038,
-      "step": 3831
-    },
-    {
-      "epoch": 0.2504411476374093,
-      "grad_norm": 0.5014188885688782,
-      "learning_rate": 9.896888710594255e-06,
-      "loss": 0.4619,
-      "step": 3832
-    },
-    {
-      "epoch": 0.25050650284295145,
-      "grad_norm": 0.4646052122116089,
-      "learning_rate": 9.896818149101352e-06,
-      "loss": 0.4399,
-      "step": 3833
-    },
-    {
-      "epoch": 0.2505718580484936,
-      "grad_norm": 0.45592236518859863,
-      "learning_rate": 9.89674756372497e-06,
-      "loss": 0.3611,
-      "step": 3834
-    },
-    {
-      "epoch": 0.2506372132540357,
-      "grad_norm": 0.47756922245025635,
-      "learning_rate": 9.896676954465454e-06,
-      "loss": 0.4089,
-      "step": 3835
-    },
-    {
-      "epoch": 0.2507025684595778,
-      "grad_norm": 0.49206778407096863,
-      "learning_rate": 9.896606321323147e-06,
-      "loss": 0.4206,
-      "step": 3836
-    },
-    {
-      "epoch": 0.2507679236651199,
-      "grad_norm": 0.48088929057121277,
-      "learning_rate": 9.896535664298396e-06,
-      "loss": 0.4064,
-      "step": 3837
-    },
-    {
-      "epoch": 0.25083327887066204,
-      "grad_norm": 0.4337596297264099,
-      "learning_rate": 9.896464983391544e-06,
-      "loss": 0.3596,
-      "step": 3838
-    },
-    {
-      "epoch": 0.25089863407620416,
-      "grad_norm": 0.45142289996147156,
-      "learning_rate": 9.896394278602937e-06,
-      "loss": 0.384,
-      "step": 3839
-    },
-    {
-      "epoch": 0.2509639892817463,
-      "grad_norm": 0.4437491297721863,
-      "learning_rate": 9.896323549932917e-06,
-      "loss": 0.357,
-      "step": 3840
-    },
-    {
-      "epoch": 0.2510293444872884,
-      "grad_norm": 0.46476033329963684,
-      "learning_rate": 9.896252797381832e-06,
-      "loss": 0.3721,
-      "step": 3841
-    },
-    {
-      "epoch": 0.25109469969283055,
-      "grad_norm": 0.5244317054748535,
-      "learning_rate": 9.896182020950026e-06,
-      "loss": 0.4606,
-      "step": 3842
-    },
-    {
-      "epoch": 0.2511600548983727,
-      "grad_norm": 0.46809300780296326,
-      "learning_rate": 9.896111220637843e-06,
-      "loss": 0.3968,
-      "step": 3843
-    },
-    {
-      "epoch": 0.2512254101039148,
-      "grad_norm": 0.44542941451072693,
-      "learning_rate": 9.89604039644563e-06,
-      "loss": 0.3994,
-      "step": 3844
-    },
-    {
-      "epoch": 0.2512907653094569,
-      "grad_norm": 0.4796372354030609,
-      "learning_rate": 9.895969548373731e-06,
-      "loss": 0.406,
-      "step": 3845
-    },
-    {
-      "epoch": 0.251356120514999,
-      "grad_norm": 0.4873756170272827,
-      "learning_rate": 9.895898676422494e-06,
-      "loss": 0.4229,
-      "step": 3846
-    },
-    {
-      "epoch": 0.25142147572054113,
-      "grad_norm": 0.483729749917984,
-      "learning_rate": 9.895827780592262e-06,
-      "loss": 0.4137,
-      "step": 3847
-    },
-    {
-      "epoch": 0.25148683092608326,
-      "grad_norm": 0.46675360202789307,
-      "learning_rate": 9.895756860883383e-06,
-      "loss": 0.3974,
-      "step": 3848
-    },
-    {
-      "epoch": 0.2515521861316254,
-      "grad_norm": 0.4536055028438568,
-      "learning_rate": 9.8956859172962e-06,
-      "loss": 0.4002,
-      "step": 3849
-    },
-    {
-      "epoch": 0.2516175413371675,
-      "grad_norm": 0.506420910358429,
-      "learning_rate": 9.895614949831063e-06,
-      "loss": 0.4112,
-      "step": 3850
-    },
-    {
-      "epoch": 0.25168289654270964,
-      "grad_norm": 0.5092169642448425,
-      "learning_rate": 9.895543958488314e-06,
-      "loss": 0.431,
-      "step": 3851
-    },
-    {
-      "epoch": 0.2517482517482518,
-      "grad_norm": 0.48873788118362427,
-      "learning_rate": 9.895472943268301e-06,
-      "loss": 0.4629,
-      "step": 3852
-    },
-    {
-      "epoch": 0.25181360695379384,
-      "grad_norm": 0.4860228896141052,
-      "learning_rate": 9.89540190417137e-06,
-      "loss": 0.3962,
-      "step": 3853
-    },
-    {
-      "epoch": 0.251878962159336,
-      "grad_norm": 0.5054576992988586,
-      "learning_rate": 9.89533084119787e-06,
-      "loss": 0.3795,
-      "step": 3854
-    },
-    {
-      "epoch": 0.2519443173648781,
-      "grad_norm": 0.49959778785705566,
-      "learning_rate": 9.895259754348145e-06,
-      "loss": 0.4733,
-      "step": 3855
-    },
-    {
-      "epoch": 0.25200967257042023,
-      "grad_norm": 0.5031869411468506,
-      "learning_rate": 9.895188643622542e-06,
-      "loss": 0.4629,
-      "step": 3856
-    },
-    {
-      "epoch": 0.25207502777596236,
-      "grad_norm": 0.47100287675857544,
-      "learning_rate": 9.895117509021408e-06,
-      "loss": 0.3989,
-      "step": 3857
-    },
-    {
-      "epoch": 0.2521403829815045,
-      "grad_norm": 0.5107840299606323,
-      "learning_rate": 9.89504635054509e-06,
-      "loss": 0.468,
-      "step": 3858
-    },
-    {
-      "epoch": 0.2522057381870466,
-      "grad_norm": 0.49201256036758423,
-      "learning_rate": 9.894975168193937e-06,
-      "loss": 0.4938,
-      "step": 3859
-    },
-    {
-      "epoch": 0.25227109339258874,
-      "grad_norm": 0.4625264108181,
-      "learning_rate": 9.894903961968292e-06,
-      "loss": 0.429,
-      "step": 3860
-    },
-    {
-      "epoch": 0.2523364485981308,
-      "grad_norm": 0.5019761323928833,
-      "learning_rate": 9.894832731868504e-06,
-      "loss": 0.4222,
-      "step": 3861
-    },
-    {
-      "epoch": 0.25240180380367294,
-      "grad_norm": 0.5039170980453491,
-      "learning_rate": 9.894761477894924e-06,
-      "loss": 0.4561,
-      "step": 3862
-    },
-    {
-      "epoch": 0.25246715900921507,
-      "grad_norm": 0.4441440999507904,
-      "learning_rate": 9.894690200047894e-06,
-      "loss": 0.368,
-      "step": 3863
-    },
-    {
-      "epoch": 0.2525325142147572,
-      "grad_norm": 0.473471999168396,
-      "learning_rate": 9.894618898327766e-06,
-      "loss": 0.3622,
-      "step": 3864
-    },
-    {
-      "epoch": 0.2525978694202993,
-      "grad_norm": 0.47707849740982056,
-      "learning_rate": 9.894547572734885e-06,
-      "loss": 0.4149,
-      "step": 3865
-    },
-    {
-      "epoch": 0.25266322462584145,
-      "grad_norm": 0.45889490842819214,
-      "learning_rate": 9.894476223269598e-06,
-      "loss": 0.3483,
-      "step": 3866
-    },
-    {
-      "epoch": 0.2527285798313836,
-      "grad_norm": 0.4801705777645111,
-      "learning_rate": 9.894404849932257e-06,
-      "loss": 0.4214,
-      "step": 3867
-    },
-    {
-      "epoch": 0.2527939350369257,
-      "grad_norm": 0.46492329239845276,
-      "learning_rate": 9.894333452723208e-06,
-      "loss": 0.4121,
-      "step": 3868
-    },
-    {
-      "epoch": 0.25285929024246784,
-      "grad_norm": 0.4617963135242462,
-      "learning_rate": 9.8942620316428e-06,
-      "loss": 0.3818,
-      "step": 3869
-    },
-    {
-      "epoch": 0.2529246454480099,
-      "grad_norm": 0.4506014585494995,
-      "learning_rate": 9.89419058669138e-06,
-      "loss": 0.3736,
-      "step": 3870
-    },
-    {
-      "epoch": 0.25299000065355204,
-      "grad_norm": 0.4737391769886017,
-      "learning_rate": 9.894119117869296e-06,
-      "loss": 0.4185,
-      "step": 3871
-    },
-    {
-      "epoch": 0.25305535585909417,
-      "grad_norm": 0.4927527606487274,
-      "learning_rate": 9.894047625176898e-06,
-      "loss": 0.4224,
-      "step": 3872
-    },
-    {
-      "epoch": 0.2531207110646363,
-      "grad_norm": 0.4633755087852478,
-      "learning_rate": 9.893976108614533e-06,
-      "loss": 0.4166,
-      "step": 3873
-    },
-    {
-      "epoch": 0.2531860662701784,
-      "grad_norm": 0.4423373341560364,
-      "learning_rate": 9.893904568182553e-06,
-      "loss": 0.4053,
-      "step": 3874
-    },
-    {
-      "epoch": 0.25325142147572055,
-      "grad_norm": 0.4814908504486084,
-      "learning_rate": 9.893833003881305e-06,
-      "loss": 0.4643,
-      "step": 3875
-    },
-    {
-      "epoch": 0.2533167766812627,
-      "grad_norm": 0.4767465591430664,
-      "learning_rate": 9.893761415711136e-06,
-      "loss": 0.4093,
-      "step": 3876
-    },
-    {
-      "epoch": 0.2533821318868048,
-      "grad_norm": 0.4405258297920227,
-      "learning_rate": 9.893689803672399e-06,
-      "loss": 0.3618,
-      "step": 3877
-    },
-    {
-      "epoch": 0.2534474870923469,
-      "grad_norm": 0.4864327907562256,
-      "learning_rate": 9.893618167765442e-06,
-      "loss": 0.4206,
-      "step": 3878
-    },
-    {
-      "epoch": 0.253512842297889,
-      "grad_norm": 0.46083828806877136,
-      "learning_rate": 9.893546507990612e-06,
-      "loss": 0.3794,
-      "step": 3879
-    },
-    {
-      "epoch": 0.25357819750343114,
-      "grad_norm": 0.46849796175956726,
-      "learning_rate": 9.893474824348261e-06,
-      "loss": 0.3849,
-      "step": 3880
-    },
-    {
-      "epoch": 0.25364355270897326,
-      "grad_norm": 0.5232370495796204,
-      "learning_rate": 9.89340311683874e-06,
-      "loss": 0.4427,
-      "step": 3881
-    },
-    {
-      "epoch": 0.2537089079145154,
-      "grad_norm": 0.46081361174583435,
-      "learning_rate": 9.893331385462394e-06,
-      "loss": 0.4173,
-      "step": 3882
-    },
-    {
-      "epoch": 0.2537742631200575,
-      "grad_norm": 0.4474320113658905,
-      "learning_rate": 9.893259630219579e-06,
-      "loss": 0.3701,
-      "step": 3883
-    },
-    {
-      "epoch": 0.25383961832559965,
-      "grad_norm": 0.454286128282547,
-      "learning_rate": 9.893187851110637e-06,
-      "loss": 0.3909,
-      "step": 3884
-    },
-    {
-      "epoch": 0.2539049735311418,
-      "grad_norm": 0.47931092977523804,
-      "learning_rate": 9.893116048135928e-06,
-      "loss": 0.4102,
-      "step": 3885
-    },
-    {
-      "epoch": 0.2539703287366839,
-      "grad_norm": 0.4402013421058655,
-      "learning_rate": 9.893044221295793e-06,
-      "loss": 0.3714,
-      "step": 3886
-    },
-    {
-      "epoch": 0.254035683942226,
-      "grad_norm": 0.44256025552749634,
-      "learning_rate": 9.892972370590586e-06,
-      "loss": 0.3918,
-      "step": 3887
-    },
-    {
-      "epoch": 0.2541010391477681,
-      "grad_norm": 0.48306509852409363,
-      "learning_rate": 9.892900496020659e-06,
-      "loss": 0.4127,
-      "step": 3888
-    },
-    {
-      "epoch": 0.25416639435331023,
-      "grad_norm": 0.4957873523235321,
-      "learning_rate": 9.892828597586362e-06,
-      "loss": 0.4352,
-      "step": 3889
-    },
-    {
-      "epoch": 0.25423174955885236,
-      "grad_norm": 0.4706031084060669,
-      "learning_rate": 9.892756675288043e-06,
-      "loss": 0.4346,
-      "step": 3890
-    },
-    {
-      "epoch": 0.2542971047643945,
-      "grad_norm": 0.4572066068649292,
-      "learning_rate": 9.892684729126056e-06,
-      "loss": 0.3892,
-      "step": 3891
-    },
-    {
-      "epoch": 0.2543624599699366,
-      "grad_norm": 0.5330418348312378,
-      "learning_rate": 9.89261275910075e-06,
-      "loss": 0.5388,
-      "step": 3892
-    },
-    {
-      "epoch": 0.25442781517547874,
-      "grad_norm": 0.49673694372177124,
-      "learning_rate": 9.892540765212477e-06,
-      "loss": 0.4226,
-      "step": 3893
-    },
-    {
-      "epoch": 0.2544931703810209,
-      "grad_norm": 0.46769097447395325,
-      "learning_rate": 9.892468747461588e-06,
-      "loss": 0.4398,
-      "step": 3894
-    },
-    {
-      "epoch": 0.25455852558656294,
-      "grad_norm": 0.46777594089508057,
-      "learning_rate": 9.892396705848433e-06,
-      "loss": 0.3878,
-      "step": 3895
-    },
-    {
-      "epoch": 0.2546238807921051,
-      "grad_norm": 0.46772271394729614,
-      "learning_rate": 9.892324640373365e-06,
-      "loss": 0.4045,
-      "step": 3896
-    },
-    {
-      "epoch": 0.2546892359976472,
-      "grad_norm": 0.4632786512374878,
-      "learning_rate": 9.892252551036735e-06,
-      "loss": 0.4049,
-      "step": 3897
-    },
-    {
-      "epoch": 0.25475459120318933,
-      "grad_norm": 0.4853006899356842,
-      "learning_rate": 9.892180437838895e-06,
-      "loss": 0.4005,
-      "step": 3898
-    },
-    {
-      "epoch": 0.25481994640873146,
-      "grad_norm": 0.46542122960090637,
-      "learning_rate": 9.892108300780195e-06,
-      "loss": 0.3882,
-      "step": 3899
-    },
-    {
-      "epoch": 0.2548853016142736,
-      "grad_norm": 0.4371604025363922,
-      "learning_rate": 9.892036139860987e-06,
-      "loss": 0.38,
-      "step": 3900
-    },
-    {
-      "epoch": 0.2549506568198157,
-      "grad_norm": 0.44974663853645325,
-      "learning_rate": 9.891963955081627e-06,
-      "loss": 0.372,
-      "step": 3901
-    },
-    {
-      "epoch": 0.25501601202535784,
-      "grad_norm": 0.4604519009590149,
-      "learning_rate": 9.891891746442462e-06,
-      "loss": 0.3589,
-      "step": 3902
-    },
-    {
-      "epoch": 0.2550813672308999,
-      "grad_norm": 0.5047156810760498,
-      "learning_rate": 9.891819513943847e-06,
-      "loss": 0.3929,
-      "step": 3903
-    },
-    {
-      "epoch": 0.25514672243644204,
-      "grad_norm": 0.4475330114364624,
-      "learning_rate": 9.891747257586134e-06,
-      "loss": 0.3784,
-      "step": 3904
-    },
-    {
-      "epoch": 0.25521207764198417,
-      "grad_norm": 0.4892031252384186,
-      "learning_rate": 9.891674977369674e-06,
-      "loss": 0.4405,
-      "step": 3905
-    },
-    {
-      "epoch": 0.2552774328475263,
-      "grad_norm": 0.49480122327804565,
-      "learning_rate": 9.89160267329482e-06,
-      "loss": 0.4084,
-      "step": 3906
-    },
-    {
-      "epoch": 0.2553427880530684,
-      "grad_norm": 0.4493216872215271,
-      "learning_rate": 9.891530345361927e-06,
-      "loss": 0.3819,
-      "step": 3907
-    },
-    {
-      "epoch": 0.25540814325861055,
-      "grad_norm": 0.5069324970245361,
-      "learning_rate": 9.891457993571345e-06,
-      "loss": 0.4356,
-      "step": 3908
-    },
-    {
-      "epoch": 0.2554734984641527,
-      "grad_norm": 0.5098019242286682,
-      "learning_rate": 9.891385617923427e-06,
-      "loss": 0.3603,
-      "step": 3909
-    },
-    {
-      "epoch": 0.2555388536696948,
-      "grad_norm": 0.4521212577819824,
-      "learning_rate": 9.891313218418528e-06,
-      "loss": 0.3858,
-      "step": 3910
-    },
-    {
-      "epoch": 0.25560420887523694,
-      "grad_norm": 0.49558156728744507,
-      "learning_rate": 9.891240795057e-06,
-      "loss": 0.4393,
-      "step": 3911
-    },
-    {
-      "epoch": 0.255669564080779,
-      "grad_norm": 0.4960598349571228,
-      "learning_rate": 9.891168347839194e-06,
-      "loss": 0.4176,
-      "step": 3912
-    },
-    {
-      "epoch": 0.25573491928632114,
-      "grad_norm": 0.4858163297176361,
-      "learning_rate": 9.891095876765468e-06,
-      "loss": 0.4654,
-      "step": 3913
-    },
-    {
-      "epoch": 0.25580027449186327,
-      "grad_norm": 0.45587536692619324,
-      "learning_rate": 9.891023381836171e-06,
-      "loss": 0.3675,
-      "step": 3914
-    },
-    {
-      "epoch": 0.2558656296974054,
-      "grad_norm": 0.437977135181427,
-      "learning_rate": 9.89095086305166e-06,
-      "loss": 0.3425,
-      "step": 3915
-    },
-    {
-      "epoch": 0.2559309849029475,
-      "grad_norm": 0.472843199968338,
-      "learning_rate": 9.890878320412288e-06,
-      "loss": 0.3945,
-      "step": 3916
-    },
-    {
-      "epoch": 0.25599634010848965,
-      "grad_norm": 0.4441700279712677,
-      "learning_rate": 9.890805753918406e-06,
-      "loss": 0.3089,
-      "step": 3917
-    },
-    {
-      "epoch": 0.2560616953140318,
-      "grad_norm": 0.44410449266433716,
-      "learning_rate": 9.890733163570372e-06,
-      "loss": 0.3577,
-      "step": 3918
-    },
-    {
-      "epoch": 0.2561270505195739,
-      "grad_norm": 0.4786568880081177,
-      "learning_rate": 9.890660549368536e-06,
-      "loss": 0.3797,
-      "step": 3919
-    },
-    {
-      "epoch": 0.256192405725116,
-      "grad_norm": 0.47811007499694824,
-      "learning_rate": 9.890587911313255e-06,
-      "loss": 0.3986,
-      "step": 3920
-    },
-    {
-      "epoch": 0.2562577609306581,
-      "grad_norm": 0.47644883394241333,
-      "learning_rate": 9.890515249404883e-06,
-      "loss": 0.429,
-      "step": 3921
-    },
-    {
-      "epoch": 0.25632311613620024,
-      "grad_norm": 0.4829001724720001,
-      "learning_rate": 9.890442563643774e-06,
-      "loss": 0.4139,
-      "step": 3922
-    },
-    {
-      "epoch": 0.25638847134174236,
-      "grad_norm": 0.4659916162490845,
-      "learning_rate": 9.890369854030281e-06,
-      "loss": 0.4007,
-      "step": 3923
-    },
-    {
-      "epoch": 0.2564538265472845,
-      "grad_norm": 0.4580618441104889,
-      "learning_rate": 9.890297120564761e-06,
-      "loss": 0.4054,
-      "step": 3924
-    },
-    {
-      "epoch": 0.2565191817528266,
-      "grad_norm": 0.490489661693573,
-      "learning_rate": 9.890224363247568e-06,
-      "loss": 0.3949,
-      "step": 3925
-    },
-    {
-      "epoch": 0.25658453695836875,
-      "grad_norm": 0.4814055263996124,
-      "learning_rate": 9.890151582079058e-06,
-      "loss": 0.436,
-      "step": 3926
-    },
-    {
-      "epoch": 0.2566498921639109,
-      "grad_norm": 0.5186078548431396,
-      "learning_rate": 9.890078777059581e-06,
-      "loss": 0.4993,
-      "step": 3927
-    },
-    {
-      "epoch": 0.256715247369453,
-      "grad_norm": 0.47267815470695496,
-      "learning_rate": 9.890005948189498e-06,
-      "loss": 0.3656,
-      "step": 3928
-    },
-    {
-      "epoch": 0.2567806025749951,
-      "grad_norm": 0.4519590735435486,
-      "learning_rate": 9.889933095469162e-06,
-      "loss": 0.3715,
-      "step": 3929
-    },
-    {
-      "epoch": 0.2568459577805372,
-      "grad_norm": 0.4927625060081482,
-      "learning_rate": 9.889860218898928e-06,
-      "loss": 0.4043,
-      "step": 3930
-    },
-    {
-      "epoch": 0.25691131298607933,
-      "grad_norm": 0.49229928851127625,
-      "learning_rate": 9.889787318479151e-06,
-      "loss": 0.3785,
-      "step": 3931
-    },
-    {
-      "epoch": 0.25697666819162146,
-      "grad_norm": 0.4878690242767334,
-      "learning_rate": 9.889714394210189e-06,
-      "loss": 0.4205,
-      "step": 3932
-    },
-    {
-      "epoch": 0.2570420233971636,
-      "grad_norm": 0.4788166582584381,
-      "learning_rate": 9.889641446092395e-06,
-      "loss": 0.4137,
-      "step": 3933
-    },
-    {
-      "epoch": 0.2571073786027057,
-      "grad_norm": 0.4857322871685028,
-      "learning_rate": 9.889568474126125e-06,
-      "loss": 0.444,
-      "step": 3934
-    },
-    {
-      "epoch": 0.25717273380824784,
-      "grad_norm": 0.500454306602478,
-      "learning_rate": 9.889495478311737e-06,
-      "loss": 0.4084,
-      "step": 3935
-    },
-    {
-      "epoch": 0.25723808901379,
-      "grad_norm": 0.43148401379585266,
-      "learning_rate": 9.889422458649585e-06,
-      "loss": 0.3234,
-      "step": 3936
-    },
-    {
-      "epoch": 0.25730344421933204,
-      "grad_norm": 0.5101863145828247,
-      "learning_rate": 9.889349415140025e-06,
-      "loss": 0.4184,
-      "step": 3937
-    },
-    {
-      "epoch": 0.2573687994248742,
-      "grad_norm": 0.47251182794570923,
-      "learning_rate": 9.889276347783413e-06,
-      "loss": 0.3775,
-      "step": 3938
-    },
-    {
-      "epoch": 0.2574341546304163,
-      "grad_norm": 0.518918514251709,
-      "learning_rate": 9.88920325658011e-06,
-      "loss": 0.4736,
-      "step": 3939
-    },
-    {
-      "epoch": 0.25749950983595843,
-      "grad_norm": 0.4914916455745697,
-      "learning_rate": 9.889130141530468e-06,
-      "loss": 0.409,
-      "step": 3940
-    },
-    {
-      "epoch": 0.25756486504150056,
-      "grad_norm": 0.48664769530296326,
-      "learning_rate": 9.889057002634844e-06,
-      "loss": 0.4299,
-      "step": 3941
-    },
-    {
-      "epoch": 0.2576302202470427,
-      "grad_norm": 0.5491824150085449,
-      "learning_rate": 9.888983839893593e-06,
-      "loss": 0.4988,
-      "step": 3942
-    },
-    {
-      "epoch": 0.2576955754525848,
-      "grad_norm": 0.48742780089378357,
-      "learning_rate": 9.888910653307078e-06,
-      "loss": 0.4271,
-      "step": 3943
-    },
-    {
-      "epoch": 0.25776093065812694,
-      "grad_norm": 0.47784173488616943,
-      "learning_rate": 9.88883744287565e-06,
-      "loss": 0.4035,
-      "step": 3944
-    },
-    {
-      "epoch": 0.257826285863669,
-      "grad_norm": 0.4807237684726715,
-      "learning_rate": 9.888764208599669e-06,
-      "loss": 0.3989,
-      "step": 3945
-    },
-    {
-      "epoch": 0.25789164106921114,
-      "grad_norm": 0.4672047197818756,
-      "learning_rate": 9.888690950479489e-06,
-      "loss": 0.3765,
-      "step": 3946
-    },
-    {
-      "epoch": 0.25795699627475327,
-      "grad_norm": 0.5563791394233704,
-      "learning_rate": 9.888617668515472e-06,
-      "loss": 0.4377,
-      "step": 3947
-    },
-    {
-      "epoch": 0.2580223514802954,
-      "grad_norm": 0.4825108051300049,
-      "learning_rate": 9.88854436270797e-06,
-      "loss": 0.445,
-      "step": 3948
-    },
-    {
-      "epoch": 0.2580877066858375,
-      "grad_norm": 0.438420832157135,
-      "learning_rate": 9.888471033057347e-06,
-      "loss": 0.3558,
-      "step": 3949
-    },
-    {
-      "epoch": 0.25815306189137965,
-      "grad_norm": 0.4145852327346802,
-      "learning_rate": 9.888397679563958e-06,
-      "loss": 0.3175,
-      "step": 3950
-    },
-    {
-      "epoch": 0.2582184170969218,
-      "grad_norm": 0.4814380407333374,
-      "learning_rate": 9.888324302228156e-06,
-      "loss": 0.3622,
-      "step": 3951
-    },
-    {
-      "epoch": 0.2582837723024639,
-      "grad_norm": 0.5345417261123657,
-      "learning_rate": 9.888250901050306e-06,
-      "loss": 0.4457,
-      "step": 3952
-    },
-    {
-      "epoch": 0.25834912750800604,
-      "grad_norm": 0.5023251175880432,
-      "learning_rate": 9.888177476030761e-06,
-      "loss": 0.4571,
-      "step": 3953
-    },
-    {
-      "epoch": 0.2584144827135481,
-      "grad_norm": 0.464595764875412,
-      "learning_rate": 9.888104027169883e-06,
-      "loss": 0.3636,
-      "step": 3954
-    },
-    {
-      "epoch": 0.25847983791909024,
-      "grad_norm": 0.48747649788856506,
-      "learning_rate": 9.888030554468026e-06,
-      "loss": 0.4364,
-      "step": 3955
-    },
-    {
-      "epoch": 0.25854519312463237,
-      "grad_norm": 0.5128535032272339,
-      "learning_rate": 9.887957057925552e-06,
-      "loss": 0.4106,
-      "step": 3956
-    },
-    {
-      "epoch": 0.2586105483301745,
-      "grad_norm": 0.49727413058280945,
-      "learning_rate": 9.887883537542818e-06,
-      "loss": 0.4135,
-      "step": 3957
-    },
-    {
-      "epoch": 0.2586759035357166,
-      "grad_norm": 0.46694761514663696,
-      "learning_rate": 9.887809993320182e-06,
-      "loss": 0.4462,
-      "step": 3958
-    },
-    {
-      "epoch": 0.25874125874125875,
-      "grad_norm": 0.4622994065284729,
-      "learning_rate": 9.887736425258006e-06,
-      "loss": 0.379,
-      "step": 3959
-    },
-    {
-      "epoch": 0.2588066139468009,
-      "grad_norm": 0.5834900736808777,
-      "learning_rate": 9.887662833356644e-06,
-      "loss": 0.4102,
-      "step": 3960
-    },
-    {
-      "epoch": 0.258871969152343,
-      "grad_norm": 0.4487383961677551,
-      "learning_rate": 9.887589217616455e-06,
-      "loss": 0.391,
-      "step": 3961
-    },
-    {
-      "epoch": 0.2589373243578851,
-      "grad_norm": 0.4785480499267578,
-      "learning_rate": 9.887515578037803e-06,
-      "loss": 0.3996,
-      "step": 3962
-    },
-    {
-      "epoch": 0.2590026795634272,
-      "grad_norm": 0.4960395097732544,
-      "learning_rate": 9.887441914621043e-06,
-      "loss": 0.3742,
-      "step": 3963
-    },
-    {
-      "epoch": 0.25906803476896934,
-      "grad_norm": 0.5078006386756897,
-      "learning_rate": 9.887368227366539e-06,
-      "loss": 0.4238,
-      "step": 3964
-    },
-    {
-      "epoch": 0.25913338997451146,
-      "grad_norm": 0.5127818584442139,
-      "learning_rate": 9.887294516274643e-06,
-      "loss": 0.4571,
-      "step": 3965
-    },
-    {
-      "epoch": 0.2591987451800536,
-      "grad_norm": 0.4530894458293915,
-      "learning_rate": 9.88722078134572e-06,
-      "loss": 0.3529,
-      "step": 3966
-    },
-    {
-      "epoch": 0.2592641003855957,
-      "grad_norm": 0.46037912368774414,
-      "learning_rate": 9.887147022580127e-06,
-      "loss": 0.4012,
-      "step": 3967
-    },
-    {
-      "epoch": 0.25932945559113785,
-      "grad_norm": 0.5028442144393921,
-      "learning_rate": 9.887073239978227e-06,
-      "loss": 0.4172,
-      "step": 3968
-    },
-    {
-      "epoch": 0.25939481079668,
-      "grad_norm": 0.5270559191703796,
-      "learning_rate": 9.886999433540376e-06,
-      "loss": 0.4431,
-      "step": 3969
-    },
-    {
-      "epoch": 0.2594601660022221,
-      "grad_norm": 0.522784411907196,
-      "learning_rate": 9.886925603266936e-06,
-      "loss": 0.401,
-      "step": 3970
-    },
-    {
-      "epoch": 0.2595255212077642,
-      "grad_norm": 0.46748587489128113,
-      "learning_rate": 9.886851749158268e-06,
-      "loss": 0.3861,
-      "step": 3971
-    },
-    {
-      "epoch": 0.2595908764133063,
-      "grad_norm": 0.5080457329750061,
-      "learning_rate": 9.88677787121473e-06,
-      "loss": 0.4376,
-      "step": 3972
-    },
-    {
-      "epoch": 0.25965623161884843,
-      "grad_norm": 0.5248509049415588,
-      "learning_rate": 9.886703969436684e-06,
-      "loss": 0.4379,
-      "step": 3973
-    },
-    {
-      "epoch": 0.25972158682439056,
-      "grad_norm": 0.4547731876373291,
-      "learning_rate": 9.88663004382449e-06,
-      "loss": 0.3392,
-      "step": 3974
-    },
-    {
-      "epoch": 0.2597869420299327,
-      "grad_norm": 0.4710044264793396,
-      "learning_rate": 9.886556094378507e-06,
-      "loss": 0.4111,
-      "step": 3975
-    },
-    {
-      "epoch": 0.2598522972354748,
-      "grad_norm": 0.4873550236225128,
-      "learning_rate": 9.8864821210991e-06,
-      "loss": 0.4262,
-      "step": 3976
-    },
-    {
-      "epoch": 0.25991765244101694,
-      "grad_norm": 0.47116461396217346,
-      "learning_rate": 9.886408123986624e-06,
-      "loss": 0.3706,
-      "step": 3977
-    },
-    {
-      "epoch": 0.2599830076465591,
-      "grad_norm": 0.506894052028656,
-      "learning_rate": 9.886334103041443e-06,
-      "loss": 0.4173,
-      "step": 3978
-    },
-    {
-      "epoch": 0.26004836285210114,
-      "grad_norm": 0.5134207010269165,
-      "learning_rate": 9.886260058263918e-06,
-      "loss": 0.4242,
-      "step": 3979
-    },
-    {
-      "epoch": 0.2601137180576433,
-      "grad_norm": 0.4570533037185669,
-      "learning_rate": 9.886185989654411e-06,
-      "loss": 0.401,
-      "step": 3980
-    },
-    {
-      "epoch": 0.2601790732631854,
-      "grad_norm": 0.46695584058761597,
-      "learning_rate": 9.886111897213282e-06,
-      "loss": 0.4072,
-      "step": 3981
-    },
-    {
-      "epoch": 0.26024442846872753,
-      "grad_norm": 0.5276580452919006,
-      "learning_rate": 9.886037780940892e-06,
-      "loss": 0.4822,
-      "step": 3982
-    },
-    {
-      "epoch": 0.26030978367426966,
-      "grad_norm": 0.47777944803237915,
-      "learning_rate": 9.885963640837601e-06,
-      "loss": 0.3822,
-      "step": 3983
-    },
-    {
-      "epoch": 0.2603751388798118,
-      "grad_norm": 0.5472515225410461,
-      "learning_rate": 9.885889476903776e-06,
-      "loss": 0.4825,
-      "step": 3984
-    },
-    {
-      "epoch": 0.2604404940853539,
-      "grad_norm": 0.47866106033325195,
-      "learning_rate": 9.885815289139774e-06,
-      "loss": 0.4205,
-      "step": 3985
-    },
-    {
-      "epoch": 0.26050584929089604,
-      "grad_norm": 0.6274198293685913,
-      "learning_rate": 9.885741077545958e-06,
-      "loss": 0.382,
-      "step": 3986
-    },
-    {
-      "epoch": 0.2605712044964381,
-      "grad_norm": 0.5079156756401062,
-      "learning_rate": 9.885666842122692e-06,
-      "loss": 0.4011,
-      "step": 3987
-    },
-    {
-      "epoch": 0.26063655970198024,
-      "grad_norm": 0.6067430973052979,
-      "learning_rate": 9.885592582870334e-06,
-      "loss": 0.4514,
-      "step": 3988
-    },
-    {
-      "epoch": 0.26070191490752237,
-      "grad_norm": 0.4657166302204132,
-      "learning_rate": 9.88551829978925e-06,
-      "loss": 0.4008,
-      "step": 3989
-    },
-    {
-      "epoch": 0.2607672701130645,
-      "grad_norm": 0.46768006682395935,
-      "learning_rate": 9.8854439928798e-06,
-      "loss": 0.3943,
-      "step": 3990
-    },
-    {
-      "epoch": 0.2608326253186066,
-      "grad_norm": 0.5387409329414368,
-      "learning_rate": 9.885369662142345e-06,
-      "loss": 0.4716,
-      "step": 3991
-    },
-    {
-      "epoch": 0.26089798052414875,
-      "grad_norm": 0.5316016674041748,
-      "learning_rate": 9.885295307577253e-06,
-      "loss": 0.4515,
-      "step": 3992
-    },
-    {
-      "epoch": 0.2609633357296909,
-      "grad_norm": 0.5189493894577026,
-      "learning_rate": 9.885220929184882e-06,
-      "loss": 0.4268,
-      "step": 3993
-    },
-    {
-      "epoch": 0.261028690935233,
-      "grad_norm": 0.4890076220035553,
-      "learning_rate": 9.885146526965597e-06,
-      "loss": 0.4038,
-      "step": 3994
-    },
-    {
-      "epoch": 0.26109404614077514,
-      "grad_norm": 0.4809994697570801,
-      "learning_rate": 9.885072100919759e-06,
-      "loss": 0.4277,
-      "step": 3995
-    },
-    {
-      "epoch": 0.2611594013463172,
-      "grad_norm": 0.5016999244689941,
-      "learning_rate": 9.884997651047732e-06,
-      "loss": 0.4215,
-      "step": 3996
-    },
-    {
-      "epoch": 0.26122475655185934,
-      "grad_norm": 0.4756239652633667,
-      "learning_rate": 9.88492317734988e-06,
-      "loss": 0.3985,
-      "step": 3997
-    },
-    {
-      "epoch": 0.26129011175740147,
-      "grad_norm": 0.52363520860672,
-      "learning_rate": 9.884848679826563e-06,
-      "loss": 0.4323,
-      "step": 3998
-    },
-    {
-      "epoch": 0.2613554669629436,
-      "grad_norm": 0.46867337822914124,
-      "learning_rate": 9.88477415847815e-06,
-      "loss": 0.3827,
-      "step": 3999
-    },
-    {
-      "epoch": 0.2614208221684857,
-      "grad_norm": 0.5153798460960388,
-      "learning_rate": 9.884699613305e-06,
-      "loss": 0.4539,
-      "step": 4000
-    },
-    {
-      "epoch": 0.26148617737402785,
-      "grad_norm": 0.486310750246048,
-      "learning_rate": 9.884625044307477e-06,
-      "loss": 0.3942,
-      "step": 4001
-    },
-    {
-      "epoch": 0.26155153257957,
-      "grad_norm": 0.5417717695236206,
-      "learning_rate": 9.884550451485945e-06,
-      "loss": 0.4121,
-      "step": 4002
-    },
-    {
-      "epoch": 0.2616168877851121,
-      "grad_norm": 0.46015897393226624,
-      "learning_rate": 9.88447583484077e-06,
-      "loss": 0.3829,
-      "step": 4003
-    },
-    {
-      "epoch": 0.2616822429906542,
-      "grad_norm": 0.5171236991882324,
-      "learning_rate": 9.884401194372316e-06,
-      "loss": 0.488,
-      "step": 4004
-    },
-    {
-      "epoch": 0.2617475981961963,
-      "grad_norm": 0.5257484912872314,
-      "learning_rate": 9.88432653008094e-06,
-      "loss": 0.458,
-      "step": 4005
-    },
-    {
-      "epoch": 0.26181295340173844,
-      "grad_norm": 0.5574596524238586,
-      "learning_rate": 9.884251841967015e-06,
-      "loss": 0.4217,
-      "step": 4006
-    },
-    {
-      "epoch": 0.26187830860728056,
-      "grad_norm": 0.4806183874607086,
-      "learning_rate": 9.8841771300309e-06,
-      "loss": 0.3959,
-      "step": 4007
-    },
-    {
-      "epoch": 0.2619436638128227,
-      "grad_norm": 0.49253955483436584,
-      "learning_rate": 9.884102394272962e-06,
-      "loss": 0.4402,
-      "step": 4008
-    },
-    {
-      "epoch": 0.2620090190183648,
-      "grad_norm": 0.47940564155578613,
-      "learning_rate": 9.884027634693566e-06,
-      "loss": 0.4011,
-      "step": 4009
-    },
-    {
-      "epoch": 0.26207437422390695,
-      "grad_norm": 0.47881099581718445,
-      "learning_rate": 9.883952851293073e-06,
-      "loss": 0.3793,
-      "step": 4010
-    },
-    {
-      "epoch": 0.2621397294294491,
-      "grad_norm": 0.48134446144104004,
-      "learning_rate": 9.883878044071851e-06,
-      "loss": 0.377,
-      "step": 4011
-    },
-    {
-      "epoch": 0.2622050846349912,
-      "grad_norm": 0.4901047348976135,
-      "learning_rate": 9.883803213030263e-06,
-      "loss": 0.4556,
-      "step": 4012
-    },
-    {
-      "epoch": 0.2622704398405333,
-      "grad_norm": 0.4936966001987457,
-      "learning_rate": 9.883728358168676e-06,
-      "loss": 0.4527,
-      "step": 4013
-    },
-    {
-      "epoch": 0.2623357950460754,
-      "grad_norm": 0.44317910075187683,
-      "learning_rate": 9.883653479487453e-06,
-      "loss": 0.3333,
-      "step": 4014
-    },
-    {
-      "epoch": 0.26240115025161753,
-      "grad_norm": 0.4643218219280243,
-      "learning_rate": 9.883578576986961e-06,
-      "loss": 0.4127,
-      "step": 4015
-    },
-    {
-      "epoch": 0.26246650545715966,
-      "grad_norm": 0.4814830422401428,
-      "learning_rate": 9.883503650667563e-06,
-      "loss": 0.4337,
-      "step": 4016
-    },
-    {
-      "epoch": 0.2625318606627018,
-      "grad_norm": 0.48062700033187866,
-      "learning_rate": 9.883428700529626e-06,
-      "loss": 0.3871,
-      "step": 4017
-    },
-    {
-      "epoch": 0.2625972158682439,
-      "grad_norm": 0.5002309679985046,
-      "learning_rate": 9.883353726573518e-06,
-      "loss": 0.4498,
-      "step": 4018
-    },
-    {
-      "epoch": 0.26266257107378604,
-      "grad_norm": 0.4648917615413666,
-      "learning_rate": 9.8832787287996e-06,
-      "loss": 0.379,
-      "step": 4019
-    },
-    {
-      "epoch": 0.26272792627932817,
-      "grad_norm": 0.5252535343170166,
-      "learning_rate": 9.88320370720824e-06,
-      "loss": 0.4755,
-      "step": 4020
-    },
-    {
-      "epoch": 0.26279328148487024,
-      "grad_norm": 0.4704096019268036,
-      "learning_rate": 9.883128661799805e-06,
-      "loss": 0.3981,
-      "step": 4021
-    },
-    {
-      "epoch": 0.2628586366904124,
-      "grad_norm": 0.47374221682548523,
-      "learning_rate": 9.883053592574658e-06,
-      "loss": 0.4171,
-      "step": 4022
-    },
-    {
-      "epoch": 0.2629239918959545,
-      "grad_norm": 0.4634070098400116,
-      "learning_rate": 9.88297849953317e-06,
-      "loss": 0.3805,
-      "step": 4023
-    },
-    {
-      "epoch": 0.26298934710149663,
-      "grad_norm": 0.5194708704948425,
-      "learning_rate": 9.8829033826757e-06,
-      "loss": 0.4813,
-      "step": 4024
-    },
-    {
-      "epoch": 0.26305470230703876,
-      "grad_norm": 0.4325157403945923,
-      "learning_rate": 9.882828242002622e-06,
-      "loss": 0.3485,
-      "step": 4025
-    },
-    {
-      "epoch": 0.2631200575125809,
-      "grad_norm": 0.45242777466773987,
-      "learning_rate": 9.882753077514298e-06,
-      "loss": 0.4173,
-      "step": 4026
-    },
-    {
-      "epoch": 0.263185412718123,
-      "grad_norm": 0.4679414629936218,
-      "learning_rate": 9.882677889211095e-06,
-      "loss": 0.4511,
-      "step": 4027
-    },
-    {
-      "epoch": 0.26325076792366514,
-      "grad_norm": 0.5232787132263184,
-      "learning_rate": 9.88260267709338e-06,
-      "loss": 0.4671,
-      "step": 4028
-    },
-    {
-      "epoch": 0.26331612312920727,
-      "grad_norm": 0.4798663854598999,
-      "learning_rate": 9.882527441161523e-06,
-      "loss": 0.4206,
-      "step": 4029
-    },
-    {
-      "epoch": 0.26338147833474934,
-      "grad_norm": 0.4608439803123474,
-      "learning_rate": 9.882452181415885e-06,
-      "loss": 0.4074,
-      "step": 4030
-    },
-    {
-      "epoch": 0.26344683354029147,
-      "grad_norm": 0.46279963850975037,
-      "learning_rate": 9.882376897856838e-06,
-      "loss": 0.3958,
-      "step": 4031
-    },
-    {
-      "epoch": 0.2635121887458336,
-      "grad_norm": 0.4729881286621094,
-      "learning_rate": 9.882301590484746e-06,
-      "loss": 0.4197,
-      "step": 4032
-    },
-    {
-      "epoch": 0.2635775439513757,
-      "grad_norm": 0.4474340081214905,
-      "learning_rate": 9.88222625929998e-06,
-      "loss": 0.3865,
-      "step": 4033
-    },
-    {
-      "epoch": 0.26364289915691785,
-      "grad_norm": 0.4575665593147278,
-      "learning_rate": 9.882150904302905e-06,
-      "loss": 0.3816,
-      "step": 4034
-    },
-    {
-      "epoch": 0.26370825436246,
-      "grad_norm": 0.4573649764060974,
-      "learning_rate": 9.882075525493885e-06,
-      "loss": 0.3803,
-      "step": 4035
-    },
-    {
-      "epoch": 0.2637736095680021,
-      "grad_norm": 0.5491724014282227,
-      "learning_rate": 9.882000122873296e-06,
-      "loss": 0.461,
-      "step": 4036
-    },
-    {
-      "epoch": 0.26383896477354424,
-      "grad_norm": 0.5086228251457214,
-      "learning_rate": 9.881924696441499e-06,
-      "loss": 0.3976,
-      "step": 4037
-    },
-    {
-      "epoch": 0.2639043199790863,
-      "grad_norm": 0.49077677726745605,
-      "learning_rate": 9.881849246198864e-06,
-      "loss": 0.4091,
-      "step": 4038
-    },
-    {
-      "epoch": 0.26396967518462844,
-      "grad_norm": 0.4744749367237091,
-      "learning_rate": 9.88177377214576e-06,
-      "loss": 0.3891,
-      "step": 4039
-    },
-    {
-      "epoch": 0.26403503039017057,
-      "grad_norm": 0.49219316244125366,
-      "learning_rate": 9.881698274282552e-06,
-      "loss": 0.4198,
-      "step": 4040
-    },
-    {
-      "epoch": 0.2641003855957127,
-      "grad_norm": 0.4904021918773651,
-      "learning_rate": 9.881622752609611e-06,
-      "loss": 0.4265,
-      "step": 4041
-    },
-    {
-      "epoch": 0.2641657408012548,
-      "grad_norm": 0.4952344596385956,
-      "learning_rate": 9.881547207127307e-06,
-      "loss": 0.4069,
-      "step": 4042
-    },
-    {
-      "epoch": 0.26423109600679695,
-      "grad_norm": 0.4888193905353546,
-      "learning_rate": 9.881471637836005e-06,
-      "loss": 0.3746,
-      "step": 4043
-    },
-    {
-      "epoch": 0.2642964512123391,
-      "grad_norm": 0.48952335119247437,
-      "learning_rate": 9.881396044736073e-06,
-      "loss": 0.4348,
-      "step": 4044
-    },
-    {
-      "epoch": 0.2643618064178812,
-      "grad_norm": 1.197820782661438,
-      "learning_rate": 9.881320427827883e-06,
-      "loss": 0.4015,
-      "step": 4045
-    },
-    {
-      "epoch": 0.2644271616234233,
-      "grad_norm": 0.4951687455177307,
-      "learning_rate": 9.881244787111802e-06,
-      "loss": 0.4245,
-      "step": 4046
-    },
-    {
-      "epoch": 0.2644925168289654,
-      "grad_norm": 0.4858238101005554,
-      "learning_rate": 9.881169122588198e-06,
-      "loss": 0.4021,
-      "step": 4047
-    },
-    {
-      "epoch": 0.26455787203450754,
-      "grad_norm": 0.5516292452812195,
-      "learning_rate": 9.881093434257443e-06,
-      "loss": 0.4342,
-      "step": 4048
-    },
-    {
-      "epoch": 0.26462322724004966,
-      "grad_norm": 0.4711117744445801,
-      "learning_rate": 9.881017722119903e-06,
-      "loss": 0.3943,
-      "step": 4049
-    },
-    {
-      "epoch": 0.2646885824455918,
-      "grad_norm": 0.5131600499153137,
-      "learning_rate": 9.880941986175948e-06,
-      "loss": 0.4188,
-      "step": 4050
-    },
-    {
-      "epoch": 0.2647539376511339,
-      "grad_norm": 0.46999862790107727,
-      "learning_rate": 9.88086622642595e-06,
-      "loss": 0.4066,
-      "step": 4051
-    },
-    {
-      "epoch": 0.26481929285667605,
-      "grad_norm": 0.5009545683860779,
-      "learning_rate": 9.880790442870277e-06,
-      "loss": 0.4341,
-      "step": 4052
-    },
-    {
-      "epoch": 0.2648846480622182,
-      "grad_norm": 0.5397742986679077,
-      "learning_rate": 9.880714635509295e-06,
-      "loss": 0.4512,
-      "step": 4053
-    },
-    {
-      "epoch": 0.2649500032677603,
-      "grad_norm": 0.45424947142601013,
-      "learning_rate": 9.880638804343378e-06,
-      "loss": 0.3894,
-      "step": 4054
-    },
-    {
-      "epoch": 0.2650153584733024,
-      "grad_norm": 0.48217862844467163,
-      "learning_rate": 9.880562949372895e-06,
-      "loss": 0.4123,
-      "step": 4055
-    },
-    {
-      "epoch": 0.2650807136788445,
-      "grad_norm": 0.49495089054107666,
-      "learning_rate": 9.880487070598217e-06,
-      "loss": 0.4691,
-      "step": 4056
-    },
-    {
-      "epoch": 0.26514606888438663,
-      "grad_norm": 0.4753105044364929,
-      "learning_rate": 9.880411168019713e-06,
-      "loss": 0.4241,
-      "step": 4057
-    },
-    {
-      "epoch": 0.26521142408992876,
-      "grad_norm": 0.47640448808670044,
-      "learning_rate": 9.880335241637751e-06,
-      "loss": 0.4326,
-      "step": 4058
-    },
-    {
-      "epoch": 0.2652767792954709,
-      "grad_norm": 0.4479662775993347,
-      "learning_rate": 9.880259291452704e-06,
-      "loss": 0.3333,
-      "step": 4059
-    },
-    {
-      "epoch": 0.265342134501013,
-      "grad_norm": 0.4611780345439911,
-      "learning_rate": 9.880183317464943e-06,
-      "loss": 0.4045,
-      "step": 4060
-    },
-    {
-      "epoch": 0.26540748970655514,
-      "grad_norm": 0.44803959131240845,
-      "learning_rate": 9.880107319674835e-06,
-      "loss": 0.3835,
-      "step": 4061
-    },
-    {
-      "epoch": 0.26547284491209727,
-      "grad_norm": 0.5119698643684387,
-      "learning_rate": 9.880031298082754e-06,
-      "loss": 0.4628,
-      "step": 4062
-    },
-    {
-      "epoch": 0.26553820011763934,
-      "grad_norm": 0.47646498680114746,
-      "learning_rate": 9.87995525268907e-06,
-      "loss": 0.3763,
-      "step": 4063
-    },
-    {
-      "epoch": 0.2656035553231815,
-      "grad_norm": 0.49275916814804077,
-      "learning_rate": 9.879879183494154e-06,
-      "loss": 0.4224,
-      "step": 4064
-    },
-    {
-      "epoch": 0.2656689105287236,
-      "grad_norm": 0.5169394016265869,
-      "learning_rate": 9.879803090498377e-06,
-      "loss": 0.4363,
-      "step": 4065
-    },
-    {
-      "epoch": 0.26573426573426573,
-      "grad_norm": 0.46649786829948425,
-      "learning_rate": 9.879726973702109e-06,
-      "loss": 0.4457,
-      "step": 4066
-    },
-    {
-      "epoch": 0.26579962093980786,
-      "grad_norm": 1.0375394821166992,
-      "learning_rate": 9.879650833105721e-06,
-      "loss": 0.4381,
-      "step": 4067
-    },
-    {
-      "epoch": 0.26586497614535,
-      "grad_norm": 0.5473775863647461,
-      "learning_rate": 9.879574668709588e-06,
-      "loss": 0.4868,
-      "step": 4068
-    },
-    {
-      "epoch": 0.2659303313508921,
-      "grad_norm": 0.4799162745475769,
-      "learning_rate": 9.879498480514077e-06,
-      "loss": 0.3871,
-      "step": 4069
-    },
-    {
-      "epoch": 0.26599568655643424,
-      "grad_norm": 0.4750443398952484,
-      "learning_rate": 9.879422268519562e-06,
-      "loss": 0.4327,
-      "step": 4070
-    },
-    {
-      "epoch": 0.26606104176197637,
-      "grad_norm": 0.46318313479423523,
-      "learning_rate": 9.879346032726413e-06,
-      "loss": 0.405,
-      "step": 4071
-    },
-    {
-      "epoch": 0.26612639696751844,
-      "grad_norm": 0.4690793752670288,
-      "learning_rate": 9.879269773135005e-06,
-      "loss": 0.3928,
-      "step": 4072
-    },
-    {
-      "epoch": 0.26619175217306057,
-      "grad_norm": 0.5178049206733704,
-      "learning_rate": 9.879193489745706e-06,
-      "loss": 0.4278,
-      "step": 4073
-    },
-    {
-      "epoch": 0.2662571073786027,
-      "grad_norm": 0.46420466899871826,
-      "learning_rate": 9.879117182558893e-06,
-      "loss": 0.3865,
-      "step": 4074
-    },
-    {
-      "epoch": 0.2663224625841448,
-      "grad_norm": 0.5093705058097839,
-      "learning_rate": 9.879040851574932e-06,
-      "loss": 0.391,
-      "step": 4075
-    },
-    {
-      "epoch": 0.26638781778968695,
-      "grad_norm": 0.44918498396873474,
-      "learning_rate": 9.878964496794202e-06,
-      "loss": 0.3983,
-      "step": 4076
-    },
-    {
-      "epoch": 0.2664531729952291,
-      "grad_norm": 0.5079885721206665,
-      "learning_rate": 9.87888811821707e-06,
-      "loss": 0.4132,
-      "step": 4077
-    },
-    {
-      "epoch": 0.2665185282007712,
-      "grad_norm": 0.4735715091228485,
-      "learning_rate": 9.878811715843908e-06,
-      "loss": 0.4042,
-      "step": 4078
-    },
-    {
-      "epoch": 0.26658388340631334,
-      "grad_norm": 0.46049559116363525,
-      "learning_rate": 9.878735289675095e-06,
-      "loss": 0.4054,
-      "step": 4079
-    },
-    {
-      "epoch": 0.2666492386118554,
-      "grad_norm": 0.46098044514656067,
-      "learning_rate": 9.878658839710997e-06,
-      "loss": 0.3684,
-      "step": 4080
-    },
-    {
-      "epoch": 0.26671459381739754,
-      "grad_norm": 0.4658360481262207,
-      "learning_rate": 9.87858236595199e-06,
-      "loss": 0.3914,
-      "step": 4081
-    },
-    {
-      "epoch": 0.26677994902293967,
-      "grad_norm": 0.48726463317871094,
-      "learning_rate": 9.87850586839845e-06,
-      "loss": 0.4097,
-      "step": 4082
-    },
-    {
-      "epoch": 0.2668453042284818,
-      "grad_norm": 0.4387027621269226,
-      "learning_rate": 9.878429347050743e-06,
-      "loss": 0.3648,
-      "step": 4083
-    },
-    {
-      "epoch": 0.2669106594340239,
-      "grad_norm": 0.4726237952709198,
-      "learning_rate": 9.878352801909248e-06,
-      "loss": 0.401,
-      "step": 4084
-    },
-    {
-      "epoch": 0.26697601463956605,
-      "grad_norm": 0.47512876987457275,
-      "learning_rate": 9.878276232974336e-06,
-      "loss": 0.3675,
-      "step": 4085
-    },
-    {
-      "epoch": 0.2670413698451082,
-      "grad_norm": 0.49428948760032654,
-      "learning_rate": 9.878199640246379e-06,
-      "loss": 0.4089,
-      "step": 4086
-    },
-    {
-      "epoch": 0.2671067250506503,
-      "grad_norm": 0.48950693011283875,
-      "learning_rate": 9.878123023725754e-06,
-      "loss": 0.423,
-      "step": 4087
-    },
-    {
-      "epoch": 0.2671720802561924,
-      "grad_norm": 0.5121861100196838,
-      "learning_rate": 9.878046383412833e-06,
-      "loss": 0.4184,
-      "step": 4088
-    },
-    {
-      "epoch": 0.2672374354617345,
-      "grad_norm": 0.4943595230579376,
-      "learning_rate": 9.87796971930799e-06,
-      "loss": 0.4107,
-      "step": 4089
-    },
-    {
-      "epoch": 0.26730279066727664,
-      "grad_norm": 0.506575882434845,
-      "learning_rate": 9.8778930314116e-06,
-      "loss": 0.3956,
-      "step": 4090
-    },
-    {
-      "epoch": 0.26736814587281876,
-      "grad_norm": 0.4697633981704712,
-      "learning_rate": 9.877816319724034e-06,
-      "loss": 0.3959,
-      "step": 4091
-    },
-    {
-      "epoch": 0.2674335010783609,
-      "grad_norm": 0.47665300965309143,
-      "learning_rate": 9.87773958424567e-06,
-      "loss": 0.3928,
-      "step": 4092
-    },
-    {
-      "epoch": 0.267498856283903,
-      "grad_norm": 0.5202171206474304,
-      "learning_rate": 9.877662824976876e-06,
-      "loss": 0.4358,
-      "step": 4093
-    },
-    {
-      "epoch": 0.26756421148944515,
-      "grad_norm": 0.44679561257362366,
-      "learning_rate": 9.877586041918034e-06,
-      "loss": 0.374,
-      "step": 4094
-    },
-    {
-      "epoch": 0.2676295666949873,
-      "grad_norm": 0.5135511755943298,
-      "learning_rate": 9.877509235069516e-06,
-      "loss": 0.4368,
-      "step": 4095
-    },
-    {
-      "epoch": 0.2676949219005294,
-      "grad_norm": 0.4886908531188965,
-      "learning_rate": 9.877432404431692e-06,
-      "loss": 0.4077,
-      "step": 4096
-    },
-    {
-      "epoch": 0.2677602771060715,
-      "grad_norm": 0.43191686272621155,
-      "learning_rate": 9.877355550004944e-06,
-      "loss": 0.3836,
-      "step": 4097
-    },
-    {
-      "epoch": 0.2678256323116136,
-      "grad_norm": 0.4852792024612427,
-      "learning_rate": 9.877278671789641e-06,
-      "loss": 0.4156,
-      "step": 4098
-    },
-    {
-      "epoch": 0.26789098751715573,
-      "grad_norm": 0.47688236832618713,
-      "learning_rate": 9.877201769786162e-06,
-      "loss": 0.3864,
-      "step": 4099
-    },
-    {
-      "epoch": 0.26795634272269786,
-      "grad_norm": 0.45649418234825134,
-      "learning_rate": 9.877124843994879e-06,
-      "loss": 0.3622,
-      "step": 4100
-    },
-    {
-      "epoch": 0.26802169792824,
-      "grad_norm": 0.47562023997306824,
-      "learning_rate": 9.87704789441617e-06,
-      "loss": 0.4457,
-      "step": 4101
-    },
-    {
-      "epoch": 0.2680870531337821,
-      "grad_norm": 0.48525452613830566,
-      "learning_rate": 9.876970921050406e-06,
-      "loss": 0.3903,
-      "step": 4102
-    },
-    {
-      "epoch": 0.26815240833932424,
-      "grad_norm": 0.45969781279563904,
-      "learning_rate": 9.87689392389797e-06,
-      "loss": 0.3931,
-      "step": 4103
-    },
-    {
-      "epoch": 0.26821776354486637,
-      "grad_norm": 0.4747847318649292,
-      "learning_rate": 9.876816902959228e-06,
-      "loss": 0.4442,
-      "step": 4104
-    },
-    {
-      "epoch": 0.26828311875040844,
-      "grad_norm": 0.5107711553573608,
-      "learning_rate": 9.876739858234563e-06,
-      "loss": 0.4395,
-      "step": 4105
-    },
-    {
-      "epoch": 0.2683484739559506,
-      "grad_norm": 0.5010309219360352,
-      "learning_rate": 9.876662789724347e-06,
-      "loss": 0.434,
-      "step": 4106
-    },
-    {
-      "epoch": 0.2684138291614927,
-      "grad_norm": 0.4896714389324188,
-      "learning_rate": 9.876585697428958e-06,
-      "loss": 0.4415,
-      "step": 4107
-    },
-    {
-      "epoch": 0.26847918436703483,
-      "grad_norm": 0.47342249751091003,
-      "learning_rate": 9.87650858134877e-06,
-      "loss": 0.4137,
-      "step": 4108
-    },
-    {
-      "epoch": 0.26854453957257696,
-      "grad_norm": 0.4698140323162079,
-      "learning_rate": 9.876431441484164e-06,
-      "loss": 0.396,
-      "step": 4109
-    },
-    {
-      "epoch": 0.2686098947781191,
-      "grad_norm": 0.4311375617980957,
-      "learning_rate": 9.876354277835509e-06,
-      "loss": 0.3754,
-      "step": 4110
-    },
-    {
-      "epoch": 0.2686752499836612,
-      "grad_norm": 0.47583404183387756,
-      "learning_rate": 9.876277090403185e-06,
-      "loss": 0.3759,
-      "step": 4111
-    },
-    {
-      "epoch": 0.26874060518920334,
-      "grad_norm": 0.5071882605552673,
-      "learning_rate": 9.87619987918757e-06,
-      "loss": 0.4031,
-      "step": 4112
-    },
-    {
-      "epoch": 0.26880596039474547,
-      "grad_norm": 0.43392661213874817,
-      "learning_rate": 9.876122644189036e-06,
-      "loss": 0.3539,
-      "step": 4113
-    },
-    {
-      "epoch": 0.26887131560028754,
-      "grad_norm": 0.4953223466873169,
-      "learning_rate": 9.876045385407966e-06,
-      "loss": 0.4125,
-      "step": 4114
-    },
-    {
-      "epoch": 0.26893667080582967,
-      "grad_norm": 0.4818981885910034,
-      "learning_rate": 9.875968102844732e-06,
-      "loss": 0.4236,
-      "step": 4115
-    },
-    {
-      "epoch": 0.2690020260113718,
-      "grad_norm": 0.5168367028236389,
-      "learning_rate": 9.875890796499711e-06,
-      "loss": 0.4124,
-      "step": 4116
-    },
-    {
-      "epoch": 0.2690673812169139,
-      "grad_norm": 0.505105197429657,
-      "learning_rate": 9.875813466373285e-06,
-      "loss": 0.4784,
-      "step": 4117
-    },
-    {
-      "epoch": 0.26913273642245605,
-      "grad_norm": 0.46311938762664795,
-      "learning_rate": 9.875736112465824e-06,
-      "loss": 0.3907,
-      "step": 4118
-    },
-    {
-      "epoch": 0.2691980916279982,
-      "grad_norm": 0.4399479627609253,
-      "learning_rate": 9.875658734777712e-06,
-      "loss": 0.3895,
-      "step": 4119
-    },
-    {
-      "epoch": 0.2692634468335403,
-      "grad_norm": 0.4627532660961151,
-      "learning_rate": 9.87558133330932e-06,
-      "loss": 0.3478,
-      "step": 4120
-    },
-    {
-      "epoch": 0.26932880203908244,
-      "grad_norm": 0.4844004213809967,
-      "learning_rate": 9.875503908061031e-06,
-      "loss": 0.3943,
-      "step": 4121
-    },
-    {
-      "epoch": 0.2693941572446245,
-      "grad_norm": 0.4906649589538574,
-      "learning_rate": 9.875426459033219e-06,
-      "loss": 0.4322,
-      "step": 4122
-    },
-    {
-      "epoch": 0.26945951245016664,
-      "grad_norm": 0.43836209177970886,
-      "learning_rate": 9.875348986226263e-06,
-      "loss": 0.3721,
-      "step": 4123
-    },
-    {
-      "epoch": 0.26952486765570877,
-      "grad_norm": 0.46436160802841187,
-      "learning_rate": 9.875271489640542e-06,
-      "loss": 0.4343,
-      "step": 4124
-    },
-    {
-      "epoch": 0.2695902228612509,
-      "grad_norm": 0.45985153317451477,
-      "learning_rate": 9.875193969276433e-06,
-      "loss": 0.4208,
-      "step": 4125
-    },
-    {
-      "epoch": 0.269655578066793,
-      "grad_norm": 0.5256823897361755,
-      "learning_rate": 9.875116425134313e-06,
-      "loss": 0.47,
-      "step": 4126
-    },
-    {
-      "epoch": 0.26972093327233515,
-      "grad_norm": 0.47644487023353577,
-      "learning_rate": 9.875038857214563e-06,
-      "loss": 0.4195,
-      "step": 4127
-    },
-    {
-      "epoch": 0.2697862884778773,
-      "grad_norm": 0.4636766314506531,
-      "learning_rate": 9.874961265517557e-06,
-      "loss": 0.4012,
-      "step": 4128
-    },
-    {
-      "epoch": 0.2698516436834194,
-      "grad_norm": 0.4734828770160675,
-      "learning_rate": 9.874883650043678e-06,
-      "loss": 0.4096,
-      "step": 4129
-    },
-    {
-      "epoch": 0.2699169988889615,
-      "grad_norm": 0.4910510778427124,
-      "learning_rate": 9.874806010793303e-06,
-      "loss": 0.4372,
-      "step": 4130
-    },
-    {
-      "epoch": 0.2699823540945036,
-      "grad_norm": 0.47214993834495544,
-      "learning_rate": 9.87472834776681e-06,
-      "loss": 0.4324,
-      "step": 4131
-    },
-    {
-      "epoch": 0.27004770930004574,
-      "grad_norm": 0.46082454919815063,
-      "learning_rate": 9.874650660964578e-06,
-      "loss": 0.4119,
-      "step": 4132
-    },
-    {
-      "epoch": 0.27011306450558786,
-      "grad_norm": 0.4675063490867615,
-      "learning_rate": 9.874572950386986e-06,
-      "loss": 0.4403,
-      "step": 4133
-    },
-    {
-      "epoch": 0.27017841971113,
-      "grad_norm": 0.4442681670188904,
-      "learning_rate": 9.874495216034413e-06,
-      "loss": 0.3491,
-      "step": 4134
-    },
-    {
-      "epoch": 0.2702437749166721,
-      "grad_norm": 0.46750369668006897,
-      "learning_rate": 9.874417457907237e-06,
-      "loss": 0.3811,
-      "step": 4135
-    },
-    {
-      "epoch": 0.27030913012221425,
-      "grad_norm": 0.4324701726436615,
-      "learning_rate": 9.87433967600584e-06,
-      "loss": 0.381,
-      "step": 4136
-    },
-    {
-      "epoch": 0.2703744853277564,
-      "grad_norm": 0.449904203414917,
-      "learning_rate": 9.8742618703306e-06,
-      "loss": 0.4025,
-      "step": 4137
-    },
-    {
-      "epoch": 0.2704398405332985,
-      "grad_norm": 0.4237454831600189,
-      "learning_rate": 9.874184040881893e-06,
-      "loss": 0.3427,
-      "step": 4138
-    },
-    {
-      "epoch": 0.2705051957388406,
-      "grad_norm": 0.4685540497303009,
-      "learning_rate": 9.874106187660106e-06,
-      "loss": 0.3583,
-      "step": 4139
-    },
-    {
-      "epoch": 0.2705705509443827,
-      "grad_norm": 0.4869711399078369,
-      "learning_rate": 9.874028310665612e-06,
-      "loss": 0.4449,
-      "step": 4140
-    },
-    {
-      "epoch": 0.27063590614992483,
-      "grad_norm": 0.4771939516067505,
-      "learning_rate": 9.873950409898793e-06,
-      "loss": 0.4557,
-      "step": 4141
-    },
-    {
-      "epoch": 0.27070126135546696,
-      "grad_norm": 0.4451706111431122,
-      "learning_rate": 9.873872485360032e-06,
-      "loss": 0.3941,
-      "step": 4142
-    },
-    {
-      "epoch": 0.2707666165610091,
-      "grad_norm": 0.4551406502723694,
-      "learning_rate": 9.873794537049704e-06,
-      "loss": 0.3592,
-      "step": 4143
-    },
-    {
-      "epoch": 0.2708319717665512,
-      "grad_norm": 0.5270684361457825,
-      "learning_rate": 9.873716564968193e-06,
-      "loss": 0.4924,
-      "step": 4144
-    },
-    {
-      "epoch": 0.27089732697209334,
-      "grad_norm": 0.5258855819702148,
-      "learning_rate": 9.873638569115878e-06,
-      "loss": 0.4887,
-      "step": 4145
-    },
-    {
-      "epoch": 0.27096268217763547,
-      "grad_norm": 0.4573988914489746,
-      "learning_rate": 9.873560549493138e-06,
-      "loss": 0.4094,
-      "step": 4146
-    },
-    {
-      "epoch": 0.27102803738317754,
-      "grad_norm": 0.42023995518684387,
-      "learning_rate": 9.873482506100355e-06,
-      "loss": 0.3675,
-      "step": 4147
-    },
-    {
-      "epoch": 0.2710933925887197,
-      "grad_norm": 0.4430708587169647,
-      "learning_rate": 9.87340443893791e-06,
-      "loss": 0.4083,
-      "step": 4148
-    },
-    {
-      "epoch": 0.2711587477942618,
-      "grad_norm": 0.47372132539749146,
-      "learning_rate": 9.873326348006185e-06,
-      "loss": 0.4293,
-      "step": 4149
-    },
-    {
-      "epoch": 0.27122410299980393,
-      "grad_norm": 0.5177667140960693,
-      "learning_rate": 9.873248233305558e-06,
-      "loss": 0.485,
-      "step": 4150
-    },
-    {
-      "epoch": 0.27128945820534606,
-      "grad_norm": 0.4773666560649872,
-      "learning_rate": 9.873170094836408e-06,
-      "loss": 0.4055,
-      "step": 4151
-    },
-    {
-      "epoch": 0.2713548134108882,
-      "grad_norm": 0.5047542452812195,
-      "learning_rate": 9.873091932599124e-06,
-      "loss": 0.4817,
-      "step": 4152
-    },
-    {
-      "epoch": 0.2714201686164303,
-      "grad_norm": 0.4635542035102844,
-      "learning_rate": 9.873013746594078e-06,
-      "loss": 0.441,
-      "step": 4153
-    },
-    {
-      "epoch": 0.27148552382197244,
-      "grad_norm": 0.4301548898220062,
-      "learning_rate": 9.87293553682166e-06,
-      "loss": 0.3518,
-      "step": 4154
-    },
-    {
-      "epoch": 0.27155087902751457,
-      "grad_norm": 0.46144917607307434,
-      "learning_rate": 9.872857303282245e-06,
-      "loss": 0.3923,
-      "step": 4155
-    },
-    {
-      "epoch": 0.27161623423305664,
-      "grad_norm": 0.5201066136360168,
-      "learning_rate": 9.872779045976215e-06,
-      "loss": 0.5005,
-      "step": 4156
-    },
-    {
-      "epoch": 0.27168158943859877,
-      "grad_norm": 0.4255172312259674,
-      "learning_rate": 9.872700764903958e-06,
-      "loss": 0.3595,
-      "step": 4157
-    },
-    {
-      "epoch": 0.2717469446441409,
-      "grad_norm": 0.4870705306529999,
-      "learning_rate": 9.872622460065848e-06,
-      "loss": 0.453,
-      "step": 4158
-    },
-    {
-      "epoch": 0.271812299849683,
-      "grad_norm": 0.4662631154060364,
-      "learning_rate": 9.87254413146227e-06,
-      "loss": 0.4058,
-      "step": 4159
-    },
-    {
-      "epoch": 0.27187765505522515,
-      "grad_norm": 0.4541662931442261,
-      "learning_rate": 9.872465779093607e-06,
-      "loss": 0.3861,
-      "step": 4160
-    },
-    {
-      "epoch": 0.2719430102607673,
-      "grad_norm": 0.44846227765083313,
-      "learning_rate": 9.872387402960241e-06,
-      "loss": 0.3581,
-      "step": 4161
-    },
-    {
-      "epoch": 0.2720083654663094,
-      "grad_norm": 0.47655147314071655,
-      "learning_rate": 9.872309003062554e-06,
-      "loss": 0.4104,
-      "step": 4162
-    },
-    {
-      "epoch": 0.27207372067185154,
-      "grad_norm": 0.46559497714042664,
-      "learning_rate": 9.872230579400928e-06,
-      "loss": 0.3903,
-      "step": 4163
-    },
-    {
-      "epoch": 0.2721390758773936,
-      "grad_norm": 0.4417462646961212,
-      "learning_rate": 9.872152131975745e-06,
-      "loss": 0.3891,
-      "step": 4164
-    },
-    {
-      "epoch": 0.27220443108293574,
-      "grad_norm": 0.4599807858467102,
-      "learning_rate": 9.872073660787388e-06,
-      "loss": 0.3899,
-      "step": 4165
-    },
-    {
-      "epoch": 0.27226978628847787,
-      "grad_norm": 0.4935910105705261,
-      "learning_rate": 9.87199516583624e-06,
-      "loss": 0.438,
-      "step": 4166
-    },
-    {
-      "epoch": 0.27233514149402,
-      "grad_norm": 0.4480958878993988,
-      "learning_rate": 9.871916647122684e-06,
-      "loss": 0.407,
-      "step": 4167
-    },
-    {
-      "epoch": 0.2724004966995621,
-      "grad_norm": 0.4256836771965027,
-      "learning_rate": 9.871838104647102e-06,
-      "loss": 0.3393,
-      "step": 4168
-    },
-    {
-      "epoch": 0.27246585190510425,
-      "grad_norm": 0.45450159907341003,
-      "learning_rate": 9.871759538409878e-06,
-      "loss": 0.4047,
-      "step": 4169
-    },
-    {
-      "epoch": 0.2725312071106464,
-      "grad_norm": 0.5286385416984558,
-      "learning_rate": 9.871680948411396e-06,
-      "loss": 0.4408,
-      "step": 4170
-    },
-    {
-      "epoch": 0.2725965623161885,
-      "grad_norm": 0.4764297604560852,
-      "learning_rate": 9.871602334652037e-06,
-      "loss": 0.4316,
-      "step": 4171
-    },
-    {
-      "epoch": 0.2726619175217306,
-      "grad_norm": 0.5015479326248169,
-      "learning_rate": 9.871523697132186e-06,
-      "loss": 0.4222,
-      "step": 4172
-    },
-    {
-      "epoch": 0.2727272727272727,
-      "grad_norm": 0.4523698389530182,
-      "learning_rate": 9.871445035852228e-06,
-      "loss": 0.3854,
-      "step": 4173
-    },
-    {
-      "epoch": 0.27279262793281484,
-      "grad_norm": 0.4936469495296478,
-      "learning_rate": 9.871366350812543e-06,
-      "loss": 0.4206,
-      "step": 4174
-    },
-    {
-      "epoch": 0.27285798313835696,
-      "grad_norm": 0.48123857378959656,
-      "learning_rate": 9.87128764201352e-06,
-      "loss": 0.3939,
-      "step": 4175
-    },
-    {
-      "epoch": 0.2729233383438991,
-      "grad_norm": 0.48141565918922424,
-      "learning_rate": 9.871208909455535e-06,
-      "loss": 0.4143,
-      "step": 4176
-    },
-    {
-      "epoch": 0.2729886935494412,
-      "grad_norm": 0.43874800205230713,
-      "learning_rate": 9.871130153138978e-06,
-      "loss": 0.3398,
-      "step": 4177
-    },
-    {
-      "epoch": 0.27305404875498335,
-      "grad_norm": 0.4822591245174408,
-      "learning_rate": 9.871051373064232e-06,
-      "loss": 0.4502,
-      "step": 4178
-    },
-    {
-      "epoch": 0.2731194039605255,
-      "grad_norm": 0.47263237833976746,
-      "learning_rate": 9.870972569231681e-06,
-      "loss": 0.3933,
-      "step": 4179
-    },
-    {
-      "epoch": 0.2731847591660676,
-      "grad_norm": 0.4788112938404083,
-      "learning_rate": 9.87089374164171e-06,
-      "loss": 0.4106,
-      "step": 4180
-    },
-    {
-      "epoch": 0.2732501143716097,
-      "grad_norm": 0.4840521812438965,
-      "learning_rate": 9.870814890294701e-06,
-      "loss": 0.4249,
-      "step": 4181
-    },
-    {
-      "epoch": 0.2733154695771518,
-      "grad_norm": 0.4387449622154236,
-      "learning_rate": 9.870736015191043e-06,
-      "loss": 0.3768,
-      "step": 4182
-    },
-    {
-      "epoch": 0.27338082478269393,
-      "grad_norm": 0.49455657601356506,
-      "learning_rate": 9.870657116331118e-06,
-      "loss": 0.4425,
-      "step": 4183
-    },
-    {
-      "epoch": 0.27344617998823606,
-      "grad_norm": 0.47682246565818787,
-      "learning_rate": 9.870578193715308e-06,
-      "loss": 0.411,
-      "step": 4184
-    },
-    {
-      "epoch": 0.2735115351937782,
-      "grad_norm": 0.5592783093452454,
-      "learning_rate": 9.870499247344004e-06,
-      "loss": 0.4408,
-      "step": 4185
-    },
-    {
-      "epoch": 0.2735768903993203,
-      "grad_norm": 0.4375693202018738,
-      "learning_rate": 9.870420277217584e-06,
-      "loss": 0.3513,
-      "step": 4186
-    },
-    {
-      "epoch": 0.27364224560486244,
-      "grad_norm": 0.5457838177680969,
-      "learning_rate": 9.870341283336439e-06,
-      "loss": 0.4368,
-      "step": 4187
-    },
-    {
-      "epoch": 0.27370760081040457,
-      "grad_norm": 0.48499536514282227,
-      "learning_rate": 9.870262265700954e-06,
-      "loss": 0.4413,
-      "step": 4188
-    },
-    {
-      "epoch": 0.27377295601594664,
-      "grad_norm": 0.49165967106819153,
-      "learning_rate": 9.870183224311512e-06,
-      "loss": 0.4388,
-      "step": 4189
-    },
-    {
-      "epoch": 0.2738383112214888,
-      "grad_norm": 0.47874099016189575,
-      "learning_rate": 9.870104159168497e-06,
-      "loss": 0.3902,
-      "step": 4190
-    },
-    {
-      "epoch": 0.2739036664270309,
-      "grad_norm": 0.49625450372695923,
-      "learning_rate": 9.870025070272298e-06,
-      "loss": 0.4004,
-      "step": 4191
-    },
-    {
-      "epoch": 0.27396902163257303,
-      "grad_norm": 0.46407896280288696,
-      "learning_rate": 9.869945957623302e-06,
-      "loss": 0.4058,
-      "step": 4192
-    },
-    {
-      "epoch": 0.27403437683811516,
-      "grad_norm": 0.474025160074234,
-      "learning_rate": 9.869866821221889e-06,
-      "loss": 0.3872,
-      "step": 4193
-    },
-    {
-      "epoch": 0.2740997320436573,
-      "grad_norm": 0.47441983222961426,
-      "learning_rate": 9.86978766106845e-06,
-      "loss": 0.4028,
-      "step": 4194
-    },
-    {
-      "epoch": 0.2741650872491994,
-      "grad_norm": 0.44902557134628296,
-      "learning_rate": 9.86970847716337e-06,
-      "loss": 0.3714,
-      "step": 4195
-    },
-    {
-      "epoch": 0.27423044245474154,
-      "grad_norm": 0.43087777495384216,
-      "learning_rate": 9.869629269507034e-06,
-      "loss": 0.3599,
-      "step": 4196
-    },
-    {
-      "epoch": 0.27429579766028367,
-      "grad_norm": 0.43613943457603455,
-      "learning_rate": 9.86955003809983e-06,
-      "loss": 0.3849,
-      "step": 4197
-    },
-    {
-      "epoch": 0.27436115286582574,
-      "grad_norm": 0.502096951007843,
-      "learning_rate": 9.86947078294214e-06,
-      "loss": 0.422,
-      "step": 4198
-    },
-    {
-      "epoch": 0.27442650807136787,
-      "grad_norm": 0.44126638770103455,
-      "learning_rate": 9.869391504034358e-06,
-      "loss": 0.3816,
-      "step": 4199
-    },
-    {
-      "epoch": 0.27449186327691,
-      "grad_norm": 0.485454261302948,
-      "learning_rate": 9.869312201376865e-06,
-      "loss": 0.4453,
-      "step": 4200
-    },
-    {
-      "epoch": 0.2745572184824521,
-      "grad_norm": 0.5053019523620605,
-      "learning_rate": 9.869232874970052e-06,
-      "loss": 0.5149,
-      "step": 4201
-    },
-    {
-      "epoch": 0.27462257368799425,
-      "grad_norm": 0.4587983191013336,
-      "learning_rate": 9.8691535248143e-06,
-      "loss": 0.4059,
-      "step": 4202
-    },
-    {
-      "epoch": 0.2746879288935364,
-      "grad_norm": 0.5347425937652588,
-      "learning_rate": 9.869074150910001e-06,
-      "loss": 0.4822,
-      "step": 4203
-    },
-    {
-      "epoch": 0.2747532840990785,
-      "grad_norm": 0.4826345443725586,
-      "learning_rate": 9.86899475325754e-06,
-      "loss": 0.4088,
-      "step": 4204
-    },
-    {
-      "epoch": 0.27481863930462064,
-      "grad_norm": 0.48696738481521606,
-      "learning_rate": 9.868915331857304e-06,
-      "loss": 0.4142,
-      "step": 4205
-    },
-    {
-      "epoch": 0.2748839945101627,
-      "grad_norm": 0.48205018043518066,
-      "learning_rate": 9.868835886709685e-06,
-      "loss": 0.4301,
-      "step": 4206
-    },
-    {
-      "epoch": 0.27494934971570484,
-      "grad_norm": 0.5135433077812195,
-      "learning_rate": 9.868756417815062e-06,
-      "loss": 0.4694,
-      "step": 4207
-    },
-    {
-      "epoch": 0.27501470492124697,
-      "grad_norm": 0.49597039818763733,
-      "learning_rate": 9.86867692517383e-06,
-      "loss": 0.4468,
-      "step": 4208
-    },
-    {
-      "epoch": 0.2750800601267891,
-      "grad_norm": 0.46526581048965454,
-      "learning_rate": 9.868597408786373e-06,
-      "loss": 0.3739,
-      "step": 4209
-    },
-    {
-      "epoch": 0.2751454153323312,
-      "grad_norm": 0.5034768581390381,
-      "learning_rate": 9.86851786865308e-06,
-      "loss": 0.4577,
-      "step": 4210
-    },
-    {
-      "epoch": 0.27521077053787335,
-      "grad_norm": 0.48159655928611755,
-      "learning_rate": 9.86843830477434e-06,
-      "loss": 0.4274,
-      "step": 4211
-    },
-    {
-      "epoch": 0.2752761257434155,
-      "grad_norm": 0.4275132119655609,
-      "learning_rate": 9.868358717150537e-06,
-      "loss": 0.3662,
-      "step": 4212
-    },
-    {
-      "epoch": 0.2753414809489576,
-      "grad_norm": 0.45612967014312744,
-      "learning_rate": 9.868279105782063e-06,
-      "loss": 0.4201,
-      "step": 4213
-    },
-    {
-      "epoch": 0.2754068361544997,
-      "grad_norm": 0.4631737172603607,
-      "learning_rate": 9.868199470669306e-06,
-      "loss": 0.3874,
-      "step": 4214
-    },
-    {
-      "epoch": 0.2754721913600418,
-      "grad_norm": 0.4741147756576538,
-      "learning_rate": 9.868119811812653e-06,
-      "loss": 0.4184,
-      "step": 4215
-    },
-    {
-      "epoch": 0.27553754656558394,
-      "grad_norm": 0.4729333221912384,
-      "learning_rate": 9.868040129212495e-06,
-      "loss": 0.4127,
-      "step": 4216
-    },
-    {
-      "epoch": 0.27560290177112606,
-      "grad_norm": 0.441802978515625,
-      "learning_rate": 9.867960422869217e-06,
-      "loss": 0.3622,
-      "step": 4217
-    },
-    {
-      "epoch": 0.2756682569766682,
-      "grad_norm": 0.47777706384658813,
-      "learning_rate": 9.867880692783209e-06,
-      "loss": 0.4121,
-      "step": 4218
-    },
-    {
-      "epoch": 0.2757336121822103,
-      "grad_norm": 0.483521044254303,
-      "learning_rate": 9.867800938954862e-06,
-      "loss": 0.4265,
-      "step": 4219
-    },
-    {
-      "epoch": 0.27579896738775245,
-      "grad_norm": 0.48338282108306885,
-      "learning_rate": 9.867721161384564e-06,
-      "loss": 0.4245,
-      "step": 4220
-    },
-    {
-      "epoch": 0.2758643225932946,
-      "grad_norm": 0.518380880355835,
-      "learning_rate": 9.867641360072702e-06,
-      "loss": 0.4834,
-      "step": 4221
-    },
-    {
-      "epoch": 0.2759296777988367,
-      "grad_norm": 0.4838882386684418,
-      "learning_rate": 9.867561535019667e-06,
-      "loss": 0.424,
-      "step": 4222
-    },
-    {
-      "epoch": 0.2759950330043788,
-      "grad_norm": 0.4742378890514374,
-      "learning_rate": 9.867481686225848e-06,
-      "loss": 0.4034,
-      "step": 4223
-    },
-    {
-      "epoch": 0.2760603882099209,
-      "grad_norm": 0.45290738344192505,
-      "learning_rate": 9.867401813691636e-06,
-      "loss": 0.3871,
-      "step": 4224
-    },
-    {
-      "epoch": 0.27612574341546303,
-      "grad_norm": 0.463894784450531,
-      "learning_rate": 9.867321917417418e-06,
-      "loss": 0.4015,
-      "step": 4225
-    },
-    {
-      "epoch": 0.27619109862100516,
-      "grad_norm": 0.660118043422699,
-      "learning_rate": 9.867241997403586e-06,
-      "loss": 0.3977,
-      "step": 4226
-    },
-    {
-      "epoch": 0.2762564538265473,
-      "grad_norm": 0.5125059485435486,
-      "learning_rate": 9.867162053650525e-06,
-      "loss": 0.4303,
-      "step": 4227
-    },
-    {
-      "epoch": 0.2763218090320894,
-      "grad_norm": 0.4675084948539734,
-      "learning_rate": 9.867082086158633e-06,
-      "loss": 0.3835,
-      "step": 4228
-    },
-    {
-      "epoch": 0.27638716423763154,
-      "grad_norm": 0.4916883111000061,
-      "learning_rate": 9.867002094928293e-06,
-      "loss": 0.4358,
-      "step": 4229
-    },
-    {
-      "epoch": 0.27645251944317367,
-      "grad_norm": 0.45190033316612244,
-      "learning_rate": 9.866922079959897e-06,
-      "loss": 0.3938,
-      "step": 4230
-    },
-    {
-      "epoch": 0.27651787464871574,
-      "grad_norm": 0.44110557436943054,
-      "learning_rate": 9.866842041253838e-06,
-      "loss": 0.3494,
-      "step": 4231
-    },
-    {
-      "epoch": 0.2765832298542579,
-      "grad_norm": 0.45761820673942566,
-      "learning_rate": 9.866761978810505e-06,
-      "loss": 0.3583,
-      "step": 4232
-    },
-    {
-      "epoch": 0.2766485850598,
-      "grad_norm": 0.5181942582130432,
-      "learning_rate": 9.866681892630286e-06,
-      "loss": 0.4689,
-      "step": 4233
-    },
-    {
-      "epoch": 0.27671394026534213,
-      "grad_norm": 0.4559297263622284,
-      "learning_rate": 9.866601782713572e-06,
-      "loss": 0.3863,
-      "step": 4234
-    },
-    {
-      "epoch": 0.27677929547088426,
-      "grad_norm": 0.4838227927684784,
-      "learning_rate": 9.866521649060758e-06,
-      "loss": 0.4521,
-      "step": 4235
-    },
-    {
-      "epoch": 0.2768446506764264,
-      "grad_norm": 0.5002492666244507,
-      "learning_rate": 9.86644149167223e-06,
-      "loss": 0.4604,
-      "step": 4236
-    },
-    {
-      "epoch": 0.2769100058819685,
-      "grad_norm": 0.4975980818271637,
-      "learning_rate": 9.866361310548383e-06,
-      "loss": 0.4169,
-      "step": 4237
-    },
-    {
-      "epoch": 0.27697536108751064,
-      "grad_norm": 0.4638250470161438,
-      "learning_rate": 9.866281105689605e-06,
-      "loss": 0.3938,
-      "step": 4238
-    },
-    {
-      "epoch": 0.27704071629305277,
-      "grad_norm": 0.491578608751297,
-      "learning_rate": 9.866200877096288e-06,
-      "loss": 0.4545,
-      "step": 4239
-    },
-    {
-      "epoch": 0.27710607149859484,
-      "grad_norm": 0.4563657343387604,
-      "learning_rate": 9.866120624768822e-06,
-      "loss": 0.3852,
-      "step": 4240
-    },
-    {
-      "epoch": 0.27717142670413697,
-      "grad_norm": 0.4132401645183563,
-      "learning_rate": 9.866040348707602e-06,
-      "loss": 0.327,
-      "step": 4241
-    },
-    {
-      "epoch": 0.2772367819096791,
-      "grad_norm": 0.5281173586845398,
-      "learning_rate": 9.865960048913018e-06,
-      "loss": 0.4479,
-      "step": 4242
-    },
-    {
-      "epoch": 0.2773021371152212,
-      "grad_norm": 0.43360963463783264,
-      "learning_rate": 9.86587972538546e-06,
-      "loss": 0.3497,
-      "step": 4243
-    },
-    {
-      "epoch": 0.27736749232076335,
-      "grad_norm": 0.49240660667419434,
-      "learning_rate": 9.86579937812532e-06,
-      "loss": 0.4213,
-      "step": 4244
-    },
-    {
-      "epoch": 0.2774328475263055,
-      "grad_norm": 0.49276968836784363,
-      "learning_rate": 9.865719007132993e-06,
-      "loss": 0.439,
-      "step": 4245
-    },
-    {
-      "epoch": 0.2774982027318476,
-      "grad_norm": 0.461213618516922,
-      "learning_rate": 9.865638612408868e-06,
-      "loss": 0.4273,
-      "step": 4246
-    },
-    {
-      "epoch": 0.27756355793738974,
-      "grad_norm": 0.4940822422504425,
-      "learning_rate": 9.865558193953336e-06,
-      "loss": 0.4213,
-      "step": 4247
-    },
-    {
-      "epoch": 0.2776289131429318,
-      "grad_norm": 0.46520429849624634,
-      "learning_rate": 9.865477751766792e-06,
-      "loss": 0.423,
-      "step": 4248
-    },
-    {
-      "epoch": 0.27769426834847394,
-      "grad_norm": 0.4513480067253113,
-      "learning_rate": 9.865397285849629e-06,
-      "loss": 0.3945,
-      "step": 4249
-    },
-    {
-      "epoch": 0.27775962355401607,
-      "grad_norm": 0.4619412422180176,
-      "learning_rate": 9.865316796202236e-06,
-      "loss": 0.3979,
-      "step": 4250
-    },
-    {
-      "epoch": 0.2778249787595582,
-      "grad_norm": 0.4449423849582672,
-      "learning_rate": 9.865236282825008e-06,
-      "loss": 0.3931,
-      "step": 4251
-    },
-    {
-      "epoch": 0.2778903339651003,
-      "grad_norm": 0.4934066832065582,
-      "learning_rate": 9.865155745718337e-06,
-      "loss": 0.4538,
-      "step": 4252
-    },
-    {
-      "epoch": 0.27795568917064245,
-      "grad_norm": 0.4296741783618927,
-      "learning_rate": 9.865075184882618e-06,
-      "loss": 0.3764,
-      "step": 4253
-    },
-    {
-      "epoch": 0.2780210443761846,
-      "grad_norm": 0.4742310345172882,
-      "learning_rate": 9.86499460031824e-06,
-      "loss": 0.4072,
-      "step": 4254
-    },
-    {
-      "epoch": 0.2780863995817267,
-      "grad_norm": 0.48595312237739563,
-      "learning_rate": 9.864913992025597e-06,
-      "loss": 0.4098,
-      "step": 4255
-    },
-    {
-      "epoch": 0.2781517547872688,
-      "grad_norm": 0.4828436076641083,
-      "learning_rate": 9.864833360005085e-06,
-      "loss": 0.4151,
-      "step": 4256
-    },
-    {
-      "epoch": 0.2782171099928109,
-      "grad_norm": 0.471900075674057,
-      "learning_rate": 9.864752704257095e-06,
-      "loss": 0.384,
-      "step": 4257
-    },
-    {
-      "epoch": 0.27828246519835304,
-      "grad_norm": 0.4691624939441681,
-      "learning_rate": 9.86467202478202e-06,
-      "loss": 0.3715,
-      "step": 4258
-    },
-    {
-      "epoch": 0.27834782040389516,
-      "grad_norm": 0.4548409581184387,
-      "learning_rate": 9.864591321580255e-06,
-      "loss": 0.4181,
-      "step": 4259
-    },
-    {
-      "epoch": 0.2784131756094373,
-      "grad_norm": 0.47972768545150757,
-      "learning_rate": 9.864510594652194e-06,
-      "loss": 0.4132,
-      "step": 4260
-    },
-    {
-      "epoch": 0.2784785308149794,
-      "grad_norm": 0.48893147706985474,
-      "learning_rate": 9.864429843998227e-06,
-      "loss": 0.4143,
-      "step": 4261
-    },
-    {
-      "epoch": 0.27854388602052155,
-      "grad_norm": 0.48815682530403137,
-      "learning_rate": 9.864349069618753e-06,
-      "loss": 0.416,
-      "step": 4262
-    },
-    {
-      "epoch": 0.2786092412260637,
-      "grad_norm": 0.5057495832443237,
-      "learning_rate": 9.864268271514162e-06,
-      "loss": 0.4514,
-      "step": 4263
-    },
-    {
-      "epoch": 0.2786745964316058,
-      "grad_norm": 0.5050725936889648,
-      "learning_rate": 9.864187449684849e-06,
-      "loss": 0.4184,
-      "step": 4264
-    },
-    {
-      "epoch": 0.2787399516371479,
-      "grad_norm": 0.4830845892429352,
-      "learning_rate": 9.864106604131209e-06,
-      "loss": 0.4709,
-      "step": 4265
-    },
-    {
-      "epoch": 0.27880530684269,
-      "grad_norm": 0.4551175534725189,
-      "learning_rate": 9.864025734853636e-06,
-      "loss": 0.3539,
-      "step": 4266
-    },
-    {
-      "epoch": 0.27887066204823213,
-      "grad_norm": 0.48389312624931335,
-      "learning_rate": 9.863944841852523e-06,
-      "loss": 0.4114,
-      "step": 4267
-    },
-    {
-      "epoch": 0.27893601725377426,
-      "grad_norm": 0.49458593130111694,
-      "learning_rate": 9.86386392512827e-06,
-      "loss": 0.4188,
-      "step": 4268
-    },
-    {
-      "epoch": 0.2790013724593164,
-      "grad_norm": 0.45328977704048157,
-      "learning_rate": 9.863782984681266e-06,
-      "loss": 0.4042,
-      "step": 4269
-    },
-    {
-      "epoch": 0.2790667276648585,
-      "grad_norm": 0.4529956877231598,
-      "learning_rate": 9.863702020511905e-06,
-      "loss": 0.4071,
-      "step": 4270
-    },
-    {
-      "epoch": 0.27913208287040064,
-      "grad_norm": 0.48014965653419495,
-      "learning_rate": 9.863621032620588e-06,
-      "loss": 0.406,
-      "step": 4271
-    },
-    {
-      "epoch": 0.27919743807594277,
-      "grad_norm": 0.5194218158721924,
-      "learning_rate": 9.863540021007702e-06,
-      "loss": 0.4139,
-      "step": 4272
-    },
-    {
-      "epoch": 0.27926279328148484,
-      "grad_norm": 0.4963468909263611,
-      "learning_rate": 9.86345898567365e-06,
-      "loss": 0.3846,
-      "step": 4273
-    },
-    {
-      "epoch": 0.279328148487027,
-      "grad_norm": 0.4484500586986542,
-      "learning_rate": 9.863377926618823e-06,
-      "loss": 0.3742,
-      "step": 4274
-    },
-    {
-      "epoch": 0.2793935036925691,
-      "grad_norm": 0.4989372193813324,
-      "learning_rate": 9.863296843843616e-06,
-      "loss": 0.4348,
-      "step": 4275
-    },
-    {
-      "epoch": 0.27945885889811123,
-      "grad_norm": 0.5235374569892883,
-      "learning_rate": 9.863215737348425e-06,
-      "loss": 0.4709,
-      "step": 4276
-    },
-    {
-      "epoch": 0.27952421410365336,
-      "grad_norm": 0.4583902657032013,
-      "learning_rate": 9.863134607133647e-06,
-      "loss": 0.368,
-      "step": 4277
-    },
-    {
-      "epoch": 0.2795895693091955,
-      "grad_norm": 0.4683593213558197,
-      "learning_rate": 9.863053453199676e-06,
-      "loss": 0.4128,
-      "step": 4278
-    },
-    {
-      "epoch": 0.2796549245147376,
-      "grad_norm": 0.49003276228904724,
-      "learning_rate": 9.862972275546911e-06,
-      "loss": 0.3988,
-      "step": 4279
-    },
-    {
-      "epoch": 0.27972027972027974,
-      "grad_norm": 0.4580129384994507,
-      "learning_rate": 9.862891074175743e-06,
-      "loss": 0.3855,
-      "step": 4280
-    },
-    {
-      "epoch": 0.27978563492582187,
-      "grad_norm": 0.4886868894100189,
-      "learning_rate": 9.862809849086571e-06,
-      "loss": 0.4067,
-      "step": 4281
-    },
-    {
-      "epoch": 0.27985099013136394,
-      "grad_norm": 0.47899141907691956,
-      "learning_rate": 9.862728600279791e-06,
-      "loss": 0.4376,
-      "step": 4282
-    },
-    {
-      "epoch": 0.27991634533690607,
-      "grad_norm": 0.5096350908279419,
-      "learning_rate": 9.8626473277558e-06,
-      "loss": 0.3962,
-      "step": 4283
-    },
-    {
-      "epoch": 0.2799817005424482,
-      "grad_norm": 0.48719924688339233,
-      "learning_rate": 9.862566031514992e-06,
-      "loss": 0.3973,
-      "step": 4284
-    },
-    {
-      "epoch": 0.2800470557479903,
-      "grad_norm": 0.4794047176837921,
-      "learning_rate": 9.862484711557765e-06,
-      "loss": 0.4489,
-      "step": 4285
-    },
-    {
-      "epoch": 0.28011241095353245,
-      "grad_norm": 0.44751083850860596,
-      "learning_rate": 9.862403367884517e-06,
-      "loss": 0.3554,
-      "step": 4286
-    },
-    {
-      "epoch": 0.2801777661590746,
-      "grad_norm": 0.4844992160797119,
-      "learning_rate": 9.862322000495642e-06,
-      "loss": 0.4344,
-      "step": 4287
-    },
-    {
-      "epoch": 0.2802431213646167,
-      "grad_norm": 0.5009134411811829,
-      "learning_rate": 9.862240609391538e-06,
-      "loss": 0.4331,
-      "step": 4288
-    },
-    {
-      "epoch": 0.28030847657015884,
-      "grad_norm": 0.49703457951545715,
-      "learning_rate": 9.862159194572602e-06,
-      "loss": 0.4175,
-      "step": 4289
-    },
-    {
-      "epoch": 0.2803738317757009,
-      "grad_norm": 0.4641529619693756,
-      "learning_rate": 9.862077756039232e-06,
-      "loss": 0.3843,
-      "step": 4290
-    },
-    {
-      "epoch": 0.28043918698124304,
-      "grad_norm": 0.490360289812088,
-      "learning_rate": 9.861996293791825e-06,
-      "loss": 0.4238,
-      "step": 4291
-    },
-    {
-      "epoch": 0.28050454218678517,
-      "grad_norm": 0.44240522384643555,
-      "learning_rate": 9.861914807830776e-06,
-      "loss": 0.3901,
-      "step": 4292
-    },
-    {
-      "epoch": 0.2805698973923273,
-      "grad_norm": 0.4521576166152954,
-      "learning_rate": 9.861833298156485e-06,
-      "loss": 0.3576,
-      "step": 4293
-    },
-    {
-      "epoch": 0.2806352525978694,
-      "grad_norm": 0.45959368348121643,
-      "learning_rate": 9.86175176476935e-06,
-      "loss": 0.3847,
-      "step": 4294
-    },
-    {
-      "epoch": 0.28070060780341155,
-      "grad_norm": 0.506125271320343,
-      "learning_rate": 9.861670207669765e-06,
-      "loss": 0.4687,
-      "step": 4295
-    },
-    {
-      "epoch": 0.2807659630089537,
-      "grad_norm": 0.47286534309387207,
-      "learning_rate": 9.861588626858131e-06,
-      "loss": 0.3955,
-      "step": 4296
-    },
-    {
-      "epoch": 0.2808313182144958,
-      "grad_norm": 0.4786333441734314,
-      "learning_rate": 9.861507022334845e-06,
-      "loss": 0.397,
-      "step": 4297
-    },
-    {
-      "epoch": 0.2808966734200379,
-      "grad_norm": 0.47633546590805054,
-      "learning_rate": 9.861425394100305e-06,
-      "loss": 0.3835,
-      "step": 4298
-    },
-    {
-      "epoch": 0.28096202862558,
-      "grad_norm": 0.4873596131801605,
-      "learning_rate": 9.86134374215491e-06,
-      "loss": 0.4445,
-      "step": 4299
-    },
-    {
-      "epoch": 0.28102738383112214,
-      "grad_norm": 0.45364174246788025,
-      "learning_rate": 9.861262066499058e-06,
-      "loss": 0.3648,
-      "step": 4300
-    },
-    {
-      "epoch": 0.28109273903666426,
-      "grad_norm": 0.5587484836578369,
-      "learning_rate": 9.861180367133144e-06,
-      "loss": 0.4146,
-      "step": 4301
-    },
-    {
-      "epoch": 0.2811580942422064,
-      "grad_norm": 0.47603994607925415,
-      "learning_rate": 9.861098644057572e-06,
-      "loss": 0.3902,
-      "step": 4302
-    },
-    {
-      "epoch": 0.2812234494477485,
-      "grad_norm": 0.5273681879043579,
-      "learning_rate": 9.861016897272738e-06,
-      "loss": 0.4315,
-      "step": 4303
-    },
-    {
-      "epoch": 0.28128880465329065,
-      "grad_norm": 0.4664003252983093,
-      "learning_rate": 9.86093512677904e-06,
-      "loss": 0.3831,
-      "step": 4304
-    },
-    {
-      "epoch": 0.2813541598588328,
-      "grad_norm": 0.4565950930118561,
-      "learning_rate": 9.860853332576876e-06,
-      "loss": 0.4142,
-      "step": 4305
-    },
-    {
-      "epoch": 0.2814195150643749,
-      "grad_norm": 0.5035836696624756,
-      "learning_rate": 9.860771514666646e-06,
-      "loss": 0.4552,
-      "step": 4306
-    },
-    {
-      "epoch": 0.281484870269917,
-      "grad_norm": 0.527451753616333,
-      "learning_rate": 9.860689673048751e-06,
-      "loss": 0.4876,
-      "step": 4307
-    },
-    {
-      "epoch": 0.2815502254754591,
-      "grad_norm": 0.4922797679901123,
-      "learning_rate": 9.860607807723587e-06,
-      "loss": 0.4235,
-      "step": 4308
-    },
-    {
-      "epoch": 0.28161558068100123,
-      "grad_norm": 0.42618000507354736,
-      "learning_rate": 9.860525918691557e-06,
-      "loss": 0.3471,
-      "step": 4309
-    },
-    {
-      "epoch": 0.28168093588654336,
-      "grad_norm": 0.5181910991668701,
-      "learning_rate": 9.860444005953058e-06,
-      "loss": 0.4926,
-      "step": 4310
-    },
-    {
-      "epoch": 0.2817462910920855,
-      "grad_norm": 0.5081905722618103,
-      "learning_rate": 9.860362069508488e-06,
-      "loss": 0.4827,
-      "step": 4311
-    },
-    {
-      "epoch": 0.2818116462976276,
-      "grad_norm": 0.49703356623649597,
-      "learning_rate": 9.860280109358248e-06,
-      "loss": 0.4451,
-      "step": 4312
-    },
-    {
-      "epoch": 0.28187700150316974,
-      "grad_norm": 0.4495966136455536,
-      "learning_rate": 9.86019812550274e-06,
-      "loss": 0.3759,
-      "step": 4313
-    },
-    {
-      "epoch": 0.28194235670871187,
-      "grad_norm": 0.5054322481155396,
-      "learning_rate": 9.860116117942363e-06,
-      "loss": 0.3962,
-      "step": 4314
-    },
-    {
-      "epoch": 0.28200771191425394,
-      "grad_norm": 0.5199347138404846,
-      "learning_rate": 9.860034086677515e-06,
-      "loss": 0.4643,
-      "step": 4315
-    },
-    {
-      "epoch": 0.2820730671197961,
-      "grad_norm": 0.45203331112861633,
-      "learning_rate": 9.859952031708595e-06,
-      "loss": 0.3936,
-      "step": 4316
-    },
-    {
-      "epoch": 0.2821384223253382,
-      "grad_norm": 0.4817637503147125,
-      "learning_rate": 9.859869953036007e-06,
-      "loss": 0.4314,
-      "step": 4317
-    },
-    {
-      "epoch": 0.28220377753088033,
-      "grad_norm": 0.4901078939437866,
-      "learning_rate": 9.85978785066015e-06,
-      "loss": 0.4126,
-      "step": 4318
-    },
-    {
-      "epoch": 0.28226913273642246,
-      "grad_norm": 0.46651607751846313,
-      "learning_rate": 9.859705724581423e-06,
-      "loss": 0.3995,
-      "step": 4319
-    },
-    {
-      "epoch": 0.2823344879419646,
-      "grad_norm": 0.4911923110485077,
-      "learning_rate": 9.859623574800228e-06,
-      "loss": 0.4519,
-      "step": 4320
-    },
-    {
-      "epoch": 0.2823998431475067,
-      "grad_norm": 0.46471184492111206,
-      "learning_rate": 9.859541401316965e-06,
-      "loss": 0.3877,
-      "step": 4321
-    },
-    {
-      "epoch": 0.28246519835304884,
-      "grad_norm": 0.4655369222164154,
-      "learning_rate": 9.859459204132037e-06,
-      "loss": 0.3872,
-      "step": 4322
-    },
-    {
-      "epoch": 0.28253055355859097,
-      "grad_norm": 0.48006582260131836,
-      "learning_rate": 9.85937698324584e-06,
-      "loss": 0.4473,
-      "step": 4323
-    },
-    {
-      "epoch": 0.28259590876413304,
-      "grad_norm": 0.48849478363990784,
-      "learning_rate": 9.85929473865878e-06,
-      "loss": 0.4537,
-      "step": 4324
-    },
-    {
-      "epoch": 0.28266126396967517,
-      "grad_norm": 0.4851961135864258,
-      "learning_rate": 9.859212470371256e-06,
-      "loss": 0.4223,
-      "step": 4325
-    },
-    {
-      "epoch": 0.2827266191752173,
-      "grad_norm": 0.44070538878440857,
-      "learning_rate": 9.859130178383669e-06,
-      "loss": 0.3319,
-      "step": 4326
-    },
-    {
-      "epoch": 0.2827919743807594,
-      "grad_norm": 0.45314934849739075,
-      "learning_rate": 9.859047862696421e-06,
-      "loss": 0.3728,
-      "step": 4327
-    },
-    {
-      "epoch": 0.28285732958630155,
-      "grad_norm": 0.45467445254325867,
-      "learning_rate": 9.858965523309914e-06,
-      "loss": 0.3492,
-      "step": 4328
-    },
-    {
-      "epoch": 0.2829226847918437,
-      "grad_norm": 0.4738505780696869,
-      "learning_rate": 9.858883160224547e-06,
-      "loss": 0.4278,
-      "step": 4329
-    },
-    {
-      "epoch": 0.2829880399973858,
-      "grad_norm": 0.49519234895706177,
-      "learning_rate": 9.858800773440724e-06,
-      "loss": 0.4361,
-      "step": 4330
-    },
-    {
-      "epoch": 0.28305339520292794,
-      "grad_norm": 0.48089075088500977,
-      "learning_rate": 9.858718362958848e-06,
-      "loss": 0.4303,
-      "step": 4331
-    },
-    {
-      "epoch": 0.28311875040847,
-      "grad_norm": 0.5376557111740112,
-      "learning_rate": 9.858635928779318e-06,
-      "loss": 0.4428,
-      "step": 4332
-    },
-    {
-      "epoch": 0.28318410561401214,
-      "grad_norm": 0.4594705402851105,
-      "learning_rate": 9.858553470902536e-06,
-      "loss": 0.4034,
-      "step": 4333
-    },
-    {
-      "epoch": 0.28324946081955427,
-      "grad_norm": 0.4774813950061798,
-      "learning_rate": 9.858470989328907e-06,
-      "loss": 0.4075,
-      "step": 4334
-    },
-    {
-      "epoch": 0.2833148160250964,
-      "grad_norm": 0.4887755513191223,
-      "learning_rate": 9.858388484058834e-06,
-      "loss": 0.3967,
-      "step": 4335
-    },
-    {
-      "epoch": 0.2833801712306385,
-      "grad_norm": 0.4676692485809326,
-      "learning_rate": 9.858305955092715e-06,
-      "loss": 0.3833,
-      "step": 4336
-    },
-    {
-      "epoch": 0.28344552643618065,
-      "grad_norm": 0.4158990681171417,
-      "learning_rate": 9.858223402430955e-06,
-      "loss": 0.3218,
-      "step": 4337
-    },
-    {
-      "epoch": 0.2835108816417228,
-      "grad_norm": 0.4769608974456787,
-      "learning_rate": 9.858140826073956e-06,
-      "loss": 0.4069,
-      "step": 4338
-    },
-    {
-      "epoch": 0.2835762368472649,
-      "grad_norm": 0.4703862965106964,
-      "learning_rate": 9.85805822602212e-06,
-      "loss": 0.4004,
-      "step": 4339
-    },
-    {
-      "epoch": 0.283641592052807,
-      "grad_norm": 0.4532063603401184,
-      "learning_rate": 9.857975602275853e-06,
-      "loss": 0.3712,
-      "step": 4340
-    },
-    {
-      "epoch": 0.2837069472583491,
-      "grad_norm": 0.5027262568473816,
-      "learning_rate": 9.857892954835558e-06,
-      "loss": 0.4135,
-      "step": 4341
-    },
-    {
-      "epoch": 0.28377230246389124,
-      "grad_norm": 0.4493769109249115,
-      "learning_rate": 9.857810283701632e-06,
-      "loss": 0.3551,
-      "step": 4342
-    },
-    {
-      "epoch": 0.28383765766943336,
-      "grad_norm": 0.4735272228717804,
-      "learning_rate": 9.857727588874484e-06,
-      "loss": 0.43,
-      "step": 4343
-    },
-    {
-      "epoch": 0.2839030128749755,
-      "grad_norm": 0.4992130696773529,
-      "learning_rate": 9.857644870354516e-06,
-      "loss": 0.4497,
-      "step": 4344
-    },
-    {
-      "epoch": 0.2839683680805176,
-      "grad_norm": 0.5045772194862366,
-      "learning_rate": 9.85756212814213e-06,
-      "loss": 0.4368,
-      "step": 4345
-    },
-    {
-      "epoch": 0.28403372328605975,
-      "grad_norm": 0.4504072666168213,
-      "learning_rate": 9.857479362237732e-06,
-      "loss": 0.3682,
-      "step": 4346
-    },
-    {
-      "epoch": 0.2840990784916019,
-      "grad_norm": 0.44460558891296387,
-      "learning_rate": 9.857396572641724e-06,
-      "loss": 0.3883,
-      "step": 4347
-    },
-    {
-      "epoch": 0.284164433697144,
-      "grad_norm": 0.4436810612678528,
-      "learning_rate": 9.85731375935451e-06,
-      "loss": 0.3719,
-      "step": 4348
-    },
-    {
-      "epoch": 0.2842297889026861,
-      "grad_norm": 0.47199153900146484,
-      "learning_rate": 9.857230922376496e-06,
-      "loss": 0.4315,
-      "step": 4349
-    },
-    {
-      "epoch": 0.2842951441082282,
-      "grad_norm": 0.4648796319961548,
-      "learning_rate": 9.857148061708082e-06,
-      "loss": 0.4245,
-      "step": 4350
-    },
-    {
-      "epoch": 0.28436049931377033,
-      "grad_norm": 0.5298538208007812,
-      "learning_rate": 9.857065177349673e-06,
-      "loss": 0.5075,
-      "step": 4351
-    },
-    {
-      "epoch": 0.28442585451931246,
-      "grad_norm": 0.47171550989151,
-      "learning_rate": 9.856982269301676e-06,
-      "loss": 0.409,
-      "step": 4352
-    },
-    {
-      "epoch": 0.2844912097248546,
-      "grad_norm": 0.4968525171279907,
-      "learning_rate": 9.856899337564494e-06,
-      "loss": 0.4702,
-      "step": 4353
-    },
-    {
-      "epoch": 0.2845565649303967,
-      "grad_norm": 0.4832991361618042,
-      "learning_rate": 9.85681638213853e-06,
-      "loss": 0.3866,
-      "step": 4354
-    },
-    {
-      "epoch": 0.28462192013593884,
-      "grad_norm": 0.47334814071655273,
-      "learning_rate": 9.856733403024192e-06,
-      "loss": 0.4389,
-      "step": 4355
-    },
-    {
-      "epoch": 0.28468727534148097,
-      "grad_norm": 0.4745638072490692,
-      "learning_rate": 9.856650400221882e-06,
-      "loss": 0.4068,
-      "step": 4356
-    },
-    {
-      "epoch": 0.28475263054702304,
-      "grad_norm": 0.4593832790851593,
-      "learning_rate": 9.856567373732005e-06,
-      "loss": 0.404,
-      "step": 4357
-    },
-    {
-      "epoch": 0.2848179857525652,
-      "grad_norm": 0.4992309510707855,
-      "learning_rate": 9.856484323554967e-06,
-      "loss": 0.4387,
-      "step": 4358
-    },
-    {
-      "epoch": 0.2848833409581073,
-      "grad_norm": 0.46862438321113586,
-      "learning_rate": 9.856401249691171e-06,
-      "loss": 0.4036,
-      "step": 4359
-    },
-    {
-      "epoch": 0.28494869616364943,
-      "grad_norm": 0.5406008362770081,
-      "learning_rate": 9.856318152141026e-06,
-      "loss": 0.4794,
-      "step": 4360
-    },
-    {
-      "epoch": 0.28501405136919156,
-      "grad_norm": 0.45046672224998474,
-      "learning_rate": 9.856235030904934e-06,
-      "loss": 0.3533,
-      "step": 4361
-    },
-    {
-      "epoch": 0.2850794065747337,
-      "grad_norm": 0.5082617998123169,
-      "learning_rate": 9.8561518859833e-06,
-      "loss": 0.4268,
-      "step": 4362
-    },
-    {
-      "epoch": 0.2851447617802758,
-      "grad_norm": 0.4527800977230072,
-      "learning_rate": 9.856068717376533e-06,
-      "loss": 0.3765,
-      "step": 4363
-    },
-    {
-      "epoch": 0.28521011698581794,
-      "grad_norm": 0.45166146755218506,
-      "learning_rate": 9.855985525085035e-06,
-      "loss": 0.3369,
-      "step": 4364
-    },
-    {
-      "epoch": 0.28527547219136007,
-      "grad_norm": 0.508328914642334,
-      "learning_rate": 9.855902309109214e-06,
-      "loss": 0.4233,
-      "step": 4365
-    },
-    {
-      "epoch": 0.28534082739690214,
-      "grad_norm": 0.47016677260398865,
-      "learning_rate": 9.855819069449475e-06,
-      "loss": 0.3705,
-      "step": 4366
-    },
-    {
-      "epoch": 0.28540618260244427,
-      "grad_norm": 0.4435829818248749,
-      "learning_rate": 9.855735806106226e-06,
-      "loss": 0.3818,
-      "step": 4367
-    },
-    {
-      "epoch": 0.2854715378079864,
-      "grad_norm": 0.4735620319843292,
-      "learning_rate": 9.855652519079867e-06,
-      "loss": 0.4362,
-      "step": 4368
-    },
-    {
-      "epoch": 0.2855368930135285,
-      "grad_norm": 0.4764656722545624,
-      "learning_rate": 9.855569208370813e-06,
-      "loss": 0.3775,
-      "step": 4369
-    },
-    {
-      "epoch": 0.28560224821907065,
-      "grad_norm": 0.4751443862915039,
-      "learning_rate": 9.855485873979464e-06,
-      "loss": 0.3796,
-      "step": 4370
-    },
-    {
-      "epoch": 0.2856676034246128,
-      "grad_norm": 0.5028964281082153,
-      "learning_rate": 9.855402515906229e-06,
-      "loss": 0.4524,
-      "step": 4371
-    },
-    {
-      "epoch": 0.2857329586301549,
-      "grad_norm": 0.4496062099933624,
-      "learning_rate": 9.855319134151514e-06,
-      "loss": 0.3677,
-      "step": 4372
-    },
-    {
-      "epoch": 0.28579831383569704,
-      "grad_norm": 0.47241446375846863,
-      "learning_rate": 9.855235728715723e-06,
-      "loss": 0.4358,
-      "step": 4373
-    },
-    {
-      "epoch": 0.2858636690412391,
-      "grad_norm": 0.44672802090644836,
-      "learning_rate": 9.855152299599267e-06,
-      "loss": 0.4042,
-      "step": 4374
-    },
-    {
-      "epoch": 0.28592902424678124,
-      "grad_norm": 0.6679685115814209,
-      "learning_rate": 9.855068846802552e-06,
-      "loss": 0.3636,
-      "step": 4375
-    },
-    {
-      "epoch": 0.28599437945232337,
-      "grad_norm": 0.4665829539299011,
-      "learning_rate": 9.854985370325983e-06,
-      "loss": 0.4069,
-      "step": 4376
-    },
-    {
-      "epoch": 0.2860597346578655,
-      "grad_norm": 0.5016047358512878,
-      "learning_rate": 9.854901870169968e-06,
-      "loss": 0.4619,
-      "step": 4377
-    },
-    {
-      "epoch": 0.2861250898634076,
-      "grad_norm": 0.536938488483429,
-      "learning_rate": 9.854818346334916e-06,
-      "loss": 0.4359,
-      "step": 4378
-    },
-    {
-      "epoch": 0.28619044506894975,
-      "grad_norm": 0.4891755282878876,
-      "learning_rate": 9.854734798821233e-06,
-      "loss": 0.3784,
-      "step": 4379
-    },
-    {
-      "epoch": 0.2862558002744919,
-      "grad_norm": 0.44296059012413025,
-      "learning_rate": 9.854651227629325e-06,
-      "loss": 0.3682,
-      "step": 4380
-    },
-    {
-      "epoch": 0.286321155480034,
-      "grad_norm": 0.5143328905105591,
-      "learning_rate": 9.854567632759604e-06,
-      "loss": 0.4594,
-      "step": 4381
-    },
-    {
-      "epoch": 0.2863865106855761,
-      "grad_norm": 0.48886820673942566,
-      "learning_rate": 9.854484014212472e-06,
-      "loss": 0.4223,
-      "step": 4382
-    },
-    {
-      "epoch": 0.2864518658911182,
-      "grad_norm": 0.49212801456451416,
-      "learning_rate": 9.854400371988342e-06,
-      "loss": 0.4465,
-      "step": 4383
-    },
-    {
-      "epoch": 0.28651722109666034,
-      "grad_norm": 0.49502620100975037,
-      "learning_rate": 9.854316706087619e-06,
-      "loss": 0.4315,
-      "step": 4384
-    },
-    {
-      "epoch": 0.28658257630220246,
-      "grad_norm": 0.5504209995269775,
-      "learning_rate": 9.854233016510712e-06,
-      "loss": 0.5186,
-      "step": 4385
-    },
-    {
-      "epoch": 0.2866479315077446,
-      "grad_norm": 0.49391207098960876,
-      "learning_rate": 9.854149303258027e-06,
-      "loss": 0.434,
-      "step": 4386
-    },
-    {
-      "epoch": 0.2867132867132867,
-      "grad_norm": 0.4450891613960266,
-      "learning_rate": 9.854065566329976e-06,
-      "loss": 0.3643,
-      "step": 4387
-    },
-    {
-      "epoch": 0.28677864191882885,
-      "grad_norm": 0.4998491406440735,
-      "learning_rate": 9.853981805726966e-06,
-      "loss": 0.443,
-      "step": 4388
-    },
-    {
-      "epoch": 0.286843997124371,
-      "grad_norm": 0.42358624935150146,
-      "learning_rate": 9.853898021449404e-06,
-      "loss": 0.3353,
-      "step": 4389
-    },
-    {
-      "epoch": 0.2869093523299131,
-      "grad_norm": 0.47219541668891907,
-      "learning_rate": 9.853814213497699e-06,
-      "loss": 0.4451,
-      "step": 4390
-    },
-    {
-      "epoch": 0.2869747075354552,
-      "grad_norm": 0.46333885192871094,
-      "learning_rate": 9.853730381872262e-06,
-      "loss": 0.4005,
-      "step": 4391
-    },
-    {
-      "epoch": 0.2870400627409973,
-      "grad_norm": 0.4804172217845917,
-      "learning_rate": 9.853646526573501e-06,
-      "loss": 0.3934,
-      "step": 4392
-    },
-    {
-      "epoch": 0.28710541794653943,
-      "grad_norm": 0.5372201800346375,
-      "learning_rate": 9.853562647601823e-06,
-      "loss": 0.5348,
-      "step": 4393
-    },
-    {
-      "epoch": 0.28717077315208156,
-      "grad_norm": 0.4997507631778717,
-      "learning_rate": 9.853478744957638e-06,
-      "loss": 0.4746,
-      "step": 4394
-    },
-    {
-      "epoch": 0.2872361283576237,
-      "grad_norm": 0.5150099992752075,
-      "learning_rate": 9.853394818641358e-06,
-      "loss": 0.4844,
-      "step": 4395
-    },
-    {
-      "epoch": 0.2873014835631658,
-      "grad_norm": 0.4945598840713501,
-      "learning_rate": 9.853310868653389e-06,
-      "loss": 0.4405,
-      "step": 4396
-    },
-    {
-      "epoch": 0.28736683876870794,
-      "grad_norm": 0.45849746465682983,
-      "learning_rate": 9.853226894994142e-06,
-      "loss": 0.3829,
-      "step": 4397
-    },
-    {
-      "epoch": 0.28743219397425007,
-      "grad_norm": 0.48079851269721985,
-      "learning_rate": 9.853142897664024e-06,
-      "loss": 0.4288,
-      "step": 4398
-    },
-    {
-      "epoch": 0.28749754917979214,
-      "grad_norm": 0.4650821089744568,
-      "learning_rate": 9.853058876663448e-06,
-      "loss": 0.371,
-      "step": 4399
-    },
-    {
-      "epoch": 0.2875629043853343,
-      "grad_norm": 0.49603012204170227,
-      "learning_rate": 9.852974831992823e-06,
-      "loss": 0.4216,
-      "step": 4400
-    },
-    {
-      "epoch": 0.2876282595908764,
-      "grad_norm": 0.45978420972824097,
-      "learning_rate": 9.852890763652558e-06,
-      "loss": 0.4004,
-      "step": 4401
-    },
-    {
-      "epoch": 0.28769361479641853,
-      "grad_norm": 0.41766291856765747,
-      "learning_rate": 9.852806671643064e-06,
-      "loss": 0.3755,
-      "step": 4402
-    },
-    {
-      "epoch": 0.28775897000196066,
-      "grad_norm": 0.4829985201358795,
-      "learning_rate": 9.85272255596475e-06,
-      "loss": 0.379,
-      "step": 4403
-    },
-    {
-      "epoch": 0.2878243252075028,
-      "grad_norm": 0.44976744055747986,
-      "learning_rate": 9.852638416618029e-06,
-      "loss": 0.392,
-      "step": 4404
-    },
-    {
-      "epoch": 0.2878896804130449,
-      "grad_norm": 0.49071407318115234,
-      "learning_rate": 9.852554253603308e-06,
-      "loss": 0.5037,
-      "step": 4405
-    },
-    {
-      "epoch": 0.28795503561858704,
-      "grad_norm": 0.522971510887146,
-      "learning_rate": 9.852470066920999e-06,
-      "loss": 0.4007,
-      "step": 4406
-    },
-    {
-      "epoch": 0.28802039082412917,
-      "grad_norm": 0.4969806969165802,
-      "learning_rate": 9.852385856571512e-06,
-      "loss": 0.4441,
-      "step": 4407
-    },
-    {
-      "epoch": 0.28808574602967124,
-      "grad_norm": 0.44576096534729004,
-      "learning_rate": 9.852301622555259e-06,
-      "loss": 0.3568,
-      "step": 4408
-    },
-    {
-      "epoch": 0.28815110123521337,
-      "grad_norm": 0.49539312720298767,
-      "learning_rate": 9.852217364872649e-06,
-      "loss": 0.4528,
-      "step": 4409
-    },
-    {
-      "epoch": 0.2882164564407555,
-      "grad_norm": 0.46742770075798035,
-      "learning_rate": 9.852133083524093e-06,
-      "loss": 0.3828,
-      "step": 4410
-    },
-    {
-      "epoch": 0.2882818116462976,
-      "grad_norm": 0.4757823348045349,
-      "learning_rate": 9.852048778510006e-06,
-      "loss": 0.3925,
-      "step": 4411
-    },
-    {
-      "epoch": 0.28834716685183975,
-      "grad_norm": 0.44201409816741943,
-      "learning_rate": 9.851964449830794e-06,
-      "loss": 0.3768,
-      "step": 4412
-    },
-    {
-      "epoch": 0.2884125220573819,
-      "grad_norm": 0.47092169523239136,
-      "learning_rate": 9.85188009748687e-06,
-      "loss": 0.4285,
-      "step": 4413
-    },
-    {
-      "epoch": 0.288477877262924,
-      "grad_norm": 0.4804210066795349,
-      "learning_rate": 9.851795721478647e-06,
-      "loss": 0.4168,
-      "step": 4414
-    },
-    {
-      "epoch": 0.28854323246846614,
-      "grad_norm": 0.43654242157936096,
-      "learning_rate": 9.851711321806537e-06,
-      "loss": 0.3518,
-      "step": 4415
-    },
-    {
-      "epoch": 0.2886085876740082,
-      "grad_norm": 0.4612194001674652,
-      "learning_rate": 9.851626898470948e-06,
-      "loss": 0.3957,
-      "step": 4416
-    },
-    {
-      "epoch": 0.28867394287955034,
-      "grad_norm": 0.46339964866638184,
-      "learning_rate": 9.851542451472293e-06,
-      "loss": 0.4083,
-      "step": 4417
-    },
-    {
-      "epoch": 0.28873929808509247,
-      "grad_norm": 0.4413035213947296,
-      "learning_rate": 9.851457980810988e-06,
-      "loss": 0.3679,
-      "step": 4418
-    },
-    {
-      "epoch": 0.2888046532906346,
-      "grad_norm": 0.4670655131340027,
-      "learning_rate": 9.851373486487439e-06,
-      "loss": 0.3959,
-      "step": 4419
-    },
-    {
-      "epoch": 0.2888700084961767,
-      "grad_norm": 0.5646473169326782,
-      "learning_rate": 9.85128896850206e-06,
-      "loss": 0.4123,
-      "step": 4420
-    },
-    {
-      "epoch": 0.28893536370171885,
-      "grad_norm": 0.49724411964416504,
-      "learning_rate": 9.851204426855267e-06,
-      "loss": 0.4474,
-      "step": 4421
-    },
-    {
-      "epoch": 0.289000718907261,
-      "grad_norm": 0.5162045955657959,
-      "learning_rate": 9.851119861547467e-06,
-      "loss": 0.4092,
-      "step": 4422
-    },
-    {
-      "epoch": 0.2890660741128031,
-      "grad_norm": 0.47607100009918213,
-      "learning_rate": 9.851035272579077e-06,
-      "loss": 0.4045,
-      "step": 4423
-    },
-    {
-      "epoch": 0.2891314293183452,
-      "grad_norm": 0.4476981461048126,
-      "learning_rate": 9.850950659950506e-06,
-      "loss": 0.375,
-      "step": 4424
-    },
-    {
-      "epoch": 0.2891967845238873,
-      "grad_norm": 0.4492538273334503,
-      "learning_rate": 9.850866023662168e-06,
-      "loss": 0.3872,
-      "step": 4425
-    },
-    {
-      "epoch": 0.28926213972942943,
-      "grad_norm": 0.45858004689216614,
-      "learning_rate": 9.850781363714477e-06,
-      "loss": 0.3755,
-      "step": 4426
-    },
-    {
-      "epoch": 0.28932749493497156,
-      "grad_norm": 0.4616815447807312,
-      "learning_rate": 9.850696680107844e-06,
-      "loss": 0.3768,
-      "step": 4427
-    },
-    {
-      "epoch": 0.2893928501405137,
-      "grad_norm": 0.42143869400024414,
-      "learning_rate": 9.850611972842682e-06,
-      "loss": 0.3498,
-      "step": 4428
-    },
-    {
-      "epoch": 0.2894582053460558,
-      "grad_norm": 0.4461817443370819,
-      "learning_rate": 9.850527241919405e-06,
-      "loss": 0.4022,
-      "step": 4429
-    },
-    {
-      "epoch": 0.28952356055159795,
-      "grad_norm": 0.48425930738449097,
-      "learning_rate": 9.850442487338427e-06,
-      "loss": 0.4367,
-      "step": 4430
-    },
-    {
-      "epoch": 0.2895889157571401,
-      "grad_norm": 0.4955015480518341,
-      "learning_rate": 9.850357709100163e-06,
-      "loss": 0.3992,
-      "step": 4431
-    },
-    {
-      "epoch": 0.2896542709626822,
-      "grad_norm": 0.47167977690696716,
-      "learning_rate": 9.85027290720502e-06,
-      "loss": 0.3774,
-      "step": 4432
-    },
-    {
-      "epoch": 0.2897196261682243,
-      "grad_norm": 0.4849473834037781,
-      "learning_rate": 9.850188081653419e-06,
-      "loss": 0.4454,
-      "step": 4433
-    },
-    {
-      "epoch": 0.2897849813737664,
-      "grad_norm": 0.5112374424934387,
-      "learning_rate": 9.850103232445769e-06,
-      "loss": 0.4332,
-      "step": 4434
-    },
-    {
-      "epoch": 0.28985033657930853,
-      "grad_norm": 0.507053792476654,
-      "learning_rate": 9.850018359582484e-06,
-      "loss": 0.3938,
-      "step": 4435
-    },
-    {
-      "epoch": 0.28991569178485066,
-      "grad_norm": 0.5383029580116272,
-      "learning_rate": 9.849933463063982e-06,
-      "loss": 0.4272,
-      "step": 4436
-    },
-    {
-      "epoch": 0.2899810469903928,
-      "grad_norm": 0.4883407652378082,
-      "learning_rate": 9.849848542890673e-06,
-      "loss": 0.4145,
-      "step": 4437
-    },
-    {
-      "epoch": 0.2900464021959349,
-      "grad_norm": 0.4537391662597656,
-      "learning_rate": 9.849763599062972e-06,
-      "loss": 0.3592,
-      "step": 4438
-    },
-    {
-      "epoch": 0.29011175740147704,
-      "grad_norm": 0.4909239113330841,
-      "learning_rate": 9.849678631581294e-06,
-      "loss": 0.398,
-      "step": 4439
-    },
-    {
-      "epoch": 0.29017711260701917,
-      "grad_norm": 0.4724974036216736,
-      "learning_rate": 9.849593640446054e-06,
-      "loss": 0.3663,
-      "step": 4440
-    },
-    {
-      "epoch": 0.29024246781256124,
-      "grad_norm": 0.5158087015151978,
-      "learning_rate": 9.849508625657666e-06,
-      "loss": 0.4682,
-      "step": 4441
-    },
-    {
-      "epoch": 0.2903078230181034,
-      "grad_norm": 0.4384050667285919,
-      "learning_rate": 9.849423587216543e-06,
-      "loss": 0.3911,
-      "step": 4442
-    },
-    {
-      "epoch": 0.2903731782236455,
-      "grad_norm": 0.4931519627571106,
-      "learning_rate": 9.849338525123102e-06,
-      "loss": 0.4265,
-      "step": 4443
-    },
-    {
-      "epoch": 0.29043853342918763,
-      "grad_norm": 0.4780293405056,
-      "learning_rate": 9.849253439377757e-06,
-      "loss": 0.4529,
-      "step": 4444
-    },
-    {
-      "epoch": 0.29050388863472976,
-      "grad_norm": 0.5410692095756531,
-      "learning_rate": 9.849168329980922e-06,
-      "loss": 0.489,
-      "step": 4445
-    },
-    {
-      "epoch": 0.2905692438402719,
-      "grad_norm": 0.5165309309959412,
-      "learning_rate": 9.849083196933014e-06,
-      "loss": 0.3583,
-      "step": 4446
-    },
-    {
-      "epoch": 0.290634599045814,
-      "grad_norm": 0.49044737219810486,
-      "learning_rate": 9.848998040234449e-06,
-      "loss": 0.4111,
-      "step": 4447
-    },
-    {
-      "epoch": 0.29069995425135614,
-      "grad_norm": 0.457153856754303,
-      "learning_rate": 9.848912859885638e-06,
-      "loss": 0.3878,
-      "step": 4448
-    },
-    {
-      "epoch": 0.29076530945689827,
-      "grad_norm": 0.4699672758579254,
-      "learning_rate": 9.848827655887002e-06,
-      "loss": 0.3797,
-      "step": 4449
-    },
-    {
-      "epoch": 0.29083066466244034,
-      "grad_norm": 0.4563548266887665,
-      "learning_rate": 9.848742428238951e-06,
-      "loss": 0.3935,
-      "step": 4450
-    },
-    {
-      "epoch": 0.29089601986798247,
-      "grad_norm": 0.4894541800022125,
-      "learning_rate": 9.848657176941905e-06,
-      "loss": 0.4398,
-      "step": 4451
-    },
-    {
-      "epoch": 0.2909613750735246,
-      "grad_norm": 0.4865151047706604,
-      "learning_rate": 9.848571901996278e-06,
-      "loss": 0.4178,
-      "step": 4452
-    },
-    {
-      "epoch": 0.2910267302790667,
-      "grad_norm": 0.45140793919563293,
-      "learning_rate": 9.848486603402484e-06,
-      "loss": 0.3697,
-      "step": 4453
-    },
-    {
-      "epoch": 0.29109208548460885,
-      "grad_norm": 0.4754602909088135,
-      "learning_rate": 9.848401281160943e-06,
-      "loss": 0.3992,
-      "step": 4454
-    },
-    {
-      "epoch": 0.291157440690151,
-      "grad_norm": 0.44573870301246643,
-      "learning_rate": 9.84831593527207e-06,
-      "loss": 0.4038,
-      "step": 4455
-    },
-    {
-      "epoch": 0.2912227958956931,
-      "grad_norm": 0.4916362464427948,
-      "learning_rate": 9.848230565736279e-06,
-      "loss": 0.4157,
-      "step": 4456
-    },
-    {
-      "epoch": 0.29128815110123524,
-      "grad_norm": 0.4606829881668091,
-      "learning_rate": 9.848145172553989e-06,
-      "loss": 0.4002,
-      "step": 4457
-    },
-    {
-      "epoch": 0.2913535063067773,
-      "grad_norm": 0.5081844329833984,
-      "learning_rate": 9.848059755725617e-06,
-      "loss": 0.4217,
-      "step": 4458
-    },
-    {
-      "epoch": 0.29141886151231944,
-      "grad_norm": 0.48040980100631714,
-      "learning_rate": 9.847974315251575e-06,
-      "loss": 0.4508,
-      "step": 4459
-    },
-    {
-      "epoch": 0.29148421671786157,
-      "grad_norm": 0.4761035740375519,
-      "learning_rate": 9.847888851132283e-06,
-      "loss": 0.3871,
-      "step": 4460
-    },
-    {
-      "epoch": 0.2915495719234037,
-      "grad_norm": 0.5370839238166809,
-      "learning_rate": 9.84780336336816e-06,
-      "loss": 0.4498,
-      "step": 4461
-    },
-    {
-      "epoch": 0.2916149271289458,
-      "grad_norm": 0.4452376961708069,
-      "learning_rate": 9.84771785195962e-06,
-      "loss": 0.4065,
-      "step": 4462
-    },
-    {
-      "epoch": 0.29168028233448795,
-      "grad_norm": 0.45047760009765625,
-      "learning_rate": 9.847632316907079e-06,
-      "loss": 0.3519,
-      "step": 4463
-    },
-    {
-      "epoch": 0.2917456375400301,
-      "grad_norm": 0.48897784948349,
-      "learning_rate": 9.847546758210956e-06,
-      "loss": 0.4454,
-      "step": 4464
-    },
-    {
-      "epoch": 0.2918109927455722,
-      "grad_norm": 0.49426940083503723,
-      "learning_rate": 9.847461175871669e-06,
-      "loss": 0.3911,
-      "step": 4465
-    },
-    {
-      "epoch": 0.2918763479511143,
-      "grad_norm": 0.4327999949455261,
-      "learning_rate": 9.847375569889635e-06,
-      "loss": 0.34,
-      "step": 4466
-    },
-    {
-      "epoch": 0.2919417031566564,
-      "grad_norm": 0.4581160545349121,
-      "learning_rate": 9.84728994026527e-06,
-      "loss": 0.3525,
-      "step": 4467
-    },
-    {
-      "epoch": 0.29200705836219853,
-      "grad_norm": 0.4722375273704529,
-      "learning_rate": 9.847204286998993e-06,
-      "loss": 0.3694,
-      "step": 4468
-    },
-    {
-      "epoch": 0.29207241356774066,
-      "grad_norm": 0.4604976773262024,
-      "learning_rate": 9.84711861009122e-06,
-      "loss": 0.3839,
-      "step": 4469
-    },
-    {
-      "epoch": 0.2921377687732828,
-      "grad_norm": 0.49679136276245117,
-      "learning_rate": 9.847032909542373e-06,
-      "loss": 0.4079,
-      "step": 4470
-    },
-    {
-      "epoch": 0.2922031239788249,
-      "grad_norm": 0.4918653964996338,
-      "learning_rate": 9.846947185352865e-06,
-      "loss": 0.4231,
-      "step": 4471
-    },
-    {
-      "epoch": 0.29226847918436705,
-      "grad_norm": 0.4945702850818634,
-      "learning_rate": 9.84686143752312e-06,
-      "loss": 0.4672,
-      "step": 4472
-    },
-    {
-      "epoch": 0.2923338343899092,
-      "grad_norm": 0.4729475975036621,
-      "learning_rate": 9.846775666053549e-06,
-      "loss": 0.4161,
-      "step": 4473
-    },
-    {
-      "epoch": 0.2923991895954513,
-      "grad_norm": 0.48046860098838806,
-      "learning_rate": 9.846689870944574e-06,
-      "loss": 0.4461,
-      "step": 4474
-    },
-    {
-      "epoch": 0.2924645448009934,
-      "grad_norm": 0.4358729422092438,
-      "learning_rate": 9.846604052196616e-06,
-      "loss": 0.3567,
-      "step": 4475
-    },
-    {
-      "epoch": 0.2925299000065355,
-      "grad_norm": 0.4827226400375366,
-      "learning_rate": 9.846518209810089e-06,
-      "loss": 0.4204,
-      "step": 4476
-    },
-    {
-      "epoch": 0.29259525521207763,
-      "grad_norm": 0.5996066927909851,
-      "learning_rate": 9.846432343785415e-06,
-      "loss": 0.4532,
-      "step": 4477
-    },
-    {
-      "epoch": 0.29266061041761976,
-      "grad_norm": 0.4725746512413025,
-      "learning_rate": 9.84634645412301e-06,
-      "loss": 0.4161,
-      "step": 4478
-    },
-    {
-      "epoch": 0.2927259656231619,
-      "grad_norm": 0.47878170013427734,
-      "learning_rate": 9.846260540823296e-06,
-      "loss": 0.4186,
-      "step": 4479
-    },
-    {
-      "epoch": 0.292791320828704,
-      "grad_norm": 0.49192699790000916,
-      "learning_rate": 9.84617460388669e-06,
-      "loss": 0.4375,
-      "step": 4480
-    },
-    {
-      "epoch": 0.29285667603424614,
-      "grad_norm": 0.48404741287231445,
-      "learning_rate": 9.84608864331361e-06,
-      "loss": 0.4356,
-      "step": 4481
-    },
-    {
-      "epoch": 0.29292203123978827,
-      "grad_norm": 0.4758375287055969,
-      "learning_rate": 9.846002659104479e-06,
-      "loss": 0.4377,
-      "step": 4482
-    },
-    {
-      "epoch": 0.29298738644533034,
-      "grad_norm": 0.4543372392654419,
-      "learning_rate": 9.845916651259713e-06,
-      "loss": 0.4154,
-      "step": 4483
-    },
-    {
-      "epoch": 0.2930527416508725,
-      "grad_norm": 0.48143458366394043,
-      "learning_rate": 9.845830619779732e-06,
-      "loss": 0.4141,
-      "step": 4484
-    },
-    {
-      "epoch": 0.2931180968564146,
-      "grad_norm": 0.4619728922843933,
-      "learning_rate": 9.845744564664958e-06,
-      "loss": 0.3974,
-      "step": 4485
-    },
-    {
-      "epoch": 0.29318345206195673,
-      "grad_norm": 0.4468969702720642,
-      "learning_rate": 9.845658485915808e-06,
-      "loss": 0.3851,
-      "step": 4486
-    },
-    {
-      "epoch": 0.29324880726749886,
-      "grad_norm": 0.4667202830314636,
-      "learning_rate": 9.845572383532703e-06,
-      "loss": 0.3952,
-      "step": 4487
-    },
-    {
-      "epoch": 0.293314162473041,
-      "grad_norm": 0.49085402488708496,
-      "learning_rate": 9.845486257516064e-06,
-      "loss": 0.446,
-      "step": 4488
-    },
-    {
-      "epoch": 0.2933795176785831,
-      "grad_norm": 0.48247113823890686,
-      "learning_rate": 9.845400107866307e-06,
-      "loss": 0.4061,
-      "step": 4489
-    },
-    {
-      "epoch": 0.29344487288412524,
-      "grad_norm": 0.5165753364562988,
-      "learning_rate": 9.845313934583858e-06,
-      "loss": 0.4231,
-      "step": 4490
-    },
-    {
-      "epoch": 0.29351022808966737,
-      "grad_norm": 0.4774547815322876,
-      "learning_rate": 9.845227737669134e-06,
-      "loss": 0.4169,
-      "step": 4491
-    },
-    {
-      "epoch": 0.29357558329520944,
-      "grad_norm": 0.4724103510379791,
-      "learning_rate": 9.845141517122554e-06,
-      "loss": 0.3999,
-      "step": 4492
-    },
-    {
-      "epoch": 0.29364093850075157,
-      "grad_norm": 0.468107670545578,
-      "learning_rate": 9.845055272944541e-06,
-      "loss": 0.3857,
-      "step": 4493
-    },
-    {
-      "epoch": 0.2937062937062937,
-      "grad_norm": 0.4065115451812744,
-      "learning_rate": 9.844969005135517e-06,
-      "loss": 0.3096,
-      "step": 4494
-    },
-    {
-      "epoch": 0.2937716489118358,
-      "grad_norm": 0.48373425006866455,
-      "learning_rate": 9.844882713695898e-06,
-      "loss": 0.4499,
-      "step": 4495
-    },
-    {
-      "epoch": 0.29383700411737795,
-      "grad_norm": 0.4965156316757202,
-      "learning_rate": 9.844796398626109e-06,
-      "loss": 0.4481,
-      "step": 4496
-    },
-    {
-      "epoch": 0.2939023593229201,
-      "grad_norm": 0.4595758020877838,
-      "learning_rate": 9.84471005992657e-06,
-      "loss": 0.3943,
-      "step": 4497
-    },
-    {
-      "epoch": 0.2939677145284622,
-      "grad_norm": 0.4767528474330902,
-      "learning_rate": 9.8446236975977e-06,
-      "loss": 0.3675,
-      "step": 4498
-    },
-    {
-      "epoch": 0.29403306973400434,
-      "grad_norm": 0.6278597712516785,
-      "learning_rate": 9.844537311639923e-06,
-      "loss": 0.429,
-      "step": 4499
-    },
-    {
-      "epoch": 0.2940984249395464,
-      "grad_norm": 0.4658062756061554,
-      "learning_rate": 9.84445090205366e-06,
-      "loss": 0.4168,
-      "step": 4500
-    },
-    {
-      "epoch": 0.29416378014508854,
-      "grad_norm": 0.45288029313087463,
-      "learning_rate": 9.84436446883933e-06,
-      "loss": 0.3918,
-      "step": 4501
-    },
-    {
-      "epoch": 0.29422913535063067,
-      "grad_norm": 0.5236877799034119,
-      "learning_rate": 9.844278011997357e-06,
-      "loss": 0.4589,
-      "step": 4502
-    },
-    {
-      "epoch": 0.2942944905561728,
-      "grad_norm": 0.47641849517822266,
-      "learning_rate": 9.844191531528162e-06,
-      "loss": 0.4035,
-      "step": 4503
-    },
-    {
-      "epoch": 0.2943598457617149,
-      "grad_norm": 0.47187209129333496,
-      "learning_rate": 9.844105027432166e-06,
-      "loss": 0.3838,
-      "step": 4504
-    },
-    {
-      "epoch": 0.29442520096725705,
-      "grad_norm": 0.5041341185569763,
-      "learning_rate": 9.844018499709793e-06,
-      "loss": 0.401,
-      "step": 4505
-    },
-    {
-      "epoch": 0.2944905561727992,
-      "grad_norm": 0.4318012297153473,
-      "learning_rate": 9.843931948361463e-06,
-      "loss": 0.3826,
-      "step": 4506
-    },
-    {
-      "epoch": 0.2945559113783413,
-      "grad_norm": 0.4853137135505676,
-      "learning_rate": 9.843845373387597e-06,
-      "loss": 0.3823,
-      "step": 4507
-    },
-    {
-      "epoch": 0.2946212665838834,
-      "grad_norm": 0.4941728711128235,
-      "learning_rate": 9.84375877478862e-06,
-      "loss": 0.3931,
-      "step": 4508
-    },
-    {
-      "epoch": 0.2946866217894255,
-      "grad_norm": 0.49251747131347656,
-      "learning_rate": 9.843672152564956e-06,
-      "loss": 0.4658,
-      "step": 4509
-    },
-    {
-      "epoch": 0.29475197699496763,
-      "grad_norm": 0.4598727226257324,
-      "learning_rate": 9.84358550671702e-06,
-      "loss": 0.4298,
-      "step": 4510
-    },
-    {
-      "epoch": 0.29481733220050976,
-      "grad_norm": 0.464657187461853,
-      "learning_rate": 9.843498837245244e-06,
-      "loss": 0.3822,
-      "step": 4511
-    },
-    {
-      "epoch": 0.2948826874060519,
-      "grad_norm": 0.4527546167373657,
-      "learning_rate": 9.843412144150045e-06,
-      "loss": 0.4042,
-      "step": 4512
-    },
-    {
-      "epoch": 0.294948042611594,
-      "grad_norm": 0.46141108870506287,
-      "learning_rate": 9.843325427431847e-06,
-      "loss": 0.4183,
-      "step": 4513
-    },
-    {
-      "epoch": 0.29501339781713615,
-      "grad_norm": 0.48721104860305786,
-      "learning_rate": 9.843238687091072e-06,
-      "loss": 0.4016,
-      "step": 4514
-    },
-    {
-      "epoch": 0.2950787530226783,
-      "grad_norm": 0.458452433347702,
-      "learning_rate": 9.843151923128146e-06,
-      "loss": 0.3895,
-      "step": 4515
-    },
-    {
-      "epoch": 0.2951441082282204,
-      "grad_norm": 0.46637964248657227,
-      "learning_rate": 9.843065135543488e-06,
-      "loss": 0.416,
-      "step": 4516
-    },
-    {
-      "epoch": 0.2952094634337625,
-      "grad_norm": 0.4475843012332916,
-      "learning_rate": 9.842978324337525e-06,
-      "loss": 0.3719,
-      "step": 4517
-    },
-    {
-      "epoch": 0.2952748186393046,
-      "grad_norm": 0.4849666357040405,
-      "learning_rate": 9.842891489510678e-06,
-      "loss": 0.4102,
-      "step": 4518
-    },
-    {
-      "epoch": 0.29534017384484673,
-      "grad_norm": 0.4784989655017853,
-      "learning_rate": 9.842804631063374e-06,
-      "loss": 0.3904,
-      "step": 4519
-    },
-    {
-      "epoch": 0.29540552905038886,
-      "grad_norm": 0.4642713665962219,
-      "learning_rate": 9.84271774899603e-06,
-      "loss": 0.3766,
-      "step": 4520
-    },
-    {
-      "epoch": 0.295470884255931,
-      "grad_norm": 0.466028094291687,
-      "learning_rate": 9.842630843309077e-06,
-      "loss": 0.3761,
-      "step": 4521
-    },
-    {
-      "epoch": 0.2955362394614731,
-      "grad_norm": 0.4432559907436371,
-      "learning_rate": 9.842543914002935e-06,
-      "loss": 0.3765,
-      "step": 4522
-    },
-    {
-      "epoch": 0.29560159466701524,
-      "grad_norm": 0.4879373610019684,
-      "learning_rate": 9.842456961078028e-06,
-      "loss": 0.4192,
-      "step": 4523
-    },
-    {
-      "epoch": 0.29566694987255737,
-      "grad_norm": 0.4954962134361267,
-      "learning_rate": 9.842369984534781e-06,
-      "loss": 0.3676,
-      "step": 4524
-    },
-    {
-      "epoch": 0.29573230507809944,
-      "grad_norm": 0.5037001967430115,
-      "learning_rate": 9.842282984373618e-06,
-      "loss": 0.456,
-      "step": 4525
-    },
-    {
-      "epoch": 0.2957976602836416,
-      "grad_norm": 0.4720968008041382,
-      "learning_rate": 9.842195960594965e-06,
-      "loss": 0.3599,
-      "step": 4526
-    },
-    {
-      "epoch": 0.2958630154891837,
-      "grad_norm": 0.47066187858581543,
-      "learning_rate": 9.842108913199241e-06,
-      "loss": 0.389,
-      "step": 4527
-    },
-    {
-      "epoch": 0.29592837069472583,
-      "grad_norm": 0.4656980335712433,
-      "learning_rate": 9.842021842186878e-06,
-      "loss": 0.4234,
-      "step": 4528
-    },
-    {
-      "epoch": 0.29599372590026796,
-      "grad_norm": 0.49611374735832214,
-      "learning_rate": 9.841934747558295e-06,
-      "loss": 0.4452,
-      "step": 4529
-    },
-    {
-      "epoch": 0.2960590811058101,
-      "grad_norm": 0.4645419716835022,
-      "learning_rate": 9.84184762931392e-06,
-      "loss": 0.369,
-      "step": 4530
-    },
-    {
-      "epoch": 0.2961244363113522,
-      "grad_norm": 0.4592958390712738,
-      "learning_rate": 9.841760487454176e-06,
-      "loss": 0.3984,
-      "step": 4531
-    },
-    {
-      "epoch": 0.29618979151689434,
-      "grad_norm": 0.47864311933517456,
-      "learning_rate": 9.84167332197949e-06,
-      "loss": 0.4141,
-      "step": 4532
-    },
-    {
-      "epoch": 0.29625514672243647,
-      "grad_norm": 0.44153285026550293,
-      "learning_rate": 9.841586132890285e-06,
-      "loss": 0.3648,
-      "step": 4533
-    },
-    {
-      "epoch": 0.29632050192797854,
-      "grad_norm": 0.44436541199684143,
-      "learning_rate": 9.841498920186987e-06,
-      "loss": 0.3942,
-      "step": 4534
-    },
-    {
-      "epoch": 0.29638585713352067,
-      "grad_norm": 0.4569571614265442,
-      "learning_rate": 9.841411683870021e-06,
-      "loss": 0.393,
-      "step": 4535
-    },
-    {
-      "epoch": 0.2964512123390628,
-      "grad_norm": 0.4499158263206482,
-      "learning_rate": 9.841324423939815e-06,
-      "loss": 0.3645,
-      "step": 4536
-    },
-    {
-      "epoch": 0.2965165675446049,
-      "grad_norm": 0.5056281685829163,
-      "learning_rate": 9.841237140396791e-06,
-      "loss": 0.4122,
-      "step": 4537
-    },
-    {
-      "epoch": 0.29658192275014705,
-      "grad_norm": 0.5346924662590027,
-      "learning_rate": 9.841149833241378e-06,
-      "loss": 0.4697,
-      "step": 4538
-    },
-    {
-      "epoch": 0.2966472779556892,
-      "grad_norm": 0.4869162440299988,
-      "learning_rate": 9.841062502473997e-06,
-      "loss": 0.4437,
-      "step": 4539
-    },
-    {
-      "epoch": 0.2967126331612313,
-      "grad_norm": 0.4643200635910034,
-      "learning_rate": 9.840975148095079e-06,
-      "loss": 0.4113,
-      "step": 4540
-    },
-    {
-      "epoch": 0.29677798836677344,
-      "grad_norm": 0.451917827129364,
-      "learning_rate": 9.840887770105048e-06,
-      "loss": 0.3693,
-      "step": 4541
-    },
-    {
-      "epoch": 0.2968433435723155,
-      "grad_norm": 0.5004224181175232,
-      "learning_rate": 9.840800368504329e-06,
-      "loss": 0.4307,
-      "step": 4542
-    },
-    {
-      "epoch": 0.29690869877785764,
-      "grad_norm": 0.4540042579174042,
-      "learning_rate": 9.840712943293351e-06,
-      "loss": 0.4094,
-      "step": 4543
-    },
-    {
-      "epoch": 0.29697405398339977,
-      "grad_norm": 0.4906425178050995,
-      "learning_rate": 9.840625494472539e-06,
-      "loss": 0.4418,
-      "step": 4544
-    },
-    {
-      "epoch": 0.2970394091889419,
-      "grad_norm": 0.45951879024505615,
-      "learning_rate": 9.84053802204232e-06,
-      "loss": 0.4282,
-      "step": 4545
-    },
-    {
-      "epoch": 0.297104764394484,
-      "grad_norm": 0.5001217126846313,
-      "learning_rate": 9.840450526003118e-06,
-      "loss": 0.4212,
-      "step": 4546
-    },
-    {
-      "epoch": 0.29717011960002615,
-      "grad_norm": 0.4755907952785492,
-      "learning_rate": 9.840363006355365e-06,
-      "loss": 0.4088,
-      "step": 4547
-    },
-    {
-      "epoch": 0.2972354748055683,
-      "grad_norm": 0.4836277961730957,
-      "learning_rate": 9.840275463099482e-06,
-      "loss": 0.3981,
-      "step": 4548
-    },
-    {
-      "epoch": 0.2973008300111104,
-      "grad_norm": 0.44367456436157227,
-      "learning_rate": 9.840187896235899e-06,
-      "loss": 0.3569,
-      "step": 4549
-    },
-    {
-      "epoch": 0.2973661852166525,
-      "grad_norm": 0.49634233117103577,
-      "learning_rate": 9.840100305765044e-06,
-      "loss": 0.405,
-      "step": 4550
-    },
-    {
-      "epoch": 0.2974315404221946,
-      "grad_norm": 0.4834412932395935,
-      "learning_rate": 9.840012691687344e-06,
-      "loss": 0.3799,
-      "step": 4551
-    },
-    {
-      "epoch": 0.29749689562773673,
-      "grad_norm": 0.46757590770721436,
-      "learning_rate": 9.839925054003223e-06,
-      "loss": 0.3779,
-      "step": 4552
-    },
-    {
-      "epoch": 0.29756225083327886,
-      "grad_norm": 0.42607083916664124,
-      "learning_rate": 9.839837392713112e-06,
-      "loss": 0.3244,
-      "step": 4553
-    },
-    {
-      "epoch": 0.297627606038821,
-      "grad_norm": 0.4805174469947815,
-      "learning_rate": 9.839749707817437e-06,
-      "loss": 0.4165,
-      "step": 4554
-    },
-    {
-      "epoch": 0.2976929612443631,
-      "grad_norm": 0.5129186511039734,
-      "learning_rate": 9.839661999316627e-06,
-      "loss": 0.4274,
-      "step": 4555
-    },
-    {
-      "epoch": 0.29775831644990525,
-      "grad_norm": 0.5060441493988037,
-      "learning_rate": 9.839574267211107e-06,
-      "loss": 0.4402,
-      "step": 4556
-    },
-    {
-      "epoch": 0.2978236716554474,
-      "grad_norm": 0.4707791209220886,
-      "learning_rate": 9.839486511501309e-06,
-      "loss": 0.3965,
-      "step": 4557
-    },
-    {
-      "epoch": 0.2978890268609895,
-      "grad_norm": 0.4697127342224121,
-      "learning_rate": 9.839398732187657e-06,
-      "loss": 0.3934,
-      "step": 4558
-    },
-    {
-      "epoch": 0.2979543820665316,
-      "grad_norm": 0.5240148901939392,
-      "learning_rate": 9.839310929270581e-06,
-      "loss": 0.4728,
-      "step": 4559
-    },
-    {
-      "epoch": 0.2980197372720737,
-      "grad_norm": 0.5014421939849854,
-      "learning_rate": 9.839223102750511e-06,
-      "loss": 0.4011,
-      "step": 4560
-    },
-    {
-      "epoch": 0.29808509247761583,
-      "grad_norm": 0.47523826360702515,
-      "learning_rate": 9.83913525262787e-06,
-      "loss": 0.4048,
-      "step": 4561
-    },
-    {
-      "epoch": 0.29815044768315796,
-      "grad_norm": 0.5312103033065796,
-      "learning_rate": 9.839047378903093e-06,
-      "loss": 0.5076,
-      "step": 4562
-    },
-    {
-      "epoch": 0.2982158028887001,
-      "grad_norm": 0.460380494594574,
-      "learning_rate": 9.838959481576602e-06,
-      "loss": 0.3707,
-      "step": 4563
-    },
-    {
-      "epoch": 0.2982811580942422,
-      "grad_norm": 0.47403544187545776,
-      "learning_rate": 9.838871560648833e-06,
-      "loss": 0.3929,
-      "step": 4564
-    },
-    {
-      "epoch": 0.29834651329978434,
-      "grad_norm": 0.45916664600372314,
-      "learning_rate": 9.838783616120209e-06,
-      "loss": 0.3943,
-      "step": 4565
-    },
-    {
-      "epoch": 0.29841186850532647,
-      "grad_norm": 0.45200034976005554,
-      "learning_rate": 9.838695647991163e-06,
-      "loss": 0.3804,
-      "step": 4566
-    },
-    {
-      "epoch": 0.29847722371086854,
-      "grad_norm": 0.4933592975139618,
-      "learning_rate": 9.83860765626212e-06,
-      "loss": 0.4077,
-      "step": 4567
-    },
-    {
-      "epoch": 0.2985425789164107,
-      "grad_norm": 0.5133188962936401,
-      "learning_rate": 9.838519640933512e-06,
-      "loss": 0.4201,
-      "step": 4568
-    },
-    {
-      "epoch": 0.2986079341219528,
-      "grad_norm": 0.47661587595939636,
-      "learning_rate": 9.838431602005767e-06,
-      "loss": 0.4297,
-      "step": 4569
-    },
-    {
-      "epoch": 0.29867328932749493,
-      "grad_norm": 0.4640342593193054,
-      "learning_rate": 9.838343539479316e-06,
-      "loss": 0.4357,
-      "step": 4570
-    },
-    {
-      "epoch": 0.29873864453303706,
-      "grad_norm": 0.47774818539619446,
-      "learning_rate": 9.838255453354585e-06,
-      "loss": 0.4113,
-      "step": 4571
-    },
-    {
-      "epoch": 0.2988039997385792,
-      "grad_norm": 0.47215497493743896,
-      "learning_rate": 9.838167343632008e-06,
-      "loss": 0.3907,
-      "step": 4572
-    },
-    {
-      "epoch": 0.2988693549441213,
-      "grad_norm": 0.5664907693862915,
-      "learning_rate": 9.838079210312012e-06,
-      "loss": 0.4052,
-      "step": 4573
-    },
-    {
-      "epoch": 0.29893471014966344,
-      "grad_norm": 0.4794827997684479,
-      "learning_rate": 9.83799105339503e-06,
-      "loss": 0.4169,
-      "step": 4574
-    },
-    {
-      "epoch": 0.29900006535520557,
-      "grad_norm": 0.48394709825515747,
-      "learning_rate": 9.837902872881486e-06,
-      "loss": 0.4334,
-      "step": 4575
-    },
-    {
-      "epoch": 0.29906542056074764,
-      "grad_norm": 0.45113661885261536,
-      "learning_rate": 9.837814668771815e-06,
-      "loss": 0.3719,
-      "step": 4576
-    },
-    {
-      "epoch": 0.29913077576628977,
-      "grad_norm": 0.47531017661094666,
-      "learning_rate": 9.837726441066446e-06,
-      "loss": 0.3827,
-      "step": 4577
-    },
-    {
-      "epoch": 0.2991961309718319,
-      "grad_norm": 0.4978850185871124,
-      "learning_rate": 9.83763818976581e-06,
-      "loss": 0.448,
-      "step": 4578
-    },
-    {
-      "epoch": 0.299261486177374,
-      "grad_norm": 0.5086488127708435,
-      "learning_rate": 9.837549914870336e-06,
-      "loss": 0.3996,
-      "step": 4579
-    },
-    {
-      "epoch": 0.29932684138291615,
-      "grad_norm": 0.5452041625976562,
-      "learning_rate": 9.837461616380455e-06,
-      "loss": 0.3795,
-      "step": 4580
-    },
-    {
-      "epoch": 0.2993921965884583,
-      "grad_norm": 0.46452298760414124,
-      "learning_rate": 9.837373294296598e-06,
-      "loss": 0.3824,
-      "step": 4581
-    },
-    {
-      "epoch": 0.2994575517940004,
-      "grad_norm": 0.4715021550655365,
-      "learning_rate": 9.837284948619195e-06,
-      "loss": 0.39,
-      "step": 4582
-    },
-    {
-      "epoch": 0.29952290699954254,
-      "grad_norm": 0.4850403666496277,
-      "learning_rate": 9.83719657934868e-06,
-      "loss": 0.4377,
-      "step": 4583
-    },
-    {
-      "epoch": 0.2995882622050846,
-      "grad_norm": 0.4575905203819275,
-      "learning_rate": 9.837108186485477e-06,
-      "loss": 0.3871,
-      "step": 4584
-    },
-    {
-      "epoch": 0.29965361741062674,
-      "grad_norm": 0.4533096253871918,
-      "learning_rate": 9.837019770030025e-06,
-      "loss": 0.4087,
-      "step": 4585
-    },
-    {
-      "epoch": 0.29971897261616887,
-      "grad_norm": 0.497069388628006,
-      "learning_rate": 9.836931329982752e-06,
-      "loss": 0.4166,
-      "step": 4586
-    },
-    {
-      "epoch": 0.299784327821711,
-      "grad_norm": 0.4475323557853699,
-      "learning_rate": 9.83684286634409e-06,
-      "loss": 0.3588,
-      "step": 4587
-    },
-    {
-      "epoch": 0.2998496830272531,
-      "grad_norm": 0.4541557729244232,
-      "learning_rate": 9.836754379114466e-06,
-      "loss": 0.3737,
-      "step": 4588
-    },
-    {
-      "epoch": 0.29991503823279525,
-      "grad_norm": 0.45018231868743896,
-      "learning_rate": 9.836665868294317e-06,
-      "loss": 0.3783,
-      "step": 4589
-    },
-    {
-      "epoch": 0.2999803934383374,
-      "grad_norm": 0.48943397402763367,
-      "learning_rate": 9.836577333884074e-06,
-      "loss": 0.3958,
-      "step": 4590
-    },
-    {
-      "epoch": 0.3000457486438795,
-      "grad_norm": 0.4738661050796509,
-      "learning_rate": 9.836488775884167e-06,
-      "loss": 0.4434,
-      "step": 4591
-    },
-    {
-      "epoch": 0.3001111038494216,
-      "grad_norm": 0.4493043124675751,
-      "learning_rate": 9.83640019429503e-06,
-      "loss": 0.383,
-      "step": 4592
-    },
-    {
-      "epoch": 0.3001764590549637,
-      "grad_norm": 0.4620586037635803,
-      "learning_rate": 9.836311589117091e-06,
-      "loss": 0.3707,
-      "step": 4593
-    },
-    {
-      "epoch": 0.30024181426050583,
-      "grad_norm": 0.4813469648361206,
-      "learning_rate": 9.836222960350788e-06,
-      "loss": 0.4225,
-      "step": 4594
-    },
-    {
-      "epoch": 0.30030716946604796,
-      "grad_norm": 0.48044130206108093,
-      "learning_rate": 9.836134307996548e-06,
-      "loss": 0.4355,
-      "step": 4595
-    },
-    {
-      "epoch": 0.3003725246715901,
-      "grad_norm": 0.47041261196136475,
-      "learning_rate": 9.836045632054806e-06,
-      "loss": 0.3756,
-      "step": 4596
-    },
-    {
-      "epoch": 0.3004378798771322,
-      "grad_norm": 0.47787100076675415,
-      "learning_rate": 9.835956932525993e-06,
-      "loss": 0.4207,
-      "step": 4597
-    },
-    {
-      "epoch": 0.30050323508267435,
-      "grad_norm": 0.4532429575920105,
-      "learning_rate": 9.835868209410547e-06,
-      "loss": 0.3889,
-      "step": 4598
-    },
-    {
-      "epoch": 0.3005685902882165,
-      "grad_norm": 0.46816444396972656,
-      "learning_rate": 9.835779462708892e-06,
-      "loss": 0.4109,
-      "step": 4599
-    },
-    {
-      "epoch": 0.3006339454937586,
-      "grad_norm": 0.4704676866531372,
-      "learning_rate": 9.835690692421466e-06,
-      "loss": 0.3553,
-      "step": 4600
-    },
-    {
-      "epoch": 0.3006993006993007,
-      "grad_norm": 0.5116649270057678,
-      "learning_rate": 9.835601898548704e-06,
-      "loss": 0.4238,
-      "step": 4601
-    },
-    {
-      "epoch": 0.3007646559048428,
-      "grad_norm": 0.47347110509872437,
-      "learning_rate": 9.835513081091034e-06,
-      "loss": 0.4105,
-      "step": 4602
-    },
-    {
-      "epoch": 0.30083001111038493,
-      "grad_norm": 0.44327715039253235,
-      "learning_rate": 9.835424240048891e-06,
-      "loss": 0.3709,
-      "step": 4603
-    },
-    {
-      "epoch": 0.30089536631592706,
-      "grad_norm": 0.4147056043148041,
-      "learning_rate": 9.83533537542271e-06,
-      "loss": 0.313,
-      "step": 4604
-    },
-    {
-      "epoch": 0.3009607215214692,
-      "grad_norm": 0.474161297082901,
-      "learning_rate": 9.835246487212924e-06,
-      "loss": 0.4169,
-      "step": 4605
-    },
-    {
-      "epoch": 0.3010260767270113,
-      "grad_norm": 0.4815036356449127,
-      "learning_rate": 9.835157575419965e-06,
-      "loss": 0.4072,
-      "step": 4606
-    },
-    {
-      "epoch": 0.30109143193255344,
-      "grad_norm": 0.49013909697532654,
-      "learning_rate": 9.835068640044266e-06,
-      "loss": 0.4277,
-      "step": 4607
-    },
-    {
-      "epoch": 0.30115678713809557,
-      "grad_norm": 0.4272536337375641,
-      "learning_rate": 9.834979681086265e-06,
-      "loss": 0.3614,
-      "step": 4608
-    },
-    {
-      "epoch": 0.30122214234363764,
-      "grad_norm": 0.45833322405815125,
-      "learning_rate": 9.834890698546392e-06,
-      "loss": 0.3711,
-      "step": 4609
-    },
-    {
-      "epoch": 0.3012874975491798,
-      "grad_norm": 0.4576624035835266,
-      "learning_rate": 9.83480169242508e-06,
-      "loss": 0.3751,
-      "step": 4610
-    },
-    {
-      "epoch": 0.3013528527547219,
-      "grad_norm": 0.5056855082511902,
-      "learning_rate": 9.834712662722768e-06,
-      "loss": 0.5115,
-      "step": 4611
-    },
-    {
-      "epoch": 0.30141820796026403,
-      "grad_norm": 0.42207542061805725,
-      "learning_rate": 9.834623609439886e-06,
-      "loss": 0.3673,
-      "step": 4612
-    },
-    {
-      "epoch": 0.30148356316580616,
-      "grad_norm": 0.504401445388794,
-      "learning_rate": 9.83453453257687e-06,
-      "loss": 0.4818,
-      "step": 4613
-    },
-    {
-      "epoch": 0.3015489183713483,
-      "grad_norm": 0.47074586153030396,
-      "learning_rate": 9.834445432134155e-06,
-      "loss": 0.3904,
-      "step": 4614
-    },
-    {
-      "epoch": 0.3016142735768904,
-      "grad_norm": 0.4869149327278137,
-      "learning_rate": 9.834356308112173e-06,
-      "loss": 0.436,
-      "step": 4615
-    },
-    {
-      "epoch": 0.30167962878243254,
-      "grad_norm": 0.4553590416908264,
-      "learning_rate": 9.834267160511361e-06,
-      "loss": 0.3819,
-      "step": 4616
-    },
-    {
-      "epoch": 0.30174498398797467,
-      "grad_norm": 0.4626240134239197,
-      "learning_rate": 9.834177989332155e-06,
-      "loss": 0.4362,
-      "step": 4617
-    },
-    {
-      "epoch": 0.30181033919351674,
-      "grad_norm": 0.47052204608917236,
-      "learning_rate": 9.834088794574986e-06,
-      "loss": 0.4109,
-      "step": 4618
-    },
-    {
-      "epoch": 0.30187569439905887,
-      "grad_norm": 0.4849553406238556,
-      "learning_rate": 9.833999576240293e-06,
-      "loss": 0.4515,
-      "step": 4619
-    },
-    {
-      "epoch": 0.301941049604601,
-      "grad_norm": 0.43929845094680786,
-      "learning_rate": 9.833910334328509e-06,
-      "loss": 0.3671,
-      "step": 4620
-    },
-    {
-      "epoch": 0.3020064048101431,
-      "grad_norm": 0.48614925146102905,
-      "learning_rate": 9.83382106884007e-06,
-      "loss": 0.4505,
-      "step": 4621
-    },
-    {
-      "epoch": 0.30207176001568525,
-      "grad_norm": 0.43567654490470886,
-      "learning_rate": 9.833731779775411e-06,
-      "loss": 0.3468,
-      "step": 4622
-    },
-    {
-      "epoch": 0.3021371152212274,
-      "grad_norm": 0.4525315463542938,
-      "learning_rate": 9.833642467134966e-06,
-      "loss": 0.403,
-      "step": 4623
-    },
-    {
-      "epoch": 0.3022024704267695,
-      "grad_norm": 0.5000825524330139,
-      "learning_rate": 9.833553130919174e-06,
-      "loss": 0.4383,
-      "step": 4624
-    },
-    {
-      "epoch": 0.30226782563231164,
-      "grad_norm": 0.5177122950553894,
-      "learning_rate": 9.833463771128467e-06,
-      "loss": 0.479,
-      "step": 4625
-    },
-    {
-      "epoch": 0.3023331808378537,
-      "grad_norm": 0.4748156666755676,
-      "learning_rate": 9.833374387763284e-06,
-      "loss": 0.4102,
-      "step": 4626
-    },
-    {
-      "epoch": 0.30239853604339584,
-      "grad_norm": 0.5039382576942444,
-      "learning_rate": 9.83328498082406e-06,
-      "loss": 0.4264,
-      "step": 4627
-    },
-    {
-      "epoch": 0.30246389124893797,
-      "grad_norm": 0.4976131319999695,
-      "learning_rate": 9.83319555031123e-06,
-      "loss": 0.4775,
-      "step": 4628
-    },
-    {
-      "epoch": 0.3025292464544801,
-      "grad_norm": 0.47715598344802856,
-      "learning_rate": 9.83310609622523e-06,
-      "loss": 0.4193,
-      "step": 4629
-    },
-    {
-      "epoch": 0.3025946016600222,
-      "grad_norm": 0.4935559332370758,
-      "learning_rate": 9.8330166185665e-06,
-      "loss": 0.3982,
-      "step": 4630
-    },
-    {
-      "epoch": 0.30265995686556435,
-      "grad_norm": 0.4852454662322998,
-      "learning_rate": 9.832927117335472e-06,
-      "loss": 0.4354,
-      "step": 4631
-    },
-    {
-      "epoch": 0.3027253120711065,
-      "grad_norm": 0.47433051466941833,
-      "learning_rate": 9.832837592532584e-06,
-      "loss": 0.3778,
-      "step": 4632
-    },
-    {
-      "epoch": 0.3027906672766486,
-      "grad_norm": 0.4809684753417969,
-      "learning_rate": 9.832748044158273e-06,
-      "loss": 0.4245,
-      "step": 4633
-    },
-    {
-      "epoch": 0.3028560224821907,
-      "grad_norm": 0.42704394459724426,
-      "learning_rate": 9.832658472212975e-06,
-      "loss": 0.3272,
-      "step": 4634
-    },
-    {
-      "epoch": 0.3029213776877328,
-      "grad_norm": 0.515950620174408,
-      "learning_rate": 9.832568876697129e-06,
-      "loss": 0.4492,
-      "step": 4635
-    },
-    {
-      "epoch": 0.30298673289327493,
-      "grad_norm": 0.46974262595176697,
-      "learning_rate": 9.83247925761117e-06,
-      "loss": 0.4386,
-      "step": 4636
-    },
-    {
-      "epoch": 0.30305208809881706,
-      "grad_norm": 0.48093926906585693,
-      "learning_rate": 9.832389614955533e-06,
-      "loss": 0.3841,
-      "step": 4637
-    },
-    {
-      "epoch": 0.3031174433043592,
-      "grad_norm": 0.49033021926879883,
-      "learning_rate": 9.83229994873066e-06,
-      "loss": 0.447,
-      "step": 4638
-    },
-    {
-      "epoch": 0.3031827985099013,
-      "grad_norm": 0.4828386902809143,
-      "learning_rate": 9.832210258936986e-06,
-      "loss": 0.4219,
-      "step": 4639
-    },
-    {
-      "epoch": 0.30324815371544345,
-      "grad_norm": 0.43323618173599243,
-      "learning_rate": 9.832120545574948e-06,
-      "loss": 0.3594,
-      "step": 4640
-    },
-    {
-      "epoch": 0.3033135089209856,
-      "grad_norm": 0.4587923288345337,
-      "learning_rate": 9.832030808644986e-06,
-      "loss": 0.3951,
-      "step": 4641
-    },
-    {
-      "epoch": 0.3033788641265277,
-      "grad_norm": 0.49970561265945435,
-      "learning_rate": 9.831941048147533e-06,
-      "loss": 0.4541,
-      "step": 4642
-    },
-    {
-      "epoch": 0.3034442193320698,
-      "grad_norm": 0.45958563685417175,
-      "learning_rate": 9.83185126408303e-06,
-      "loss": 0.3875,
-      "step": 4643
-    },
-    {
-      "epoch": 0.3035095745376119,
-      "grad_norm": 0.4726570248603821,
-      "learning_rate": 9.831761456451916e-06,
-      "loss": 0.4299,
-      "step": 4644
-    },
-    {
-      "epoch": 0.30357492974315403,
-      "grad_norm": 0.49867337942123413,
-      "learning_rate": 9.831671625254626e-06,
-      "loss": 0.4522,
-      "step": 4645
-    },
-    {
-      "epoch": 0.30364028494869616,
-      "grad_norm": 0.5065587162971497,
-      "learning_rate": 9.8315817704916e-06,
-      "loss": 0.4377,
-      "step": 4646
-    },
-    {
-      "epoch": 0.3037056401542383,
-      "grad_norm": 0.5036814212799072,
-      "learning_rate": 9.831491892163277e-06,
-      "loss": 0.3949,
-      "step": 4647
-    },
-    {
-      "epoch": 0.3037709953597804,
-      "grad_norm": 0.4690750241279602,
-      "learning_rate": 9.831401990270094e-06,
-      "loss": 0.4522,
-      "step": 4648
-    },
-    {
-      "epoch": 0.30383635056532254,
-      "grad_norm": 0.48045486211776733,
-      "learning_rate": 9.83131206481249e-06,
-      "loss": 0.3932,
-      "step": 4649
-    },
-    {
-      "epoch": 0.30390170577086467,
-      "grad_norm": 0.47589701414108276,
-      "learning_rate": 9.831222115790902e-06,
-      "loss": 0.3997,
-      "step": 4650
-    },
-    {
-      "epoch": 0.30396706097640674,
-      "grad_norm": 0.48952358961105347,
-      "learning_rate": 9.83113214320577e-06,
-      "loss": 0.4137,
-      "step": 4651
-    },
-    {
-      "epoch": 0.3040324161819489,
-      "grad_norm": 0.4988320469856262,
-      "learning_rate": 9.831042147057533e-06,
-      "loss": 0.4335,
-      "step": 4652
-    },
-    {
-      "epoch": 0.304097771387491,
-      "grad_norm": 0.4423309862613678,
-      "learning_rate": 9.830952127346631e-06,
-      "loss": 0.3441,
-      "step": 4653
-    },
-    {
-      "epoch": 0.30416312659303313,
-      "grad_norm": 0.4505099058151245,
-      "learning_rate": 9.8308620840735e-06,
-      "loss": 0.3588,
-      "step": 4654
-    },
-    {
-      "epoch": 0.30422848179857526,
-      "grad_norm": 0.48584893345832825,
-      "learning_rate": 9.830772017238583e-06,
-      "loss": 0.4113,
-      "step": 4655
-    },
-    {
-      "epoch": 0.3042938370041174,
-      "grad_norm": 0.4860285818576813,
-      "learning_rate": 9.830681926842316e-06,
-      "loss": 0.4269,
-      "step": 4656
-    },
-    {
-      "epoch": 0.3043591922096595,
-      "grad_norm": 0.45671284198760986,
-      "learning_rate": 9.830591812885142e-06,
-      "loss": 0.3842,
-      "step": 4657
-    },
-    {
-      "epoch": 0.30442454741520164,
-      "grad_norm": 0.49379563331604004,
-      "learning_rate": 9.830501675367497e-06,
-      "loss": 0.4139,
-      "step": 4658
-    },
-    {
-      "epoch": 0.30448990262074377,
-      "grad_norm": 0.5117769837379456,
-      "learning_rate": 9.83041151428982e-06,
-      "loss": 0.4672,
-      "step": 4659
-    },
-    {
-      "epoch": 0.30455525782628584,
-      "grad_norm": 0.4411393702030182,
-      "learning_rate": 9.830321329652555e-06,
-      "loss": 0.3989,
-      "step": 4660
-    },
-    {
-      "epoch": 0.30462061303182797,
-      "grad_norm": 0.47283071279525757,
-      "learning_rate": 9.830231121456137e-06,
-      "loss": 0.409,
-      "step": 4661
-    },
-    {
-      "epoch": 0.3046859682373701,
-      "grad_norm": 0.4507950246334076,
-      "learning_rate": 9.83014088970101e-06,
-      "loss": 0.3753,
-      "step": 4662
-    },
-    {
-      "epoch": 0.3047513234429122,
-      "grad_norm": 0.46355167031288147,
-      "learning_rate": 9.830050634387614e-06,
-      "loss": 0.3925,
-      "step": 4663
-    },
-    {
-      "epoch": 0.30481667864845435,
-      "grad_norm": 0.49535447359085083,
-      "learning_rate": 9.829960355516385e-06,
-      "loss": 0.4166,
-      "step": 4664
-    },
-    {
-      "epoch": 0.3048820338539965,
-      "grad_norm": 0.4795665740966797,
-      "learning_rate": 9.829870053087768e-06,
-      "loss": 0.4013,
-      "step": 4665
-    },
-    {
-      "epoch": 0.3049473890595386,
-      "grad_norm": 0.46602609753608704,
-      "learning_rate": 9.8297797271022e-06,
-      "loss": 0.3946,
-      "step": 4666
-    },
-    {
-      "epoch": 0.30501274426508074,
-      "grad_norm": 0.4819977879524231,
-      "learning_rate": 9.829689377560125e-06,
-      "loss": 0.4248,
-      "step": 4667
-    },
-    {
-      "epoch": 0.3050780994706228,
-      "grad_norm": 0.4841451048851013,
-      "learning_rate": 9.82959900446198e-06,
-      "loss": 0.4017,
-      "step": 4668
-    },
-    {
-      "epoch": 0.30514345467616494,
-      "grad_norm": 0.47611433267593384,
-      "learning_rate": 9.829508607808208e-06,
-      "loss": 0.4294,
-      "step": 4669
-    },
-    {
-      "epoch": 0.30520880988170707,
-      "grad_norm": 0.44730934500694275,
-      "learning_rate": 9.829418187599252e-06,
-      "loss": 0.378,
-      "step": 4670
-    },
-    {
-      "epoch": 0.3052741650872492,
-      "grad_norm": 0.46959617733955383,
-      "learning_rate": 9.829327743835547e-06,
-      "loss": 0.3983,
-      "step": 4671
-    },
-    {
-      "epoch": 0.3053395202927913,
-      "grad_norm": 0.585527777671814,
-      "learning_rate": 9.829237276517538e-06,
-      "loss": 0.4016,
-      "step": 4672
-    },
-    {
-      "epoch": 0.30540487549833345,
-      "grad_norm": 0.4545510411262512,
-      "learning_rate": 9.829146785645667e-06,
-      "loss": 0.3583,
-      "step": 4673
-    },
-    {
-      "epoch": 0.3054702307038756,
-      "grad_norm": 0.4610280394554138,
-      "learning_rate": 9.829056271220373e-06,
-      "loss": 0.3798,
-      "step": 4674
-    },
-    {
-      "epoch": 0.3055355859094177,
-      "grad_norm": 0.4837823212146759,
-      "learning_rate": 9.8289657332421e-06,
-      "loss": 0.3746,
-      "step": 4675
-    },
-    {
-      "epoch": 0.30560094111495983,
-      "grad_norm": 0.4811020493507385,
-      "learning_rate": 9.828875171711288e-06,
-      "loss": 0.4325,
-      "step": 4676
-    },
-    {
-      "epoch": 0.3056662963205019,
-      "grad_norm": 0.481588214635849,
-      "learning_rate": 9.828784586628378e-06,
-      "loss": 0.4401,
-      "step": 4677
-    },
-    {
-      "epoch": 0.30573165152604403,
-      "grad_norm": 0.4185912013053894,
-      "learning_rate": 9.828693977993813e-06,
-      "loss": 0.3133,
-      "step": 4678
-    },
-    {
-      "epoch": 0.30579700673158616,
-      "grad_norm": 0.4645090103149414,
-      "learning_rate": 9.828603345808034e-06,
-      "loss": 0.3459,
-      "step": 4679
-    },
-    {
-      "epoch": 0.3058623619371283,
-      "grad_norm": 0.5124992728233337,
-      "learning_rate": 9.828512690071485e-06,
-      "loss": 0.4257,
-      "step": 4680
-    },
-    {
-      "epoch": 0.3059277171426704,
-      "grad_norm": 0.5183594226837158,
-      "learning_rate": 9.828422010784606e-06,
-      "loss": 0.4466,
-      "step": 4681
-    },
-    {
-      "epoch": 0.30599307234821255,
-      "grad_norm": 0.48616012930870056,
-      "learning_rate": 9.82833130794784e-06,
-      "loss": 0.4588,
-      "step": 4682
-    },
-    {
-      "epoch": 0.3060584275537547,
-      "grad_norm": 0.4824143648147583,
-      "learning_rate": 9.82824058156163e-06,
-      "loss": 0.4333,
-      "step": 4683
-    },
-    {
-      "epoch": 0.3061237827592968,
-      "grad_norm": 0.4992258548736572,
-      "learning_rate": 9.828149831626418e-06,
-      "loss": 0.3931,
-      "step": 4684
-    },
-    {
-      "epoch": 0.3061891379648389,
-      "grad_norm": 0.5806849598884583,
-      "learning_rate": 9.828059058142647e-06,
-      "loss": 0.4993,
-      "step": 4685
-    },
-    {
-      "epoch": 0.306254493170381,
-      "grad_norm": 0.4938088059425354,
-      "learning_rate": 9.827968261110758e-06,
-      "loss": 0.3776,
-      "step": 4686
-    },
-    {
-      "epoch": 0.30631984837592313,
-      "grad_norm": 0.5305953025817871,
-      "learning_rate": 9.827877440531195e-06,
-      "loss": 0.4366,
-      "step": 4687
-    },
-    {
-      "epoch": 0.30638520358146526,
-      "grad_norm": 0.4520416557788849,
-      "learning_rate": 9.827786596404403e-06,
-      "loss": 0.3701,
-      "step": 4688
-    },
-    {
-      "epoch": 0.3064505587870074,
-      "grad_norm": 0.48587965965270996,
-      "learning_rate": 9.827695728730824e-06,
-      "loss": 0.4261,
-      "step": 4689
-    },
-    {
-      "epoch": 0.3065159139925495,
-      "grad_norm": 0.5094895958900452,
-      "learning_rate": 9.827604837510897e-06,
-      "loss": 0.4354,
-      "step": 4690
-    },
-    {
-      "epoch": 0.30658126919809164,
-      "grad_norm": 0.4881570041179657,
-      "learning_rate": 9.827513922745073e-06,
-      "loss": 0.4471,
-      "step": 4691
-    },
-    {
-      "epoch": 0.30664662440363377,
-      "grad_norm": 0.4493294060230255,
-      "learning_rate": 9.827422984433787e-06,
-      "loss": 0.3971,
-      "step": 4692
-    },
-    {
-      "epoch": 0.30671197960917584,
-      "grad_norm": 0.41316869854927063,
-      "learning_rate": 9.82733202257749e-06,
-      "loss": 0.3215,
-      "step": 4693
-    },
-    {
-      "epoch": 0.306777334814718,
-      "grad_norm": 0.4954182207584381,
-      "learning_rate": 9.82724103717662e-06,
-      "loss": 0.4057,
-      "step": 4694
-    },
-    {
-      "epoch": 0.3068426900202601,
-      "grad_norm": 0.45912855863571167,
-      "learning_rate": 9.827150028231625e-06,
-      "loss": 0.3981,
-      "step": 4695
-    },
-    {
-      "epoch": 0.30690804522580223,
-      "grad_norm": 0.46424421668052673,
-      "learning_rate": 9.827058995742947e-06,
-      "loss": 0.3586,
-      "step": 4696
-    },
-    {
-      "epoch": 0.30697340043134436,
-      "grad_norm": 0.4777919352054596,
-      "learning_rate": 9.826967939711029e-06,
-      "loss": 0.4328,
-      "step": 4697
-    },
-    {
-      "epoch": 0.3070387556368865,
-      "grad_norm": 0.4849260151386261,
-      "learning_rate": 9.826876860136317e-06,
-      "loss": 0.4256,
-      "step": 4698
-    },
-    {
-      "epoch": 0.3071041108424286,
-      "grad_norm": 0.4411482512950897,
-      "learning_rate": 9.826785757019253e-06,
-      "loss": 0.3589,
-      "step": 4699
-    },
-    {
-      "epoch": 0.30716946604797074,
-      "grad_norm": 0.45035502314567566,
-      "learning_rate": 9.826694630360283e-06,
-      "loss": 0.3917,
-      "step": 4700
-    },
-    {
-      "epoch": 0.30723482125351287,
-      "grad_norm": 0.48693060874938965,
-      "learning_rate": 9.826603480159852e-06,
-      "loss": 0.4292,
-      "step": 4701
-    },
-    {
-      "epoch": 0.30730017645905494,
-      "grad_norm": 0.4641180634498596,
-      "learning_rate": 9.826512306418403e-06,
-      "loss": 0.3931,
-      "step": 4702
-    },
-    {
-      "epoch": 0.30736553166459707,
-      "grad_norm": 0.4459921717643738,
-      "learning_rate": 9.82642110913638e-06,
-      "loss": 0.3673,
-      "step": 4703
-    },
-    {
-      "epoch": 0.3074308868701392,
-      "grad_norm": 0.4809280037879944,
-      "learning_rate": 9.826329888314233e-06,
-      "loss": 0.4166,
-      "step": 4704
-    },
-    {
-      "epoch": 0.3074962420756813,
-      "grad_norm": 0.4905911684036255,
-      "learning_rate": 9.826238643952402e-06,
-      "loss": 0.4126,
-      "step": 4705
-    },
-    {
-      "epoch": 0.30756159728122345,
-      "grad_norm": 0.47121360898017883,
-      "learning_rate": 9.826147376051332e-06,
-      "loss": 0.3723,
-      "step": 4706
-    },
-    {
-      "epoch": 0.3076269524867656,
-      "grad_norm": 0.7335187792778015,
-      "learning_rate": 9.82605608461147e-06,
-      "loss": 0.4497,
-      "step": 4707
-    },
-    {
-      "epoch": 0.3076923076923077,
-      "grad_norm": 0.45816537737846375,
-      "learning_rate": 9.825964769633259e-06,
-      "loss": 0.3828,
-      "step": 4708
-    },
-    {
-      "epoch": 0.30775766289784984,
-      "grad_norm": 0.46843087673187256,
-      "learning_rate": 9.825873431117147e-06,
-      "loss": 0.4246,
-      "step": 4709
-    },
-    {
-      "epoch": 0.3078230181033919,
-      "grad_norm": 0.4835190773010254,
-      "learning_rate": 9.82578206906358e-06,
-      "loss": 0.4053,
-      "step": 4710
-    },
-    {
-      "epoch": 0.30788837330893404,
-      "grad_norm": 0.4642230272293091,
-      "learning_rate": 9.825690683472998e-06,
-      "loss": 0.4198,
-      "step": 4711
-    },
-    {
-      "epoch": 0.30795372851447617,
-      "grad_norm": 0.5255396962165833,
-      "learning_rate": 9.825599274345856e-06,
-      "loss": 0.4735,
-      "step": 4712
-    },
-    {
-      "epoch": 0.3080190837200183,
-      "grad_norm": 0.4632430374622345,
-      "learning_rate": 9.825507841682591e-06,
-      "loss": 0.431,
-      "step": 4713
-    },
-    {
-      "epoch": 0.3080844389255604,
-      "grad_norm": 0.4993661344051361,
-      "learning_rate": 9.825416385483654e-06,
-      "loss": 0.3825,
-      "step": 4714
-    },
-    {
-      "epoch": 0.30814979413110255,
-      "grad_norm": 0.49783140420913696,
-      "learning_rate": 9.82532490574949e-06,
-      "loss": 0.4408,
-      "step": 4715
-    },
-    {
-      "epoch": 0.3082151493366447,
-      "grad_norm": 0.5053375959396362,
-      "learning_rate": 9.825233402480543e-06,
-      "loss": 0.4418,
-      "step": 4716
-    },
-    {
-      "epoch": 0.3082805045421868,
-      "grad_norm": 0.4538673162460327,
-      "learning_rate": 9.825141875677263e-06,
-      "loss": 0.3689,
-      "step": 4717
-    },
-    {
-      "epoch": 0.30834585974772893,
-      "grad_norm": 0.4795911908149719,
-      "learning_rate": 9.825050325340092e-06,
-      "loss": 0.4425,
-      "step": 4718
-    },
-    {
-      "epoch": 0.308411214953271,
-      "grad_norm": 0.4877456724643707,
-      "learning_rate": 9.82495875146948e-06,
-      "loss": 0.4171,
-      "step": 4719
-    },
-    {
-      "epoch": 0.30847657015881313,
-      "grad_norm": 0.4846137464046478,
-      "learning_rate": 9.824867154065874e-06,
-      "loss": 0.4039,
-      "step": 4720
-    },
-    {
-      "epoch": 0.30854192536435526,
-      "grad_norm": 0.4265865981578827,
-      "learning_rate": 9.824775533129718e-06,
-      "loss": 0.3286,
-      "step": 4721
-    },
-    {
-      "epoch": 0.3086072805698974,
-      "grad_norm": 0.4717889428138733,
-      "learning_rate": 9.824683888661462e-06,
-      "loss": 0.4207,
-      "step": 4722
-    },
-    {
-      "epoch": 0.3086726357754395,
-      "grad_norm": 0.4649326205253601,
-      "learning_rate": 9.824592220661548e-06,
-      "loss": 0.4077,
-      "step": 4723
-    },
-    {
-      "epoch": 0.30873799098098165,
-      "grad_norm": 0.4909663200378418,
-      "learning_rate": 9.824500529130429e-06,
-      "loss": 0.4668,
-      "step": 4724
-    },
-    {
-      "epoch": 0.3088033461865238,
-      "grad_norm": 0.4453728497028351,
-      "learning_rate": 9.824408814068548e-06,
-      "loss": 0.3678,
-      "step": 4725
-    },
-    {
-      "epoch": 0.3088687013920659,
-      "grad_norm": 0.5093079805374146,
-      "learning_rate": 9.824317075476356e-06,
-      "loss": 0.4338,
-      "step": 4726
-    },
-    {
-      "epoch": 0.308934056597608,
-      "grad_norm": 0.48250722885131836,
-      "learning_rate": 9.824225313354296e-06,
-      "loss": 0.4095,
-      "step": 4727
-    },
-    {
-      "epoch": 0.3089994118031501,
-      "grad_norm": 0.4546104967594147,
-      "learning_rate": 9.824133527702819e-06,
-      "loss": 0.4029,
-      "step": 4728
-    },
-    {
-      "epoch": 0.30906476700869223,
-      "grad_norm": 0.48363611102104187,
-      "learning_rate": 9.824041718522372e-06,
-      "loss": 0.4135,
-      "step": 4729
-    },
-    {
-      "epoch": 0.30913012221423436,
-      "grad_norm": 0.47212186455726624,
-      "learning_rate": 9.823949885813402e-06,
-      "loss": 0.3982,
-      "step": 4730
-    },
-    {
-      "epoch": 0.3091954774197765,
-      "grad_norm": 0.5019171237945557,
-      "learning_rate": 9.823858029576357e-06,
-      "loss": 0.4371,
-      "step": 4731
-    },
-    {
-      "epoch": 0.3092608326253186,
-      "grad_norm": 0.4634392559528351,
-      "learning_rate": 9.823766149811686e-06,
-      "loss": 0.381,
-      "step": 4732
-    },
-    {
-      "epoch": 0.30932618783086074,
-      "grad_norm": 0.4996906518936157,
-      "learning_rate": 9.823674246519835e-06,
-      "loss": 0.4128,
-      "step": 4733
-    },
-    {
-      "epoch": 0.30939154303640287,
-      "grad_norm": 0.4786403477191925,
-      "learning_rate": 9.823582319701255e-06,
-      "loss": 0.3792,
-      "step": 4734
-    },
-    {
-      "epoch": 0.30945689824194494,
-      "grad_norm": 0.4715338349342346,
-      "learning_rate": 9.823490369356392e-06,
-      "loss": 0.3604,
-      "step": 4735
-    },
-    {
-      "epoch": 0.3095222534474871,
-      "grad_norm": 0.5077422857284546,
-      "learning_rate": 9.823398395485696e-06,
-      "loss": 0.4201,
-      "step": 4736
-    },
-    {
-      "epoch": 0.3095876086530292,
-      "grad_norm": 0.4325348138809204,
-      "learning_rate": 9.823306398089615e-06,
-      "loss": 0.3879,
-      "step": 4737
-    },
-    {
-      "epoch": 0.30965296385857133,
-      "grad_norm": 0.47307687997817993,
-      "learning_rate": 9.823214377168597e-06,
-      "loss": 0.4156,
-      "step": 4738
-    },
-    {
-      "epoch": 0.30971831906411346,
-      "grad_norm": 0.5081484913825989,
-      "learning_rate": 9.823122332723091e-06,
-      "loss": 0.4202,
-      "step": 4739
-    },
-    {
-      "epoch": 0.3097836742696556,
-      "grad_norm": 0.4860573410987854,
-      "learning_rate": 9.823030264753549e-06,
-      "loss": 0.4427,
-      "step": 4740
-    },
-    {
-      "epoch": 0.3098490294751977,
-      "grad_norm": 0.5136462450027466,
-      "learning_rate": 9.822938173260416e-06,
-      "loss": 0.4832,
-      "step": 4741
-    },
-    {
-      "epoch": 0.30991438468073984,
-      "grad_norm": 0.49386855959892273,
-      "learning_rate": 9.822846058244141e-06,
-      "loss": 0.4266,
-      "step": 4742
-    },
-    {
-      "epoch": 0.30997973988628197,
-      "grad_norm": 0.4761289060115814,
-      "learning_rate": 9.822753919705179e-06,
-      "loss": 0.4064,
-      "step": 4743
-    },
-    {
-      "epoch": 0.31004509509182404,
-      "grad_norm": 0.4351711869239807,
-      "learning_rate": 9.822661757643971e-06,
-      "loss": 0.3351,
-      "step": 4744
-    },
-    {
-      "epoch": 0.31011045029736617,
-      "grad_norm": 0.4576933681964874,
-      "learning_rate": 9.822569572060975e-06,
-      "loss": 0.3785,
-      "step": 4745
-    },
-    {
-      "epoch": 0.3101758055029083,
-      "grad_norm": 0.48872002959251404,
-      "learning_rate": 9.822477362956635e-06,
-      "loss": 0.4176,
-      "step": 4746
-    },
-    {
-      "epoch": 0.3102411607084504,
-      "grad_norm": 0.5006043910980225,
-      "learning_rate": 9.822385130331401e-06,
-      "loss": 0.4231,
-      "step": 4747
-    },
-    {
-      "epoch": 0.31030651591399255,
-      "grad_norm": 0.4624032974243164,
-      "learning_rate": 9.822292874185726e-06,
-      "loss": 0.4088,
-      "step": 4748
-    },
-    {
-      "epoch": 0.3103718711195347,
-      "grad_norm": 0.43168407678604126,
-      "learning_rate": 9.822200594520055e-06,
-      "loss": 0.3781,
-      "step": 4749
-    },
-    {
-      "epoch": 0.3104372263250768,
-      "grad_norm": 0.5045598745346069,
-      "learning_rate": 9.822108291334845e-06,
-      "loss": 0.4475,
-      "step": 4750
-    },
-    {
-      "epoch": 0.31050258153061894,
-      "grad_norm": 0.45648258924484253,
-      "learning_rate": 9.822015964630539e-06,
-      "loss": 0.3695,
-      "step": 4751
-    },
-    {
-      "epoch": 0.310567936736161,
-      "grad_norm": 0.4925973415374756,
-      "learning_rate": 9.821923614407593e-06,
-      "loss": 0.4386,
-      "step": 4752
-    },
-    {
-      "epoch": 0.31063329194170314,
-      "grad_norm": 0.4795874059200287,
-      "learning_rate": 9.821831240666453e-06,
-      "loss": 0.3643,
-      "step": 4753
-    },
-    {
-      "epoch": 0.31069864714724527,
-      "grad_norm": 0.46666625142097473,
-      "learning_rate": 9.821738843407574e-06,
-      "loss": 0.4192,
-      "step": 4754
-    },
-    {
-      "epoch": 0.3107640023527874,
-      "grad_norm": 0.4786038100719452,
-      "learning_rate": 9.8216464226314e-06,
-      "loss": 0.4122,
-      "step": 4755
-    },
-    {
-      "epoch": 0.3108293575583295,
-      "grad_norm": 0.4647231698036194,
-      "learning_rate": 9.821553978338391e-06,
-      "loss": 0.3796,
-      "step": 4756
-    },
-    {
-      "epoch": 0.31089471276387165,
-      "grad_norm": 0.4673691987991333,
-      "learning_rate": 9.821461510528991e-06,
-      "loss": 0.4054,
-      "step": 4757
-    },
-    {
-      "epoch": 0.3109600679694138,
-      "grad_norm": 0.4552198648452759,
-      "learning_rate": 9.821369019203652e-06,
-      "loss": 0.362,
-      "step": 4758
-    },
-    {
-      "epoch": 0.3110254231749559,
-      "grad_norm": 0.42929479479789734,
-      "learning_rate": 9.821276504362827e-06,
-      "loss": 0.3518,
-      "step": 4759
-    },
-    {
-      "epoch": 0.31109077838049803,
-      "grad_norm": 0.45886924862861633,
-      "learning_rate": 9.821183966006967e-06,
-      "loss": 0.4119,
-      "step": 4760
-    },
-    {
-      "epoch": 0.3111561335860401,
-      "grad_norm": 0.485416442155838,
-      "learning_rate": 9.821091404136521e-06,
-      "loss": 0.4136,
-      "step": 4761
-    },
-    {
-      "epoch": 0.31122148879158223,
-      "grad_norm": 0.4691421091556549,
-      "learning_rate": 9.820998818751943e-06,
-      "loss": 0.4121,
-      "step": 4762
-    },
-    {
-      "epoch": 0.31128684399712436,
-      "grad_norm": 0.525909423828125,
-      "learning_rate": 9.820906209853682e-06,
-      "loss": 0.4613,
-      "step": 4763
-    },
-    {
-      "epoch": 0.3113521992026665,
-      "grad_norm": 0.5122382640838623,
-      "learning_rate": 9.820813577442192e-06,
-      "loss": 0.4174,
-      "step": 4764
-    },
-    {
-      "epoch": 0.3114175544082086,
-      "grad_norm": 0.46602943539619446,
-      "learning_rate": 9.820720921517925e-06,
-      "loss": 0.3721,
-      "step": 4765
-    },
-    {
-      "epoch": 0.31148290961375075,
-      "grad_norm": 0.5424133539199829,
-      "learning_rate": 9.820628242081332e-06,
-      "loss": 0.462,
-      "step": 4766
-    },
-    {
-      "epoch": 0.3115482648192929,
-      "grad_norm": 0.49611037969589233,
-      "learning_rate": 9.820535539132865e-06,
-      "loss": 0.4226,
-      "step": 4767
-    },
-    {
-      "epoch": 0.311613620024835,
-      "grad_norm": 0.4741397500038147,
-      "learning_rate": 9.820442812672974e-06,
-      "loss": 0.3786,
-      "step": 4768
-    },
-    {
-      "epoch": 0.3116789752303771,
-      "grad_norm": 0.5043719410896301,
-      "learning_rate": 9.820350062702117e-06,
-      "loss": 0.4327,
-      "step": 4769
-    },
-    {
-      "epoch": 0.3117443304359192,
-      "grad_norm": 0.47938835620880127,
-      "learning_rate": 9.82025728922074e-06,
-      "loss": 0.397,
-      "step": 4770
-    },
-    {
-      "epoch": 0.31180968564146133,
-      "grad_norm": 0.6575962901115417,
-      "learning_rate": 9.8201644922293e-06,
-      "loss": 0.4668,
-      "step": 4771
-    },
-    {
-      "epoch": 0.31187504084700346,
-      "grad_norm": 0.4484061598777771,
-      "learning_rate": 9.820071671728247e-06,
-      "loss": 0.3998,
-      "step": 4772
-    },
-    {
-      "epoch": 0.3119403960525456,
-      "grad_norm": 0.4660794734954834,
-      "learning_rate": 9.819978827718035e-06,
-      "loss": 0.3946,
-      "step": 4773
-    },
-    {
-      "epoch": 0.3120057512580877,
-      "grad_norm": 0.42437443137168884,
-      "learning_rate": 9.819885960199115e-06,
-      "loss": 0.3356,
-      "step": 4774
-    },
-    {
-      "epoch": 0.31207110646362984,
-      "grad_norm": 0.47052139043807983,
-      "learning_rate": 9.819793069171944e-06,
-      "loss": 0.3964,
-      "step": 4775
-    },
-    {
-      "epoch": 0.31213646166917197,
-      "grad_norm": 0.46570703387260437,
-      "learning_rate": 9.819700154636971e-06,
-      "loss": 0.4351,
-      "step": 4776
-    },
-    {
-      "epoch": 0.31220181687471404,
-      "grad_norm": 0.4962845742702484,
-      "learning_rate": 9.819607216594652e-06,
-      "loss": 0.4024,
-      "step": 4777
-    },
-    {
-      "epoch": 0.3122671720802562,
-      "grad_norm": 0.4872830808162689,
-      "learning_rate": 9.819514255045437e-06,
-      "loss": 0.4276,
-      "step": 4778
-    },
-    {
-      "epoch": 0.3123325272857983,
-      "grad_norm": 0.5150973200798035,
-      "learning_rate": 9.819421269989782e-06,
-      "loss": 0.4137,
-      "step": 4779
-    },
-    {
-      "epoch": 0.31239788249134043,
-      "grad_norm": 0.46941718459129333,
-      "learning_rate": 9.81932826142814e-06,
-      "loss": 0.3733,
-      "step": 4780
-    },
-    {
-      "epoch": 0.31246323769688256,
-      "grad_norm": 0.47068560123443604,
-      "learning_rate": 9.819235229360964e-06,
-      "loss": 0.3926,
-      "step": 4781
-    },
-    {
-      "epoch": 0.3125285929024247,
-      "grad_norm": 0.5268428325653076,
-      "learning_rate": 9.81914217378871e-06,
-      "loss": 0.4788,
-      "step": 4782
-    },
-    {
-      "epoch": 0.3125939481079668,
-      "grad_norm": 0.44260939955711365,
-      "learning_rate": 9.819049094711827e-06,
-      "loss": 0.3621,
-      "step": 4783
-    },
-    {
-      "epoch": 0.31265930331350894,
-      "grad_norm": 0.44305866956710815,
-      "learning_rate": 9.818955992130776e-06,
-      "loss": 0.3822,
-      "step": 4784
-    },
-    {
-      "epoch": 0.31272465851905107,
-      "grad_norm": 0.48154309391975403,
-      "learning_rate": 9.818862866046004e-06,
-      "loss": 0.4069,
-      "step": 4785
-    },
-    {
-      "epoch": 0.31279001372459314,
-      "grad_norm": 0.4962350130081177,
-      "learning_rate": 9.81876971645797e-06,
-      "loss": 0.4843,
-      "step": 4786
-    },
-    {
-      "epoch": 0.31285536893013527,
-      "grad_norm": 0.4804168939590454,
-      "learning_rate": 9.818676543367127e-06,
-      "loss": 0.4579,
-      "step": 4787
-    },
-    {
-      "epoch": 0.3129207241356774,
-      "grad_norm": 0.4673965275287628,
-      "learning_rate": 9.818583346773926e-06,
-      "loss": 0.4024,
-      "step": 4788
-    },
-    {
-      "epoch": 0.3129860793412195,
-      "grad_norm": 0.49661630392074585,
-      "learning_rate": 9.818490126678828e-06,
-      "loss": 0.4329,
-      "step": 4789
-    },
-    {
-      "epoch": 0.31305143454676165,
-      "grad_norm": 0.45884013175964355,
-      "learning_rate": 9.818396883082285e-06,
-      "loss": 0.3635,
-      "step": 4790
-    },
-    {
-      "epoch": 0.3131167897523038,
-      "grad_norm": 0.43619024753570557,
-      "learning_rate": 9.818303615984748e-06,
-      "loss": 0.3677,
-      "step": 4791
-    },
-    {
-      "epoch": 0.3131821449578459,
-      "grad_norm": 0.4732099771499634,
-      "learning_rate": 9.818210325386676e-06,
-      "loss": 0.4172,
-      "step": 4792
-    },
-    {
-      "epoch": 0.31324750016338804,
-      "grad_norm": 0.4724849462509155,
-      "learning_rate": 9.818117011288523e-06,
-      "loss": 0.4241,
-      "step": 4793
-    },
-    {
-      "epoch": 0.3133128553689301,
-      "grad_norm": 0.4618249237537384,
-      "learning_rate": 9.818023673690746e-06,
-      "loss": 0.3868,
-      "step": 4794
-    },
-    {
-      "epoch": 0.31337821057447224,
-      "grad_norm": 0.49833354353904724,
-      "learning_rate": 9.817930312593798e-06,
-      "loss": 0.4183,
-      "step": 4795
-    },
-    {
-      "epoch": 0.31344356578001437,
-      "grad_norm": 0.5455418229103088,
-      "learning_rate": 9.817836927998134e-06,
-      "loss": 0.4504,
-      "step": 4796
-    },
-    {
-      "epoch": 0.3135089209855565,
-      "grad_norm": 0.49523916840553284,
-      "learning_rate": 9.81774351990421e-06,
-      "loss": 0.4108,
-      "step": 4797
-    },
-    {
-      "epoch": 0.3135742761910986,
-      "grad_norm": 0.4911780059337616,
-      "learning_rate": 9.817650088312483e-06,
-      "loss": 0.3926,
-      "step": 4798
-    },
-    {
-      "epoch": 0.31363963139664075,
-      "grad_norm": 0.4689830243587494,
-      "learning_rate": 9.817556633223407e-06,
-      "loss": 0.3995,
-      "step": 4799
-    },
-    {
-      "epoch": 0.3137049866021829,
-      "grad_norm": 0.471271276473999,
-      "learning_rate": 9.817463154637436e-06,
-      "loss": 0.4288,
-      "step": 4800
-    },
-    {
-      "epoch": 0.313770341807725,
-      "grad_norm": 0.46512287855148315,
-      "learning_rate": 9.817369652555032e-06,
-      "loss": 0.404,
-      "step": 4801
-    },
-    {
-      "epoch": 0.31383569701326713,
-      "grad_norm": 0.44516071677207947,
-      "learning_rate": 9.817276126976645e-06,
-      "loss": 0.3895,
-      "step": 4802
-    },
-    {
-      "epoch": 0.3139010522188092,
-      "grad_norm": 0.4859333038330078,
-      "learning_rate": 9.817182577902733e-06,
-      "loss": 0.4241,
-      "step": 4803
-    },
-    {
-      "epoch": 0.31396640742435133,
-      "grad_norm": 0.44133713841438293,
-      "learning_rate": 9.817089005333754e-06,
-      "loss": 0.368,
-      "step": 4804
-    },
-    {
-      "epoch": 0.31403176262989346,
-      "grad_norm": 0.43846553564071655,
-      "learning_rate": 9.816995409270163e-06,
-      "loss": 0.3794,
-      "step": 4805
-    },
-    {
-      "epoch": 0.3140971178354356,
-      "grad_norm": 0.4645024538040161,
-      "learning_rate": 9.816901789712417e-06,
-      "loss": 0.3909,
-      "step": 4806
-    },
-    {
-      "epoch": 0.3141624730409777,
-      "grad_norm": 0.4686172902584076,
-      "learning_rate": 9.81680814666097e-06,
-      "loss": 0.3942,
-      "step": 4807
-    },
-    {
-      "epoch": 0.31422782824651985,
-      "grad_norm": 0.481277734041214,
-      "learning_rate": 9.816714480116284e-06,
-      "loss": 0.4025,
-      "step": 4808
-    },
-    {
-      "epoch": 0.314293183452062,
-      "grad_norm": 0.4675150513648987,
-      "learning_rate": 9.816620790078811e-06,
-      "loss": 0.3855,
-      "step": 4809
-    },
-    {
-      "epoch": 0.3143585386576041,
-      "grad_norm": 0.528398334980011,
-      "learning_rate": 9.81652707654901e-06,
-      "loss": 0.4621,
-      "step": 4810
-    },
-    {
-      "epoch": 0.3144238938631462,
-      "grad_norm": 0.4680522680282593,
-      "learning_rate": 9.816433339527338e-06,
-      "loss": 0.3746,
-      "step": 4811
-    },
-    {
-      "epoch": 0.3144892490686883,
-      "grad_norm": 0.4627244174480438,
-      "learning_rate": 9.816339579014253e-06,
-      "loss": 0.4109,
-      "step": 4812
-    },
-    {
-      "epoch": 0.31455460427423043,
-      "grad_norm": 0.43803465366363525,
-      "learning_rate": 9.81624579501021e-06,
-      "loss": 0.3649,
-      "step": 4813
-    },
-    {
-      "epoch": 0.31461995947977256,
-      "grad_norm": 0.49013015627861023,
-      "learning_rate": 9.816151987515669e-06,
-      "loss": 0.435,
-      "step": 4814
-    },
-    {
-      "epoch": 0.3146853146853147,
-      "grad_norm": 0.4839861989021301,
-      "learning_rate": 9.816058156531085e-06,
-      "loss": 0.4232,
-      "step": 4815
-    },
-    {
-      "epoch": 0.3147506698908568,
-      "grad_norm": 0.5070106387138367,
-      "learning_rate": 9.815964302056918e-06,
-      "loss": 0.4771,
-      "step": 4816
-    },
-    {
-      "epoch": 0.31481602509639894,
-      "grad_norm": 0.525214672088623,
-      "learning_rate": 9.815870424093623e-06,
-      "loss": 0.47,
-      "step": 4817
-    },
-    {
-      "epoch": 0.31488138030194107,
-      "grad_norm": 0.468043714761734,
-      "learning_rate": 9.81577652264166e-06,
-      "loss": 0.4163,
-      "step": 4818
-    },
-    {
-      "epoch": 0.31494673550748314,
-      "grad_norm": 0.4506476819515228,
-      "learning_rate": 9.815682597701488e-06,
-      "loss": 0.3898,
-      "step": 4819
-    },
-    {
-      "epoch": 0.3150120907130253,
-      "grad_norm": 0.45676279067993164,
-      "learning_rate": 9.815588649273565e-06,
-      "loss": 0.3912,
-      "step": 4820
-    },
-    {
-      "epoch": 0.3150774459185674,
-      "grad_norm": 0.4628327488899231,
-      "learning_rate": 9.815494677358344e-06,
-      "loss": 0.3812,
-      "step": 4821
-    },
-    {
-      "epoch": 0.31514280112410953,
-      "grad_norm": 0.4558892548084259,
-      "learning_rate": 9.815400681956288e-06,
-      "loss": 0.415,
-      "step": 4822
-    },
-    {
-      "epoch": 0.31520815632965166,
-      "grad_norm": 0.4446291923522949,
-      "learning_rate": 9.815306663067856e-06,
-      "loss": 0.3809,
-      "step": 4823
-    },
-    {
-      "epoch": 0.3152735115351938,
-      "grad_norm": 0.48152267932891846,
-      "learning_rate": 9.815212620693506e-06,
-      "loss": 0.4539,
-      "step": 4824
-    },
-    {
-      "epoch": 0.3153388667407359,
-      "grad_norm": 0.4603506326675415,
-      "learning_rate": 9.815118554833695e-06,
-      "loss": 0.3669,
-      "step": 4825
-    },
-    {
-      "epoch": 0.31540422194627804,
-      "grad_norm": 0.41335731744766235,
-      "learning_rate": 9.815024465488883e-06,
-      "loss": 0.3086,
-      "step": 4826
-    },
-    {
-      "epoch": 0.31546957715182017,
-      "grad_norm": 0.4538346827030182,
-      "learning_rate": 9.814930352659527e-06,
-      "loss": 0.4002,
-      "step": 4827
-    },
-    {
-      "epoch": 0.31553493235736224,
-      "grad_norm": 0.49037274718284607,
-      "learning_rate": 9.814836216346089e-06,
-      "loss": 0.4029,
-      "step": 4828
-    },
-    {
-      "epoch": 0.31560028756290437,
-      "grad_norm": 0.4771523177623749,
-      "learning_rate": 9.814742056549025e-06,
-      "loss": 0.3865,
-      "step": 4829
-    },
-    {
-      "epoch": 0.3156656427684465,
-      "grad_norm": 0.48241978883743286,
-      "learning_rate": 9.8146478732688e-06,
-      "loss": 0.4385,
-      "step": 4830
-    },
-    {
-      "epoch": 0.3157309979739886,
-      "grad_norm": 0.46045243740081787,
-      "learning_rate": 9.814553666505864e-06,
-      "loss": 0.4259,
-      "step": 4831
-    },
-    {
-      "epoch": 0.31579635317953075,
-      "grad_norm": 0.4759175777435303,
-      "learning_rate": 9.814459436260686e-06,
-      "loss": 0.4349,
-      "step": 4832
-    },
-    {
-      "epoch": 0.3158617083850729,
-      "grad_norm": 0.47514161467552185,
-      "learning_rate": 9.814365182533721e-06,
-      "loss": 0.4178,
-      "step": 4833
-    },
-    {
-      "epoch": 0.315927063590615,
-      "grad_norm": 0.44572609663009644,
-      "learning_rate": 9.814270905325428e-06,
-      "loss": 0.4036,
-      "step": 4834
-    },
-    {
-      "epoch": 0.31599241879615714,
-      "grad_norm": 0.44046252965927124,
-      "learning_rate": 9.814176604636268e-06,
-      "loss": 0.3579,
-      "step": 4835
-    },
-    {
-      "epoch": 0.3160577740016992,
-      "grad_norm": 0.5021373629570007,
-      "learning_rate": 9.8140822804667e-06,
-      "loss": 0.4373,
-      "step": 4836
-    },
-    {
-      "epoch": 0.31612312920724134,
-      "grad_norm": 0.46788740158081055,
-      "learning_rate": 9.813987932817185e-06,
-      "loss": 0.4139,
-      "step": 4837
-    },
-    {
-      "epoch": 0.31618848441278347,
-      "grad_norm": 0.4405035078525543,
-      "learning_rate": 9.813893561688186e-06,
-      "loss": 0.3801,
-      "step": 4838
-    },
-    {
-      "epoch": 0.3162538396183256,
-      "grad_norm": 0.4530717134475708,
-      "learning_rate": 9.813799167080157e-06,
-      "loss": 0.3852,
-      "step": 4839
-    },
-    {
-      "epoch": 0.3163191948238677,
-      "grad_norm": 0.43838727474212646,
-      "learning_rate": 9.813704748993564e-06,
-      "loss": 0.3637,
-      "step": 4840
-    },
-    {
-      "epoch": 0.31638455002940985,
-      "grad_norm": 0.46214601397514343,
-      "learning_rate": 9.813610307428866e-06,
-      "loss": 0.4053,
-      "step": 4841
-    },
-    {
-      "epoch": 0.316449905234952,
-      "grad_norm": 0.516463577747345,
-      "learning_rate": 9.81351584238652e-06,
-      "loss": 0.472,
-      "step": 4842
-    },
-    {
-      "epoch": 0.3165152604404941,
-      "grad_norm": 0.4470294117927551,
-      "learning_rate": 9.813421353866991e-06,
-      "loss": 0.3857,
-      "step": 4843
-    },
-    {
-      "epoch": 0.31658061564603623,
-      "grad_norm": 0.4604097902774811,
-      "learning_rate": 9.813326841870741e-06,
-      "loss": 0.3819,
-      "step": 4844
-    },
-    {
-      "epoch": 0.3166459708515783,
-      "grad_norm": 0.4638279378414154,
-      "learning_rate": 9.813232306398226e-06,
-      "loss": 0.3988,
-      "step": 4845
-    },
-    {
-      "epoch": 0.31671132605712043,
-      "grad_norm": 0.4969153106212616,
-      "learning_rate": 9.81313774744991e-06,
-      "loss": 0.4113,
-      "step": 4846
-    },
-    {
-      "epoch": 0.31677668126266256,
-      "grad_norm": 0.4891085922718048,
-      "learning_rate": 9.813043165026252e-06,
-      "loss": 0.3624,
-      "step": 4847
-    },
-    {
-      "epoch": 0.3168420364682047,
-      "grad_norm": 0.45764076709747314,
-      "learning_rate": 9.812948559127717e-06,
-      "loss": 0.3808,
-      "step": 4848
-    },
-    {
-      "epoch": 0.3169073916737468,
-      "grad_norm": 0.5100204348564148,
-      "learning_rate": 9.812853929754765e-06,
-      "loss": 0.4807,
-      "step": 4849
-    },
-    {
-      "epoch": 0.31697274687928895,
-      "grad_norm": 0.5149915814399719,
-      "learning_rate": 9.812759276907857e-06,
-      "loss": 0.4546,
-      "step": 4850
-    },
-    {
-      "epoch": 0.3170381020848311,
-      "grad_norm": 0.4715445339679718,
-      "learning_rate": 9.812664600587454e-06,
-      "loss": 0.4152,
-      "step": 4851
-    },
-    {
-      "epoch": 0.3171034572903732,
-      "grad_norm": 0.4651075303554535,
-      "learning_rate": 9.812569900794018e-06,
-      "loss": 0.3809,
-      "step": 4852
-    },
-    {
-      "epoch": 0.3171688124959153,
-      "grad_norm": 0.5031759142875671,
-      "learning_rate": 9.812475177528012e-06,
-      "loss": 0.4265,
-      "step": 4853
-    },
-    {
-      "epoch": 0.3172341677014574,
-      "grad_norm": 0.4432404637336731,
-      "learning_rate": 9.812380430789898e-06,
-      "loss": 0.402,
-      "step": 4854
-    },
-    {
-      "epoch": 0.31729952290699953,
-      "grad_norm": 0.4887496829032898,
-      "learning_rate": 9.812285660580136e-06,
-      "loss": 0.4299,
-      "step": 4855
-    },
-    {
-      "epoch": 0.31736487811254166,
-      "grad_norm": 0.5075775980949402,
-      "learning_rate": 9.81219086689919e-06,
-      "loss": 0.4039,
-      "step": 4856
-    },
-    {
-      "epoch": 0.3174302333180838,
-      "grad_norm": 0.4657699167728424,
-      "learning_rate": 9.812096049747524e-06,
-      "loss": 0.4107,
-      "step": 4857
-    },
-    {
-      "epoch": 0.3174955885236259,
-      "grad_norm": 0.4746883511543274,
-      "learning_rate": 9.812001209125597e-06,
-      "loss": 0.402,
-      "step": 4858
-    },
-    {
-      "epoch": 0.31756094372916804,
-      "grad_norm": 0.4992033541202545,
-      "learning_rate": 9.811906345033873e-06,
-      "loss": 0.4126,
-      "step": 4859
-    },
-    {
-      "epoch": 0.31762629893471017,
-      "grad_norm": 0.480905681848526,
-      "learning_rate": 9.811811457472813e-06,
-      "loss": 0.4132,
-      "step": 4860
-    },
-    {
-      "epoch": 0.31769165414025224,
-      "grad_norm": 0.4425283670425415,
-      "learning_rate": 9.811716546442884e-06,
-      "loss": 0.3699,
-      "step": 4861
-    },
-    {
-      "epoch": 0.3177570093457944,
-      "grad_norm": 0.46087726950645447,
-      "learning_rate": 9.811621611944547e-06,
-      "loss": 0.4009,
-      "step": 4862
-    },
-    {
-      "epoch": 0.3178223645513365,
-      "grad_norm": 0.4641392230987549,
-      "learning_rate": 9.811526653978262e-06,
-      "loss": 0.3651,
-      "step": 4863
-    },
-    {
-      "epoch": 0.31788771975687863,
-      "grad_norm": 0.49839305877685547,
-      "learning_rate": 9.811431672544496e-06,
-      "loss": 0.4535,
-      "step": 4864
-    },
-    {
-      "epoch": 0.31795307496242076,
-      "grad_norm": 0.4499477744102478,
-      "learning_rate": 9.811336667643711e-06,
-      "loss": 0.3932,
-      "step": 4865
-    },
-    {
-      "epoch": 0.3180184301679629,
-      "grad_norm": 0.4783080816268921,
-      "learning_rate": 9.81124163927637e-06,
-      "loss": 0.4182,
-      "step": 4866
-    },
-    {
-      "epoch": 0.318083785373505,
-      "grad_norm": 0.4732609689235687,
-      "learning_rate": 9.811146587442937e-06,
-      "loss": 0.3861,
-      "step": 4867
-    },
-    {
-      "epoch": 0.31814914057904714,
-      "grad_norm": 0.46453970670700073,
-      "learning_rate": 9.811051512143875e-06,
-      "loss": 0.3576,
-      "step": 4868
-    },
-    {
-      "epoch": 0.31821449578458927,
-      "grad_norm": 0.44703081250190735,
-      "learning_rate": 9.810956413379649e-06,
-      "loss": 0.349,
-      "step": 4869
-    },
-    {
-      "epoch": 0.31827985099013134,
-      "grad_norm": 0.4656713604927063,
-      "learning_rate": 9.81086129115072e-06,
-      "loss": 0.3806,
-      "step": 4870
-    },
-    {
-      "epoch": 0.31834520619567347,
-      "grad_norm": 0.46229737997055054,
-      "learning_rate": 9.810766145457556e-06,
-      "loss": 0.4023,
-      "step": 4871
-    },
-    {
-      "epoch": 0.3184105614012156,
-      "grad_norm": 0.4506048560142517,
-      "learning_rate": 9.810670976300618e-06,
-      "loss": 0.4009,
-      "step": 4872
-    },
-    {
-      "epoch": 0.3184759166067577,
-      "grad_norm": 0.5028387308120728,
-      "learning_rate": 9.810575783680369e-06,
-      "loss": 0.4266,
-      "step": 4873
-    },
-    {
-      "epoch": 0.31854127181229985,
-      "grad_norm": 0.47096943855285645,
-      "learning_rate": 9.810480567597278e-06,
-      "loss": 0.4004,
-      "step": 4874
-    },
-    {
-      "epoch": 0.318606627017842,
-      "grad_norm": 0.5066052079200745,
-      "learning_rate": 9.810385328051806e-06,
-      "loss": 0.3837,
-      "step": 4875
-    },
-    {
-      "epoch": 0.3186719822233841,
-      "grad_norm": 0.4741377830505371,
-      "learning_rate": 9.810290065044418e-06,
-      "loss": 0.3798,
-      "step": 4876
-    },
-    {
-      "epoch": 0.31873733742892624,
-      "grad_norm": 0.504368245601654,
-      "learning_rate": 9.81019477857558e-06,
-      "loss": 0.4111,
-      "step": 4877
-    },
-    {
-      "epoch": 0.3188026926344683,
-      "grad_norm": 0.44283339381217957,
-      "learning_rate": 9.810099468645756e-06,
-      "loss": 0.3882,
-      "step": 4878
-    },
-    {
-      "epoch": 0.31886804784001044,
-      "grad_norm": 0.44365501403808594,
-      "learning_rate": 9.810004135255409e-06,
-      "loss": 0.3829,
-      "step": 4879
-    },
-    {
-      "epoch": 0.31893340304555257,
-      "grad_norm": 0.4875386953353882,
-      "learning_rate": 9.809908778405006e-06,
-      "loss": 0.4594,
-      "step": 4880
-    },
-    {
-      "epoch": 0.3189987582510947,
-      "grad_norm": 0.49244803190231323,
-      "learning_rate": 9.80981339809501e-06,
-      "loss": 0.4057,
-      "step": 4881
-    },
-    {
-      "epoch": 0.3190641134566368,
-      "grad_norm": 0.4813857674598694,
-      "learning_rate": 9.80971799432589e-06,
-      "loss": 0.4603,
-      "step": 4882
-    },
-    {
-      "epoch": 0.31912946866217895,
-      "grad_norm": 0.47425293922424316,
-      "learning_rate": 9.809622567098108e-06,
-      "loss": 0.456,
-      "step": 4883
-    },
-    {
-      "epoch": 0.3191948238677211,
-      "grad_norm": 0.5302835702896118,
-      "learning_rate": 9.809527116412133e-06,
-      "loss": 0.4752,
-      "step": 4884
-    },
-    {
-      "epoch": 0.3192601790732632,
-      "grad_norm": 0.45904242992401123,
-      "learning_rate": 9.809431642268424e-06,
-      "loss": 0.4233,
-      "step": 4885
-    },
-    {
-      "epoch": 0.31932553427880533,
-      "grad_norm": 0.48194384574890137,
-      "learning_rate": 9.809336144667454e-06,
-      "loss": 0.4301,
-      "step": 4886
-    },
-    {
-      "epoch": 0.3193908894843474,
-      "grad_norm": 0.45269575715065,
-      "learning_rate": 9.809240623609683e-06,
-      "loss": 0.4144,
-      "step": 4887
-    },
-    {
-      "epoch": 0.31945624468988953,
-      "grad_norm": 0.43395453691482544,
-      "learning_rate": 9.809145079095581e-06,
-      "loss": 0.395,
-      "step": 4888
-    },
-    {
-      "epoch": 0.31952159989543166,
-      "grad_norm": 0.48187124729156494,
-      "learning_rate": 9.809049511125613e-06,
-      "loss": 0.4329,
-      "step": 4889
-    },
-    {
-      "epoch": 0.3195869551009738,
-      "grad_norm": 0.5337011218070984,
-      "learning_rate": 9.808953919700243e-06,
-      "loss": 0.4293,
-      "step": 4890
-    },
-    {
-      "epoch": 0.3196523103065159,
-      "grad_norm": 0.4339333176612854,
-      "learning_rate": 9.80885830481994e-06,
-      "loss": 0.3563,
-      "step": 4891
-    },
-    {
-      "epoch": 0.31971766551205805,
-      "grad_norm": 0.4830499291419983,
-      "learning_rate": 9.808762666485167e-06,
-      "loss": 0.4327,
-      "step": 4892
-    },
-    {
-      "epoch": 0.3197830207176002,
-      "grad_norm": 0.476817786693573,
-      "learning_rate": 9.808667004696394e-06,
-      "loss": 0.3727,
-      "step": 4893
-    },
-    {
-      "epoch": 0.3198483759231423,
-      "grad_norm": 0.5114434957504272,
-      "learning_rate": 9.808571319454085e-06,
-      "loss": 0.4815,
-      "step": 4894
-    },
-    {
-      "epoch": 0.3199137311286844,
-      "grad_norm": 0.46511292457580566,
-      "learning_rate": 9.80847561075871e-06,
-      "loss": 0.4011,
-      "step": 4895
-    },
-    {
-      "epoch": 0.3199790863342265,
-      "grad_norm": 0.49821165204048157,
-      "learning_rate": 9.808379878610732e-06,
-      "loss": 0.4239,
-      "step": 4896
-    },
-    {
-      "epoch": 0.32004444153976863,
-      "grad_norm": 0.4866395890712738,
-      "learning_rate": 9.80828412301062e-06,
-      "loss": 0.39,
-      "step": 4897
-    },
-    {
-      "epoch": 0.32010979674531076,
-      "grad_norm": 0.43412354588508606,
-      "learning_rate": 9.80818834395884e-06,
-      "loss": 0.3541,
-      "step": 4898
-    },
-    {
-      "epoch": 0.3201751519508529,
-      "grad_norm": 0.4595932960510254,
-      "learning_rate": 9.808092541455862e-06,
-      "loss": 0.4036,
-      "step": 4899
-    },
-    {
-      "epoch": 0.320240507156395,
-      "grad_norm": 0.46590059995651245,
-      "learning_rate": 9.807996715502148e-06,
-      "loss": 0.3805,
-      "step": 4900
-    },
-    {
-      "epoch": 0.32030586236193714,
-      "grad_norm": 0.5135805606842041,
-      "learning_rate": 9.80790086609817e-06,
-      "loss": 0.4093,
-      "step": 4901
-    },
-    {
-      "epoch": 0.32037121756747927,
-      "grad_norm": 0.4773502051830292,
-      "learning_rate": 9.807804993244394e-06,
-      "loss": 0.3762,
-      "step": 4902
-    },
-    {
-      "epoch": 0.32043657277302134,
-      "grad_norm": 0.44620808959007263,
-      "learning_rate": 9.807709096941287e-06,
-      "loss": 0.3836,
-      "step": 4903
-    },
-    {
-      "epoch": 0.3205019279785635,
-      "grad_norm": 0.4804670810699463,
-      "learning_rate": 9.807613177189318e-06,
-      "loss": 0.3777,
-      "step": 4904
-    },
-    {
-      "epoch": 0.3205672831841056,
-      "grad_norm": 0.44309714436531067,
-      "learning_rate": 9.807517233988952e-06,
-      "loss": 0.3586,
-      "step": 4905
-    },
-    {
-      "epoch": 0.32063263838964773,
-      "grad_norm": 0.5175981521606445,
-      "learning_rate": 9.80742126734066e-06,
-      "loss": 0.4161,
-      "step": 4906
-    },
-    {
-      "epoch": 0.32069799359518986,
-      "grad_norm": 0.514324426651001,
-      "learning_rate": 9.80732527724491e-06,
-      "loss": 0.5022,
-      "step": 4907
-    },
-    {
-      "epoch": 0.320763348800732,
-      "grad_norm": 0.4319520890712738,
-      "learning_rate": 9.807229263702169e-06,
-      "loss": 0.3607,
-      "step": 4908
-    },
-    {
-      "epoch": 0.3208287040062741,
-      "grad_norm": 0.5073283314704895,
-      "learning_rate": 9.807133226712905e-06,
-      "loss": 0.4502,
-      "step": 4909
-    },
-    {
-      "epoch": 0.32089405921181624,
-      "grad_norm": 0.4918001890182495,
-      "learning_rate": 9.807037166277586e-06,
-      "loss": 0.3678,
-      "step": 4910
-    },
-    {
-      "epoch": 0.32095941441735837,
-      "grad_norm": 0.4978146553039551,
-      "learning_rate": 9.806941082396683e-06,
-      "loss": 0.4138,
-      "step": 4911
-    },
-    {
-      "epoch": 0.32102476962290044,
-      "grad_norm": 0.47458863258361816,
-      "learning_rate": 9.806844975070662e-06,
-      "loss": 0.3617,
-      "step": 4912
-    },
-    {
-      "epoch": 0.32109012482844257,
-      "grad_norm": 0.5195441246032715,
-      "learning_rate": 9.806748844299994e-06,
-      "loss": 0.4032,
-      "step": 4913
-    },
-    {
-      "epoch": 0.3211554800339847,
-      "grad_norm": 0.4969855546951294,
-      "learning_rate": 9.806652690085146e-06,
-      "loss": 0.4334,
-      "step": 4914
-    },
-    {
-      "epoch": 0.3212208352395268,
-      "grad_norm": 0.48649662733078003,
-      "learning_rate": 9.806556512426586e-06,
-      "loss": 0.4287,
-      "step": 4915
-    },
-    {
-      "epoch": 0.32128619044506895,
-      "grad_norm": 0.49306175112724304,
-      "learning_rate": 9.806460311324787e-06,
-      "loss": 0.3836,
-      "step": 4916
-    },
-    {
-      "epoch": 0.3213515456506111,
-      "grad_norm": 0.5063831806182861,
-      "learning_rate": 9.806364086780216e-06,
-      "loss": 0.4104,
-      "step": 4917
-    },
-    {
-      "epoch": 0.3214169008561532,
-      "grad_norm": 0.4891044497489929,
-      "learning_rate": 9.80626783879334e-06,
-      "loss": 0.3936,
-      "step": 4918
-    },
-    {
-      "epoch": 0.32148225606169534,
-      "grad_norm": 0.5033397078514099,
-      "learning_rate": 9.806171567364633e-06,
-      "loss": 0.4287,
-      "step": 4919
-    },
-    {
-      "epoch": 0.3215476112672374,
-      "grad_norm": 0.5145201086997986,
-      "learning_rate": 9.806075272494562e-06,
-      "loss": 0.4038,
-      "step": 4920
-    },
-    {
-      "epoch": 0.32161296647277954,
-      "grad_norm": 0.4916360080242157,
-      "learning_rate": 9.805978954183595e-06,
-      "loss": 0.4277,
-      "step": 4921
-    },
-    {
-      "epoch": 0.32167832167832167,
-      "grad_norm": 0.5213127732276917,
-      "learning_rate": 9.805882612432205e-06,
-      "loss": 0.4565,
-      "step": 4922
-    },
-    {
-      "epoch": 0.3217436768838638,
-      "grad_norm": 0.508414626121521,
-      "learning_rate": 9.80578624724086e-06,
-      "loss": 0.3896,
-      "step": 4923
-    },
-    {
-      "epoch": 0.3218090320894059,
-      "grad_norm": 0.4595962464809418,
-      "learning_rate": 9.805689858610033e-06,
-      "loss": 0.4066,
-      "step": 4924
-    },
-    {
-      "epoch": 0.32187438729494805,
-      "grad_norm": 0.5002645254135132,
-      "learning_rate": 9.805593446540188e-06,
-      "loss": 0.4005,
-      "step": 4925
-    },
-    {
-      "epoch": 0.3219397425004902,
-      "grad_norm": 0.47116753458976746,
-      "learning_rate": 9.8054970110318e-06,
-      "loss": 0.3768,
-      "step": 4926
-    },
-    {
-      "epoch": 0.3220050977060323,
-      "grad_norm": 0.5051274299621582,
-      "learning_rate": 9.80540055208534e-06,
-      "loss": 0.4134,
-      "step": 4927
-    },
-    {
-      "epoch": 0.32207045291157443,
-      "grad_norm": 0.5116768479347229,
-      "learning_rate": 9.805304069701276e-06,
-      "loss": 0.4379,
-      "step": 4928
-    },
-    {
-      "epoch": 0.3221358081171165,
-      "grad_norm": 0.4674070477485657,
-      "learning_rate": 9.805207563880078e-06,
-      "loss": 0.4012,
-      "step": 4929
-    },
-    {
-      "epoch": 0.32220116332265863,
-      "grad_norm": 0.46312177181243896,
-      "learning_rate": 9.805111034622221e-06,
-      "loss": 0.4031,
-      "step": 4930
-    },
-    {
-      "epoch": 0.32226651852820076,
-      "grad_norm": 0.5625869035720825,
-      "learning_rate": 9.80501448192817e-06,
-      "loss": 0.4368,
-      "step": 4931
-    },
-    {
-      "epoch": 0.3223318737337429,
-      "grad_norm": 0.5150803923606873,
-      "learning_rate": 9.8049179057984e-06,
-      "loss": 0.4453,
-      "step": 4932
-    },
-    {
-      "epoch": 0.322397228939285,
-      "grad_norm": 0.49893876910209656,
-      "learning_rate": 9.80482130623338e-06,
-      "loss": 0.4186,
-      "step": 4933
-    },
-    {
-      "epoch": 0.32246258414482715,
-      "grad_norm": 0.5060268640518188,
-      "learning_rate": 9.804724683233584e-06,
-      "loss": 0.4821,
-      "step": 4934
-    },
-    {
-      "epoch": 0.3225279393503693,
-      "grad_norm": 0.4880949854850769,
-      "learning_rate": 9.80462803679948e-06,
-      "loss": 0.4248,
-      "step": 4935
-    },
-    {
-      "epoch": 0.3225932945559114,
-      "grad_norm": 0.445774644613266,
-      "learning_rate": 9.80453136693154e-06,
-      "loss": 0.3584,
-      "step": 4936
-    },
-    {
-      "epoch": 0.3226586497614535,
-      "grad_norm": 0.48850998282432556,
-      "learning_rate": 9.804434673630236e-06,
-      "loss": 0.4177,
-      "step": 4937
-    },
-    {
-      "epoch": 0.3227240049669956,
-      "grad_norm": 0.46730414032936096,
-      "learning_rate": 9.80433795689604e-06,
-      "loss": 0.4052,
-      "step": 4938
-    },
-    {
-      "epoch": 0.32278936017253773,
-      "grad_norm": 0.5012783408164978,
-      "learning_rate": 9.804241216729425e-06,
-      "loss": 0.4372,
-      "step": 4939
-    },
-    {
-      "epoch": 0.32285471537807986,
-      "grad_norm": 0.4370277523994446,
-      "learning_rate": 9.804144453130858e-06,
-      "loss": 0.3667,
-      "step": 4940
-    },
-    {
-      "epoch": 0.322920070583622,
-      "grad_norm": 0.521051287651062,
-      "learning_rate": 9.804047666100816e-06,
-      "loss": 0.4247,
-      "step": 4941
-    },
-    {
-      "epoch": 0.3229854257891641,
-      "grad_norm": 0.4966512620449066,
-      "learning_rate": 9.80395085563977e-06,
-      "loss": 0.3422,
-      "step": 4942
-    },
-    {
-      "epoch": 0.32305078099470624,
-      "grad_norm": 0.4762769639492035,
-      "learning_rate": 9.80385402174819e-06,
-      "loss": 0.4253,
-      "step": 4943
-    },
-    {
-      "epoch": 0.32311613620024837,
-      "grad_norm": 0.46476513147354126,
-      "learning_rate": 9.80375716442655e-06,
-      "loss": 0.3973,
-      "step": 4944
-    },
-    {
-      "epoch": 0.32318149140579044,
-      "grad_norm": 0.5150056481361389,
-      "learning_rate": 9.803660283675323e-06,
-      "loss": 0.4335,
-      "step": 4945
-    },
-    {
-      "epoch": 0.32324684661133257,
-      "grad_norm": 0.45104771852493286,
-      "learning_rate": 9.80356337949498e-06,
-      "loss": 0.3653,
-      "step": 4946
-    },
-    {
-      "epoch": 0.3233122018168747,
-      "grad_norm": 0.4779456555843353,
-      "learning_rate": 9.803466451885995e-06,
-      "loss": 0.3699,
-      "step": 4947
-    },
-    {
-      "epoch": 0.32337755702241683,
-      "grad_norm": 0.49230143427848816,
-      "learning_rate": 9.803369500848839e-06,
-      "loss": 0.4168,
-      "step": 4948
-    },
-    {
-      "epoch": 0.32344291222795896,
-      "grad_norm": 0.554871678352356,
-      "learning_rate": 9.803272526383985e-06,
-      "loss": 0.5157,
-      "step": 4949
-    },
-    {
-      "epoch": 0.3235082674335011,
-      "grad_norm": 0.45842504501342773,
-      "learning_rate": 9.803175528491909e-06,
-      "loss": 0.3702,
-      "step": 4950
-    },
-    {
-      "epoch": 0.3235736226390432,
-      "grad_norm": 0.4780445694923401,
-      "learning_rate": 9.80307850717308e-06,
-      "loss": 0.4227,
-      "step": 4951
-    },
-    {
-      "epoch": 0.32363897784458534,
-      "grad_norm": 0.4503099024295807,
-      "learning_rate": 9.802981462427975e-06,
-      "loss": 0.3619,
-      "step": 4952
-    },
-    {
-      "epoch": 0.32370433305012747,
-      "grad_norm": 0.970115602016449,
-      "learning_rate": 9.802884394257066e-06,
-      "loss": 0.3804,
-      "step": 4953
-    },
-    {
-      "epoch": 0.32376968825566954,
-      "grad_norm": 0.5114675760269165,
-      "learning_rate": 9.802787302660823e-06,
-      "loss": 0.3734,
-      "step": 4954
-    },
-    {
-      "epoch": 0.32383504346121167,
-      "grad_norm": 0.5145630836486816,
-      "learning_rate": 9.802690187639725e-06,
-      "loss": 0.434,
-      "step": 4955
-    },
-    {
-      "epoch": 0.3239003986667538,
-      "grad_norm": 0.46561309695243835,
-      "learning_rate": 9.802593049194243e-06,
-      "loss": 0.3827,
-      "step": 4956
-    },
-    {
-      "epoch": 0.3239657538722959,
-      "grad_norm": 0.5032932758331299,
-      "learning_rate": 9.80249588732485e-06,
-      "loss": 0.3679,
-      "step": 4957
-    },
-    {
-      "epoch": 0.32403110907783805,
-      "grad_norm": 0.4980396628379822,
-      "learning_rate": 9.80239870203202e-06,
-      "loss": 0.416,
-      "step": 4958
-    },
-    {
-      "epoch": 0.3240964642833802,
-      "grad_norm": 0.4916515052318573,
-      "learning_rate": 9.802301493316229e-06,
-      "loss": 0.3937,
-      "step": 4959
-    },
-    {
-      "epoch": 0.3241618194889223,
-      "grad_norm": 0.48291802406311035,
-      "learning_rate": 9.80220426117795e-06,
-      "loss": 0.4344,
-      "step": 4960
-    },
-    {
-      "epoch": 0.32422717469446444,
-      "grad_norm": 0.5286703705787659,
-      "learning_rate": 9.802107005617658e-06,
-      "loss": 0.4356,
-      "step": 4961
-    },
-    {
-      "epoch": 0.3242925299000065,
-      "grad_norm": 0.4947613775730133,
-      "learning_rate": 9.802009726635825e-06,
-      "loss": 0.4141,
-      "step": 4962
-    },
-    {
-      "epoch": 0.32435788510554864,
-      "grad_norm": 0.4657216966152191,
-      "learning_rate": 9.801912424232928e-06,
-      "loss": 0.4074,
-      "step": 4963
-    },
-    {
-      "epoch": 0.32442324031109077,
-      "grad_norm": 0.4662453830242157,
-      "learning_rate": 9.801815098409439e-06,
-      "loss": 0.3998,
-      "step": 4964
-    },
-    {
-      "epoch": 0.3244885955166329,
-      "grad_norm": 0.46382343769073486,
-      "learning_rate": 9.801717749165835e-06,
-      "loss": 0.371,
-      "step": 4965
-    },
-    {
-      "epoch": 0.324553950722175,
-      "grad_norm": 0.46157434582710266,
-      "learning_rate": 9.801620376502592e-06,
-      "loss": 0.3621,
-      "step": 4966
-    },
-    {
-      "epoch": 0.32461930592771715,
-      "grad_norm": 0.47322601079940796,
-      "learning_rate": 9.80152298042018e-06,
-      "loss": 0.4423,
-      "step": 4967
-    },
-    {
-      "epoch": 0.3246846611332593,
-      "grad_norm": 0.5009423494338989,
-      "learning_rate": 9.80142556091908e-06,
-      "loss": 0.4561,
-      "step": 4968
-    },
-    {
-      "epoch": 0.3247500163388014,
-      "grad_norm": 0.4587727189064026,
-      "learning_rate": 9.801328117999762e-06,
-      "loss": 0.3871,
-      "step": 4969
-    },
-    {
-      "epoch": 0.32481537154434353,
-      "grad_norm": 0.46929559111595154,
-      "learning_rate": 9.801230651662703e-06,
-      "loss": 0.4181,
-      "step": 4970
-    },
-    {
-      "epoch": 0.3248807267498856,
-      "grad_norm": 0.5121802687644958,
-      "learning_rate": 9.80113316190838e-06,
-      "loss": 0.4297,
-      "step": 4971
-    },
-    {
-      "epoch": 0.32494608195542773,
-      "grad_norm": 0.5123001933097839,
-      "learning_rate": 9.801035648737266e-06,
-      "loss": 0.4747,
-      "step": 4972
-    },
-    {
-      "epoch": 0.32501143716096986,
-      "grad_norm": 0.461913526058197,
-      "learning_rate": 9.80093811214984e-06,
-      "loss": 0.3989,
-      "step": 4973
-    },
-    {
-      "epoch": 0.325076792366512,
-      "grad_norm": 0.441133975982666,
-      "learning_rate": 9.800840552146576e-06,
-      "loss": 0.3403,
-      "step": 4974
-    },
-    {
-      "epoch": 0.3251421475720541,
-      "grad_norm": 0.4553835988044739,
-      "learning_rate": 9.800742968727947e-06,
-      "loss": 0.3905,
-      "step": 4975
-    },
-    {
-      "epoch": 0.32520750277759625,
-      "grad_norm": 0.4045025706291199,
-      "learning_rate": 9.800645361894432e-06,
-      "loss": 0.3317,
-      "step": 4976
-    },
-    {
-      "epoch": 0.3252728579831384,
-      "grad_norm": 0.4598342478275299,
-      "learning_rate": 9.800547731646508e-06,
-      "loss": 0.3968,
-      "step": 4977
-    },
-    {
-      "epoch": 0.3253382131886805,
-      "grad_norm": 0.45731550455093384,
-      "learning_rate": 9.800450077984648e-06,
-      "loss": 0.3762,
-      "step": 4978
-    },
-    {
-      "epoch": 0.3254035683942226,
-      "grad_norm": 0.504328727722168,
-      "learning_rate": 9.800352400909331e-06,
-      "loss": 0.4328,
-      "step": 4979
-    },
-    {
-      "epoch": 0.3254689235997647,
-      "grad_norm": 0.4817037582397461,
-      "learning_rate": 9.800254700421032e-06,
-      "loss": 0.4082,
-      "step": 4980
-    },
-    {
-      "epoch": 0.32553427880530683,
-      "grad_norm": 0.5107463002204895,
-      "learning_rate": 9.800156976520227e-06,
-      "loss": 0.4378,
-      "step": 4981
-    },
-    {
-      "epoch": 0.32559963401084896,
-      "grad_norm": 0.44809725880622864,
-      "learning_rate": 9.800059229207394e-06,
-      "loss": 0.3974,
-      "step": 4982
-    },
-    {
-      "epoch": 0.3256649892163911,
-      "grad_norm": 0.4267743229866028,
-      "learning_rate": 9.799961458483011e-06,
-      "loss": 0.3727,
-      "step": 4983
-    },
-    {
-      "epoch": 0.3257303444219332,
-      "grad_norm": 0.4678683876991272,
-      "learning_rate": 9.79986366434755e-06,
-      "loss": 0.41,
-      "step": 4984
-    },
-    {
-      "epoch": 0.32579569962747534,
-      "grad_norm": 0.4843780994415283,
-      "learning_rate": 9.799765846801494e-06,
-      "loss": 0.4251,
-      "step": 4985
-    },
-    {
-      "epoch": 0.32586105483301747,
-      "grad_norm": 0.4673929512500763,
-      "learning_rate": 9.799668005845315e-06,
-      "loss": 0.3929,
-      "step": 4986
-    },
-    {
-      "epoch": 0.32592641003855954,
-      "grad_norm": 0.47374188899993896,
-      "learning_rate": 9.799570141479493e-06,
-      "loss": 0.4159,
-      "step": 4987
-    },
-    {
-      "epoch": 0.32599176524410167,
-      "grad_norm": 0.443036288022995,
-      "learning_rate": 9.799472253704504e-06,
-      "loss": 0.3666,
-      "step": 4988
-    },
-    {
-      "epoch": 0.3260571204496438,
-      "grad_norm": 0.44358721375465393,
-      "learning_rate": 9.799374342520829e-06,
-      "loss": 0.3487,
-      "step": 4989
-    },
-    {
-      "epoch": 0.32612247565518593,
-      "grad_norm": 0.4454592764377594,
-      "learning_rate": 9.799276407928938e-06,
-      "loss": 0.3666,
-      "step": 4990
-    },
-    {
-      "epoch": 0.32618783086072806,
-      "grad_norm": 0.47238078713417053,
-      "learning_rate": 9.799178449929318e-06,
-      "loss": 0.3851,
-      "step": 4991
-    },
-    {
-      "epoch": 0.3262531860662702,
-      "grad_norm": 0.5328107476234436,
-      "learning_rate": 9.799080468522439e-06,
-      "loss": 0.4467,
-      "step": 4992
-    },
-    {
-      "epoch": 0.3263185412718123,
-      "grad_norm": 0.4789188504219055,
-      "learning_rate": 9.798982463708785e-06,
-      "loss": 0.4206,
-      "step": 4993
-    },
-    {
-      "epoch": 0.32638389647735444,
-      "grad_norm": 0.4428007900714874,
-      "learning_rate": 9.798884435488829e-06,
-      "loss": 0.3313,
-      "step": 4994
-    },
-    {
-      "epoch": 0.32644925168289657,
-      "grad_norm": 0.4746893346309662,
-      "learning_rate": 9.79878638386305e-06,
-      "loss": 0.4084,
-      "step": 4995
-    },
-    {
-      "epoch": 0.32651460688843864,
-      "grad_norm": 0.4389726221561432,
-      "learning_rate": 9.79868830883193e-06,
-      "loss": 0.314,
-      "step": 4996
-    },
-    {
-      "epoch": 0.32657996209398077,
-      "grad_norm": 0.45682990550994873,
-      "learning_rate": 9.798590210395943e-06,
-      "loss": 0.3771,
-      "step": 4997
-    },
-    {
-      "epoch": 0.3266453172995229,
-      "grad_norm": 0.48574742674827576,
-      "learning_rate": 9.79849208855557e-06,
-      "loss": 0.409,
-      "step": 4998
-    },
-    {
-      "epoch": 0.326710672505065,
-      "grad_norm": 0.7528083324432373,
-      "learning_rate": 9.798393943311286e-06,
-      "loss": 0.3808,
-      "step": 4999
-    },
-    {
-      "epoch": 0.32677602771060715,
-      "grad_norm": 0.5345405340194702,
-      "learning_rate": 9.798295774663576e-06,
-      "loss": 0.4426,
-      "step": 5000
-    },
-    {
-      "epoch": 0.3268413829161493,
-      "grad_norm": 0.5063375234603882,
-      "learning_rate": 9.798197582612914e-06,
-      "loss": 0.4368,
-      "step": 5001
-    },
-    {
-      "epoch": 0.3269067381216914,
-      "grad_norm": 0.4531327188014984,
-      "learning_rate": 9.79809936715978e-06,
-      "loss": 0.3938,
-      "step": 5002
-    },
-    {
-      "epoch": 0.32697209332723354,
-      "grad_norm": 0.4788791835308075,
-      "learning_rate": 9.798001128304652e-06,
-      "loss": 0.4262,
-      "step": 5003
-    },
-    {
-      "epoch": 0.3270374485327756,
-      "grad_norm": 0.4856942296028137,
-      "learning_rate": 9.79790286604801e-06,
-      "loss": 0.407,
-      "step": 5004
-    },
-    {
-      "epoch": 0.32710280373831774,
-      "grad_norm": 0.48944294452667236,
-      "learning_rate": 9.797804580390337e-06,
-      "loss": 0.3982,
-      "step": 5005
-    },
-    {
-      "epoch": 0.32716815894385987,
-      "grad_norm": 0.45362338423728943,
-      "learning_rate": 9.797706271332106e-06,
-      "loss": 0.3587,
-      "step": 5006
-    },
-    {
-      "epoch": 0.327233514149402,
-      "grad_norm": 0.4958278238773346,
-      "learning_rate": 9.7976079388738e-06,
-      "loss": 0.4072,
-      "step": 5007
-    },
-    {
-      "epoch": 0.3272988693549441,
-      "grad_norm": 0.48262861371040344,
-      "learning_rate": 9.797509583015898e-06,
-      "loss": 0.3975,
-      "step": 5008
-    },
-    {
-      "epoch": 0.32736422456048625,
-      "grad_norm": 0.48696309328079224,
-      "learning_rate": 9.79741120375888e-06,
-      "loss": 0.4068,
-      "step": 5009
-    },
-    {
-      "epoch": 0.3274295797660284,
-      "grad_norm": 0.44781172275543213,
-      "learning_rate": 9.797312801103227e-06,
-      "loss": 0.3859,
-      "step": 5010
-    },
-    {
-      "epoch": 0.3274949349715705,
-      "grad_norm": 0.4755784273147583,
-      "learning_rate": 9.797214375049416e-06,
-      "loss": 0.4057,
-      "step": 5011
-    },
-    {
-      "epoch": 0.32756029017711263,
-      "grad_norm": 0.4229740500450134,
-      "learning_rate": 9.797115925597929e-06,
-      "loss": 0.3708,
-      "step": 5012
-    },
-    {
-      "epoch": 0.3276256453826547,
-      "grad_norm": 0.4717099368572235,
-      "learning_rate": 9.797017452749245e-06,
-      "loss": 0.4248,
-      "step": 5013
-    },
-    {
-      "epoch": 0.32769100058819683,
-      "grad_norm": 0.5107245445251465,
-      "learning_rate": 9.796918956503845e-06,
-      "loss": 0.429,
-      "step": 5014
-    },
-    {
-      "epoch": 0.32775635579373896,
-      "grad_norm": 0.4377608299255371,
-      "learning_rate": 9.796820436862212e-06,
-      "loss": 0.363,
-      "step": 5015
-    },
-    {
-      "epoch": 0.3278217109992811,
-      "grad_norm": 0.45328426361083984,
-      "learning_rate": 9.79672189382482e-06,
-      "loss": 0.3815,
-      "step": 5016
-    },
-    {
-      "epoch": 0.3278870662048232,
-      "grad_norm": 0.47197818756103516,
-      "learning_rate": 9.796623327392156e-06,
-      "loss": 0.4267,
-      "step": 5017
-    },
-    {
-      "epoch": 0.32795242141036535,
-      "grad_norm": 0.5229854583740234,
-      "learning_rate": 9.796524737564697e-06,
-      "loss": 0.4421,
-      "step": 5018
-    },
-    {
-      "epoch": 0.3280177766159075,
-      "grad_norm": 0.522213876247406,
-      "learning_rate": 9.796426124342927e-06,
-      "loss": 0.4714,
-      "step": 5019
-    },
-    {
-      "epoch": 0.3280831318214496,
-      "grad_norm": 0.4774869680404663,
-      "learning_rate": 9.796327487727324e-06,
-      "loss": 0.4045,
-      "step": 5020
-    },
-    {
-      "epoch": 0.3281484870269917,
-      "grad_norm": 0.4742977023124695,
-      "learning_rate": 9.796228827718371e-06,
-      "loss": 0.4321,
-      "step": 5021
-    },
-    {
-      "epoch": 0.3282138422325338,
-      "grad_norm": 0.4220803678035736,
-      "learning_rate": 9.796130144316547e-06,
-      "loss": 0.3275,
-      "step": 5022
-    },
-    {
-      "epoch": 0.32827919743807593,
-      "grad_norm": 0.5076951384544373,
-      "learning_rate": 9.796031437522335e-06,
-      "loss": 0.4535,
-      "step": 5023
-    },
-    {
-      "epoch": 0.32834455264361806,
-      "grad_norm": 0.47955322265625,
-      "learning_rate": 9.795932707336218e-06,
-      "loss": 0.4473,
-      "step": 5024
-    },
-    {
-      "epoch": 0.3284099078491602,
-      "grad_norm": 0.4727722406387329,
-      "learning_rate": 9.795833953758674e-06,
-      "loss": 0.3898,
-      "step": 5025
-    },
-    {
-      "epoch": 0.3284752630547023,
-      "grad_norm": 1.5456161499023438,
-      "learning_rate": 9.795735176790187e-06,
-      "loss": 0.417,
-      "step": 5026
-    },
-    {
-      "epoch": 0.32854061826024444,
-      "grad_norm": 0.43937140703201294,
-      "learning_rate": 9.795636376431239e-06,
-      "loss": 0.3395,
-      "step": 5027
-    },
-    {
-      "epoch": 0.32860597346578657,
-      "grad_norm": 0.44527506828308105,
-      "learning_rate": 9.795537552682307e-06,
-      "loss": 0.3617,
-      "step": 5028
-    },
-    {
-      "epoch": 0.32867132867132864,
-      "grad_norm": 0.45572924613952637,
-      "learning_rate": 9.795438705543883e-06,
-      "loss": 0.3687,
-      "step": 5029
-    },
-    {
-      "epoch": 0.32873668387687077,
-      "grad_norm": 0.4578344523906708,
-      "learning_rate": 9.795339835016439e-06,
-      "loss": 0.3938,
-      "step": 5030
-    },
-    {
-      "epoch": 0.3288020390824129,
-      "grad_norm": 0.43646878004074097,
-      "learning_rate": 9.795240941100462e-06,
-      "loss": 0.356,
-      "step": 5031
-    },
-    {
-      "epoch": 0.32886739428795503,
-      "grad_norm": 0.4848819375038147,
-      "learning_rate": 9.795142023796434e-06,
-      "loss": 0.392,
-      "step": 5032
-    },
-    {
-      "epoch": 0.32893274949349716,
-      "grad_norm": 0.4681786298751831,
-      "learning_rate": 9.795043083104838e-06,
-      "loss": 0.3809,
-      "step": 5033
-    },
-    {
-      "epoch": 0.3289981046990393,
-      "grad_norm": 0.45695218443870544,
-      "learning_rate": 9.794944119026154e-06,
-      "loss": 0.3961,
-      "step": 5034
-    },
-    {
-      "epoch": 0.3290634599045814,
-      "grad_norm": 0.43824487924575806,
-      "learning_rate": 9.794845131560869e-06,
-      "loss": 0.3477,
-      "step": 5035
-    },
-    {
-      "epoch": 0.32912881511012354,
-      "grad_norm": 0.4756670594215393,
-      "learning_rate": 9.794746120709461e-06,
-      "loss": 0.4306,
-      "step": 5036
-    },
-    {
-      "epoch": 0.32919417031566567,
-      "grad_norm": 0.4912284016609192,
-      "learning_rate": 9.794647086472416e-06,
-      "loss": 0.4074,
-      "step": 5037
-    },
-    {
-      "epoch": 0.32925952552120774,
-      "grad_norm": 0.4648467004299164,
-      "learning_rate": 9.794548028850215e-06,
-      "loss": 0.3952,
-      "step": 5038
-    },
-    {
-      "epoch": 0.32932488072674987,
-      "grad_norm": 0.42651990056037903,
-      "learning_rate": 9.794448947843345e-06,
-      "loss": 0.3829,
-      "step": 5039
-    },
-    {
-      "epoch": 0.329390235932292,
-      "grad_norm": 0.42729347944259644,
-      "learning_rate": 9.794349843452284e-06,
-      "loss": 0.3613,
-      "step": 5040
-    },
-    {
-      "epoch": 0.3294555911378341,
-      "grad_norm": 0.4726579487323761,
-      "learning_rate": 9.794250715677518e-06,
-      "loss": 0.3726,
-      "step": 5041
-    },
-    {
-      "epoch": 0.32952094634337625,
-      "grad_norm": 0.45628806948661804,
-      "learning_rate": 9.794151564519532e-06,
-      "loss": 0.4173,
-      "step": 5042
-    },
-    {
-      "epoch": 0.3295863015489184,
-      "grad_norm": 0.43488210439682007,
-      "learning_rate": 9.794052389978806e-06,
-      "loss": 0.3653,
-      "step": 5043
-    },
-    {
-      "epoch": 0.3296516567544605,
-      "grad_norm": 0.4603349566459656,
-      "learning_rate": 9.793953192055826e-06,
-      "loss": 0.3667,
-      "step": 5044
-    },
-    {
-      "epoch": 0.32971701196000264,
-      "grad_norm": 0.4581151604652405,
-      "learning_rate": 9.793853970751077e-06,
-      "loss": 0.3619,
-      "step": 5045
-    },
-    {
-      "epoch": 0.3297823671655447,
-      "grad_norm": 0.4290536642074585,
-      "learning_rate": 9.793754726065042e-06,
-      "loss": 0.3468,
-      "step": 5046
-    },
-    {
-      "epoch": 0.32984772237108684,
-      "grad_norm": 0.4620216488838196,
-      "learning_rate": 9.793655457998202e-06,
-      "loss": 0.3679,
-      "step": 5047
-    },
-    {
-      "epoch": 0.32991307757662897,
-      "grad_norm": 0.4647311568260193,
-      "learning_rate": 9.793556166551045e-06,
-      "loss": 0.3801,
-      "step": 5048
-    },
-    {
-      "epoch": 0.3299784327821711,
-      "grad_norm": 0.4497399628162384,
-      "learning_rate": 9.793456851724053e-06,
-      "loss": 0.3709,
-      "step": 5049
-    },
-    {
-      "epoch": 0.3300437879877132,
-      "grad_norm": 0.4660644829273224,
-      "learning_rate": 9.793357513517711e-06,
-      "loss": 0.4166,
-      "step": 5050
-    },
-    {
-      "epoch": 0.33010914319325535,
-      "grad_norm": 0.4483477473258972,
-      "learning_rate": 9.793258151932505e-06,
-      "loss": 0.4081,
-      "step": 5051
-    },
-    {
-      "epoch": 0.3301744983987975,
-      "grad_norm": 0.4496282637119293,
-      "learning_rate": 9.79315876696892e-06,
-      "loss": 0.3476,
-      "step": 5052
-    },
-    {
-      "epoch": 0.3302398536043396,
-      "grad_norm": 0.45826923847198486,
-      "learning_rate": 9.793059358627437e-06,
-      "loss": 0.4207,
-      "step": 5053
-    },
-    {
-      "epoch": 0.33030520880988173,
-      "grad_norm": 0.47262704372406006,
-      "learning_rate": 9.792959926908543e-06,
-      "loss": 0.4522,
-      "step": 5054
-    },
-    {
-      "epoch": 0.3303705640154238,
-      "grad_norm": 0.4803999662399292,
-      "learning_rate": 9.792860471812723e-06,
-      "loss": 0.4425,
-      "step": 5055
-    },
-    {
-      "epoch": 0.33043591922096593,
-      "grad_norm": 0.5056366920471191,
-      "learning_rate": 9.792760993340463e-06,
-      "loss": 0.4312,
-      "step": 5056
-    },
-    {
-      "epoch": 0.33050127442650806,
-      "grad_norm": 0.480058491230011,
-      "learning_rate": 9.792661491492247e-06,
-      "loss": 0.4309,
-      "step": 5057
-    },
-    {
-      "epoch": 0.3305666296320502,
-      "grad_norm": 0.45647209882736206,
-      "learning_rate": 9.79256196626856e-06,
-      "loss": 0.4007,
-      "step": 5058
-    },
-    {
-      "epoch": 0.3306319848375923,
-      "grad_norm": 0.4690300524234772,
-      "learning_rate": 9.792462417669887e-06,
-      "loss": 0.4277,
-      "step": 5059
-    },
-    {
-      "epoch": 0.33069734004313445,
-      "grad_norm": 0.4632362127304077,
-      "learning_rate": 9.792362845696716e-06,
-      "loss": 0.3772,
-      "step": 5060
-    },
-    {
-      "epoch": 0.3307626952486766,
-      "grad_norm": 0.45691102743148804,
-      "learning_rate": 9.792263250349532e-06,
-      "loss": 0.4073,
-      "step": 5061
-    },
-    {
-      "epoch": 0.3308280504542187,
-      "grad_norm": 0.47819358110427856,
-      "learning_rate": 9.79216363162882e-06,
-      "loss": 0.4256,
-      "step": 5062
-    },
-    {
-      "epoch": 0.3308934056597608,
-      "grad_norm": 0.508348286151886,
-      "learning_rate": 9.792063989535064e-06,
-      "loss": 0.4182,
-      "step": 5063
-    },
-    {
-      "epoch": 0.3309587608653029,
-      "grad_norm": 0.48340311646461487,
-      "learning_rate": 9.791964324068753e-06,
-      "loss": 0.3991,
-      "step": 5064
-    },
-    {
-      "epoch": 0.33102411607084503,
-      "grad_norm": 0.49286407232284546,
-      "learning_rate": 9.791864635230372e-06,
-      "loss": 0.429,
-      "step": 5065
-    },
-    {
-      "epoch": 0.33108947127638716,
-      "grad_norm": 0.4674474895000458,
-      "learning_rate": 9.791764923020407e-06,
-      "loss": 0.3437,
-      "step": 5066
-    },
-    {
-      "epoch": 0.3311548264819293,
-      "grad_norm": 0.45047426223754883,
-      "learning_rate": 9.791665187439344e-06,
-      "loss": 0.3666,
-      "step": 5067
-    },
-    {
-      "epoch": 0.3312201816874714,
-      "grad_norm": 0.5228461027145386,
-      "learning_rate": 9.791565428487668e-06,
-      "loss": 0.4622,
-      "step": 5068
-    },
-    {
-      "epoch": 0.33128553689301354,
-      "grad_norm": 0.48304784297943115,
-      "learning_rate": 9.79146564616587e-06,
-      "loss": 0.4281,
-      "step": 5069
-    },
-    {
-      "epoch": 0.33135089209855567,
-      "grad_norm": 0.466929167509079,
-      "learning_rate": 9.791365840474434e-06,
-      "loss": 0.3524,
-      "step": 5070
-    },
-    {
-      "epoch": 0.33141624730409774,
-      "grad_norm": 0.5755605697631836,
-      "learning_rate": 9.791266011413846e-06,
-      "loss": 0.4478,
-      "step": 5071
-    },
-    {
-      "epoch": 0.33148160250963987,
-      "grad_norm": 0.5303314328193665,
-      "learning_rate": 9.791166158984593e-06,
-      "loss": 0.4259,
-      "step": 5072
-    },
-    {
-      "epoch": 0.331546957715182,
-      "grad_norm": 0.48098957538604736,
-      "learning_rate": 9.791066283187165e-06,
-      "loss": 0.3562,
-      "step": 5073
-    },
-    {
-      "epoch": 0.33161231292072413,
-      "grad_norm": 0.46910491585731506,
-      "learning_rate": 9.790966384022047e-06,
-      "loss": 0.3722,
-      "step": 5074
-    },
-    {
-      "epoch": 0.33167766812626626,
-      "grad_norm": 0.4961332380771637,
-      "learning_rate": 9.790866461489725e-06,
-      "loss": 0.3984,
-      "step": 5075
-    },
-    {
-      "epoch": 0.3317430233318084,
-      "grad_norm": 0.49478527903556824,
-      "learning_rate": 9.790766515590688e-06,
-      "loss": 0.4149,
-      "step": 5076
-    },
-    {
-      "epoch": 0.3318083785373505,
-      "grad_norm": 0.47687938809394836,
-      "learning_rate": 9.790666546325422e-06,
-      "loss": 0.4127,
-      "step": 5077
-    },
-    {
-      "epoch": 0.33187373374289264,
-      "grad_norm": 0.4761548340320587,
-      "learning_rate": 9.790566553694415e-06,
-      "loss": 0.3927,
-      "step": 5078
-    },
-    {
-      "epoch": 0.33193908894843477,
-      "grad_norm": 0.49564751982688904,
-      "learning_rate": 9.790466537698157e-06,
-      "loss": 0.4073,
-      "step": 5079
-    },
-    {
-      "epoch": 0.33200444415397684,
-      "grad_norm": 0.46131080389022827,
-      "learning_rate": 9.790366498337134e-06,
-      "loss": 0.368,
-      "step": 5080
-    },
-    {
-      "epoch": 0.33206979935951897,
-      "grad_norm": 0.46849748492240906,
-      "learning_rate": 9.790266435611835e-06,
-      "loss": 0.3266,
-      "step": 5081
-    },
-    {
-      "epoch": 0.3321351545650611,
-      "grad_norm": 0.45079654455184937,
-      "learning_rate": 9.790166349522745e-06,
-      "loss": 0.3825,
-      "step": 5082
-    },
-    {
-      "epoch": 0.3322005097706032,
-      "grad_norm": 0.49253472685813904,
-      "learning_rate": 9.790066240070355e-06,
-      "loss": 0.4122,
-      "step": 5083
-    },
-    {
-      "epoch": 0.33226586497614535,
-      "grad_norm": 0.45975279808044434,
-      "learning_rate": 9.789966107255154e-06,
-      "loss": 0.3643,
-      "step": 5084
-    },
-    {
-      "epoch": 0.3323312201816875,
-      "grad_norm": 0.48659080266952515,
-      "learning_rate": 9.789865951077626e-06,
-      "loss": 0.4521,
-      "step": 5085
-    },
-    {
-      "epoch": 0.3323965753872296,
-      "grad_norm": 0.5242331027984619,
-      "learning_rate": 9.789765771538264e-06,
-      "loss": 0.4395,
-      "step": 5086
-    },
-    {
-      "epoch": 0.33246193059277174,
-      "grad_norm": 0.5002502799034119,
-      "learning_rate": 9.789665568637556e-06,
-      "loss": 0.4354,
-      "step": 5087
-    },
-    {
-      "epoch": 0.3325272857983138,
-      "grad_norm": 0.4841715395450592,
-      "learning_rate": 9.789565342375989e-06,
-      "loss": 0.379,
-      "step": 5088
-    },
-    {
-      "epoch": 0.33259264100385594,
-      "grad_norm": 0.4892784357070923,
-      "learning_rate": 9.78946509275405e-06,
-      "loss": 0.4251,
-      "step": 5089
-    },
-    {
-      "epoch": 0.33265799620939807,
-      "grad_norm": 0.4734096825122833,
-      "learning_rate": 9.789364819772233e-06,
-      "loss": 0.4023,
-      "step": 5090
-    },
-    {
-      "epoch": 0.3327233514149402,
-      "grad_norm": 0.515963077545166,
-      "learning_rate": 9.789264523431026e-06,
-      "loss": 0.4709,
-      "step": 5091
-    },
-    {
-      "epoch": 0.3327887066204823,
-      "grad_norm": 0.48209792375564575,
-      "learning_rate": 9.789164203730915e-06,
-      "loss": 0.4056,
-      "step": 5092
-    },
-    {
-      "epoch": 0.33285406182602445,
-      "grad_norm": 0.5084713697433472,
-      "learning_rate": 9.78906386067239e-06,
-      "loss": 0.4286,
-      "step": 5093
-    },
-    {
-      "epoch": 0.3329194170315666,
-      "grad_norm": 0.4530640244483948,
-      "learning_rate": 9.788963494255943e-06,
-      "loss": 0.3542,
-      "step": 5094
-    },
-    {
-      "epoch": 0.3329847722371087,
-      "grad_norm": 0.4868583083152771,
-      "learning_rate": 9.788863104482062e-06,
-      "loss": 0.4345,
-      "step": 5095
-    },
-    {
-      "epoch": 0.33305012744265083,
-      "grad_norm": 0.45594459772109985,
-      "learning_rate": 9.788762691351235e-06,
-      "loss": 0.3618,
-      "step": 5096
-    },
-    {
-      "epoch": 0.3331154826481929,
-      "grad_norm": 0.4857982397079468,
-      "learning_rate": 9.788662254863955e-06,
-      "loss": 0.4055,
-      "step": 5097
-    },
-    {
-      "epoch": 0.33318083785373503,
-      "grad_norm": 0.4911488890647888,
-      "learning_rate": 9.78856179502071e-06,
-      "loss": 0.3894,
-      "step": 5098
-    },
-    {
-      "epoch": 0.33324619305927716,
-      "grad_norm": 0.5101039409637451,
-      "learning_rate": 9.788461311821989e-06,
-      "loss": 0.4088,
-      "step": 5099
-    },
-    {
-      "epoch": 0.3333115482648193,
-      "grad_norm": 0.4890715777873993,
-      "learning_rate": 9.788360805268285e-06,
-      "loss": 0.4507,
-      "step": 5100
-    },
-    {
-      "epoch": 0.3333769034703614,
-      "grad_norm": 0.4909462034702301,
-      "learning_rate": 9.788260275360087e-06,
-      "loss": 0.437,
-      "step": 5101
-    },
-    {
-      "epoch": 0.33344225867590355,
-      "grad_norm": 0.4736488163471222,
-      "learning_rate": 9.788159722097883e-06,
-      "loss": 0.4016,
-      "step": 5102
-    },
-    {
-      "epoch": 0.3335076138814457,
-      "grad_norm": 0.5087499618530273,
-      "learning_rate": 9.788059145482166e-06,
-      "loss": 0.454,
-      "step": 5103
-    },
-    {
-      "epoch": 0.3335729690869878,
-      "grad_norm": 0.4919663965702057,
-      "learning_rate": 9.787958545513425e-06,
-      "loss": 0.4448,
-      "step": 5104
-    },
-    {
-      "epoch": 0.3336383242925299,
-      "grad_norm": 0.49457287788391113,
-      "learning_rate": 9.787857922192151e-06,
-      "loss": 0.3984,
-      "step": 5105
-    },
-    {
-      "epoch": 0.333703679498072,
-      "grad_norm": 0.4914587736129761,
-      "learning_rate": 9.787757275518837e-06,
-      "loss": 0.4475,
-      "step": 5106
-    },
-    {
-      "epoch": 0.33376903470361413,
-      "grad_norm": 0.4439575672149658,
-      "learning_rate": 9.787656605493971e-06,
-      "loss": 0.3585,
-      "step": 5107
-    },
-    {
-      "epoch": 0.33383438990915626,
-      "grad_norm": 0.44091567397117615,
-      "learning_rate": 9.787555912118047e-06,
-      "loss": 0.3979,
-      "step": 5108
-    },
-    {
-      "epoch": 0.3338997451146984,
-      "grad_norm": 0.45893609523773193,
-      "learning_rate": 9.787455195391554e-06,
-      "loss": 0.3796,
-      "step": 5109
-    },
-    {
-      "epoch": 0.3339651003202405,
-      "grad_norm": 0.44441351294517517,
-      "learning_rate": 9.787354455314981e-06,
-      "loss": 0.3589,
-      "step": 5110
-    },
-    {
-      "epoch": 0.33403045552578264,
-      "grad_norm": 0.48110440373420715,
-      "learning_rate": 9.787253691888822e-06,
-      "loss": 0.3875,
-      "step": 5111
-    },
-    {
-      "epoch": 0.33409581073132477,
-      "grad_norm": 0.46581965684890747,
-      "learning_rate": 9.78715290511357e-06,
-      "loss": 0.4347,
-      "step": 5112
-    },
-    {
-      "epoch": 0.33416116593686684,
-      "grad_norm": 0.46184372901916504,
-      "learning_rate": 9.787052094989716e-06,
-      "loss": 0.3464,
-      "step": 5113
-    },
-    {
-      "epoch": 0.33422652114240897,
-      "grad_norm": 0.45507046580314636,
-      "learning_rate": 9.786951261517747e-06,
-      "loss": 0.3647,
-      "step": 5114
-    },
-    {
-      "epoch": 0.3342918763479511,
-      "grad_norm": 0.4871494174003601,
-      "learning_rate": 9.78685040469816e-06,
-      "loss": 0.3937,
-      "step": 5115
-    },
-    {
-      "epoch": 0.33435723155349323,
-      "grad_norm": 0.48447662591934204,
-      "learning_rate": 9.786749524531446e-06,
-      "loss": 0.379,
-      "step": 5116
-    },
-    {
-      "epoch": 0.33442258675903536,
-      "grad_norm": 0.5180823802947998,
-      "learning_rate": 9.786648621018096e-06,
-      "loss": 0.4582,
-      "step": 5117
-    },
-    {
-      "epoch": 0.3344879419645775,
-      "grad_norm": 0.46476200222969055,
-      "learning_rate": 9.786547694158602e-06,
-      "loss": 0.3645,
-      "step": 5118
-    },
-    {
-      "epoch": 0.3345532971701196,
-      "grad_norm": 0.46656325459480286,
-      "learning_rate": 9.786446743953457e-06,
-      "loss": 0.3696,
-      "step": 5119
-    },
-    {
-      "epoch": 0.33461865237566174,
-      "grad_norm": 0.44941627979278564,
-      "learning_rate": 9.786345770403153e-06,
-      "loss": 0.3945,
-      "step": 5120
-    },
-    {
-      "epoch": 0.33468400758120387,
-      "grad_norm": 0.483822762966156,
-      "learning_rate": 9.786244773508182e-06,
-      "loss": 0.4445,
-      "step": 5121
-    },
-    {
-      "epoch": 0.33474936278674594,
-      "grad_norm": 0.4862872362136841,
-      "learning_rate": 9.786143753269038e-06,
-      "loss": 0.3872,
-      "step": 5122
-    },
-    {
-      "epoch": 0.33481471799228807,
-      "grad_norm": 0.4959210753440857,
-      "learning_rate": 9.786042709686212e-06,
-      "loss": 0.4456,
-      "step": 5123
-    },
-    {
-      "epoch": 0.3348800731978302,
-      "grad_norm": 0.48525673151016235,
-      "learning_rate": 9.785941642760198e-06,
-      "loss": 0.4144,
-      "step": 5124
-    },
-    {
-      "epoch": 0.3349454284033723,
-      "grad_norm": 0.5273424386978149,
-      "learning_rate": 9.785840552491488e-06,
-      "loss": 0.4137,
-      "step": 5125
-    },
-    {
-      "epoch": 0.33501078360891445,
-      "grad_norm": 0.43980926275253296,
-      "learning_rate": 9.785739438880577e-06,
-      "loss": 0.3595,
-      "step": 5126
-    },
-    {
-      "epoch": 0.3350761388144566,
-      "grad_norm": 0.44354137778282166,
-      "learning_rate": 9.785638301927956e-06,
-      "loss": 0.3639,
-      "step": 5127
-    },
-    {
-      "epoch": 0.3351414940199987,
-      "grad_norm": 0.4843825101852417,
-      "learning_rate": 9.785537141634118e-06,
-      "loss": 0.3806,
-      "step": 5128
-    },
-    {
-      "epoch": 0.33520684922554084,
-      "grad_norm": 0.47106778621673584,
-      "learning_rate": 9.78543595799956e-06,
-      "loss": 0.4186,
-      "step": 5129
-    },
-    {
-      "epoch": 0.3352722044310829,
-      "grad_norm": 0.48503050208091736,
-      "learning_rate": 9.78533475102477e-06,
-      "loss": 0.3921,
-      "step": 5130
-    },
-    {
-      "epoch": 0.33533755963662504,
-      "grad_norm": 0.5354419946670532,
-      "learning_rate": 9.785233520710248e-06,
-      "loss": 0.5246,
-      "step": 5131
-    },
-    {
-      "epoch": 0.33540291484216717,
-      "grad_norm": 0.5101272463798523,
-      "learning_rate": 9.785132267056483e-06,
-      "loss": 0.44,
-      "step": 5132
-    },
-    {
-      "epoch": 0.3354682700477093,
-      "grad_norm": 0.44861114025115967,
-      "learning_rate": 9.785030990063968e-06,
-      "loss": 0.3906,
-      "step": 5133
-    },
-    {
-      "epoch": 0.3355336252532514,
-      "grad_norm": 0.47585704922676086,
-      "learning_rate": 9.784929689733202e-06,
-      "loss": 0.4409,
-      "step": 5134
-    },
-    {
-      "epoch": 0.33559898045879355,
-      "grad_norm": 0.4529799222946167,
-      "learning_rate": 9.784828366064677e-06,
-      "loss": 0.3843,
-      "step": 5135
-    },
-    {
-      "epoch": 0.3356643356643357,
-      "grad_norm": 0.427306592464447,
-      "learning_rate": 9.784727019058884e-06,
-      "loss": 0.3343,
-      "step": 5136
-    },
-    {
-      "epoch": 0.3357296908698778,
-      "grad_norm": 0.5161789655685425,
-      "learning_rate": 9.78462564871632e-06,
-      "loss": 0.4282,
-      "step": 5137
-    },
-    {
-      "epoch": 0.33579504607541993,
-      "grad_norm": 0.5111210346221924,
-      "learning_rate": 9.78452425503748e-06,
-      "loss": 0.4301,
-      "step": 5138
-    },
-    {
-      "epoch": 0.335860401280962,
-      "grad_norm": 0.4565163850784302,
-      "learning_rate": 9.784422838022855e-06,
-      "loss": 0.3852,
-      "step": 5139
-    },
-    {
-      "epoch": 0.33592575648650413,
-      "grad_norm": 0.46572020649909973,
-      "learning_rate": 9.784321397672947e-06,
-      "loss": 0.3753,
-      "step": 5140
-    },
-    {
-      "epoch": 0.33599111169204626,
-      "grad_norm": 0.5026994347572327,
-      "learning_rate": 9.784219933988242e-06,
-      "loss": 0.4427,
-      "step": 5141
-    },
-    {
-      "epoch": 0.3360564668975884,
-      "grad_norm": 0.4901333153247833,
-      "learning_rate": 9.784118446969241e-06,
-      "loss": 0.422,
-      "step": 5142
-    },
-    {
-      "epoch": 0.3361218221031305,
-      "grad_norm": 0.45889827609062195,
-      "learning_rate": 9.784016936616436e-06,
-      "loss": 0.4112,
-      "step": 5143
-    },
-    {
-      "epoch": 0.33618717730867265,
-      "grad_norm": 0.43094539642333984,
-      "learning_rate": 9.783915402930324e-06,
-      "loss": 0.3826,
-      "step": 5144
-    },
-    {
-      "epoch": 0.3362525325142148,
-      "grad_norm": 0.500842809677124,
-      "learning_rate": 9.7838138459114e-06,
-      "loss": 0.4607,
-      "step": 5145
-    },
-    {
-      "epoch": 0.3363178877197569,
-      "grad_norm": 0.5126469731330872,
-      "learning_rate": 9.783712265560157e-06,
-      "loss": 0.4148,
-      "step": 5146
-    },
-    {
-      "epoch": 0.336383242925299,
-      "grad_norm": 0.4601035714149475,
-      "learning_rate": 9.783610661877093e-06,
-      "loss": 0.4182,
-      "step": 5147
-    },
-    {
-      "epoch": 0.3364485981308411,
-      "grad_norm": 0.43518704175949097,
-      "learning_rate": 9.783509034862702e-06,
-      "loss": 0.3486,
-      "step": 5148
-    },
-    {
-      "epoch": 0.33651395333638323,
-      "grad_norm": 0.45159706473350525,
-      "learning_rate": 9.78340738451748e-06,
-      "loss": 0.3528,
-      "step": 5149
-    },
-    {
-      "epoch": 0.33657930854192536,
-      "grad_norm": 0.42190390825271606,
-      "learning_rate": 9.783305710841923e-06,
-      "loss": 0.3731,
-      "step": 5150
-    },
-    {
-      "epoch": 0.3366446637474675,
-      "grad_norm": 0.47011086344718933,
-      "learning_rate": 9.78320401383653e-06,
-      "loss": 0.4203,
-      "step": 5151
-    },
-    {
-      "epoch": 0.3367100189530096,
-      "grad_norm": 0.43599116802215576,
-      "learning_rate": 9.78310229350179e-06,
-      "loss": 0.336,
-      "step": 5152
-    },
-    {
-      "epoch": 0.33677537415855174,
-      "grad_norm": 0.4630354046821594,
-      "learning_rate": 9.783000549838205e-06,
-      "loss": 0.3942,
-      "step": 5153
-    },
-    {
-      "epoch": 0.33684072936409387,
-      "grad_norm": 0.4692355692386627,
-      "learning_rate": 9.782898782846268e-06,
-      "loss": 0.3973,
-      "step": 5154
-    },
-    {
-      "epoch": 0.33690608456963594,
-      "grad_norm": 0.45558837056159973,
-      "learning_rate": 9.782796992526478e-06,
-      "loss": 0.4409,
-      "step": 5155
-    },
-    {
-      "epoch": 0.33697143977517807,
-      "grad_norm": 0.4604448974132538,
-      "learning_rate": 9.782695178879328e-06,
-      "loss": 0.397,
-      "step": 5156
-    },
-    {
-      "epoch": 0.3370367949807202,
-      "grad_norm": 0.4689256548881531,
-      "learning_rate": 9.78259334190532e-06,
-      "loss": 0.4044,
-      "step": 5157
-    },
-    {
-      "epoch": 0.33710215018626233,
-      "grad_norm": 0.4754765033721924,
-      "learning_rate": 9.782491481604945e-06,
-      "loss": 0.4247,
-      "step": 5158
-    },
-    {
-      "epoch": 0.33716750539180446,
-      "grad_norm": 0.4838542342185974,
-      "learning_rate": 9.782389597978703e-06,
-      "loss": 0.4363,
-      "step": 5159
-    },
-    {
-      "epoch": 0.3372328605973466,
-      "grad_norm": 0.428602933883667,
-      "learning_rate": 9.782287691027092e-06,
-      "loss": 0.3475,
-      "step": 5160
-    },
-    {
-      "epoch": 0.3372982158028887,
-      "grad_norm": 0.4951384961605072,
-      "learning_rate": 9.782185760750605e-06,
-      "loss": 0.3946,
-      "step": 5161
-    },
-    {
-      "epoch": 0.33736357100843084,
-      "grad_norm": 0.47937989234924316,
-      "learning_rate": 9.782083807149741e-06,
-      "loss": 0.4453,
-      "step": 5162
-    },
-    {
-      "epoch": 0.33742892621397297,
-      "grad_norm": 0.44877153635025024,
-      "learning_rate": 9.781981830224998e-06,
-      "loss": 0.385,
-      "step": 5163
-    },
-    {
-      "epoch": 0.33749428141951504,
-      "grad_norm": 0.4709230065345764,
-      "learning_rate": 9.781879829976875e-06,
-      "loss": 0.3853,
-      "step": 5164
-    },
-    {
-      "epoch": 0.33755963662505717,
-      "grad_norm": 0.48274341225624084,
-      "learning_rate": 9.781777806405866e-06,
-      "loss": 0.4037,
-      "step": 5165
-    },
-    {
-      "epoch": 0.3376249918305993,
-      "grad_norm": 0.43445566296577454,
-      "learning_rate": 9.781675759512468e-06,
-      "loss": 0.362,
-      "step": 5166
-    },
-    {
-      "epoch": 0.3376903470361414,
-      "grad_norm": 0.4579179883003235,
-      "learning_rate": 9.781573689297183e-06,
-      "loss": 0.3854,
-      "step": 5167
-    },
-    {
-      "epoch": 0.33775570224168355,
-      "grad_norm": 0.4639510214328766,
-      "learning_rate": 9.781471595760507e-06,
-      "loss": 0.4206,
-      "step": 5168
-    },
-    {
-      "epoch": 0.3378210574472257,
-      "grad_norm": 0.4690952003002167,
-      "learning_rate": 9.781369478902936e-06,
-      "loss": 0.4138,
-      "step": 5169
-    },
-    {
-      "epoch": 0.3378864126527678,
-      "grad_norm": 0.45831334590911865,
-      "learning_rate": 9.781267338724971e-06,
-      "loss": 0.4038,
-      "step": 5170
-    },
-    {
-      "epoch": 0.33795176785830994,
-      "grad_norm": 0.46414628624916077,
-      "learning_rate": 9.781165175227108e-06,
-      "loss": 0.4025,
-      "step": 5171
-    },
-    {
-      "epoch": 0.338017123063852,
-      "grad_norm": 0.43977731466293335,
-      "learning_rate": 9.781062988409846e-06,
-      "loss": 0.374,
-      "step": 5172
-    },
-    {
-      "epoch": 0.33808247826939414,
-      "grad_norm": 0.46739593148231506,
-      "learning_rate": 9.780960778273685e-06,
-      "loss": 0.4351,
-      "step": 5173
-    },
-    {
-      "epoch": 0.33814783347493627,
-      "grad_norm": 0.4551723301410675,
-      "learning_rate": 9.78085854481912e-06,
-      "loss": 0.3894,
-      "step": 5174
-    },
-    {
-      "epoch": 0.3382131886804784,
-      "grad_norm": 0.4585324227809906,
-      "learning_rate": 9.780756288046653e-06,
-      "loss": 0.3566,
-      "step": 5175
-    },
-    {
-      "epoch": 0.3382785438860205,
-      "grad_norm": 0.4442490339279175,
-      "learning_rate": 9.78065400795678e-06,
-      "loss": 0.3983,
-      "step": 5176
-    },
-    {
-      "epoch": 0.33834389909156265,
-      "grad_norm": 0.4540204405784607,
-      "learning_rate": 9.780551704550003e-06,
-      "loss": 0.3662,
-      "step": 5177
-    },
-    {
-      "epoch": 0.3384092542971048,
-      "grad_norm": 0.5195617079734802,
-      "learning_rate": 9.780449377826818e-06,
-      "loss": 0.4317,
-      "step": 5178
-    },
-    {
-      "epoch": 0.3384746095026469,
-      "grad_norm": 0.4470427632331848,
-      "learning_rate": 9.780347027787726e-06,
-      "loss": 0.3786,
-      "step": 5179
-    },
-    {
-      "epoch": 0.33853996470818903,
-      "grad_norm": 0.469748854637146,
-      "learning_rate": 9.780244654433224e-06,
-      "loss": 0.3994,
-      "step": 5180
-    },
-    {
-      "epoch": 0.3386053199137311,
-      "grad_norm": 0.48107877373695374,
-      "learning_rate": 9.780142257763815e-06,
-      "loss": 0.436,
-      "step": 5181
-    },
-    {
-      "epoch": 0.33867067511927323,
-      "grad_norm": 0.4793350398540497,
-      "learning_rate": 9.780039837779994e-06,
-      "loss": 0.3486,
-      "step": 5182
-    },
-    {
-      "epoch": 0.33873603032481536,
-      "grad_norm": 0.4753890633583069,
-      "learning_rate": 9.779937394482263e-06,
-      "loss": 0.4229,
-      "step": 5183
-    },
-    {
-      "epoch": 0.3388013855303575,
-      "grad_norm": 0.47949421405792236,
-      "learning_rate": 9.779834927871124e-06,
-      "loss": 0.3871,
-      "step": 5184
-    },
-    {
-      "epoch": 0.3388667407358996,
-      "grad_norm": 0.4666721522808075,
-      "learning_rate": 9.779732437947072e-06,
-      "loss": 0.393,
-      "step": 5185
-    },
-    {
-      "epoch": 0.33893209594144175,
-      "grad_norm": 0.46229565143585205,
-      "learning_rate": 9.779629924710608e-06,
-      "loss": 0.4183,
-      "step": 5186
-    },
-    {
-      "epoch": 0.3389974511469839,
-      "grad_norm": 0.4590091407299042,
-      "learning_rate": 9.779527388162236e-06,
-      "loss": 0.3985,
-      "step": 5187
-    },
-    {
-      "epoch": 0.339062806352526,
-      "grad_norm": 0.4367179274559021,
-      "learning_rate": 9.77942482830245e-06,
-      "loss": 0.3498,
-      "step": 5188
-    },
-    {
-      "epoch": 0.3391281615580681,
-      "grad_norm": 0.4872339963912964,
-      "learning_rate": 9.779322245131755e-06,
-      "loss": 0.4351,
-      "step": 5189
-    },
-    {
-      "epoch": 0.3391935167636102,
-      "grad_norm": 0.43242597579956055,
-      "learning_rate": 9.77921963865065e-06,
-      "loss": 0.33,
-      "step": 5190
-    },
-    {
-      "epoch": 0.33925887196915233,
-      "grad_norm": 0.4527781009674072,
-      "learning_rate": 9.779117008859635e-06,
-      "loss": 0.3354,
-      "step": 5191
-    },
-    {
-      "epoch": 0.33932422717469446,
-      "grad_norm": 0.45935195684432983,
-      "learning_rate": 9.77901435575921e-06,
-      "loss": 0.397,
-      "step": 5192
-    },
-    {
-      "epoch": 0.3393895823802366,
-      "grad_norm": 0.4584974944591522,
-      "learning_rate": 9.778911679349877e-06,
-      "loss": 0.4017,
-      "step": 5193
-    },
-    {
-      "epoch": 0.3394549375857787,
-      "grad_norm": 0.4841710925102234,
-      "learning_rate": 9.778808979632136e-06,
-      "loss": 0.3804,
-      "step": 5194
-    },
-    {
-      "epoch": 0.33952029279132084,
-      "grad_norm": 0.46480992436408997,
-      "learning_rate": 9.778706256606488e-06,
-      "loss": 0.4033,
-      "step": 5195
-    },
-    {
-      "epoch": 0.33958564799686297,
-      "grad_norm": 0.4626947343349457,
-      "learning_rate": 9.778603510273435e-06,
-      "loss": 0.3508,
-      "step": 5196
-    },
-    {
-      "epoch": 0.33965100320240504,
-      "grad_norm": 0.48210132122039795,
-      "learning_rate": 9.778500740633476e-06,
-      "loss": 0.4048,
-      "step": 5197
-    },
-    {
-      "epoch": 0.33971635840794717,
-      "grad_norm": 0.4198559820652008,
-      "learning_rate": 9.778397947687114e-06,
-      "loss": 0.3492,
-      "step": 5198
-    },
-    {
-      "epoch": 0.3397817136134893,
-      "grad_norm": 0.4523397982120514,
-      "learning_rate": 9.77829513143485e-06,
-      "loss": 0.3888,
-      "step": 5199
-    },
-    {
-      "epoch": 0.33984706881903143,
-      "grad_norm": 0.4638594090938568,
-      "learning_rate": 9.778192291877185e-06,
-      "loss": 0.4205,
-      "step": 5200
-    },
-    {
-      "epoch": 0.33991242402457356,
-      "grad_norm": 0.436329185962677,
-      "learning_rate": 9.778089429014619e-06,
-      "loss": 0.3906,
-      "step": 5201
-    },
-    {
-      "epoch": 0.3399777792301157,
-      "grad_norm": 0.5121533274650574,
-      "learning_rate": 9.777986542847658e-06,
-      "loss": 0.3779,
-      "step": 5202
-    },
-    {
-      "epoch": 0.3400431344356578,
-      "grad_norm": 0.4427978992462158,
-      "learning_rate": 9.777883633376801e-06,
-      "loss": 0.3539,
-      "step": 5203
-    },
-    {
-      "epoch": 0.34010848964119994,
-      "grad_norm": 0.46456608176231384,
-      "learning_rate": 9.77778070060255e-06,
-      "loss": 0.4138,
-      "step": 5204
-    },
-    {
-      "epoch": 0.34017384484674207,
-      "grad_norm": 0.4669976234436035,
-      "learning_rate": 9.777677744525406e-06,
-      "loss": 0.395,
-      "step": 5205
-    },
-    {
-      "epoch": 0.34023920005228414,
-      "grad_norm": 0.47090378403663635,
-      "learning_rate": 9.777574765145874e-06,
-      "loss": 0.4263,
-      "step": 5206
-    },
-    {
-      "epoch": 0.34030455525782627,
-      "grad_norm": 0.49403876066207886,
-      "learning_rate": 9.777471762464456e-06,
-      "loss": 0.389,
-      "step": 5207
-    },
-    {
-      "epoch": 0.3403699104633684,
-      "grad_norm": 0.4523240625858307,
-      "learning_rate": 9.777368736481652e-06,
-      "loss": 0.3816,
-      "step": 5208
-    },
-    {
-      "epoch": 0.3404352656689105,
-      "grad_norm": 0.4438014328479767,
-      "learning_rate": 9.777265687197965e-06,
-      "loss": 0.4136,
-      "step": 5209
-    },
-    {
-      "epoch": 0.34050062087445265,
-      "grad_norm": 0.45609503984451294,
-      "learning_rate": 9.777162614613898e-06,
-      "loss": 0.4134,
-      "step": 5210
-    },
-    {
-      "epoch": 0.3405659760799948,
-      "grad_norm": 0.44377031922340393,
-      "learning_rate": 9.777059518729954e-06,
-      "loss": 0.3619,
-      "step": 5211
-    },
-    {
-      "epoch": 0.3406313312855369,
-      "grad_norm": 0.4724767804145813,
-      "learning_rate": 9.776956399546638e-06,
-      "loss": 0.4506,
-      "step": 5212
-    },
-    {
-      "epoch": 0.34069668649107904,
-      "grad_norm": 0.45906752347946167,
-      "learning_rate": 9.77685325706445e-06,
-      "loss": 0.4049,
-      "step": 5213
-    },
-    {
-      "epoch": 0.3407620416966211,
-      "grad_norm": 0.455089271068573,
-      "learning_rate": 9.776750091283891e-06,
-      "loss": 0.4018,
-      "step": 5214
-    },
-    {
-      "epoch": 0.34082739690216324,
-      "grad_norm": 0.43956664204597473,
-      "learning_rate": 9.77664690220547e-06,
-      "loss": 0.4037,
-      "step": 5215
-    },
-    {
-      "epoch": 0.34089275210770537,
-      "grad_norm": 0.4752846956253052,
-      "learning_rate": 9.776543689829685e-06,
-      "loss": 0.413,
-      "step": 5216
-    },
-    {
-      "epoch": 0.3409581073132475,
-      "grad_norm": 0.48933643102645874,
-      "learning_rate": 9.776440454157043e-06,
-      "loss": 0.4334,
-      "step": 5217
-    },
-    {
-      "epoch": 0.3410234625187896,
-      "grad_norm": 0.4655178487300873,
-      "learning_rate": 9.776337195188046e-06,
-      "loss": 0.4199,
-      "step": 5218
-    },
-    {
-      "epoch": 0.34108881772433175,
-      "grad_norm": 0.45571649074554443,
-      "learning_rate": 9.776233912923198e-06,
-      "loss": 0.4059,
-      "step": 5219
-    },
-    {
-      "epoch": 0.3411541729298739,
-      "grad_norm": 0.503077507019043,
-      "learning_rate": 9.776130607363003e-06,
-      "loss": 0.4645,
-      "step": 5220
-    },
-    {
-      "epoch": 0.341219528135416,
-      "grad_norm": 0.4741753339767456,
-      "learning_rate": 9.776027278507963e-06,
-      "loss": 0.4037,
-      "step": 5221
-    },
-    {
-      "epoch": 0.34128488334095813,
-      "grad_norm": 0.43636104464530945,
-      "learning_rate": 9.775923926358584e-06,
-      "loss": 0.3667,
-      "step": 5222
-    },
-    {
-      "epoch": 0.3413502385465002,
-      "grad_norm": 0.5163292288780212,
-      "learning_rate": 9.77582055091537e-06,
-      "loss": 0.4511,
-      "step": 5223
-    },
-    {
-      "epoch": 0.34141559375204233,
-      "grad_norm": 0.4682936370372772,
-      "learning_rate": 9.775717152178825e-06,
-      "loss": 0.4061,
-      "step": 5224
-    },
-    {
-      "epoch": 0.34148094895758446,
-      "grad_norm": 0.449651300907135,
-      "learning_rate": 9.775613730149452e-06,
-      "loss": 0.3349,
-      "step": 5225
-    },
-    {
-      "epoch": 0.3415463041631266,
-      "grad_norm": 0.45347052812576294,
-      "learning_rate": 9.775510284827756e-06,
-      "loss": 0.3614,
-      "step": 5226
-    },
-    {
-      "epoch": 0.3416116593686687,
-      "grad_norm": 0.4409697949886322,
-      "learning_rate": 9.775406816214244e-06,
-      "loss": 0.3522,
-      "step": 5227
-    },
-    {
-      "epoch": 0.34167701457421085,
-      "grad_norm": 0.45578670501708984,
-      "learning_rate": 9.775303324309416e-06,
-      "loss": 0.4011,
-      "step": 5228
-    },
-    {
-      "epoch": 0.341742369779753,
-      "grad_norm": 0.5071733593940735,
-      "learning_rate": 9.77519980911378e-06,
-      "loss": 0.3959,
-      "step": 5229
-    },
-    {
-      "epoch": 0.3418077249852951,
-      "grad_norm": 0.5738910436630249,
-      "learning_rate": 9.775096270627841e-06,
-      "loss": 0.4307,
-      "step": 5230
-    },
-    {
-      "epoch": 0.3418730801908372,
-      "grad_norm": 0.44033265113830566,
-      "learning_rate": 9.774992708852104e-06,
-      "loss": 0.3795,
-      "step": 5231
-    },
-    {
-      "epoch": 0.3419384353963793,
-      "grad_norm": 0.47738906741142273,
-      "learning_rate": 9.774889123787072e-06,
-      "loss": 0.4089,
-      "step": 5232
-    },
-    {
-      "epoch": 0.34200379060192143,
-      "grad_norm": 0.4668895900249481,
-      "learning_rate": 9.774785515433252e-06,
-      "loss": 0.3823,
-      "step": 5233
-    },
-    {
-      "epoch": 0.34206914580746356,
-      "grad_norm": 0.4882426857948303,
-      "learning_rate": 9.77468188379115e-06,
-      "loss": 0.4395,
-      "step": 5234
-    },
-    {
-      "epoch": 0.3421345010130057,
-      "grad_norm": 0.44981008768081665,
-      "learning_rate": 9.77457822886127e-06,
-      "loss": 0.3734,
-      "step": 5235
-    },
-    {
-      "epoch": 0.3421998562185478,
-      "grad_norm": 0.43056395649909973,
-      "learning_rate": 9.774474550644115e-06,
-      "loss": 0.3545,
-      "step": 5236
-    },
-    {
-      "epoch": 0.34226521142408994,
-      "grad_norm": 0.4812091886997223,
-      "learning_rate": 9.774370849140196e-06,
-      "loss": 0.4662,
-      "step": 5237
-    },
-    {
-      "epoch": 0.34233056662963207,
-      "grad_norm": 0.48207148909568787,
-      "learning_rate": 9.774267124350017e-06,
-      "loss": 0.4371,
-      "step": 5238
-    },
-    {
-      "epoch": 0.34239592183517414,
-      "grad_norm": 0.47329193353652954,
-      "learning_rate": 9.774163376274083e-06,
-      "loss": 0.3981,
-      "step": 5239
-    },
-    {
-      "epoch": 0.34246127704071627,
-      "grad_norm": 0.45506712794303894,
-      "learning_rate": 9.774059604912899e-06,
-      "loss": 0.3911,
-      "step": 5240
-    },
-    {
-      "epoch": 0.3425266322462584,
-      "grad_norm": 0.47248575091362,
-      "learning_rate": 9.773955810266972e-06,
-      "loss": 0.3569,
-      "step": 5241
-    },
-    {
-      "epoch": 0.34259198745180053,
-      "grad_norm": 0.443486750125885,
-      "learning_rate": 9.773851992336812e-06,
-      "loss": 0.3458,
-      "step": 5242
-    },
-    {
-      "epoch": 0.34265734265734266,
-      "grad_norm": 0.47293639183044434,
-      "learning_rate": 9.773748151122918e-06,
-      "loss": 0.3702,
-      "step": 5243
-    },
-    {
-      "epoch": 0.3427226978628848,
-      "grad_norm": 0.492477148771286,
-      "learning_rate": 9.773644286625803e-06,
-      "loss": 0.4305,
-      "step": 5244
-    },
-    {
-      "epoch": 0.3427880530684269,
-      "grad_norm": 0.5646010637283325,
-      "learning_rate": 9.77354039884597e-06,
-      "loss": 0.5537,
-      "step": 5245
-    },
-    {
-      "epoch": 0.34285340827396904,
-      "grad_norm": 0.5110243558883667,
-      "learning_rate": 9.773436487783927e-06,
-      "loss": 0.4172,
-      "step": 5246
-    },
-    {
-      "epoch": 0.34291876347951117,
-      "grad_norm": 0.4851347506046295,
-      "learning_rate": 9.77333255344018e-06,
-      "loss": 0.4543,
-      "step": 5247
-    },
-    {
-      "epoch": 0.34298411868505324,
-      "grad_norm": 0.5463439226150513,
-      "learning_rate": 9.773228595815238e-06,
-      "loss": 0.4234,
-      "step": 5248
-    },
-    {
-      "epoch": 0.34304947389059537,
-      "grad_norm": 0.46826621890068054,
-      "learning_rate": 9.773124614909606e-06,
-      "loss": 0.4571,
-      "step": 5249
-    },
-    {
-      "epoch": 0.3431148290961375,
-      "grad_norm": 0.48085686564445496,
-      "learning_rate": 9.773020610723792e-06,
-      "loss": 0.4218,
-      "step": 5250
-    },
-    {
-      "epoch": 0.3431801843016796,
-      "grad_norm": 0.5734224915504456,
-      "learning_rate": 9.7729165832583e-06,
-      "loss": 0.4724,
-      "step": 5251
-    },
-    {
-      "epoch": 0.34324553950722175,
-      "grad_norm": 0.44442424178123474,
-      "learning_rate": 9.772812532513644e-06,
-      "loss": 0.3399,
-      "step": 5252
-    },
-    {
-      "epoch": 0.3433108947127639,
-      "grad_norm": 0.4710421562194824,
-      "learning_rate": 9.772708458490326e-06,
-      "loss": 0.4409,
-      "step": 5253
-    },
-    {
-      "epoch": 0.343376249918306,
-      "grad_norm": 0.49244454503059387,
-      "learning_rate": 9.772604361188855e-06,
-      "loss": 0.3878,
-      "step": 5254
-    },
-    {
-      "epoch": 0.34344160512384814,
-      "grad_norm": 0.5072375535964966,
-      "learning_rate": 9.772500240609741e-06,
-      "loss": 0.4529,
-      "step": 5255
-    },
-    {
-      "epoch": 0.3435069603293902,
-      "grad_norm": 0.4409193694591522,
-      "learning_rate": 9.77239609675349e-06,
-      "loss": 0.3615,
-      "step": 5256
-    },
-    {
-      "epoch": 0.34357231553493234,
-      "grad_norm": 0.45375481247901917,
-      "learning_rate": 9.772291929620608e-06,
-      "loss": 0.3863,
-      "step": 5257
-    },
-    {
-      "epoch": 0.34363767074047447,
-      "grad_norm": 0.46864810585975647,
-      "learning_rate": 9.772187739211607e-06,
-      "loss": 0.4379,
-      "step": 5258
-    },
-    {
-      "epoch": 0.3437030259460166,
-      "grad_norm": 0.46199578046798706,
-      "learning_rate": 9.77208352552699e-06,
-      "loss": 0.4082,
-      "step": 5259
-    },
-    {
-      "epoch": 0.3437683811515587,
-      "grad_norm": 0.46045181155204773,
-      "learning_rate": 9.77197928856727e-06,
-      "loss": 0.381,
-      "step": 5260
-    },
-    {
-      "epoch": 0.34383373635710085,
-      "grad_norm": 0.4706612825393677,
-      "learning_rate": 9.771875028332956e-06,
-      "loss": 0.3901,
-      "step": 5261
-    },
-    {
-      "epoch": 0.343899091562643,
-      "grad_norm": 0.48839184641838074,
-      "learning_rate": 9.771770744824553e-06,
-      "loss": 0.424,
-      "step": 5262
-    },
-    {
-      "epoch": 0.3439644467681851,
-      "grad_norm": 0.46977683901786804,
-      "learning_rate": 9.77166643804257e-06,
-      "loss": 0.3891,
-      "step": 5263
-    },
-    {
-      "epoch": 0.34402980197372723,
-      "grad_norm": 0.46044379472732544,
-      "learning_rate": 9.771562107987518e-06,
-      "loss": 0.401,
-      "step": 5264
-    },
-    {
-      "epoch": 0.3440951571792693,
-      "grad_norm": 0.4603383541107178,
-      "learning_rate": 9.771457754659903e-06,
-      "loss": 0.3716,
-      "step": 5265
-    },
-    {
-      "epoch": 0.34416051238481143,
-      "grad_norm": 0.43309587240219116,
-      "learning_rate": 9.771353378060236e-06,
-      "loss": 0.3743,
-      "step": 5266
-    },
-    {
-      "epoch": 0.34422586759035356,
-      "grad_norm": 0.44623294472694397,
-      "learning_rate": 9.771248978189027e-06,
-      "loss": 0.3862,
-      "step": 5267
-    },
-    {
-      "epoch": 0.3442912227958957,
-      "grad_norm": 0.4425095021724701,
-      "learning_rate": 9.771144555046783e-06,
-      "loss": 0.3765,
-      "step": 5268
-    },
-    {
-      "epoch": 0.3443565780014378,
-      "grad_norm": 0.48969170451164246,
-      "learning_rate": 9.771040108634013e-06,
-      "loss": 0.4295,
-      "step": 5269
-    },
-    {
-      "epoch": 0.34442193320697995,
-      "grad_norm": 0.4806336760520935,
-      "learning_rate": 9.770935638951229e-06,
-      "loss": 0.4105,
-      "step": 5270
-    },
-    {
-      "epoch": 0.3444872884125221,
-      "grad_norm": 0.517914891242981,
-      "learning_rate": 9.770831145998938e-06,
-      "loss": 0.4211,
-      "step": 5271
-    },
-    {
-      "epoch": 0.3445526436180642,
-      "grad_norm": 0.4635848104953766,
-      "learning_rate": 9.770726629777652e-06,
-      "loss": 0.4277,
-      "step": 5272
-    },
-    {
-      "epoch": 0.3446179988236063,
-      "grad_norm": 0.47082462906837463,
-      "learning_rate": 9.770622090287879e-06,
-      "loss": 0.4485,
-      "step": 5273
-    },
-    {
-      "epoch": 0.3446833540291484,
-      "grad_norm": 0.5020426511764526,
-      "learning_rate": 9.77051752753013e-06,
-      "loss": 0.4301,
-      "step": 5274
-    },
-    {
-      "epoch": 0.34474870923469053,
-      "grad_norm": 0.4663870334625244,
-      "learning_rate": 9.770412941504912e-06,
-      "loss": 0.3832,
-      "step": 5275
-    },
-    {
-      "epoch": 0.34481406444023266,
-      "grad_norm": 0.49421462416648865,
-      "learning_rate": 9.77030833221274e-06,
-      "loss": 0.437,
-      "step": 5276
-    },
-    {
-      "epoch": 0.3448794196457748,
-      "grad_norm": 0.49273785948753357,
-      "learning_rate": 9.77020369965412e-06,
-      "loss": 0.3942,
-      "step": 5277
-    },
-    {
-      "epoch": 0.3449447748513169,
-      "grad_norm": 0.4762526750564575,
-      "learning_rate": 9.770099043829564e-06,
-      "loss": 0.4323,
-      "step": 5278
-    },
-    {
-      "epoch": 0.34501013005685904,
-      "grad_norm": 0.4772988259792328,
-      "learning_rate": 9.769994364739585e-06,
-      "loss": 0.3905,
-      "step": 5279
-    },
-    {
-      "epoch": 0.34507548526240117,
-      "grad_norm": 0.4762629568576813,
-      "learning_rate": 9.769889662384689e-06,
-      "loss": 0.3995,
-      "step": 5280
-    },
-    {
-      "epoch": 0.34514084046794324,
-      "grad_norm": 0.4549536406993866,
-      "learning_rate": 9.769784936765389e-06,
-      "loss": 0.4048,
-      "step": 5281
-    },
-    {
-      "epoch": 0.34520619567348537,
-      "grad_norm": 0.511263906955719,
-      "learning_rate": 9.769680187882195e-06,
-      "loss": 0.4277,
-      "step": 5282
-    },
-    {
-      "epoch": 0.3452715508790275,
-      "grad_norm": 0.47359809279441833,
-      "learning_rate": 9.769575415735618e-06,
-      "loss": 0.403,
-      "step": 5283
-    },
-    {
-      "epoch": 0.34533690608456963,
-      "grad_norm": 0.455730676651001,
-      "learning_rate": 9.76947062032617e-06,
-      "loss": 0.3845,
-      "step": 5284
-    },
-    {
-      "epoch": 0.34540226129011176,
-      "grad_norm": 0.5127042531967163,
-      "learning_rate": 9.76936580165436e-06,
-      "loss": 0.4843,
-      "step": 5285
-    },
-    {
-      "epoch": 0.3454676164956539,
-      "grad_norm": 0.4402686059474945,
-      "learning_rate": 9.769260959720703e-06,
-      "loss": 0.3592,
-      "step": 5286
-    },
-    {
-      "epoch": 0.345532971701196,
-      "grad_norm": 0.45730164647102356,
-      "learning_rate": 9.769156094525708e-06,
-      "loss": 0.3927,
-      "step": 5287
-    },
-    {
-      "epoch": 0.34559832690673814,
-      "grad_norm": 0.4772600829601288,
-      "learning_rate": 9.769051206069886e-06,
-      "loss": 0.4324,
-      "step": 5288
-    },
-    {
-      "epoch": 0.34566368211228027,
-      "grad_norm": 0.4552757143974304,
-      "learning_rate": 9.768946294353749e-06,
-      "loss": 0.3365,
-      "step": 5289
-    },
-    {
-      "epoch": 0.34572903731782234,
-      "grad_norm": 0.4652750492095947,
-      "learning_rate": 9.768841359377808e-06,
-      "loss": 0.3756,
-      "step": 5290
-    },
-    {
-      "epoch": 0.34579439252336447,
-      "grad_norm": 0.4607808589935303,
-      "learning_rate": 9.768736401142576e-06,
-      "loss": 0.3864,
-      "step": 5291
-    },
-    {
-      "epoch": 0.3458597477289066,
-      "grad_norm": 0.4790264368057251,
-      "learning_rate": 9.768631419648565e-06,
-      "loss": 0.4045,
-      "step": 5292
-    },
-    {
-      "epoch": 0.3459251029344487,
-      "grad_norm": 0.46356624364852905,
-      "learning_rate": 9.768526414896286e-06,
-      "loss": 0.4037,
-      "step": 5293
-    },
-    {
-      "epoch": 0.34599045813999085,
-      "grad_norm": 0.46210983395576477,
-      "learning_rate": 9.768421386886253e-06,
-      "loss": 0.4008,
-      "step": 5294
-    },
-    {
-      "epoch": 0.346055813345533,
-      "grad_norm": 0.4501861035823822,
-      "learning_rate": 9.768316335618975e-06,
-      "loss": 0.3902,
-      "step": 5295
-    },
-    {
-      "epoch": 0.3461211685510751,
-      "grad_norm": 0.4635646641254425,
-      "learning_rate": 9.76821126109497e-06,
-      "loss": 0.4008,
-      "step": 5296
-    },
-    {
-      "epoch": 0.34618652375661724,
-      "grad_norm": 0.4980321228504181,
-      "learning_rate": 9.768106163314742e-06,
-      "loss": 0.407,
-      "step": 5297
-    },
-    {
-      "epoch": 0.3462518789621593,
-      "grad_norm": 0.4468640089035034,
-      "learning_rate": 9.76800104227881e-06,
-      "loss": 0.3635,
-      "step": 5298
-    },
-    {
-      "epoch": 0.34631723416770144,
-      "grad_norm": 0.4522833526134491,
-      "learning_rate": 9.767895897987688e-06,
-      "loss": 0.376,
-      "step": 5299
-    },
-    {
-      "epoch": 0.34638258937324357,
-      "grad_norm": 0.43752583861351013,
-      "learning_rate": 9.767790730441882e-06,
-      "loss": 0.3781,
-      "step": 5300
-    },
-    {
-      "epoch": 0.3464479445787857,
-      "grad_norm": 0.4958682954311371,
-      "learning_rate": 9.767685539641911e-06,
-      "loss": 0.4468,
-      "step": 5301
-    },
-    {
-      "epoch": 0.3465132997843278,
-      "grad_norm": 0.4759429693222046,
-      "learning_rate": 9.767580325588286e-06,
-      "loss": 0.4078,
-      "step": 5302
-    },
-    {
-      "epoch": 0.34657865498986995,
-      "grad_norm": 0.45632949471473694,
-      "learning_rate": 9.76747508828152e-06,
-      "loss": 0.3854,
-      "step": 5303
-    },
-    {
-      "epoch": 0.3466440101954121,
-      "grad_norm": 0.41211169958114624,
-      "learning_rate": 9.767369827722123e-06,
-      "loss": 0.325,
-      "step": 5304
-    },
-    {
-      "epoch": 0.3467093654009542,
-      "grad_norm": 0.4712884724140167,
-      "learning_rate": 9.767264543910617e-06,
-      "loss": 0.3689,
-      "step": 5305
-    },
-    {
-      "epoch": 0.34677472060649633,
-      "grad_norm": 0.44414106011390686,
-      "learning_rate": 9.767159236847508e-06,
-      "loss": 0.3944,
-      "step": 5306
-    },
-    {
-      "epoch": 0.3468400758120384,
-      "grad_norm": 0.49052873253822327,
-      "learning_rate": 9.767053906533312e-06,
-      "loss": 0.4559,
-      "step": 5307
-    },
-    {
-      "epoch": 0.34690543101758053,
-      "grad_norm": 0.45353513956069946,
-      "learning_rate": 9.766948552968542e-06,
-      "loss": 0.4019,
-      "step": 5308
-    },
-    {
-      "epoch": 0.34697078622312266,
-      "grad_norm": 0.49267831444740295,
-      "learning_rate": 9.766843176153714e-06,
-      "loss": 0.3805,
-      "step": 5309
-    },
-    {
-      "epoch": 0.3470361414286648,
-      "grad_norm": 0.468070924282074,
-      "learning_rate": 9.766737776089339e-06,
-      "loss": 0.3812,
-      "step": 5310
-    },
-    {
-      "epoch": 0.3471014966342069,
-      "grad_norm": 0.491280198097229,
-      "learning_rate": 9.766632352775932e-06,
-      "loss": 0.4057,
-      "step": 5311
-    },
-    {
-      "epoch": 0.34716685183974905,
-      "grad_norm": 0.4380781352519989,
-      "learning_rate": 9.76652690621401e-06,
-      "loss": 0.3849,
-      "step": 5312
-    },
-    {
-      "epoch": 0.3472322070452912,
-      "grad_norm": 0.441520094871521,
-      "learning_rate": 9.766421436404083e-06,
-      "loss": 0.336,
-      "step": 5313
-    },
-    {
-      "epoch": 0.3472975622508333,
-      "grad_norm": 0.45357683300971985,
-      "learning_rate": 9.766315943346668e-06,
-      "loss": 0.3855,
-      "step": 5314
-    },
-    {
-      "epoch": 0.3473629174563754,
-      "grad_norm": 0.482016384601593,
-      "learning_rate": 9.76621042704228e-06,
-      "loss": 0.4172,
-      "step": 5315
-    },
-    {
-      "epoch": 0.3474282726619175,
-      "grad_norm": 0.5012636184692383,
-      "learning_rate": 9.76610488749143e-06,
-      "loss": 0.4854,
-      "step": 5316
-    },
-    {
-      "epoch": 0.34749362786745963,
-      "grad_norm": 0.4567621648311615,
-      "learning_rate": 9.765999324694637e-06,
-      "loss": 0.3955,
-      "step": 5317
-    },
-    {
-      "epoch": 0.34755898307300176,
-      "grad_norm": 0.4705146551132202,
-      "learning_rate": 9.765893738652415e-06,
-      "loss": 0.3946,
-      "step": 5318
-    },
-    {
-      "epoch": 0.3476243382785439,
-      "grad_norm": 0.44210195541381836,
-      "learning_rate": 9.765788129365276e-06,
-      "loss": 0.3675,
-      "step": 5319
-    },
-    {
-      "epoch": 0.347689693484086,
-      "grad_norm": 0.4626901149749756,
-      "learning_rate": 9.76568249683374e-06,
-      "loss": 0.4241,
-      "step": 5320
-    },
-    {
-      "epoch": 0.34775504868962814,
-      "grad_norm": 0.45457345247268677,
-      "learning_rate": 9.765576841058317e-06,
-      "loss": 0.3814,
-      "step": 5321
-    },
-    {
-      "epoch": 0.34782040389517027,
-      "grad_norm": 0.466986745595932,
-      "learning_rate": 9.765471162039526e-06,
-      "loss": 0.4255,
-      "step": 5322
-    },
-    {
-      "epoch": 0.3478857591007124,
-      "grad_norm": 0.4808909595012665,
-      "learning_rate": 9.76536545977788e-06,
-      "loss": 0.4366,
-      "step": 5323
-    },
-    {
-      "epoch": 0.34795111430625447,
-      "grad_norm": 0.49690648913383484,
-      "learning_rate": 9.765259734273898e-06,
-      "loss": 0.4808,
-      "step": 5324
-    },
-    {
-      "epoch": 0.3480164695117966,
-      "grad_norm": 0.45402097702026367,
-      "learning_rate": 9.765153985528092e-06,
-      "loss": 0.3989,
-      "step": 5325
-    },
-    {
-      "epoch": 0.34808182471733873,
-      "grad_norm": 0.4688388407230377,
-      "learning_rate": 9.76504821354098e-06,
-      "loss": 0.3746,
-      "step": 5326
-    },
-    {
-      "epoch": 0.34814717992288086,
-      "grad_norm": 0.4402535557746887,
-      "learning_rate": 9.764942418313074e-06,
-      "loss": 0.3504,
-      "step": 5327
-    },
-    {
-      "epoch": 0.348212535128423,
-      "grad_norm": 0.46410199999809265,
-      "learning_rate": 9.764836599844896e-06,
-      "loss": 0.3876,
-      "step": 5328
-    },
-    {
-      "epoch": 0.3482778903339651,
-      "grad_norm": 0.45752081274986267,
-      "learning_rate": 9.76473075813696e-06,
-      "loss": 0.3709,
-      "step": 5329
-    },
-    {
-      "epoch": 0.34834324553950724,
-      "grad_norm": 0.5043820142745972,
-      "learning_rate": 9.76462489318978e-06,
-      "loss": 0.4149,
-      "step": 5330
-    },
-    {
-      "epoch": 0.34840860074504937,
-      "grad_norm": 0.44153255224227905,
-      "learning_rate": 9.764519005003874e-06,
-      "loss": 0.3739,
-      "step": 5331
-    },
-    {
-      "epoch": 0.34847395595059144,
-      "grad_norm": 0.4349711835384369,
-      "learning_rate": 9.76441309357976e-06,
-      "loss": 0.3354,
-      "step": 5332
-    },
-    {
-      "epoch": 0.34853931115613357,
-      "grad_norm": 0.5045499801635742,
-      "learning_rate": 9.76430715891795e-06,
-      "loss": 0.4174,
-      "step": 5333
-    },
-    {
-      "epoch": 0.3486046663616757,
-      "grad_norm": 0.46213701367378235,
-      "learning_rate": 9.764201201018963e-06,
-      "loss": 0.373,
-      "step": 5334
-    },
-    {
-      "epoch": 0.3486700215672178,
-      "grad_norm": 0.4964313507080078,
-      "learning_rate": 9.764095219883319e-06,
-      "loss": 0.4172,
-      "step": 5335
-    },
-    {
-      "epoch": 0.34873537677275995,
-      "grad_norm": 0.4488334655761719,
-      "learning_rate": 9.76398921551153e-06,
-      "loss": 0.3682,
-      "step": 5336
-    },
-    {
-      "epoch": 0.3488007319783021,
-      "grad_norm": 0.49103713035583496,
-      "learning_rate": 9.763883187904117e-06,
-      "loss": 0.4111,
-      "step": 5337
-    },
-    {
-      "epoch": 0.3488660871838442,
-      "grad_norm": 0.441040962934494,
-      "learning_rate": 9.763777137061595e-06,
-      "loss": 0.39,
-      "step": 5338
-    },
-    {
-      "epoch": 0.34893144238938634,
-      "grad_norm": 0.48513248562812805,
-      "learning_rate": 9.76367106298448e-06,
-      "loss": 0.3817,
-      "step": 5339
-    },
-    {
-      "epoch": 0.3489967975949284,
-      "grad_norm": 0.44116392731666565,
-      "learning_rate": 9.763564965673292e-06,
-      "loss": 0.3998,
-      "step": 5340
-    },
-    {
-      "epoch": 0.34906215280047054,
-      "grad_norm": 0.4690583050251007,
-      "learning_rate": 9.763458845128547e-06,
-      "loss": 0.393,
-      "step": 5341
-    },
-    {
-      "epoch": 0.34912750800601267,
-      "grad_norm": 0.4888540804386139,
-      "learning_rate": 9.763352701350764e-06,
-      "loss": 0.4198,
-      "step": 5342
-    },
-    {
-      "epoch": 0.3491928632115548,
-      "grad_norm": 0.501666784286499,
-      "learning_rate": 9.763246534340458e-06,
-      "loss": 0.3886,
-      "step": 5343
-    },
-    {
-      "epoch": 0.3492582184170969,
-      "grad_norm": 0.4757966995239258,
-      "learning_rate": 9.76314034409815e-06,
-      "loss": 0.3674,
-      "step": 5344
-    },
-    {
-      "epoch": 0.34932357362263905,
-      "grad_norm": 0.4631199538707733,
-      "learning_rate": 9.763034130624355e-06,
-      "loss": 0.3988,
-      "step": 5345
-    },
-    {
-      "epoch": 0.3493889288281812,
-      "grad_norm": 0.4730125963687897,
-      "learning_rate": 9.762927893919595e-06,
-      "loss": 0.4347,
-      "step": 5346
-    },
-    {
-      "epoch": 0.3494542840337233,
-      "grad_norm": 0.5159075260162354,
-      "learning_rate": 9.762821633984383e-06,
-      "loss": 0.4224,
-      "step": 5347
-    },
-    {
-      "epoch": 0.34951963923926543,
-      "grad_norm": 0.43075433373451233,
-      "learning_rate": 9.762715350819242e-06,
-      "loss": 0.3216,
-      "step": 5348
-    },
-    {
-      "epoch": 0.3495849944448075,
-      "grad_norm": 0.47610026597976685,
-      "learning_rate": 9.762609044424688e-06,
-      "loss": 0.3888,
-      "step": 5349
-    },
-    {
-      "epoch": 0.34965034965034963,
-      "grad_norm": 0.49334654211997986,
-      "learning_rate": 9.762502714801239e-06,
-      "loss": 0.4107,
-      "step": 5350
-    },
-    {
-      "epoch": 0.34971570485589176,
-      "grad_norm": 0.5176300406455994,
-      "learning_rate": 9.762396361949414e-06,
-      "loss": 0.4084,
-      "step": 5351
-    },
-    {
-      "epoch": 0.3497810600614339,
-      "grad_norm": 0.48287269473075867,
-      "learning_rate": 9.762289985869731e-06,
-      "loss": 0.4043,
-      "step": 5352
-    },
-    {
-      "epoch": 0.349846415266976,
-      "grad_norm": 0.4483534097671509,
-      "learning_rate": 9.762183586562713e-06,
-      "loss": 0.3574,
-      "step": 5353
-    },
-    {
-      "epoch": 0.34991177047251815,
-      "grad_norm": 0.48141026496887207,
-      "learning_rate": 9.762077164028874e-06,
-      "loss": 0.4202,
-      "step": 5354
-    },
-    {
-      "epoch": 0.3499771256780603,
-      "grad_norm": 0.47812286019325256,
-      "learning_rate": 9.761970718268734e-06,
-      "loss": 0.4009,
-      "step": 5355
-    },
-    {
-      "epoch": 0.3500424808836024,
-      "grad_norm": 0.4493675231933594,
-      "learning_rate": 9.761864249282815e-06,
-      "loss": 0.3916,
-      "step": 5356
-    },
-    {
-      "epoch": 0.3501078360891445,
-      "grad_norm": 0.4531344771385193,
-      "learning_rate": 9.761757757071632e-06,
-      "loss": 0.3684,
-      "step": 5357
-    },
-    {
-      "epoch": 0.3501731912946866,
-      "grad_norm": 0.4760850667953491,
-      "learning_rate": 9.76165124163571e-06,
-      "loss": 0.3981,
-      "step": 5358
-    },
-    {
-      "epoch": 0.35023854650022873,
-      "grad_norm": 0.48776134848594666,
-      "learning_rate": 9.761544702975562e-06,
-      "loss": 0.4452,
-      "step": 5359
-    },
-    {
-      "epoch": 0.35030390170577086,
-      "grad_norm": 0.4582652449607849,
-      "learning_rate": 9.761438141091715e-06,
-      "loss": 0.4244,
-      "step": 5360
-    },
-    {
-      "epoch": 0.350369256911313,
-      "grad_norm": 0.4685945212841034,
-      "learning_rate": 9.76133155598468e-06,
-      "loss": 0.3681,
-      "step": 5361
-    },
-    {
-      "epoch": 0.3504346121168551,
-      "grad_norm": 0.4862159788608551,
-      "learning_rate": 9.761224947654986e-06,
-      "loss": 0.4116,
-      "step": 5362
-    },
-    {
-      "epoch": 0.35049996732239724,
-      "grad_norm": 0.48712801933288574,
-      "learning_rate": 9.761118316103146e-06,
-      "loss": 0.4026,
-      "step": 5363
-    },
-    {
-      "epoch": 0.35056532252793937,
-      "grad_norm": 0.4633937180042267,
-      "learning_rate": 9.761011661329683e-06,
-      "loss": 0.3921,
-      "step": 5364
-    },
-    {
-      "epoch": 0.3506306777334815,
-      "grad_norm": 0.47881579399108887,
-      "learning_rate": 9.760904983335117e-06,
-      "loss": 0.4227,
-      "step": 5365
-    },
-    {
-      "epoch": 0.35069603293902357,
-      "grad_norm": 0.4811265170574188,
-      "learning_rate": 9.760798282119967e-06,
-      "loss": 0.4271,
-      "step": 5366
-    },
-    {
-      "epoch": 0.3507613881445657,
-      "grad_norm": 0.46356168389320374,
-      "learning_rate": 9.760691557684756e-06,
-      "loss": 0.4003,
-      "step": 5367
-    },
-    {
-      "epoch": 0.35082674335010783,
-      "grad_norm": 0.48348528146743774,
-      "learning_rate": 9.760584810030002e-06,
-      "loss": 0.3982,
-      "step": 5368
-    },
-    {
-      "epoch": 0.35089209855564996,
-      "grad_norm": 0.4457794427871704,
-      "learning_rate": 9.760478039156226e-06,
-      "loss": 0.3609,
-      "step": 5369
-    },
-    {
-      "epoch": 0.3509574537611921,
-      "grad_norm": 0.48174190521240234,
-      "learning_rate": 9.760371245063951e-06,
-      "loss": 0.4377,
-      "step": 5370
-    },
-    {
-      "epoch": 0.3510228089667342,
-      "grad_norm": 0.4791134297847748,
-      "learning_rate": 9.760264427753695e-06,
-      "loss": 0.4164,
-      "step": 5371
-    },
-    {
-      "epoch": 0.35108816417227634,
-      "grad_norm": 0.4756622910499573,
-      "learning_rate": 9.760157587225981e-06,
-      "loss": 0.3839,
-      "step": 5372
-    },
-    {
-      "epoch": 0.35115351937781847,
-      "grad_norm": 0.4487653374671936,
-      "learning_rate": 9.76005072348133e-06,
-      "loss": 0.3948,
-      "step": 5373
-    },
-    {
-      "epoch": 0.35121887458336054,
-      "grad_norm": 0.4537963569164276,
-      "learning_rate": 9.759943836520261e-06,
-      "loss": 0.3767,
-      "step": 5374
-    },
-    {
-      "epoch": 0.35128422978890267,
-      "grad_norm": 0.49372246861457825,
-      "learning_rate": 9.759836926343298e-06,
-      "loss": 0.3957,
-      "step": 5375
-    },
-    {
-      "epoch": 0.3513495849944448,
-      "grad_norm": 0.471619188785553,
-      "learning_rate": 9.75972999295096e-06,
-      "loss": 0.4235,
-      "step": 5376
-    },
-    {
-      "epoch": 0.3514149401999869,
-      "grad_norm": 0.48206186294555664,
-      "learning_rate": 9.759623036343772e-06,
-      "loss": 0.4246,
-      "step": 5377
-    },
-    {
-      "epoch": 0.35148029540552905,
-      "grad_norm": 0.45793044567108154,
-      "learning_rate": 9.759516056522252e-06,
-      "loss": 0.3924,
-      "step": 5378
-    },
-    {
-      "epoch": 0.3515456506110712,
-      "grad_norm": 0.5243789553642273,
-      "learning_rate": 9.759409053486923e-06,
-      "loss": 0.4208,
-      "step": 5379
-    },
-    {
-      "epoch": 0.3516110058166133,
-      "grad_norm": 0.4804753065109253,
-      "learning_rate": 9.759302027238308e-06,
-      "loss": 0.4052,
-      "step": 5380
-    },
-    {
-      "epoch": 0.35167636102215544,
-      "grad_norm": 0.45715075731277466,
-      "learning_rate": 9.75919497777693e-06,
-      "loss": 0.3764,
-      "step": 5381
-    },
-    {
-      "epoch": 0.3517417162276975,
-      "grad_norm": 0.4755937457084656,
-      "learning_rate": 9.759087905103307e-06,
-      "loss": 0.4235,
-      "step": 5382
-    },
-    {
-      "epoch": 0.35180707143323964,
-      "grad_norm": 0.45948106050491333,
-      "learning_rate": 9.758980809217964e-06,
-      "loss": 0.3657,
-      "step": 5383
-    },
-    {
-      "epoch": 0.35187242663878177,
-      "grad_norm": 0.47409531474113464,
-      "learning_rate": 9.758873690121424e-06,
-      "loss": 0.426,
-      "step": 5384
-    },
-    {
-      "epoch": 0.3519377818443239,
-      "grad_norm": 0.5137777924537659,
-      "learning_rate": 9.758766547814207e-06,
-      "loss": 0.4567,
-      "step": 5385
-    },
-    {
-      "epoch": 0.352003137049866,
-      "grad_norm": 0.4825906455516815,
-      "learning_rate": 9.758659382296838e-06,
-      "loss": 0.4238,
-      "step": 5386
-    },
-    {
-      "epoch": 0.35206849225540815,
-      "grad_norm": 0.4490877687931061,
-      "learning_rate": 9.758552193569838e-06,
-      "loss": 0.3907,
-      "step": 5387
-    },
-    {
-      "epoch": 0.3521338474609503,
-      "grad_norm": 0.4570304751396179,
-      "learning_rate": 9.75844498163373e-06,
-      "loss": 0.3823,
-      "step": 5388
-    },
-    {
-      "epoch": 0.3521992026664924,
-      "grad_norm": 0.46208950877189636,
-      "learning_rate": 9.758337746489038e-06,
-      "loss": 0.3856,
-      "step": 5389
-    },
-    {
-      "epoch": 0.35226455787203453,
-      "grad_norm": 0.4878591299057007,
-      "learning_rate": 9.758230488136285e-06,
-      "loss": 0.4141,
-      "step": 5390
-    },
-    {
-      "epoch": 0.3523299130775766,
-      "grad_norm": 0.44514963030815125,
-      "learning_rate": 9.758123206575993e-06,
-      "loss": 0.3846,
-      "step": 5391
-    },
-    {
-      "epoch": 0.35239526828311873,
-      "grad_norm": 0.43946024775505066,
-      "learning_rate": 9.758015901808684e-06,
-      "loss": 0.3655,
-      "step": 5392
-    },
-    {
-      "epoch": 0.35246062348866086,
-      "grad_norm": 0.4391081631183624,
-      "learning_rate": 9.757908573834886e-06,
-      "loss": 0.4043,
-      "step": 5393
-    },
-    {
-      "epoch": 0.352525978694203,
-      "grad_norm": 0.47638437151908875,
-      "learning_rate": 9.757801222655119e-06,
-      "loss": 0.4361,
-      "step": 5394
-    },
-    {
-      "epoch": 0.3525913338997451,
-      "grad_norm": 0.4875653088092804,
-      "learning_rate": 9.757693848269904e-06,
-      "loss": 0.4101,
-      "step": 5395
-    },
-    {
-      "epoch": 0.35265668910528725,
-      "grad_norm": 0.46153193712234497,
-      "learning_rate": 9.757586450679771e-06,
-      "loss": 0.3962,
-      "step": 5396
-    },
-    {
-      "epoch": 0.3527220443108294,
-      "grad_norm": 0.4558506906032562,
-      "learning_rate": 9.75747902988524e-06,
-      "loss": 0.41,
-      "step": 5397
-    },
-    {
-      "epoch": 0.3527873995163715,
-      "grad_norm": 0.4575936198234558,
-      "learning_rate": 9.757371585886836e-06,
-      "loss": 0.3452,
-      "step": 5398
-    },
-    {
-      "epoch": 0.3528527547219136,
-      "grad_norm": 0.4510190486907959,
-      "learning_rate": 9.757264118685081e-06,
-      "loss": 0.3873,
-      "step": 5399
-    },
-    {
-      "epoch": 0.3529181099274557,
-      "grad_norm": 0.4614260196685791,
-      "learning_rate": 9.757156628280504e-06,
-      "loss": 0.4343,
-      "step": 5400
-    },
-    {
-      "epoch": 0.35298346513299783,
-      "grad_norm": 0.4935864806175232,
-      "learning_rate": 9.757049114673623e-06,
-      "loss": 0.4269,
-      "step": 5401
-    },
-    {
-      "epoch": 0.35304882033853996,
-      "grad_norm": 0.47003424167633057,
-      "learning_rate": 9.756941577864967e-06,
-      "loss": 0.4262,
-      "step": 5402
-    },
-    {
-      "epoch": 0.3531141755440821,
-      "grad_norm": 0.46406206488609314,
-      "learning_rate": 9.756834017855059e-06,
-      "loss": 0.4061,
-      "step": 5403
-    },
-    {
-      "epoch": 0.3531795307496242,
-      "grad_norm": 0.47702664136886597,
-      "learning_rate": 9.756726434644424e-06,
-      "loss": 0.3961,
-      "step": 5404
-    },
-    {
-      "epoch": 0.35324488595516634,
-      "grad_norm": 0.47764095664024353,
-      "learning_rate": 9.756618828233585e-06,
-      "loss": 0.4145,
-      "step": 5405
-    },
-    {
-      "epoch": 0.35331024116070847,
-      "grad_norm": 0.45300671458244324,
-      "learning_rate": 9.756511198623067e-06,
-      "loss": 0.3881,
-      "step": 5406
-    },
-    {
-      "epoch": 0.3533755963662506,
-      "grad_norm": 0.46042224764823914,
-      "learning_rate": 9.756403545813398e-06,
-      "loss": 0.3944,
-      "step": 5407
-    },
-    {
-      "epoch": 0.35344095157179267,
-      "grad_norm": 0.5612374544143677,
-      "learning_rate": 9.7562958698051e-06,
-      "loss": 0.3907,
-      "step": 5408
-    },
-    {
-      "epoch": 0.3535063067773348,
-      "grad_norm": 0.4778275489807129,
-      "learning_rate": 9.756188170598702e-06,
-      "loss": 0.4056,
-      "step": 5409
-    },
-    {
-      "epoch": 0.35357166198287693,
-      "grad_norm": 0.4500736892223358,
-      "learning_rate": 9.756080448194724e-06,
-      "loss": 0.3736,
-      "step": 5410
-    },
-    {
-      "epoch": 0.35363701718841906,
-      "grad_norm": 0.46924859285354614,
-      "learning_rate": 9.755972702593695e-06,
-      "loss": 0.3956,
-      "step": 5411
-    },
-    {
-      "epoch": 0.3537023723939612,
-      "grad_norm": 0.4467761814594269,
-      "learning_rate": 9.755864933796139e-06,
-      "loss": 0.3717,
-      "step": 5412
-    },
-    {
-      "epoch": 0.3537677275995033,
-      "grad_norm": 0.4989909529685974,
-      "learning_rate": 9.755757141802582e-06,
-      "loss": 0.451,
-      "step": 5413
-    },
-    {
-      "epoch": 0.35383308280504544,
-      "grad_norm": 0.48000356554985046,
-      "learning_rate": 9.75564932661355e-06,
-      "loss": 0.409,
-      "step": 5414
-    },
-    {
-      "epoch": 0.35389843801058757,
-      "grad_norm": 0.5046958923339844,
-      "learning_rate": 9.75554148822957e-06,
-      "loss": 0.424,
-      "step": 5415
-    },
-    {
-      "epoch": 0.35396379321612964,
-      "grad_norm": 0.5241200923919678,
-      "learning_rate": 9.755433626651165e-06,
-      "loss": 0.46,
-      "step": 5416
-    },
-    {
-      "epoch": 0.35402914842167177,
-      "grad_norm": 0.46465495228767395,
-      "learning_rate": 9.755325741878863e-06,
-      "loss": 0.3875,
-      "step": 5417
-    },
-    {
-      "epoch": 0.3540945036272139,
-      "grad_norm": 0.4638899564743042,
-      "learning_rate": 9.75521783391319e-06,
-      "loss": 0.4184,
-      "step": 5418
-    },
-    {
-      "epoch": 0.354159858832756,
-      "grad_norm": 0.464154452085495,
-      "learning_rate": 9.755109902754673e-06,
-      "loss": 0.4188,
-      "step": 5419
-    },
-    {
-      "epoch": 0.35422521403829815,
-      "grad_norm": 0.4229236841201782,
-      "learning_rate": 9.755001948403838e-06,
-      "loss": 0.3127,
-      "step": 5420
-    },
-    {
-      "epoch": 0.3542905692438403,
-      "grad_norm": 0.5362170934677124,
-      "learning_rate": 9.754893970861208e-06,
-      "loss": 0.3835,
-      "step": 5421
-    },
-    {
-      "epoch": 0.3543559244493824,
-      "grad_norm": 0.48871511220932007,
-      "learning_rate": 9.754785970127317e-06,
-      "loss": 0.4176,
-      "step": 5422
-    },
-    {
-      "epoch": 0.35442127965492454,
-      "grad_norm": 0.46417954564094543,
-      "learning_rate": 9.754677946202686e-06,
-      "loss": 0.3755,
-      "step": 5423
-    },
-    {
-      "epoch": 0.3544866348604666,
-      "grad_norm": 0.46701881289482117,
-      "learning_rate": 9.754569899087843e-06,
-      "loss": 0.3702,
-      "step": 5424
-    },
-    {
-      "epoch": 0.35455199006600874,
-      "grad_norm": 0.5148953199386597,
-      "learning_rate": 9.754461828783315e-06,
-      "loss": 0.4643,
-      "step": 5425
-    },
-    {
-      "epoch": 0.35461734527155087,
-      "grad_norm": 0.47312691807746887,
-      "learning_rate": 9.75435373528963e-06,
-      "loss": 0.3995,
-      "step": 5426
-    },
-    {
-      "epoch": 0.354682700477093,
-      "grad_norm": 0.5017833709716797,
-      "learning_rate": 9.754245618607317e-06,
-      "loss": 0.4244,
-      "step": 5427
-    },
-    {
-      "epoch": 0.3547480556826351,
-      "grad_norm": 0.49053823947906494,
-      "learning_rate": 9.754137478736898e-06,
-      "loss": 0.4149,
-      "step": 5428
-    },
-    {
-      "epoch": 0.35481341088817725,
-      "grad_norm": 0.47810912132263184,
-      "learning_rate": 9.754029315678906e-06,
-      "loss": 0.3963,
-      "step": 5429
-    },
-    {
-      "epoch": 0.3548787660937194,
-      "grad_norm": 0.4808669984340668,
-      "learning_rate": 9.753921129433864e-06,
-      "loss": 0.3927,
-      "step": 5430
-    },
-    {
-      "epoch": 0.3549441212992615,
-      "grad_norm": 0.455217182636261,
-      "learning_rate": 9.753812920002302e-06,
-      "loss": 0.3875,
-      "step": 5431
-    },
-    {
-      "epoch": 0.35500947650480363,
-      "grad_norm": 0.4422004818916321,
-      "learning_rate": 9.753704687384749e-06,
-      "loss": 0.3604,
-      "step": 5432
-    },
-    {
-      "epoch": 0.3550748317103457,
-      "grad_norm": 0.42850714921951294,
-      "learning_rate": 9.75359643158173e-06,
-      "loss": 0.3491,
-      "step": 5433
-    },
-    {
-      "epoch": 0.35514018691588783,
-      "grad_norm": 0.5053018927574158,
-      "learning_rate": 9.753488152593774e-06,
-      "loss": 0.4384,
-      "step": 5434
-    },
-    {
-      "epoch": 0.35520554212142996,
-      "grad_norm": 0.42704612016677856,
-      "learning_rate": 9.753379850421412e-06,
-      "loss": 0.3278,
-      "step": 5435
-    },
-    {
-      "epoch": 0.3552708973269721,
-      "grad_norm": 0.4832640588283539,
-      "learning_rate": 9.753271525065166e-06,
-      "loss": 0.3469,
-      "step": 5436
-    },
-    {
-      "epoch": 0.3553362525325142,
-      "grad_norm": 0.46148040890693665,
-      "learning_rate": 9.753163176525572e-06,
-      "loss": 0.3929,
-      "step": 5437
-    },
-    {
-      "epoch": 0.35540160773805635,
-      "grad_norm": 0.46563684940338135,
-      "learning_rate": 9.753054804803153e-06,
-      "loss": 0.4351,
-      "step": 5438
-    },
-    {
-      "epoch": 0.3554669629435985,
-      "grad_norm": 0.44625213742256165,
-      "learning_rate": 9.752946409898439e-06,
-      "loss": 0.3957,
-      "step": 5439
-    },
-    {
-      "epoch": 0.3555323181491406,
-      "grad_norm": 0.4298880398273468,
-      "learning_rate": 9.752837991811958e-06,
-      "loss": 0.3268,
-      "step": 5440
-    },
-    {
-      "epoch": 0.3555976733546827,
-      "grad_norm": 0.464630663394928,
-      "learning_rate": 9.75272955054424e-06,
-      "loss": 0.4075,
-      "step": 5441
-    },
-    {
-      "epoch": 0.3556630285602248,
-      "grad_norm": 0.44898244738578796,
-      "learning_rate": 9.752621086095813e-06,
-      "loss": 0.3573,
-      "step": 5442
-    },
-    {
-      "epoch": 0.35572838376576693,
-      "grad_norm": 0.4824317693710327,
-      "learning_rate": 9.752512598467207e-06,
-      "loss": 0.4329,
-      "step": 5443
-    },
-    {
-      "epoch": 0.35579373897130906,
-      "grad_norm": 0.465991348028183,
-      "learning_rate": 9.752404087658951e-06,
-      "loss": 0.3885,
-      "step": 5444
-    },
-    {
-      "epoch": 0.3558590941768512,
-      "grad_norm": 0.4964909255504608,
-      "learning_rate": 9.752295553671574e-06,
-      "loss": 0.4342,
-      "step": 5445
-    },
-    {
-      "epoch": 0.3559244493823933,
-      "grad_norm": 0.5367789268493652,
-      "learning_rate": 9.752186996505605e-06,
-      "loss": 0.454,
-      "step": 5446
-    },
-    {
-      "epoch": 0.35598980458793544,
-      "grad_norm": 0.4750783145427704,
-      "learning_rate": 9.752078416161574e-06,
-      "loss": 0.4286,
-      "step": 5447
-    },
-    {
-      "epoch": 0.35605515979347757,
-      "grad_norm": 0.4420899748802185,
-      "learning_rate": 9.751969812640009e-06,
-      "loss": 0.3954,
-      "step": 5448
-    },
-    {
-      "epoch": 0.3561205149990197,
-      "grad_norm": 0.48597240447998047,
-      "learning_rate": 9.751861185941442e-06,
-      "loss": 0.4412,
-      "step": 5449
-    },
-    {
-      "epoch": 0.35618587020456177,
-      "grad_norm": 0.4420667588710785,
-      "learning_rate": 9.7517525360664e-06,
-      "loss": 0.3875,
-      "step": 5450
-    },
-    {
-      "epoch": 0.3562512254101039,
-      "grad_norm": 0.47928571701049805,
-      "learning_rate": 9.751643863015418e-06,
-      "loss": 0.423,
-      "step": 5451
-    },
-    {
-      "epoch": 0.35631658061564603,
-      "grad_norm": 0.5068503022193909,
-      "learning_rate": 9.751535166789021e-06,
-      "loss": 0.3496,
-      "step": 5452
-    },
-    {
-      "epoch": 0.35638193582118816,
-      "grad_norm": 0.4657083749771118,
-      "learning_rate": 9.751426447387741e-06,
-      "loss": 0.4095,
-      "step": 5453
-    },
-    {
-      "epoch": 0.3564472910267303,
-      "grad_norm": 0.5119538903236389,
-      "learning_rate": 9.751317704812108e-06,
-      "loss": 0.438,
-      "step": 5454
-    },
-    {
-      "epoch": 0.3565126462322724,
-      "grad_norm": 0.4512006938457489,
-      "learning_rate": 9.751208939062653e-06,
-      "loss": 0.3885,
-      "step": 5455
-    },
-    {
-      "epoch": 0.35657800143781454,
-      "grad_norm": 0.49278318881988525,
-      "learning_rate": 9.751100150139906e-06,
-      "loss": 0.4138,
-      "step": 5456
-    },
-    {
-      "epoch": 0.35664335664335667,
-      "grad_norm": 0.49377885460853577,
-      "learning_rate": 9.750991338044397e-06,
-      "loss": 0.4183,
-      "step": 5457
-    },
-    {
-      "epoch": 0.35670871184889874,
-      "grad_norm": 0.4811665415763855,
-      "learning_rate": 9.750882502776658e-06,
-      "loss": 0.4146,
-      "step": 5458
-    },
-    {
-      "epoch": 0.35677406705444087,
-      "grad_norm": 0.5551828742027283,
-      "learning_rate": 9.750773644337219e-06,
-      "loss": 0.4836,
-      "step": 5459
-    },
-    {
-      "epoch": 0.356839422259983,
-      "grad_norm": 0.45828860998153687,
-      "learning_rate": 9.750664762726612e-06,
-      "loss": 0.4045,
-      "step": 5460
-    },
-    {
-      "epoch": 0.3569047774655251,
-      "grad_norm": 0.44688984751701355,
-      "learning_rate": 9.750555857945366e-06,
-      "loss": 0.3596,
-      "step": 5461
-    },
-    {
-      "epoch": 0.35697013267106725,
-      "grad_norm": 0.5028802156448364,
-      "learning_rate": 9.750446929994014e-06,
-      "loss": 0.4572,
-      "step": 5462
-    },
-    {
-      "epoch": 0.3570354878766094,
-      "grad_norm": 0.4700569808483124,
-      "learning_rate": 9.750337978873085e-06,
-      "loss": 0.3513,
-      "step": 5463
-    },
-    {
-      "epoch": 0.3571008430821515,
-      "grad_norm": 0.46906787157058716,
-      "learning_rate": 9.750229004583112e-06,
-      "loss": 0.3789,
-      "step": 5464
-    },
-    {
-      "epoch": 0.35716619828769364,
-      "grad_norm": 0.47062036395072937,
-      "learning_rate": 9.750120007124628e-06,
-      "loss": 0.4085,
-      "step": 5465
-    },
-    {
-      "epoch": 0.3572315534932357,
-      "grad_norm": 0.4533155858516693,
-      "learning_rate": 9.750010986498164e-06,
-      "loss": 0.4235,
-      "step": 5466
-    },
-    {
-      "epoch": 0.35729690869877784,
-      "grad_norm": 0.47060179710388184,
-      "learning_rate": 9.74990194270425e-06,
-      "loss": 0.4232,
-      "step": 5467
-    },
-    {
-      "epoch": 0.35736226390431997,
-      "grad_norm": 0.5088547468185425,
-      "learning_rate": 9.749792875743418e-06,
-      "loss": 0.4294,
-      "step": 5468
-    },
-    {
-      "epoch": 0.3574276191098621,
-      "grad_norm": 0.4410618245601654,
-      "learning_rate": 9.7496837856162e-06,
-      "loss": 0.3867,
-      "step": 5469
-    },
-    {
-      "epoch": 0.3574929743154042,
-      "grad_norm": 0.4552725851535797,
-      "learning_rate": 9.74957467232313e-06,
-      "loss": 0.3738,
-      "step": 5470
-    },
-    {
-      "epoch": 0.35755832952094635,
-      "grad_norm": 0.4843587875366211,
-      "learning_rate": 9.749465535864738e-06,
-      "loss": 0.4186,
-      "step": 5471
-    },
-    {
-      "epoch": 0.3576236847264885,
-      "grad_norm": 0.49985405802726746,
-      "learning_rate": 9.749356376241559e-06,
-      "loss": 0.475,
-      "step": 5472
-    },
-    {
-      "epoch": 0.3576890399320306,
-      "grad_norm": 0.47680795192718506,
-      "learning_rate": 9.74924719345412e-06,
-      "loss": 0.39,
-      "step": 5473
-    },
-    {
-      "epoch": 0.35775439513757273,
-      "grad_norm": 0.4933762848377228,
-      "learning_rate": 9.74913798750296e-06,
-      "loss": 0.3891,
-      "step": 5474
-    },
-    {
-      "epoch": 0.3578197503431148,
-      "grad_norm": 0.47685137391090393,
-      "learning_rate": 9.74902875838861e-06,
-      "loss": 0.3745,
-      "step": 5475
-    },
-    {
-      "epoch": 0.35788510554865693,
-      "grad_norm": 0.507700502872467,
-      "learning_rate": 9.748919506111601e-06,
-      "loss": 0.4134,
-      "step": 5476
-    },
-    {
-      "epoch": 0.35795046075419906,
-      "grad_norm": 0.45000192523002625,
-      "learning_rate": 9.748810230672463e-06,
-      "loss": 0.3845,
-      "step": 5477
-    },
-    {
-      "epoch": 0.3580158159597412,
-      "grad_norm": 0.4318159222602844,
-      "learning_rate": 9.748700932071735e-06,
-      "loss": 0.3561,
-      "step": 5478
-    },
-    {
-      "epoch": 0.3580811711652833,
-      "grad_norm": 0.436652272939682,
-      "learning_rate": 9.748591610309948e-06,
-      "loss": 0.3441,
-      "step": 5479
-    },
-    {
-      "epoch": 0.35814652637082545,
-      "grad_norm": 0.5244379043579102,
-      "learning_rate": 9.748482265387634e-06,
-      "loss": 0.4799,
-      "step": 5480
-    },
-    {
-      "epoch": 0.3582118815763676,
-      "grad_norm": 0.4660002291202545,
-      "learning_rate": 9.748372897305327e-06,
-      "loss": 0.3899,
-      "step": 5481
-    },
-    {
-      "epoch": 0.3582772367819097,
-      "grad_norm": 0.4497160315513611,
-      "learning_rate": 9.74826350606356e-06,
-      "loss": 0.3476,
-      "step": 5482
-    },
-    {
-      "epoch": 0.3583425919874518,
-      "grad_norm": 0.45765113830566406,
-      "learning_rate": 9.748154091662867e-06,
-      "loss": 0.4087,
-      "step": 5483
-    },
-    {
-      "epoch": 0.3584079471929939,
-      "grad_norm": 0.5139644145965576,
-      "learning_rate": 9.748044654103781e-06,
-      "loss": 0.4456,
-      "step": 5484
-    },
-    {
-      "epoch": 0.35847330239853603,
-      "grad_norm": 0.4682544767856598,
-      "learning_rate": 9.747935193386837e-06,
-      "loss": 0.3922,
-      "step": 5485
-    },
-    {
-      "epoch": 0.35853865760407816,
-      "grad_norm": 0.47969430685043335,
-      "learning_rate": 9.747825709512568e-06,
-      "loss": 0.4239,
-      "step": 5486
-    },
-    {
-      "epoch": 0.3586040128096203,
-      "grad_norm": 0.4690496325492859,
-      "learning_rate": 9.747716202481507e-06,
-      "loss": 0.3948,
-      "step": 5487
-    },
-    {
-      "epoch": 0.3586693680151624,
-      "grad_norm": 0.4625760614871979,
-      "learning_rate": 9.747606672294192e-06,
-      "loss": 0.3936,
-      "step": 5488
-    },
-    {
-      "epoch": 0.35873472322070454,
-      "grad_norm": 0.4752259850502014,
-      "learning_rate": 9.747497118951152e-06,
-      "loss": 0.4128,
-      "step": 5489
-    },
-    {
-      "epoch": 0.35880007842624667,
-      "grad_norm": 0.4447445273399353,
-      "learning_rate": 9.747387542452927e-06,
-      "loss": 0.3358,
-      "step": 5490
-    },
-    {
-      "epoch": 0.3588654336317888,
-      "grad_norm": 0.45358070731163025,
-      "learning_rate": 9.747277942800045e-06,
-      "loss": 0.388,
-      "step": 5491
-    },
-    {
-      "epoch": 0.35893078883733087,
-      "grad_norm": 0.46292173862457275,
-      "learning_rate": 9.747168319993045e-06,
-      "loss": 0.3976,
-      "step": 5492
-    },
-    {
-      "epoch": 0.358996144042873,
-      "grad_norm": 0.47038424015045166,
-      "learning_rate": 9.747058674032462e-06,
-      "loss": 0.4258,
-      "step": 5493
-    },
-    {
-      "epoch": 0.35906149924841513,
-      "grad_norm": 0.4775792062282562,
-      "learning_rate": 9.746949004918826e-06,
-      "loss": 0.4294,
-      "step": 5494
-    },
-    {
-      "epoch": 0.35912685445395726,
-      "grad_norm": 0.48537901043891907,
-      "learning_rate": 9.746839312652678e-06,
-      "loss": 0.4168,
-      "step": 5495
-    },
-    {
-      "epoch": 0.3591922096594994,
-      "grad_norm": 0.4761759638786316,
-      "learning_rate": 9.74672959723455e-06,
-      "loss": 0.3888,
-      "step": 5496
-    },
-    {
-      "epoch": 0.3592575648650415,
-      "grad_norm": 0.4897666573524475,
-      "learning_rate": 9.746619858664976e-06,
-      "loss": 0.4147,
-      "step": 5497
-    },
-    {
-      "epoch": 0.35932292007058364,
-      "grad_norm": 0.4796993136405945,
-      "learning_rate": 9.746510096944494e-06,
-      "loss": 0.4033,
-      "step": 5498
-    },
-    {
-      "epoch": 0.35938827527612577,
-      "grad_norm": 0.4589497148990631,
-      "learning_rate": 9.746400312073637e-06,
-      "loss": 0.3749,
-      "step": 5499
-    },
-    {
-      "epoch": 0.35945363048166784,
-      "grad_norm": 0.46367713809013367,
-      "learning_rate": 9.746290504052942e-06,
-      "loss": 0.404,
-      "step": 5500
-    },
-    {
-      "epoch": 0.35951898568720997,
-      "grad_norm": 0.5216031074523926,
-      "learning_rate": 9.746180672882943e-06,
-      "loss": 0.4589,
-      "step": 5501
-    },
-    {
-      "epoch": 0.3595843408927521,
-      "grad_norm": 0.467579185962677,
-      "learning_rate": 9.746070818564178e-06,
-      "loss": 0.3964,
-      "step": 5502
-    },
-    {
-      "epoch": 0.3596496960982942,
-      "grad_norm": 0.44784092903137207,
-      "learning_rate": 9.74596094109718e-06,
-      "loss": 0.3994,
-      "step": 5503
-    },
-    {
-      "epoch": 0.35971505130383635,
-      "grad_norm": 0.45274531841278076,
-      "learning_rate": 9.745851040482486e-06,
-      "loss": 0.3899,
-      "step": 5504
-    },
-    {
-      "epoch": 0.3597804065093785,
-      "grad_norm": 0.504065990447998,
-      "learning_rate": 9.745741116720635e-06,
-      "loss": 0.3982,
-      "step": 5505
-    },
-    {
-      "epoch": 0.3598457617149206,
-      "grad_norm": 0.4520778954029083,
-      "learning_rate": 9.745631169812157e-06,
-      "loss": 0.3813,
-      "step": 5506
-    },
-    {
-      "epoch": 0.35991111692046274,
-      "grad_norm": 0.47651755809783936,
-      "learning_rate": 9.745521199757595e-06,
-      "loss": 0.418,
-      "step": 5507
-    },
-    {
-      "epoch": 0.3599764721260048,
-      "grad_norm": 0.4437931180000305,
-      "learning_rate": 9.74541120655748e-06,
-      "loss": 0.3728,
-      "step": 5508
-    },
-    {
-      "epoch": 0.36004182733154694,
-      "grad_norm": 0.5083013772964478,
-      "learning_rate": 9.74530119021235e-06,
-      "loss": 0.4607,
-      "step": 5509
-    },
-    {
-      "epoch": 0.36010718253708907,
-      "grad_norm": 0.49100443720817566,
-      "learning_rate": 9.745191150722745e-06,
-      "loss": 0.4,
-      "step": 5510
-    },
-    {
-      "epoch": 0.3601725377426312,
-      "grad_norm": 0.46294522285461426,
-      "learning_rate": 9.745081088089196e-06,
-      "loss": 0.3774,
-      "step": 5511
-    },
-    {
-      "epoch": 0.3602378929481733,
-      "grad_norm": 0.446426659822464,
-      "learning_rate": 9.744971002312244e-06,
-      "loss": 0.4067,
-      "step": 5512
-    },
-    {
-      "epoch": 0.36030324815371545,
-      "grad_norm": 0.466602623462677,
-      "learning_rate": 9.744860893392425e-06,
-      "loss": 0.4036,
-      "step": 5513
-    },
-    {
-      "epoch": 0.3603686033592576,
-      "grad_norm": 0.5313547849655151,
-      "learning_rate": 9.744750761330276e-06,
-      "loss": 0.4251,
-      "step": 5514
-    },
-    {
-      "epoch": 0.3604339585647997,
-      "grad_norm": 0.45921388268470764,
-      "learning_rate": 9.744640606126332e-06,
-      "loss": 0.4073,
-      "step": 5515
-    },
-    {
-      "epoch": 0.36049931377034183,
-      "grad_norm": 0.4541753828525543,
-      "learning_rate": 9.744530427781134e-06,
-      "loss": 0.4036,
-      "step": 5516
-    },
-    {
-      "epoch": 0.3605646689758839,
-      "grad_norm": 0.4802130460739136,
-      "learning_rate": 9.744420226295215e-06,
-      "loss": 0.3744,
-      "step": 5517
-    },
-    {
-      "epoch": 0.36063002418142603,
-      "grad_norm": 0.475797563791275,
-      "learning_rate": 9.744310001669117e-06,
-      "loss": 0.3781,
-      "step": 5518
-    },
-    {
-      "epoch": 0.36069537938696816,
-      "grad_norm": 0.4912639260292053,
-      "learning_rate": 9.744199753903375e-06,
-      "loss": 0.3703,
-      "step": 5519
-    },
-    {
-      "epoch": 0.3607607345925103,
-      "grad_norm": 0.45177358388900757,
-      "learning_rate": 9.744089482998526e-06,
-      "loss": 0.3866,
-      "step": 5520
-    },
-    {
-      "epoch": 0.3608260897980524,
-      "grad_norm": 0.49249595403671265,
-      "learning_rate": 9.743979188955111e-06,
-      "loss": 0.4488,
-      "step": 5521
-    },
-    {
-      "epoch": 0.36089144500359455,
-      "grad_norm": 0.49944764375686646,
-      "learning_rate": 9.743868871773666e-06,
-      "loss": 0.4194,
-      "step": 5522
-    },
-    {
-      "epoch": 0.3609568002091367,
-      "grad_norm": 0.49870505928993225,
-      "learning_rate": 9.743758531454727e-06,
-      "loss": 0.4651,
-      "step": 5523
-    },
-    {
-      "epoch": 0.3610221554146788,
-      "grad_norm": 0.4502068758010864,
-      "learning_rate": 9.743648167998837e-06,
-      "loss": 0.3812,
-      "step": 5524
-    },
-    {
-      "epoch": 0.3610875106202209,
-      "grad_norm": 0.42723122239112854,
-      "learning_rate": 9.743537781406529e-06,
-      "loss": 0.3463,
-      "step": 5525
-    },
-    {
-      "epoch": 0.361152865825763,
-      "grad_norm": 0.4512379765510559,
-      "learning_rate": 9.743427371678346e-06,
-      "loss": 0.3532,
-      "step": 5526
-    },
-    {
-      "epoch": 0.36121822103130513,
-      "grad_norm": 0.46591150760650635,
-      "learning_rate": 9.743316938814824e-06,
-      "loss": 0.3973,
-      "step": 5527
-    },
-    {
-      "epoch": 0.36128357623684726,
-      "grad_norm": 0.4805774688720703,
-      "learning_rate": 9.743206482816501e-06,
-      "loss": 0.4118,
-      "step": 5528
-    },
-    {
-      "epoch": 0.3613489314423894,
-      "grad_norm": 0.4822915196418762,
-      "learning_rate": 9.743096003683918e-06,
-      "loss": 0.4368,
-      "step": 5529
-    },
-    {
-      "epoch": 0.3614142866479315,
-      "grad_norm": 0.45295917987823486,
-      "learning_rate": 9.742985501417611e-06,
-      "loss": 0.4003,
-      "step": 5530
-    },
-    {
-      "epoch": 0.36147964185347364,
-      "grad_norm": 0.46429935097694397,
-      "learning_rate": 9.742874976018122e-06,
-      "loss": 0.3672,
-      "step": 5531
-    },
-    {
-      "epoch": 0.36154499705901577,
-      "grad_norm": 0.5184620022773743,
-      "learning_rate": 9.742764427485988e-06,
-      "loss": 0.4481,
-      "step": 5532
-    },
-    {
-      "epoch": 0.3616103522645579,
-      "grad_norm": 0.47342073917388916,
-      "learning_rate": 9.742653855821748e-06,
-      "loss": 0.3731,
-      "step": 5533
-    },
-    {
-      "epoch": 0.36167570747009997,
-      "grad_norm": 0.4863404929637909,
-      "learning_rate": 9.742543261025943e-06,
-      "loss": 0.411,
-      "step": 5534
-    },
-    {
-      "epoch": 0.3617410626756421,
-      "grad_norm": 0.47418177127838135,
-      "learning_rate": 9.742432643099112e-06,
-      "loss": 0.3984,
-      "step": 5535
-    },
-    {
-      "epoch": 0.36180641788118423,
-      "grad_norm": 0.504815936088562,
-      "learning_rate": 9.742322002041793e-06,
-      "loss": 0.4734,
-      "step": 5536
-    },
-    {
-      "epoch": 0.36187177308672636,
-      "grad_norm": 0.49866777658462524,
-      "learning_rate": 9.742211337854529e-06,
-      "loss": 0.4398,
-      "step": 5537
-    },
-    {
-      "epoch": 0.3619371282922685,
-      "grad_norm": 0.50166255235672,
-      "learning_rate": 9.742100650537856e-06,
-      "loss": 0.4316,
-      "step": 5538
-    },
-    {
-      "epoch": 0.3620024834978106,
-      "grad_norm": 0.4519596993923187,
-      "learning_rate": 9.741989940092314e-06,
-      "loss": 0.3626,
-      "step": 5539
-    },
-    {
-      "epoch": 0.36206783870335274,
-      "grad_norm": 0.46182096004486084,
-      "learning_rate": 9.741879206518447e-06,
-      "loss": 0.3627,
-      "step": 5540
-    },
-    {
-      "epoch": 0.36213319390889487,
-      "grad_norm": 0.4782857894897461,
-      "learning_rate": 9.74176844981679e-06,
-      "loss": 0.3856,
-      "step": 5541
-    },
-    {
-      "epoch": 0.36219854911443694,
-      "grad_norm": 0.46283137798309326,
-      "learning_rate": 9.741657669987887e-06,
-      "loss": 0.3823,
-      "step": 5542
-    },
-    {
-      "epoch": 0.36226390431997907,
-      "grad_norm": 0.48076778650283813,
-      "learning_rate": 9.741546867032277e-06,
-      "loss": 0.4086,
-      "step": 5543
-    },
-    {
-      "epoch": 0.3623292595255212,
-      "grad_norm": 0.44915562868118286,
-      "learning_rate": 9.741436040950499e-06,
-      "loss": 0.4086,
-      "step": 5544
-    },
-    {
-      "epoch": 0.3623946147310633,
-      "grad_norm": 0.47054797410964966,
-      "learning_rate": 9.741325191743093e-06,
-      "loss": 0.3952,
-      "step": 5545
-    },
-    {
-      "epoch": 0.36245996993660545,
-      "grad_norm": 0.488459050655365,
-      "learning_rate": 9.741214319410606e-06,
-      "loss": 0.4387,
-      "step": 5546
-    },
-    {
-      "epoch": 0.3625253251421476,
-      "grad_norm": 0.4568168520927429,
-      "learning_rate": 9.741103423953572e-06,
-      "loss": 0.3785,
-      "step": 5547
-    },
-    {
-      "epoch": 0.3625906803476897,
-      "grad_norm": 0.49804478883743286,
-      "learning_rate": 9.740992505372533e-06,
-      "loss": 0.41,
-      "step": 5548
-    },
-    {
-      "epoch": 0.36265603555323184,
-      "grad_norm": 0.49567046761512756,
-      "learning_rate": 9.740881563668032e-06,
-      "loss": 0.4289,
-      "step": 5549
-    },
-    {
-      "epoch": 0.3627213907587739,
-      "grad_norm": 0.4165554642677307,
-      "learning_rate": 9.740770598840611e-06,
-      "loss": 0.3466,
-      "step": 5550
-    },
-    {
-      "epoch": 0.36278674596431604,
-      "grad_norm": 0.5371134877204895,
-      "learning_rate": 9.740659610890808e-06,
-      "loss": 0.3951,
-      "step": 5551
-    },
-    {
-      "epoch": 0.36285210116985817,
-      "grad_norm": 0.46541106700897217,
-      "learning_rate": 9.740548599819166e-06,
-      "loss": 0.4122,
-      "step": 5552
-    },
-    {
-      "epoch": 0.3629174563754003,
-      "grad_norm": 0.48032405972480774,
-      "learning_rate": 9.740437565626226e-06,
-      "loss": 0.4446,
-      "step": 5553
-    },
-    {
-      "epoch": 0.3629828115809424,
-      "grad_norm": 0.447642058134079,
-      "learning_rate": 9.74032650831253e-06,
-      "loss": 0.3726,
-      "step": 5554
-    },
-    {
-      "epoch": 0.36304816678648455,
-      "grad_norm": 0.4682174324989319,
-      "learning_rate": 9.740215427878618e-06,
-      "loss": 0.4127,
-      "step": 5555
-    },
-    {
-      "epoch": 0.3631135219920267,
-      "grad_norm": 0.46120691299438477,
-      "learning_rate": 9.740104324325035e-06,
-      "loss": 0.4178,
-      "step": 5556
-    },
-    {
-      "epoch": 0.3631788771975688,
-      "grad_norm": 0.4727381765842438,
-      "learning_rate": 9.73999319765232e-06,
-      "loss": 0.3631,
-      "step": 5557
-    },
-    {
-      "epoch": 0.36324423240311093,
-      "grad_norm": 0.4543185532093048,
-      "learning_rate": 9.739882047861016e-06,
-      "loss": 0.4067,
-      "step": 5558
-    },
-    {
-      "epoch": 0.363309587608653,
-      "grad_norm": 0.4852119982242584,
-      "learning_rate": 9.739770874951666e-06,
-      "loss": 0.4622,
-      "step": 5559
-    },
-    {
-      "epoch": 0.36337494281419513,
-      "grad_norm": 0.5047418475151062,
-      "learning_rate": 9.73965967892481e-06,
-      "loss": 0.4463,
-      "step": 5560
-    },
-    {
-      "epoch": 0.36344029801973726,
-      "grad_norm": 0.48737582564353943,
-      "learning_rate": 9.739548459780993e-06,
-      "loss": 0.3986,
-      "step": 5561
-    },
-    {
-      "epoch": 0.3635056532252794,
-      "grad_norm": 0.4619516432285309,
-      "learning_rate": 9.739437217520758e-06,
-      "loss": 0.3787,
-      "step": 5562
-    },
-    {
-      "epoch": 0.3635710084308215,
-      "grad_norm": 0.47795623540878296,
-      "learning_rate": 9.739325952144644e-06,
-      "loss": 0.4186,
-      "step": 5563
-    },
-    {
-      "epoch": 0.36363636363636365,
-      "grad_norm": 0.47678953409194946,
-      "learning_rate": 9.739214663653196e-06,
-      "loss": 0.4466,
-      "step": 5564
-    },
-    {
-      "epoch": 0.3637017188419058,
-      "grad_norm": 0.47488972544670105,
-      "learning_rate": 9.739103352046955e-06,
-      "loss": 0.4109,
-      "step": 5565
-    },
-    {
-      "epoch": 0.3637670740474479,
-      "grad_norm": 0.4641948640346527,
-      "learning_rate": 9.738992017326465e-06,
-      "loss": 0.435,
-      "step": 5566
-    },
-    {
-      "epoch": 0.36383242925299,
-      "grad_norm": 0.48926520347595215,
-      "learning_rate": 9.73888065949227e-06,
-      "loss": 0.4337,
-      "step": 5567
-    },
-    {
-      "epoch": 0.3638977844585321,
-      "grad_norm": 0.4352390468120575,
-      "learning_rate": 9.738769278544914e-06,
-      "loss": 0.328,
-      "step": 5568
-    },
-    {
-      "epoch": 0.36396313966407423,
-      "grad_norm": 0.5158409476280212,
-      "learning_rate": 9.738657874484936e-06,
-      "loss": 0.4929,
-      "step": 5569
-    },
-    {
-      "epoch": 0.36402849486961636,
-      "grad_norm": 0.44906100630760193,
-      "learning_rate": 9.738546447312883e-06,
-      "loss": 0.3907,
-      "step": 5570
-    },
-    {
-      "epoch": 0.3640938500751585,
-      "grad_norm": 0.47618475556373596,
-      "learning_rate": 9.738434997029295e-06,
-      "loss": 0.4269,
-      "step": 5571
-    },
-    {
-      "epoch": 0.3641592052807006,
-      "grad_norm": 0.4537811577320099,
-      "learning_rate": 9.738323523634722e-06,
-      "loss": 0.3738,
-      "step": 5572
-    },
-    {
-      "epoch": 0.36422456048624274,
-      "grad_norm": 0.46539798378944397,
-      "learning_rate": 9.7382120271297e-06,
-      "loss": 0.3697,
-      "step": 5573
-    },
-    {
-      "epoch": 0.36428991569178487,
-      "grad_norm": 0.4666674733161926,
-      "learning_rate": 9.738100507514779e-06,
-      "loss": 0.3902,
-      "step": 5574
-    },
-    {
-      "epoch": 0.364355270897327,
-      "grad_norm": 0.45224064588546753,
-      "learning_rate": 9.7379889647905e-06,
-      "loss": 0.3936,
-      "step": 5575
-    },
-    {
-      "epoch": 0.36442062610286907,
-      "grad_norm": 0.48884791135787964,
-      "learning_rate": 9.737877398957406e-06,
-      "loss": 0.4348,
-      "step": 5576
-    },
-    {
-      "epoch": 0.3644859813084112,
-      "grad_norm": 0.7112709879875183,
-      "learning_rate": 9.737765810016045e-06,
-      "loss": 0.4619,
-      "step": 5577
-    },
-    {
-      "epoch": 0.36455133651395333,
-      "grad_norm": 0.4430217444896698,
-      "learning_rate": 9.737654197966957e-06,
-      "loss": 0.3548,
-      "step": 5578
-    },
-    {
-      "epoch": 0.36461669171949546,
-      "grad_norm": 0.45775002241134644,
-      "learning_rate": 9.737542562810689e-06,
-      "loss": 0.3367,
-      "step": 5579
-    },
-    {
-      "epoch": 0.3646820469250376,
-      "grad_norm": 0.48870351910591125,
-      "learning_rate": 9.737430904547785e-06,
-      "loss": 0.413,
-      "step": 5580
-    },
-    {
-      "epoch": 0.3647474021305797,
-      "grad_norm": 0.469472199678421,
-      "learning_rate": 9.737319223178788e-06,
-      "loss": 0.4088,
-      "step": 5581
-    },
-    {
-      "epoch": 0.36481275733612184,
-      "grad_norm": 0.4777522683143616,
-      "learning_rate": 9.737207518704245e-06,
-      "loss": 0.4041,
-      "step": 5582
-    },
-    {
-      "epoch": 0.36487811254166397,
-      "grad_norm": 0.5005249977111816,
-      "learning_rate": 9.7370957911247e-06,
-      "loss": 0.42,
-      "step": 5583
-    },
-    {
-      "epoch": 0.36494346774720604,
-      "grad_norm": 0.47446170449256897,
-      "learning_rate": 9.736984040440695e-06,
-      "loss": 0.3897,
-      "step": 5584
-    },
-    {
-      "epoch": 0.36500882295274817,
-      "grad_norm": 0.4743311107158661,
-      "learning_rate": 9.736872266652782e-06,
-      "loss": 0.386,
-      "step": 5585
-    },
-    {
-      "epoch": 0.3650741781582903,
-      "grad_norm": 0.4345497488975525,
-      "learning_rate": 9.7367604697615e-06,
-      "loss": 0.3466,
-      "step": 5586
-    },
-    {
-      "epoch": 0.3651395333638324,
-      "grad_norm": 0.43958693742752075,
-      "learning_rate": 9.736648649767396e-06,
-      "loss": 0.3466,
-      "step": 5587
-    },
-    {
-      "epoch": 0.36520488856937455,
-      "grad_norm": 0.48933878540992737,
-      "learning_rate": 9.736536806671015e-06,
-      "loss": 0.4191,
-      "step": 5588
-    },
-    {
-      "epoch": 0.3652702437749167,
-      "grad_norm": 0.4951789677143097,
-      "learning_rate": 9.736424940472903e-06,
-      "loss": 0.4231,
-      "step": 5589
-    },
-    {
-      "epoch": 0.3653355989804588,
-      "grad_norm": 0.5030261278152466,
-      "learning_rate": 9.736313051173606e-06,
-      "loss": 0.3794,
-      "step": 5590
-    },
-    {
-      "epoch": 0.36540095418600094,
-      "grad_norm": 0.5088279843330383,
-      "learning_rate": 9.736201138773671e-06,
-      "loss": 0.4746,
-      "step": 5591
-    },
-    {
-      "epoch": 0.365466309391543,
-      "grad_norm": 0.45412909984588623,
-      "learning_rate": 9.736089203273641e-06,
-      "loss": 0.3731,
-      "step": 5592
-    },
-    {
-      "epoch": 0.36553166459708514,
-      "grad_norm": 0.47598782181739807,
-      "learning_rate": 9.735977244674064e-06,
-      "loss": 0.4199,
-      "step": 5593
-    },
-    {
-      "epoch": 0.36559701980262727,
-      "grad_norm": 0.47414252161979675,
-      "learning_rate": 9.735865262975485e-06,
-      "loss": 0.4206,
-      "step": 5594
-    },
-    {
-      "epoch": 0.3656623750081694,
-      "grad_norm": 0.47325989603996277,
-      "learning_rate": 9.73575325817845e-06,
-      "loss": 0.3965,
-      "step": 5595
-    },
-    {
-      "epoch": 0.3657277302137115,
-      "grad_norm": 0.49983224272727966,
-      "learning_rate": 9.735641230283506e-06,
-      "loss": 0.4014,
-      "step": 5596
-    },
-    {
-      "epoch": 0.36579308541925365,
-      "grad_norm": 0.4910130202770233,
-      "learning_rate": 9.7355291792912e-06,
-      "loss": 0.4269,
-      "step": 5597
-    },
-    {
-      "epoch": 0.3658584406247958,
-      "grad_norm": 0.49281221628189087,
-      "learning_rate": 9.735417105202076e-06,
-      "loss": 0.4208,
-      "step": 5598
-    },
-    {
-      "epoch": 0.3659237958303379,
-      "grad_norm": 0.45060423016548157,
-      "learning_rate": 9.735305008016682e-06,
-      "loss": 0.3841,
-      "step": 5599
-    },
-    {
-      "epoch": 0.36598915103588003,
-      "grad_norm": 0.477408230304718,
-      "learning_rate": 9.735192887735566e-06,
-      "loss": 0.4059,
-      "step": 5600
-    },
-    {
-      "epoch": 0.3660545062414221,
-      "grad_norm": 0.4904358983039856,
-      "learning_rate": 9.735080744359274e-06,
-      "loss": 0.447,
-      "step": 5601
-    },
-    {
-      "epoch": 0.36611986144696423,
-      "grad_norm": 0.4667161703109741,
-      "learning_rate": 9.734968577888354e-06,
-      "loss": 0.4022,
-      "step": 5602
-    },
-    {
-      "epoch": 0.36618521665250636,
-      "grad_norm": 0.4091729521751404,
-      "learning_rate": 9.73485638832335e-06,
-      "loss": 0.3389,
-      "step": 5603
-    },
-    {
-      "epoch": 0.3662505718580485,
-      "grad_norm": 0.4527824819087982,
-      "learning_rate": 9.734744175664812e-06,
-      "loss": 0.3562,
-      "step": 5604
-    },
-    {
-      "epoch": 0.3663159270635906,
-      "grad_norm": 0.46282854676246643,
-      "learning_rate": 9.734631939913284e-06,
-      "loss": 0.3935,
-      "step": 5605
-    },
-    {
-      "epoch": 0.36638128226913275,
-      "grad_norm": 0.4634261131286621,
-      "learning_rate": 9.734519681069318e-06,
-      "loss": 0.3451,
-      "step": 5606
-    },
-    {
-      "epoch": 0.3664466374746749,
-      "grad_norm": 0.41418299078941345,
-      "learning_rate": 9.73440739913346e-06,
-      "loss": 0.312,
-      "step": 5607
-    },
-    {
-      "epoch": 0.366511992680217,
-      "grad_norm": 0.47577157616615295,
-      "learning_rate": 9.734295094106257e-06,
-      "loss": 0.4029,
-      "step": 5608
-    },
-    {
-      "epoch": 0.3665773478857591,
-      "grad_norm": 0.4775184094905853,
-      "learning_rate": 9.734182765988255e-06,
-      "loss": 0.4235,
-      "step": 5609
-    },
-    {
-      "epoch": 0.3666427030913012,
-      "grad_norm": 0.4890030324459076,
-      "learning_rate": 9.734070414780003e-06,
-      "loss": 0.4324,
-      "step": 5610
-    },
-    {
-      "epoch": 0.36670805829684333,
-      "grad_norm": 0.4508154094219208,
-      "learning_rate": 9.733958040482053e-06,
-      "loss": 0.3926,
-      "step": 5611
-    },
-    {
-      "epoch": 0.36677341350238546,
-      "grad_norm": 0.48513153195381165,
-      "learning_rate": 9.733845643094947e-06,
-      "loss": 0.3973,
-      "step": 5612
-    },
-    {
-      "epoch": 0.3668387687079276,
-      "grad_norm": 0.4305776357650757,
-      "learning_rate": 9.733733222619235e-06,
-      "loss": 0.3661,
-      "step": 5613
-    },
-    {
-      "epoch": 0.3669041239134697,
-      "grad_norm": 0.45274618268013,
-      "learning_rate": 9.733620779055467e-06,
-      "loss": 0.3762,
-      "step": 5614
-    },
-    {
-      "epoch": 0.36696947911901184,
-      "grad_norm": 0.4698070287704468,
-      "learning_rate": 9.733508312404192e-06,
-      "loss": 0.3982,
-      "step": 5615
-    },
-    {
-      "epoch": 0.36703483432455397,
-      "grad_norm": 0.487503319978714,
-      "learning_rate": 9.733395822665956e-06,
-      "loss": 0.4062,
-      "step": 5616
-    },
-    {
-      "epoch": 0.3671001895300961,
-      "grad_norm": 0.43929579854011536,
-      "learning_rate": 9.73328330984131e-06,
-      "loss": 0.3857,
-      "step": 5617
-    },
-    {
-      "epoch": 0.36716554473563817,
-      "grad_norm": 0.4575706720352173,
-      "learning_rate": 9.7331707739308e-06,
-      "loss": 0.4024,
-      "step": 5618
-    },
-    {
-      "epoch": 0.3672308999411803,
-      "grad_norm": 0.5243099927902222,
-      "learning_rate": 9.733058214934976e-06,
-      "loss": 0.4088,
-      "step": 5619
-    },
-    {
-      "epoch": 0.36729625514672243,
-      "grad_norm": 0.45200109481811523,
-      "learning_rate": 9.73294563285439e-06,
-      "loss": 0.4076,
-      "step": 5620
-    },
-    {
-      "epoch": 0.36736161035226456,
-      "grad_norm": 0.45660802721977234,
-      "learning_rate": 9.732833027689586e-06,
-      "loss": 0.3723,
-      "step": 5621
-    },
-    {
-      "epoch": 0.3674269655578067,
-      "grad_norm": 0.530707061290741,
-      "learning_rate": 9.732720399441116e-06,
-      "loss": 0.5005,
-      "step": 5622
-    },
-    {
-      "epoch": 0.3674923207633488,
-      "grad_norm": 0.4938926696777344,
-      "learning_rate": 9.732607748109531e-06,
-      "loss": 0.4579,
-      "step": 5623
-    },
-    {
-      "epoch": 0.36755767596889094,
-      "grad_norm": 0.46209609508514404,
-      "learning_rate": 9.732495073695377e-06,
-      "loss": 0.3743,
-      "step": 5624
-    },
-    {
-      "epoch": 0.36762303117443307,
-      "grad_norm": 0.47291889786720276,
-      "learning_rate": 9.732382376199205e-06,
-      "loss": 0.372,
-      "step": 5625
-    },
-    {
-      "epoch": 0.36768838637997514,
-      "grad_norm": 0.46801483631134033,
-      "learning_rate": 9.732269655621565e-06,
-      "loss": 0.415,
-      "step": 5626
-    },
-    {
-      "epoch": 0.36775374158551727,
-      "grad_norm": 0.4820306897163391,
-      "learning_rate": 9.732156911963006e-06,
-      "loss": 0.4191,
-      "step": 5627
-    },
-    {
-      "epoch": 0.3678190967910594,
-      "grad_norm": 0.5146209597587585,
-      "learning_rate": 9.732044145224078e-06,
-      "loss": 0.4556,
-      "step": 5628
-    },
-    {
-      "epoch": 0.3678844519966015,
-      "grad_norm": 0.42632946372032166,
-      "learning_rate": 9.731931355405334e-06,
-      "loss": 0.3393,
-      "step": 5629
-    },
-    {
-      "epoch": 0.36794980720214365,
-      "grad_norm": 0.498308926820755,
-      "learning_rate": 9.731818542507318e-06,
-      "loss": 0.3939,
-      "step": 5630
-    },
-    {
-      "epoch": 0.3680151624076858,
-      "grad_norm": 0.4742498993873596,
-      "learning_rate": 9.731705706530585e-06,
-      "loss": 0.3908,
-      "step": 5631
-    },
-    {
-      "epoch": 0.3680805176132279,
-      "grad_norm": 0.4501262307167053,
-      "learning_rate": 9.731592847475685e-06,
-      "loss": 0.3827,
-      "step": 5632
-    },
-    {
-      "epoch": 0.36814587281877004,
-      "grad_norm": 0.44378212094306946,
-      "learning_rate": 9.731479965343166e-06,
-      "loss": 0.3771,
-      "step": 5633
-    },
-    {
-      "epoch": 0.3682112280243121,
-      "grad_norm": 0.4440082013607025,
-      "learning_rate": 9.73136706013358e-06,
-      "loss": 0.3379,
-      "step": 5634
-    },
-    {
-      "epoch": 0.36827658322985424,
-      "grad_norm": 0.44855231046676636,
-      "learning_rate": 9.73125413184748e-06,
-      "loss": 0.3703,
-      "step": 5635
-    },
-    {
-      "epoch": 0.36834193843539637,
-      "grad_norm": 0.45003512501716614,
-      "learning_rate": 9.731141180485412e-06,
-      "loss": 0.3888,
-      "step": 5636
-    },
-    {
-      "epoch": 0.3684072936409385,
-      "grad_norm": 0.49998369812965393,
-      "learning_rate": 9.73102820604793e-06,
-      "loss": 0.4369,
-      "step": 5637
-    },
-    {
-      "epoch": 0.3684726488464806,
-      "grad_norm": 0.4833516478538513,
-      "learning_rate": 9.730915208535584e-06,
-      "loss": 0.4015,
-      "step": 5638
-    },
-    {
-      "epoch": 0.36853800405202275,
-      "grad_norm": 0.5183923244476318,
-      "learning_rate": 9.730802187948929e-06,
-      "loss": 0.4831,
-      "step": 5639
-    },
-    {
-      "epoch": 0.3686033592575649,
-      "grad_norm": 0.44433853030204773,
-      "learning_rate": 9.73068914428851e-06,
-      "loss": 0.3593,
-      "step": 5640
-    },
-    {
-      "epoch": 0.368668714463107,
-      "grad_norm": 0.4593752920627594,
-      "learning_rate": 9.730576077554881e-06,
-      "loss": 0.3753,
-      "step": 5641
-    },
-    {
-      "epoch": 0.36873406966864913,
-      "grad_norm": 0.4683303236961365,
-      "learning_rate": 9.730462987748595e-06,
-      "loss": 0.4123,
-      "step": 5642
-    },
-    {
-      "epoch": 0.3687994248741912,
-      "grad_norm": 0.45201343297958374,
-      "learning_rate": 9.7303498748702e-06,
-      "loss": 0.3621,
-      "step": 5643
-    },
-    {
-      "epoch": 0.36886478007973333,
-      "grad_norm": 0.4989899694919586,
-      "learning_rate": 9.730236738920253e-06,
-      "loss": 0.422,
-      "step": 5644
-    },
-    {
-      "epoch": 0.36893013528527546,
-      "grad_norm": 0.4664297103881836,
-      "learning_rate": 9.7301235798993e-06,
-      "loss": 0.3796,
-      "step": 5645
-    },
-    {
-      "epoch": 0.3689954904908176,
-      "grad_norm": 0.45861926674842834,
-      "learning_rate": 9.730010397807896e-06,
-      "loss": 0.357,
-      "step": 5646
-    },
-    {
-      "epoch": 0.3690608456963597,
-      "grad_norm": 0.4439290761947632,
-      "learning_rate": 9.729897192646593e-06,
-      "loss": 0.3652,
-      "step": 5647
-    },
-    {
-      "epoch": 0.36912620090190185,
-      "grad_norm": 0.46171101927757263,
-      "learning_rate": 9.729783964415943e-06,
-      "loss": 0.3981,
-      "step": 5648
-    },
-    {
-      "epoch": 0.369191556107444,
-      "grad_norm": 0.47597968578338623,
-      "learning_rate": 9.729670713116499e-06,
-      "loss": 0.4216,
-      "step": 5649
-    },
-    {
-      "epoch": 0.3692569113129861,
-      "grad_norm": 0.4658285975456238,
-      "learning_rate": 9.729557438748813e-06,
-      "loss": 0.407,
-      "step": 5650
-    },
-    {
-      "epoch": 0.3693222665185282,
-      "grad_norm": 0.470427006483078,
-      "learning_rate": 9.729444141313435e-06,
-      "loss": 0.3899,
-      "step": 5651
-    },
-    {
-      "epoch": 0.3693876217240703,
-      "grad_norm": 0.4995580315589905,
-      "learning_rate": 9.729330820810919e-06,
-      "loss": 0.4577,
-      "step": 5652
-    },
-    {
-      "epoch": 0.36945297692961243,
-      "grad_norm": 0.4420444071292877,
-      "learning_rate": 9.729217477241818e-06,
-      "loss": 0.3681,
-      "step": 5653
-    },
-    {
-      "epoch": 0.36951833213515456,
-      "grad_norm": 0.4462299644947052,
-      "learning_rate": 9.729104110606688e-06,
-      "loss": 0.3972,
-      "step": 5654
-    },
-    {
-      "epoch": 0.3695836873406967,
-      "grad_norm": 0.4327520728111267,
-      "learning_rate": 9.728990720906078e-06,
-      "loss": 0.3553,
-      "step": 5655
-    },
-    {
-      "epoch": 0.3696490425462388,
-      "grad_norm": 0.5445707440376282,
-      "learning_rate": 9.72887730814054e-06,
-      "loss": 0.464,
-      "step": 5656
-    },
-    {
-      "epoch": 0.36971439775178094,
-      "grad_norm": 0.4488747715950012,
-      "learning_rate": 9.728763872310631e-06,
-      "loss": 0.3879,
-      "step": 5657
-    },
-    {
-      "epoch": 0.36977975295732307,
-      "grad_norm": 0.4353260397911072,
-      "learning_rate": 9.7286504134169e-06,
-      "loss": 0.3943,
-      "step": 5658
-    },
-    {
-      "epoch": 0.3698451081628652,
-      "grad_norm": 0.4729407727718353,
-      "learning_rate": 9.728536931459906e-06,
-      "loss": 0.4071,
-      "step": 5659
-    },
-    {
-      "epoch": 0.36991046336840727,
-      "grad_norm": 0.48383960127830505,
-      "learning_rate": 9.728423426440197e-06,
-      "loss": 0.4353,
-      "step": 5660
-    },
-    {
-      "epoch": 0.3699758185739494,
-      "grad_norm": 0.5584110021591187,
-      "learning_rate": 9.728309898358329e-06,
-      "loss": 0.4949,
-      "step": 5661
-    },
-    {
-      "epoch": 0.37004117377949153,
-      "grad_norm": 0.4433134198188782,
-      "learning_rate": 9.728196347214857e-06,
-      "loss": 0.3992,
-      "step": 5662
-    },
-    {
-      "epoch": 0.37010652898503366,
-      "grad_norm": 0.46349090337753296,
-      "learning_rate": 9.728082773010331e-06,
-      "loss": 0.4185,
-      "step": 5663
-    },
-    {
-      "epoch": 0.3701718841905758,
-      "grad_norm": 0.4926023781299591,
-      "learning_rate": 9.72796917574531e-06,
-      "loss": 0.4228,
-      "step": 5664
-    },
-    {
-      "epoch": 0.3702372393961179,
-      "grad_norm": 0.4178747534751892,
-      "learning_rate": 9.727855555420345e-06,
-      "loss": 0.3268,
-      "step": 5665
-    },
-    {
-      "epoch": 0.37030259460166004,
-      "grad_norm": 0.4787285327911377,
-      "learning_rate": 9.72774191203599e-06,
-      "loss": 0.4176,
-      "step": 5666
-    },
-    {
-      "epoch": 0.37036794980720217,
-      "grad_norm": 0.4780355393886566,
-      "learning_rate": 9.727628245592798e-06,
-      "loss": 0.4169,
-      "step": 5667
-    },
-    {
-      "epoch": 0.37043330501274424,
-      "grad_norm": 0.45276856422424316,
-      "learning_rate": 9.727514556091327e-06,
-      "loss": 0.376,
-      "step": 5668
-    },
-    {
-      "epoch": 0.37049866021828637,
-      "grad_norm": 0.4688391089439392,
-      "learning_rate": 9.72740084353213e-06,
-      "loss": 0.4083,
-      "step": 5669
-    },
-    {
-      "epoch": 0.3705640154238285,
-      "grad_norm": 0.4540936052799225,
-      "learning_rate": 9.727287107915762e-06,
-      "loss": 0.3592,
-      "step": 5670
-    },
-    {
-      "epoch": 0.3706293706293706,
-      "grad_norm": 0.47923359274864197,
-      "learning_rate": 9.727173349242775e-06,
-      "loss": 0.4295,
-      "step": 5671
-    },
-    {
-      "epoch": 0.37069472583491275,
-      "grad_norm": 0.4697933495044708,
-      "learning_rate": 9.727059567513726e-06,
-      "loss": 0.4117,
-      "step": 5672
-    },
-    {
-      "epoch": 0.3707600810404549,
-      "grad_norm": 0.42383813858032227,
-      "learning_rate": 9.726945762729172e-06,
-      "loss": 0.3304,
-      "step": 5673
-    },
-    {
-      "epoch": 0.370825436245997,
-      "grad_norm": 0.44558316469192505,
-      "learning_rate": 9.726831934889664e-06,
-      "loss": 0.3653,
-      "step": 5674
-    },
-    {
-      "epoch": 0.37089079145153914,
-      "grad_norm": 0.462527334690094,
-      "learning_rate": 9.72671808399576e-06,
-      "loss": 0.3771,
-      "step": 5675
-    },
-    {
-      "epoch": 0.3709561466570812,
-      "grad_norm": 0.4809938967227936,
-      "learning_rate": 9.726604210048015e-06,
-      "loss": 0.4431,
-      "step": 5676
-    },
-    {
-      "epoch": 0.37102150186262334,
-      "grad_norm": 0.43497100472450256,
-      "learning_rate": 9.726490313046984e-06,
-      "loss": 0.3566,
-      "step": 5677
-    },
-    {
-      "epoch": 0.37108685706816547,
-      "grad_norm": 0.41215407848358154,
-      "learning_rate": 9.726376392993221e-06,
-      "loss": 0.3405,
-      "step": 5678
-    },
-    {
-      "epoch": 0.3711522122737076,
-      "grad_norm": 0.4581126868724823,
-      "learning_rate": 9.726262449887282e-06,
-      "loss": 0.378,
-      "step": 5679
-    },
-    {
-      "epoch": 0.3712175674792497,
-      "grad_norm": 0.4768775403499603,
-      "learning_rate": 9.726148483729728e-06,
-      "loss": 0.4278,
-      "step": 5680
-    },
-    {
-      "epoch": 0.37128292268479185,
-      "grad_norm": 0.45675817131996155,
-      "learning_rate": 9.726034494521105e-06,
-      "loss": 0.4159,
-      "step": 5681
-    },
-    {
-      "epoch": 0.371348277890334,
-      "grad_norm": 0.46914592385292053,
-      "learning_rate": 9.725920482261979e-06,
-      "loss": 0.4281,
-      "step": 5682
-    },
-    {
-      "epoch": 0.3714136330958761,
-      "grad_norm": 0.4493812620639801,
-      "learning_rate": 9.725806446952899e-06,
-      "loss": 0.3658,
-      "step": 5683
-    },
-    {
-      "epoch": 0.37147898830141823,
-      "grad_norm": 0.7395375967025757,
-      "learning_rate": 9.725692388594426e-06,
-      "loss": 0.3804,
-      "step": 5684
-    },
-    {
-      "epoch": 0.3715443435069603,
-      "grad_norm": 0.45588287711143494,
-      "learning_rate": 9.725578307187112e-06,
-      "loss": 0.3539,
-      "step": 5685
-    },
-    {
-      "epoch": 0.37160969871250243,
-      "grad_norm": 0.4964325726032257,
-      "learning_rate": 9.725464202731515e-06,
-      "loss": 0.3911,
-      "step": 5686
-    },
-    {
-      "epoch": 0.37167505391804456,
-      "grad_norm": 0.5292208194732666,
-      "learning_rate": 9.725350075228194e-06,
-      "loss": 0.4343,
-      "step": 5687
-    },
-    {
-      "epoch": 0.3717404091235867,
-      "grad_norm": 0.4464513063430786,
-      "learning_rate": 9.725235924677703e-06,
-      "loss": 0.3828,
-      "step": 5688
-    },
-    {
-      "epoch": 0.3718057643291288,
-      "grad_norm": 0.47259700298309326,
-      "learning_rate": 9.7251217510806e-06,
-      "loss": 0.4177,
-      "step": 5689
-    },
-    {
-      "epoch": 0.37187111953467095,
-      "grad_norm": 0.4382416903972626,
-      "learning_rate": 9.72500755443744e-06,
-      "loss": 0.3304,
-      "step": 5690
-    },
-    {
-      "epoch": 0.3719364747402131,
-      "grad_norm": 0.4824185371398926,
-      "learning_rate": 9.724893334748781e-06,
-      "loss": 0.3786,
-      "step": 5691
-    },
-    {
-      "epoch": 0.3720018299457552,
-      "grad_norm": 0.47357267141342163,
-      "learning_rate": 9.724779092015183e-06,
-      "loss": 0.3813,
-      "step": 5692
-    },
-    {
-      "epoch": 0.3720671851512973,
-      "grad_norm": 0.4315038323402405,
-      "learning_rate": 9.724664826237198e-06,
-      "loss": 0.3573,
-      "step": 5693
-    },
-    {
-      "epoch": 0.3721325403568394,
-      "grad_norm": 0.44466453790664673,
-      "learning_rate": 9.724550537415386e-06,
-      "loss": 0.3895,
-      "step": 5694
-    },
-    {
-      "epoch": 0.37219789556238153,
-      "grad_norm": 0.42833951115608215,
-      "learning_rate": 9.724436225550305e-06,
-      "loss": 0.336,
-      "step": 5695
-    },
-    {
-      "epoch": 0.37226325076792366,
-      "grad_norm": 0.4936835467815399,
-      "learning_rate": 9.724321890642512e-06,
-      "loss": 0.3872,
-      "step": 5696
-    },
-    {
-      "epoch": 0.3723286059734658,
-      "grad_norm": 0.4252793788909912,
-      "learning_rate": 9.724207532692563e-06,
-      "loss": 0.331,
-      "step": 5697
-    },
-    {
-      "epoch": 0.3723939611790079,
-      "grad_norm": 0.48302435874938965,
-      "learning_rate": 9.724093151701019e-06,
-      "loss": 0.4133,
-      "step": 5698
-    },
-    {
-      "epoch": 0.37245931638455004,
-      "grad_norm": 0.44952890276908875,
-      "learning_rate": 9.723978747668436e-06,
-      "loss": 0.3816,
-      "step": 5699
-    },
-    {
-      "epoch": 0.37252467159009217,
-      "grad_norm": 0.46215274930000305,
-      "learning_rate": 9.72386432059537e-06,
-      "loss": 0.4048,
-      "step": 5700
-    },
-    {
-      "epoch": 0.3725900267956343,
-      "grad_norm": 0.4442006051540375,
-      "learning_rate": 9.723749870482384e-06,
-      "loss": 0.3711,
-      "step": 5701
-    },
-    {
-      "epoch": 0.37265538200117637,
-      "grad_norm": 0.4747190475463867,
-      "learning_rate": 9.723635397330032e-06,
-      "loss": 0.3532,
-      "step": 5702
-    },
-    {
-      "epoch": 0.3727207372067185,
-      "grad_norm": 0.509239673614502,
-      "learning_rate": 9.723520901138875e-06,
-      "loss": 0.435,
-      "step": 5703
-    },
-    {
-      "epoch": 0.37278609241226063,
-      "grad_norm": 0.4344327449798584,
-      "learning_rate": 9.723406381909469e-06,
-      "loss": 0.3907,
-      "step": 5704
-    },
-    {
-      "epoch": 0.37285144761780276,
-      "grad_norm": 0.41855117678642273,
-      "learning_rate": 9.723291839642372e-06,
-      "loss": 0.3361,
-      "step": 5705
-    },
-    {
-      "epoch": 0.3729168028233449,
-      "grad_norm": 0.4844302833080292,
-      "learning_rate": 9.723177274338146e-06,
-      "loss": 0.4348,
-      "step": 5706
-    },
-    {
-      "epoch": 0.372982158028887,
-      "grad_norm": 0.4601530432701111,
-      "learning_rate": 9.72306268599735e-06,
-      "loss": 0.3793,
-      "step": 5707
-    },
-    {
-      "epoch": 0.37304751323442914,
-      "grad_norm": 0.5064185261726379,
-      "learning_rate": 9.722948074620539e-06,
-      "loss": 0.4081,
-      "step": 5708
-    },
-    {
-      "epoch": 0.37311286843997127,
-      "grad_norm": 0.4733896851539612,
-      "learning_rate": 9.722833440208274e-06,
-      "loss": 0.4084,
-      "step": 5709
-    },
-    {
-      "epoch": 0.37317822364551334,
-      "grad_norm": 0.4767979085445404,
-      "learning_rate": 9.722718782761116e-06,
-      "loss": 0.4331,
-      "step": 5710
-    },
-    {
-      "epoch": 0.37324357885105547,
-      "grad_norm": 0.47276976704597473,
-      "learning_rate": 9.72260410227962e-06,
-      "loss": 0.4029,
-      "step": 5711
-    },
-    {
-      "epoch": 0.3733089340565976,
-      "grad_norm": 0.4697682857513428,
-      "learning_rate": 9.72248939876435e-06,
-      "loss": 0.4054,
-      "step": 5712
-    },
-    {
-      "epoch": 0.3733742892621397,
-      "grad_norm": 0.4399544596672058,
-      "learning_rate": 9.722374672215861e-06,
-      "loss": 0.369,
-      "step": 5713
-    },
-    {
-      "epoch": 0.37343964446768185,
-      "grad_norm": 0.4537452161312103,
-      "learning_rate": 9.722259922634717e-06,
-      "loss": 0.3864,
-      "step": 5714
-    },
-    {
-      "epoch": 0.373504999673224,
-      "grad_norm": 0.4806780517101288,
-      "learning_rate": 9.722145150021474e-06,
-      "loss": 0.4301,
-      "step": 5715
-    },
-    {
-      "epoch": 0.3735703548787661,
-      "grad_norm": 0.459480881690979,
-      "learning_rate": 9.722030354376693e-06,
-      "loss": 0.3682,
-      "step": 5716
-    },
-    {
-      "epoch": 0.37363571008430824,
-      "grad_norm": 0.45941513776779175,
-      "learning_rate": 9.721915535700934e-06,
-      "loss": 0.3576,
-      "step": 5717
-    },
-    {
-      "epoch": 0.3737010652898503,
-      "grad_norm": 0.4493623375892639,
-      "learning_rate": 9.721800693994759e-06,
-      "loss": 0.3914,
-      "step": 5718
-    },
-    {
-      "epoch": 0.37376642049539244,
-      "grad_norm": 0.47806409001350403,
-      "learning_rate": 9.721685829258724e-06,
-      "loss": 0.3855,
-      "step": 5719
-    },
-    {
-      "epoch": 0.37383177570093457,
-      "grad_norm": 0.4922315180301666,
-      "learning_rate": 9.721570941493392e-06,
-      "loss": 0.4044,
-      "step": 5720
-    },
-    {
-      "epoch": 0.3738971309064767,
-      "grad_norm": 0.4188794493675232,
-      "learning_rate": 9.721456030699323e-06,
-      "loss": 0.3548,
-      "step": 5721
-    },
-    {
-      "epoch": 0.3739624861120188,
-      "grad_norm": 0.4768242835998535,
-      "learning_rate": 9.721341096877078e-06,
-      "loss": 0.3684,
-      "step": 5722
-    },
-    {
-      "epoch": 0.37402784131756095,
-      "grad_norm": 0.4828105568885803,
-      "learning_rate": 9.721226140027215e-06,
-      "loss": 0.3814,
-      "step": 5723
-    },
-    {
-      "epoch": 0.3740931965231031,
-      "grad_norm": 0.41793200373649597,
-      "learning_rate": 9.721111160150299e-06,
-      "loss": 0.3481,
-      "step": 5724
-    },
-    {
-      "epoch": 0.3741585517286452,
-      "grad_norm": 0.45060980319976807,
-      "learning_rate": 9.720996157246887e-06,
-      "loss": 0.4173,
-      "step": 5725
-    },
-    {
-      "epoch": 0.37422390693418733,
-      "grad_norm": 0.4805615544319153,
-      "learning_rate": 9.72088113131754e-06,
-      "loss": 0.4074,
-      "step": 5726
-    },
-    {
-      "epoch": 0.3742892621397294,
-      "grad_norm": 0.4488745629787445,
-      "learning_rate": 9.720766082362823e-06,
-      "loss": 0.3878,
-      "step": 5727
-    },
-    {
-      "epoch": 0.37435461734527153,
-      "grad_norm": 0.4357205331325531,
-      "learning_rate": 9.720651010383293e-06,
-      "loss": 0.3577,
-      "step": 5728
-    },
-    {
-      "epoch": 0.37441997255081366,
-      "grad_norm": 0.47308945655822754,
-      "learning_rate": 9.720535915379513e-06,
-      "loss": 0.4411,
-      "step": 5729
-    },
-    {
-      "epoch": 0.3744853277563558,
-      "grad_norm": 0.44247686862945557,
-      "learning_rate": 9.720420797352041e-06,
-      "loss": 0.3704,
-      "step": 5730
-    },
-    {
-      "epoch": 0.3745506829618979,
-      "grad_norm": 0.447162389755249,
-      "learning_rate": 9.720305656301446e-06,
-      "loss": 0.3855,
-      "step": 5731
-    },
-    {
-      "epoch": 0.37461603816744005,
-      "grad_norm": 0.4559944272041321,
-      "learning_rate": 9.720190492228283e-06,
-      "loss": 0.3663,
-      "step": 5732
-    },
-    {
-      "epoch": 0.3746813933729822,
-      "grad_norm": 0.5380062460899353,
-      "learning_rate": 9.720075305133115e-06,
-      "loss": 0.5212,
-      "step": 5733
-    },
-    {
-      "epoch": 0.3747467485785243,
-      "grad_norm": 0.45457741618156433,
-      "learning_rate": 9.719960095016506e-06,
-      "loss": 0.3365,
-      "step": 5734
-    },
-    {
-      "epoch": 0.3748121037840664,
-      "grad_norm": 0.43439793586730957,
-      "learning_rate": 9.719844861879016e-06,
-      "loss": 0.3446,
-      "step": 5735
-    },
-    {
-      "epoch": 0.3748774589896085,
-      "grad_norm": 0.43436864018440247,
-      "learning_rate": 9.719729605721206e-06,
-      "loss": 0.3738,
-      "step": 5736
-    },
-    {
-      "epoch": 0.37494281419515063,
-      "grad_norm": 0.4463585615158081,
-      "learning_rate": 9.719614326543642e-06,
-      "loss": 0.3721,
-      "step": 5737
-    },
-    {
-      "epoch": 0.37500816940069276,
-      "grad_norm": 0.5004968047142029,
-      "learning_rate": 9.719499024346883e-06,
-      "loss": 0.4225,
-      "step": 5738
-    },
-    {
-      "epoch": 0.3750735246062349,
-      "grad_norm": 0.4942430853843689,
-      "learning_rate": 9.71938369913149e-06,
-      "loss": 0.4333,
-      "step": 5739
-    },
-    {
-      "epoch": 0.375138879811777,
-      "grad_norm": 0.48074761033058167,
-      "learning_rate": 9.719268350898031e-06,
-      "loss": 0.3774,
-      "step": 5740
-    },
-    {
-      "epoch": 0.37520423501731914,
-      "grad_norm": 0.47752365469932556,
-      "learning_rate": 9.719152979647064e-06,
-      "loss": 0.4025,
-      "step": 5741
-    },
-    {
-      "epoch": 0.37526959022286127,
-      "grad_norm": 0.5058532953262329,
-      "learning_rate": 9.719037585379155e-06,
-      "loss": 0.4587,
-      "step": 5742
-    },
-    {
-      "epoch": 0.3753349454284034,
-      "grad_norm": 0.5277442336082458,
-      "learning_rate": 9.718922168094862e-06,
-      "loss": 0.4512,
-      "step": 5743
-    },
-    {
-      "epoch": 0.37540030063394547,
-      "grad_norm": 0.4554592967033386,
-      "learning_rate": 9.718806727794751e-06,
-      "loss": 0.3809,
-      "step": 5744
-    },
-    {
-      "epoch": 0.3754656558394876,
-      "grad_norm": 0.4330997169017792,
-      "learning_rate": 9.718691264479386e-06,
-      "loss": 0.335,
-      "step": 5745
-    },
-    {
-      "epoch": 0.37553101104502973,
-      "grad_norm": 0.49338728189468384,
-      "learning_rate": 9.718575778149328e-06,
-      "loss": 0.3968,
-      "step": 5746
-    },
-    {
-      "epoch": 0.37559636625057186,
-      "grad_norm": 0.4403141438961029,
-      "learning_rate": 9.718460268805143e-06,
-      "loss": 0.3625,
-      "step": 5747
-    },
-    {
-      "epoch": 0.375661721456114,
-      "grad_norm": 0.45258715748786926,
-      "learning_rate": 9.718344736447392e-06,
-      "loss": 0.3745,
-      "step": 5748
-    },
-    {
-      "epoch": 0.3757270766616561,
-      "grad_norm": 0.4778488576412201,
-      "learning_rate": 9.718229181076639e-06,
-      "loss": 0.3956,
-      "step": 5749
-    },
-    {
-      "epoch": 0.37579243186719824,
-      "grad_norm": 0.46462589502334595,
-      "learning_rate": 9.718113602693446e-06,
-      "loss": 0.3873,
-      "step": 5750
-    },
-    {
-      "epoch": 0.37585778707274037,
-      "grad_norm": 0.4922333359718323,
-      "learning_rate": 9.717998001298382e-06,
-      "loss": 0.4461,
-      "step": 5751
-    },
-    {
-      "epoch": 0.37592314227828244,
-      "grad_norm": 0.49379634857177734,
-      "learning_rate": 9.717882376892002e-06,
-      "loss": 0.4088,
-      "step": 5752
-    },
-    {
-      "epoch": 0.37598849748382457,
-      "grad_norm": 0.502359926700592,
-      "learning_rate": 9.71776672947488e-06,
-      "loss": 0.47,
-      "step": 5753
-    },
-    {
-      "epoch": 0.3760538526893667,
-      "grad_norm": 0.479813814163208,
-      "learning_rate": 9.717651059047574e-06,
-      "loss": 0.3779,
-      "step": 5754
-    },
-    {
-      "epoch": 0.3761192078949088,
-      "grad_norm": 0.4545009732246399,
-      "learning_rate": 9.717535365610649e-06,
-      "loss": 0.3494,
-      "step": 5755
-    },
-    {
-      "epoch": 0.37618456310045095,
-      "grad_norm": 0.43695205450057983,
-      "learning_rate": 9.71741964916467e-06,
-      "loss": 0.3387,
-      "step": 5756
-    },
-    {
-      "epoch": 0.3762499183059931,
-      "grad_norm": 0.4959096610546112,
-      "learning_rate": 9.717303909710201e-06,
-      "loss": 0.3893,
-      "step": 5757
-    },
-    {
-      "epoch": 0.3763152735115352,
-      "grad_norm": 0.49351975321769714,
-      "learning_rate": 9.717188147247806e-06,
-      "loss": 0.4322,
-      "step": 5758
-    },
-    {
-      "epoch": 0.37638062871707734,
-      "grad_norm": 0.49044859409332275,
-      "learning_rate": 9.71707236177805e-06,
-      "loss": 0.427,
-      "step": 5759
-    },
-    {
-      "epoch": 0.3764459839226194,
-      "grad_norm": 0.4676279127597809,
-      "learning_rate": 9.7169565533015e-06,
-      "loss": 0.3888,
-      "step": 5760
-    },
-    {
-      "epoch": 0.37651133912816154,
-      "grad_norm": 0.46494054794311523,
-      "learning_rate": 9.716840721818717e-06,
-      "loss": 0.3148,
-      "step": 5761
-    },
-    {
-      "epoch": 0.37657669433370367,
-      "grad_norm": 0.4796944260597229,
-      "learning_rate": 9.716724867330268e-06,
-      "loss": 0.4077,
-      "step": 5762
-    },
-    {
-      "epoch": 0.3766420495392458,
-      "grad_norm": 0.44766202569007874,
-      "learning_rate": 9.716608989836718e-06,
-      "loss": 0.3447,
-      "step": 5763
-    },
-    {
-      "epoch": 0.3767074047447879,
-      "grad_norm": 0.4754466712474823,
-      "learning_rate": 9.716493089338632e-06,
-      "loss": 0.417,
-      "step": 5764
-    },
-    {
-      "epoch": 0.37677275995033005,
-      "grad_norm": 0.5241832137107849,
-      "learning_rate": 9.716377165836575e-06,
-      "loss": 0.3804,
-      "step": 5765
-    },
-    {
-      "epoch": 0.3768381151558722,
-      "grad_norm": 0.5270892977714539,
-      "learning_rate": 9.716261219331113e-06,
-      "loss": 0.4039,
-      "step": 5766
-    },
-    {
-      "epoch": 0.3769034703614143,
-      "grad_norm": 0.4481326639652252,
-      "learning_rate": 9.71614524982281e-06,
-      "loss": 0.3254,
-      "step": 5767
-    },
-    {
-      "epoch": 0.37696882556695643,
-      "grad_norm": 0.44411876797676086,
-      "learning_rate": 9.716029257312234e-06,
-      "loss": 0.3497,
-      "step": 5768
-    },
-    {
-      "epoch": 0.3770341807724985,
-      "grad_norm": 0.4580141603946686,
-      "learning_rate": 9.71591324179995e-06,
-      "loss": 0.3828,
-      "step": 5769
-    },
-    {
-      "epoch": 0.37709953597804063,
-      "grad_norm": 0.7634884119033813,
-      "learning_rate": 9.715797203286523e-06,
-      "loss": 0.4167,
-      "step": 5770
-    },
-    {
-      "epoch": 0.37716489118358276,
-      "grad_norm": 0.4594669044017792,
-      "learning_rate": 9.71568114177252e-06,
-      "loss": 0.4285,
-      "step": 5771
-    },
-    {
-      "epoch": 0.3772302463891249,
-      "grad_norm": 0.4424137771129608,
-      "learning_rate": 9.715565057258506e-06,
-      "loss": 0.3743,
-      "step": 5772
-    },
-    {
-      "epoch": 0.377295601594667,
-      "grad_norm": 0.4911845624446869,
-      "learning_rate": 9.715448949745045e-06,
-      "loss": 0.391,
-      "step": 5773
-    },
-    {
-      "epoch": 0.37736095680020915,
-      "grad_norm": 0.4715944528579712,
-      "learning_rate": 9.715332819232708e-06,
-      "loss": 0.3792,
-      "step": 5774
-    },
-    {
-      "epoch": 0.3774263120057513,
-      "grad_norm": 0.46876052021980286,
-      "learning_rate": 9.715216665722059e-06,
-      "loss": 0.4158,
-      "step": 5775
-    },
-    {
-      "epoch": 0.3774916672112934,
-      "grad_norm": 0.5218589901924133,
-      "learning_rate": 9.715100489213665e-06,
-      "loss": 0.4765,
-      "step": 5776
-    },
-    {
-      "epoch": 0.3775570224168355,
-      "grad_norm": 0.45977842807769775,
-      "learning_rate": 9.714984289708093e-06,
-      "loss": 0.381,
-      "step": 5777
-    },
-    {
-      "epoch": 0.3776223776223776,
-      "grad_norm": 0.45912104845046997,
-      "learning_rate": 9.714868067205908e-06,
-      "loss": 0.3881,
-      "step": 5778
-    },
-    {
-      "epoch": 0.37768773282791973,
-      "grad_norm": 0.47479283809661865,
-      "learning_rate": 9.714751821707678e-06,
-      "loss": 0.4224,
-      "step": 5779
-    },
-    {
-      "epoch": 0.37775308803346186,
-      "grad_norm": 0.4962746202945709,
-      "learning_rate": 9.714635553213971e-06,
-      "loss": 0.4039,
-      "step": 5780
-    },
-    {
-      "epoch": 0.377818443239004,
-      "grad_norm": 0.4834185242652893,
-      "learning_rate": 9.714519261725354e-06,
-      "loss": 0.3381,
-      "step": 5781
-    },
-    {
-      "epoch": 0.3778837984445461,
-      "grad_norm": 0.4590103030204773,
-      "learning_rate": 9.714402947242392e-06,
-      "loss": 0.3527,
-      "step": 5782
-    },
-    {
-      "epoch": 0.37794915365008824,
-      "grad_norm": 0.5560005903244019,
-      "learning_rate": 9.714286609765654e-06,
-      "loss": 0.4437,
-      "step": 5783
-    },
-    {
-      "epoch": 0.37801450885563037,
-      "grad_norm": 0.47062599658966064,
-      "learning_rate": 9.714170249295705e-06,
-      "loss": 0.3903,
-      "step": 5784
-    },
-    {
-      "epoch": 0.3780798640611725,
-      "grad_norm": 0.47350549697875977,
-      "learning_rate": 9.714053865833117e-06,
-      "loss": 0.3844,
-      "step": 5785
-    },
-    {
-      "epoch": 0.37814521926671457,
-      "grad_norm": 0.46465545892715454,
-      "learning_rate": 9.713937459378456e-06,
-      "loss": 0.3819,
-      "step": 5786
-    },
-    {
-      "epoch": 0.3782105744722567,
-      "grad_norm": 0.4582565128803253,
-      "learning_rate": 9.713821029932287e-06,
-      "loss": 0.3651,
-      "step": 5787
-    },
-    {
-      "epoch": 0.37827592967779883,
-      "grad_norm": 0.48874345421791077,
-      "learning_rate": 9.71370457749518e-06,
-      "loss": 0.4033,
-      "step": 5788
-    },
-    {
-      "epoch": 0.37834128488334096,
-      "grad_norm": 0.4774669110774994,
-      "learning_rate": 9.7135881020677e-06,
-      "loss": 0.4221,
-      "step": 5789
-    },
-    {
-      "epoch": 0.3784066400888831,
-      "grad_norm": 0.45412692427635193,
-      "learning_rate": 9.713471603650422e-06,
-      "loss": 0.3775,
-      "step": 5790
-    },
-    {
-      "epoch": 0.3784719952944252,
-      "grad_norm": 0.4761379361152649,
-      "learning_rate": 9.713355082243909e-06,
-      "loss": 0.4267,
-      "step": 5791
-    },
-    {
-      "epoch": 0.37853735049996734,
-      "grad_norm": 0.536064624786377,
-      "learning_rate": 9.713238537848731e-06,
-      "loss": 0.3554,
-      "step": 5792
-    },
-    {
-      "epoch": 0.37860270570550947,
-      "grad_norm": 0.43408045172691345,
-      "learning_rate": 9.713121970465456e-06,
-      "loss": 0.393,
-      "step": 5793
-    },
-    {
-      "epoch": 0.37866806091105154,
-      "grad_norm": 0.4712388813495636,
-      "learning_rate": 9.713005380094651e-06,
-      "loss": 0.3941,
-      "step": 5794
-    },
-    {
-      "epoch": 0.37873341611659367,
-      "grad_norm": 0.5197617411613464,
-      "learning_rate": 9.712888766736887e-06,
-      "loss": 0.3907,
-      "step": 5795
-    },
-    {
-      "epoch": 0.3787987713221358,
-      "grad_norm": 0.4356193542480469,
-      "learning_rate": 9.712772130392731e-06,
-      "loss": 0.3731,
-      "step": 5796
-    },
-    {
-      "epoch": 0.3788641265276779,
-      "grad_norm": 0.4431528151035309,
-      "learning_rate": 9.712655471062753e-06,
-      "loss": 0.3671,
-      "step": 5797
-    },
-    {
-      "epoch": 0.37892948173322005,
-      "grad_norm": 0.41992664337158203,
-      "learning_rate": 9.712538788747522e-06,
-      "loss": 0.3212,
-      "step": 5798
-    },
-    {
-      "epoch": 0.3789948369387622,
-      "grad_norm": 0.5286357402801514,
-      "learning_rate": 9.712422083447606e-06,
-      "loss": 0.4632,
-      "step": 5799
-    },
-    {
-      "epoch": 0.3790601921443043,
-      "grad_norm": 0.4489547312259674,
-      "learning_rate": 9.712305355163577e-06,
-      "loss": 0.3681,
-      "step": 5800
-    },
-    {
-      "epoch": 0.37912554734984644,
-      "grad_norm": 0.43811309337615967,
-      "learning_rate": 9.712188603896e-06,
-      "loss": 0.3575,
-      "step": 5801
-    },
-    {
-      "epoch": 0.3791909025553885,
-      "grad_norm": 0.506718099117279,
-      "learning_rate": 9.712071829645447e-06,
-      "loss": 0.3645,
-      "step": 5802
-    },
-    {
-      "epoch": 0.37925625776093064,
-      "grad_norm": 0.4482395052909851,
-      "learning_rate": 9.711955032412488e-06,
-      "loss": 0.3644,
-      "step": 5803
-    },
-    {
-      "epoch": 0.37932161296647277,
-      "grad_norm": 0.46370434761047363,
-      "learning_rate": 9.71183821219769e-06,
-      "loss": 0.4268,
-      "step": 5804
-    },
-    {
-      "epoch": 0.3793869681720149,
-      "grad_norm": 0.45522060990333557,
-      "learning_rate": 9.711721369001628e-06,
-      "loss": 0.4008,
-      "step": 5805
-    },
-    {
-      "epoch": 0.379452323377557,
-      "grad_norm": 0.4726666808128357,
-      "learning_rate": 9.711604502824866e-06,
-      "loss": 0.4255,
-      "step": 5806
-    },
-    {
-      "epoch": 0.37951767858309915,
-      "grad_norm": 0.48169249296188354,
-      "learning_rate": 9.711487613667979e-06,
-      "loss": 0.4176,
-      "step": 5807
-    },
-    {
-      "epoch": 0.3795830337886413,
-      "grad_norm": 0.4644019901752472,
-      "learning_rate": 9.711370701531531e-06,
-      "loss": 0.442,
-      "step": 5808
-    },
-    {
-      "epoch": 0.3796483889941834,
-      "grad_norm": 0.5224023461341858,
-      "learning_rate": 9.711253766416098e-06,
-      "loss": 0.43,
-      "step": 5809
-    },
-    {
-      "epoch": 0.37971374419972553,
-      "grad_norm": 0.4404146373271942,
-      "learning_rate": 9.711136808322248e-06,
-      "loss": 0.3619,
-      "step": 5810
-    },
-    {
-      "epoch": 0.3797790994052676,
-      "grad_norm": 0.42197537422180176,
-      "learning_rate": 9.711019827250553e-06,
-      "loss": 0.3293,
-      "step": 5811
-    },
-    {
-      "epoch": 0.37984445461080973,
-      "grad_norm": 0.4823967218399048,
-      "learning_rate": 9.710902823201581e-06,
-      "loss": 0.429,
-      "step": 5812
-    },
-    {
-      "epoch": 0.37990980981635186,
-      "grad_norm": 0.46075764298439026,
-      "learning_rate": 9.710785796175904e-06,
-      "loss": 0.4204,
-      "step": 5813
-    },
-    {
-      "epoch": 0.379975165021894,
-      "grad_norm": 0.49664467573165894,
-      "learning_rate": 9.710668746174094e-06,
-      "loss": 0.4367,
-      "step": 5814
-    },
-    {
-      "epoch": 0.3800405202274361,
-      "grad_norm": 0.4818166196346283,
-      "learning_rate": 9.710551673196718e-06,
-      "loss": 0.4268,
-      "step": 5815
-    },
-    {
-      "epoch": 0.38010587543297825,
-      "grad_norm": 0.47597020864486694,
-      "learning_rate": 9.710434577244352e-06,
-      "loss": 0.3874,
-      "step": 5816
-    },
-    {
-      "epoch": 0.3801712306385204,
-      "grad_norm": 0.48701581358909607,
-      "learning_rate": 9.710317458317563e-06,
-      "loss": 0.4458,
-      "step": 5817
-    },
-    {
-      "epoch": 0.3802365858440625,
-      "grad_norm": 0.46156200766563416,
-      "learning_rate": 9.710200316416925e-06,
-      "loss": 0.3805,
-      "step": 5818
-    },
-    {
-      "epoch": 0.3803019410496046,
-      "grad_norm": 0.46452781558036804,
-      "learning_rate": 9.710083151543009e-06,
-      "loss": 0.3927,
-      "step": 5819
-    },
-    {
-      "epoch": 0.3803672962551467,
-      "grad_norm": 0.47801366448402405,
-      "learning_rate": 9.709965963696384e-06,
-      "loss": 0.4331,
-      "step": 5820
-    },
-    {
-      "epoch": 0.38043265146068883,
-      "grad_norm": 0.4413713216781616,
-      "learning_rate": 9.709848752877625e-06,
-      "loss": 0.3852,
-      "step": 5821
-    },
-    {
-      "epoch": 0.38049800666623096,
-      "grad_norm": 0.4716586470603943,
-      "learning_rate": 9.7097315190873e-06,
-      "loss": 0.4046,
-      "step": 5822
-    },
-    {
-      "epoch": 0.3805633618717731,
-      "grad_norm": 0.488387793302536,
-      "learning_rate": 9.709614262325984e-06,
-      "loss": 0.3903,
-      "step": 5823
-    },
-    {
-      "epoch": 0.3806287170773152,
-      "grad_norm": 0.4786669611930847,
-      "learning_rate": 9.709496982594248e-06,
-      "loss": 0.4106,
-      "step": 5824
-    },
-    {
-      "epoch": 0.38069407228285734,
-      "grad_norm": 0.49281588196754456,
-      "learning_rate": 9.709379679892664e-06,
-      "loss": 0.4434,
-      "step": 5825
-    },
-    {
-      "epoch": 0.38075942748839947,
-      "grad_norm": 0.44772833585739136,
-      "learning_rate": 9.709262354221802e-06,
-      "loss": 0.3613,
-      "step": 5826
-    },
-    {
-      "epoch": 0.3808247826939416,
-      "grad_norm": 0.4543074667453766,
-      "learning_rate": 9.709145005582236e-06,
-      "loss": 0.3803,
-      "step": 5827
-    },
-    {
-      "epoch": 0.38089013789948367,
-      "grad_norm": 0.4856089949607849,
-      "learning_rate": 9.709027633974539e-06,
-      "loss": 0.4341,
-      "step": 5828
-    },
-    {
-      "epoch": 0.3809554931050258,
-      "grad_norm": 0.46512290835380554,
-      "learning_rate": 9.708910239399285e-06,
-      "loss": 0.4293,
-      "step": 5829
-    },
-    {
-      "epoch": 0.38102084831056793,
-      "grad_norm": 0.48706063628196716,
-      "learning_rate": 9.708792821857043e-06,
-      "loss": 0.4709,
-      "step": 5830
-    },
-    {
-      "epoch": 0.38108620351611006,
-      "grad_norm": 0.5061370134353638,
-      "learning_rate": 9.708675381348386e-06,
-      "loss": 0.4216,
-      "step": 5831
-    },
-    {
-      "epoch": 0.3811515587216522,
-      "grad_norm": 0.4879632592201233,
-      "learning_rate": 9.708557917873888e-06,
-      "loss": 0.4117,
-      "step": 5832
-    },
-    {
-      "epoch": 0.3812169139271943,
-      "grad_norm": 0.44429755210876465,
-      "learning_rate": 9.708440431434124e-06,
-      "loss": 0.3531,
-      "step": 5833
-    },
-    {
-      "epoch": 0.38128226913273644,
-      "grad_norm": 0.45839354395866394,
-      "learning_rate": 9.708322922029663e-06,
-      "loss": 0.4112,
-      "step": 5834
-    },
-    {
-      "epoch": 0.38134762433827857,
-      "grad_norm": 0.4795777499675751,
-      "learning_rate": 9.70820538966108e-06,
-      "loss": 0.3855,
-      "step": 5835
-    },
-    {
-      "epoch": 0.38141297954382064,
-      "grad_norm": 0.48305612802505493,
-      "learning_rate": 9.70808783432895e-06,
-      "loss": 0.3883,
-      "step": 5836
-    },
-    {
-      "epoch": 0.38147833474936277,
-      "grad_norm": 0.48505690693855286,
-      "learning_rate": 9.707970256033842e-06,
-      "loss": 0.4406,
-      "step": 5837
-    },
-    {
-      "epoch": 0.3815436899549049,
-      "grad_norm": 0.42778703570365906,
-      "learning_rate": 9.707852654776334e-06,
-      "loss": 0.3501,
-      "step": 5838
-    },
-    {
-      "epoch": 0.381609045160447,
-      "grad_norm": 0.5006610155105591,
-      "learning_rate": 9.707735030556997e-06,
-      "loss": 0.425,
-      "step": 5839
-    },
-    {
-      "epoch": 0.38167440036598915,
-      "grad_norm": 0.46174129843711853,
-      "learning_rate": 9.707617383376405e-06,
-      "loss": 0.383,
-      "step": 5840
-    },
-    {
-      "epoch": 0.3817397555715313,
-      "grad_norm": 0.47170525789260864,
-      "learning_rate": 9.707499713235134e-06,
-      "loss": 0.373,
-      "step": 5841
-    },
-    {
-      "epoch": 0.3818051107770734,
-      "grad_norm": 0.4463726878166199,
-      "learning_rate": 9.707382020133753e-06,
-      "loss": 0.3583,
-      "step": 5842
-    },
-    {
-      "epoch": 0.38187046598261554,
-      "grad_norm": 0.49965202808380127,
-      "learning_rate": 9.707264304072842e-06,
-      "loss": 0.3849,
-      "step": 5843
-    },
-    {
-      "epoch": 0.3819358211881576,
-      "grad_norm": 0.4620155096054077,
-      "learning_rate": 9.70714656505297e-06,
-      "loss": 0.4252,
-      "step": 5844
-    },
-    {
-      "epoch": 0.38200117639369974,
-      "grad_norm": 0.478447288274765,
-      "learning_rate": 9.707028803074714e-06,
-      "loss": 0.4037,
-      "step": 5845
-    },
-    {
-      "epoch": 0.38206653159924187,
-      "grad_norm": 0.4635656177997589,
-      "learning_rate": 9.706911018138648e-06,
-      "loss": 0.4044,
-      "step": 5846
-    },
-    {
-      "epoch": 0.382131886804784,
-      "grad_norm": 0.46208301186561584,
-      "learning_rate": 9.706793210245347e-06,
-      "loss": 0.3903,
-      "step": 5847
-    },
-    {
-      "epoch": 0.3821972420103261,
-      "grad_norm": 0.4922182261943817,
-      "learning_rate": 9.706675379395384e-06,
-      "loss": 0.4278,
-      "step": 5848
-    },
-    {
-      "epoch": 0.38226259721586825,
-      "grad_norm": 0.4474487006664276,
-      "learning_rate": 9.706557525589335e-06,
-      "loss": 0.3734,
-      "step": 5849
-    },
-    {
-      "epoch": 0.3823279524214104,
-      "grad_norm": 0.47487691044807434,
-      "learning_rate": 9.706439648827774e-06,
-      "loss": 0.4198,
-      "step": 5850
-    },
-    {
-      "epoch": 0.3823933076269525,
-      "grad_norm": 0.4491572380065918,
-      "learning_rate": 9.706321749111278e-06,
-      "loss": 0.3854,
-      "step": 5851
-    },
-    {
-      "epoch": 0.38245866283249463,
-      "grad_norm": 0.45293572545051575,
-      "learning_rate": 9.70620382644042e-06,
-      "loss": 0.3901,
-      "step": 5852
-    },
-    {
-      "epoch": 0.3825240180380367,
-      "grad_norm": 0.48248592019081116,
-      "learning_rate": 9.706085880815775e-06,
-      "loss": 0.3819,
-      "step": 5853
-    },
-    {
-      "epoch": 0.38258937324357883,
-      "grad_norm": 0.4660474956035614,
-      "learning_rate": 9.705967912237918e-06,
-      "loss": 0.39,
-      "step": 5854
-    },
-    {
-      "epoch": 0.38265472844912096,
-      "grad_norm": 0.44777587056159973,
-      "learning_rate": 9.705849920707426e-06,
-      "loss": 0.4059,
-      "step": 5855
-    },
-    {
-      "epoch": 0.3827200836546631,
-      "grad_norm": 0.47871822118759155,
-      "learning_rate": 9.705731906224874e-06,
-      "loss": 0.3948,
-      "step": 5856
-    },
-    {
-      "epoch": 0.3827854388602052,
-      "grad_norm": 0.47725963592529297,
-      "learning_rate": 9.705613868790836e-06,
-      "loss": 0.4298,
-      "step": 5857
-    },
-    {
-      "epoch": 0.38285079406574735,
-      "grad_norm": 0.4882162809371948,
-      "learning_rate": 9.70549580840589e-06,
-      "loss": 0.4004,
-      "step": 5858
-    },
-    {
-      "epoch": 0.3829161492712895,
-      "grad_norm": 0.43429046869277954,
-      "learning_rate": 9.70537772507061e-06,
-      "loss": 0.3604,
-      "step": 5859
-    },
-    {
-      "epoch": 0.3829815044768316,
-      "grad_norm": 0.5152312517166138,
-      "learning_rate": 9.705259618785574e-06,
-      "loss": 0.4511,
-      "step": 5860
-    },
-    {
-      "epoch": 0.3830468596823737,
-      "grad_norm": 0.4840807318687439,
-      "learning_rate": 9.705141489551355e-06,
-      "loss": 0.4271,
-      "step": 5861
-    },
-    {
-      "epoch": 0.3831122148879158,
-      "grad_norm": 0.4814096987247467,
-      "learning_rate": 9.705023337368533e-06,
-      "loss": 0.3728,
-      "step": 5862
-    },
-    {
-      "epoch": 0.38317757009345793,
-      "grad_norm": 0.4504605531692505,
-      "learning_rate": 9.704905162237682e-06,
-      "loss": 0.3762,
-      "step": 5863
-    },
-    {
-      "epoch": 0.38324292529900006,
-      "grad_norm": 0.48239997029304504,
-      "learning_rate": 9.704786964159378e-06,
-      "loss": 0.3957,
-      "step": 5864
-    },
-    {
-      "epoch": 0.3833082805045422,
-      "grad_norm": 0.47395145893096924,
-      "learning_rate": 9.704668743134198e-06,
-      "loss": 0.369,
-      "step": 5865
-    },
-    {
-      "epoch": 0.3833736357100843,
-      "grad_norm": 0.4495809078216553,
-      "learning_rate": 9.704550499162718e-06,
-      "loss": 0.3869,
-      "step": 5866
-    },
-    {
-      "epoch": 0.38343899091562644,
-      "grad_norm": 0.4741559326648712,
-      "learning_rate": 9.704432232245515e-06,
-      "loss": 0.4328,
-      "step": 5867
-    },
-    {
-      "epoch": 0.38350434612116857,
-      "grad_norm": 0.4773492217063904,
-      "learning_rate": 9.704313942383168e-06,
-      "loss": 0.3766,
-      "step": 5868
-    },
-    {
-      "epoch": 0.3835697013267107,
-      "grad_norm": 0.5167033076286316,
-      "learning_rate": 9.704195629576251e-06,
-      "loss": 0.441,
-      "step": 5869
-    },
-    {
-      "epoch": 0.38363505653225277,
-      "grad_norm": 0.42016810178756714,
-      "learning_rate": 9.704077293825344e-06,
-      "loss": 0.3633,
-      "step": 5870
-    },
-    {
-      "epoch": 0.3837004117377949,
-      "grad_norm": 0.4640307128429413,
-      "learning_rate": 9.70395893513102e-06,
-      "loss": 0.3903,
-      "step": 5871
-    },
-    {
-      "epoch": 0.38376576694333703,
-      "grad_norm": 0.4658341109752655,
-      "learning_rate": 9.70384055349386e-06,
-      "loss": 0.3917,
-      "step": 5872
-    },
-    {
-      "epoch": 0.38383112214887916,
-      "grad_norm": 0.4318199157714844,
-      "learning_rate": 9.70372214891444e-06,
-      "loss": 0.3425,
-      "step": 5873
-    },
-    {
-      "epoch": 0.3838964773544213,
-      "grad_norm": 0.4524925649166107,
-      "learning_rate": 9.703603721393338e-06,
-      "loss": 0.3607,
-      "step": 5874
-    },
-    {
-      "epoch": 0.3839618325599634,
-      "grad_norm": 0.44121888279914856,
-      "learning_rate": 9.70348527093113e-06,
-      "loss": 0.3294,
-      "step": 5875
-    },
-    {
-      "epoch": 0.38402718776550554,
-      "grad_norm": 0.46296006441116333,
-      "learning_rate": 9.703366797528396e-06,
-      "loss": 0.3842,
-      "step": 5876
-    },
-    {
-      "epoch": 0.38409254297104767,
-      "grad_norm": 0.44925475120544434,
-      "learning_rate": 9.703248301185712e-06,
-      "loss": 0.3972,
-      "step": 5877
-    },
-    {
-      "epoch": 0.38415789817658974,
-      "grad_norm": 0.44322964549064636,
-      "learning_rate": 9.703129781903657e-06,
-      "loss": 0.3795,
-      "step": 5878
-    },
-    {
-      "epoch": 0.38422325338213187,
-      "grad_norm": 0.5163261294364929,
-      "learning_rate": 9.703011239682808e-06,
-      "loss": 0.492,
-      "step": 5879
-    },
-    {
-      "epoch": 0.384288608587674,
-      "grad_norm": 0.485990047454834,
-      "learning_rate": 9.702892674523744e-06,
-      "loss": 0.4458,
-      "step": 5880
-    },
-    {
-      "epoch": 0.3843539637932161,
-      "grad_norm": 0.4604759216308594,
-      "learning_rate": 9.702774086427044e-06,
-      "loss": 0.3971,
-      "step": 5881
-    },
-    {
-      "epoch": 0.38441931899875825,
-      "grad_norm": 0.46555641293525696,
-      "learning_rate": 9.702655475393286e-06,
-      "loss": 0.4059,
-      "step": 5882
-    },
-    {
-      "epoch": 0.3844846742043004,
-      "grad_norm": 0.6112158298492432,
-      "learning_rate": 9.702536841423047e-06,
-      "loss": 0.3652,
-      "step": 5883
-    },
-    {
-      "epoch": 0.3845500294098425,
-      "grad_norm": 0.4829863905906677,
-      "learning_rate": 9.702418184516906e-06,
-      "loss": 0.4305,
-      "step": 5884
-    },
-    {
-      "epoch": 0.38461538461538464,
-      "grad_norm": 0.4427691400051117,
-      "learning_rate": 9.702299504675443e-06,
-      "loss": 0.402,
-      "step": 5885
-    },
-    {
-      "epoch": 0.3846807398209267,
-      "grad_norm": 0.44634321331977844,
-      "learning_rate": 9.702180801899237e-06,
-      "loss": 0.3748,
-      "step": 5886
-    },
-    {
-      "epoch": 0.38474609502646884,
-      "grad_norm": 0.46292808651924133,
-      "learning_rate": 9.702062076188866e-06,
-      "loss": 0.4121,
-      "step": 5887
-    },
-    {
-      "epoch": 0.38481145023201097,
-      "grad_norm": 0.47690489888191223,
-      "learning_rate": 9.701943327544909e-06,
-      "loss": 0.3898,
-      "step": 5888
-    },
-    {
-      "epoch": 0.3848768054375531,
-      "grad_norm": 0.468337744474411,
-      "learning_rate": 9.701824555967947e-06,
-      "loss": 0.387,
-      "step": 5889
-    },
-    {
-      "epoch": 0.3849421606430952,
-      "grad_norm": 0.4708351194858551,
-      "learning_rate": 9.701705761458555e-06,
-      "loss": 0.4185,
-      "step": 5890
-    },
-    {
-      "epoch": 0.38500751584863735,
-      "grad_norm": 0.4758375883102417,
-      "learning_rate": 9.701586944017317e-06,
-      "loss": 0.3729,
-      "step": 5891
-    },
-    {
-      "epoch": 0.3850728710541795,
-      "grad_norm": 0.485735148191452,
-      "learning_rate": 9.70146810364481e-06,
-      "loss": 0.434,
-      "step": 5892
-    },
-    {
-      "epoch": 0.3851382262597216,
-      "grad_norm": 0.4785767197608948,
-      "learning_rate": 9.701349240341615e-06,
-      "loss": 0.3815,
-      "step": 5893
-    },
-    {
-      "epoch": 0.38520358146526373,
-      "grad_norm": 0.4771265685558319,
-      "learning_rate": 9.70123035410831e-06,
-      "loss": 0.4088,
-      "step": 5894
-    },
-    {
-      "epoch": 0.3852689366708058,
-      "grad_norm": 0.48006612062454224,
-      "learning_rate": 9.701111444945478e-06,
-      "loss": 0.4555,
-      "step": 5895
-    },
-    {
-      "epoch": 0.38533429187634793,
-      "grad_norm": 0.4835205376148224,
-      "learning_rate": 9.700992512853695e-06,
-      "loss": 0.4414,
-      "step": 5896
-    },
-    {
-      "epoch": 0.38539964708189006,
-      "grad_norm": 0.4276254177093506,
-      "learning_rate": 9.700873557833543e-06,
-      "loss": 0.3414,
-      "step": 5897
-    },
-    {
-      "epoch": 0.3854650022874322,
-      "grad_norm": 0.4507283568382263,
-      "learning_rate": 9.700754579885603e-06,
-      "loss": 0.3464,
-      "step": 5898
-    },
-    {
-      "epoch": 0.3855303574929743,
-      "grad_norm": 0.4980127513408661,
-      "learning_rate": 9.700635579010454e-06,
-      "loss": 0.447,
-      "step": 5899
-    },
-    {
-      "epoch": 0.38559571269851645,
-      "grad_norm": 0.4952795207500458,
-      "learning_rate": 9.700516555208677e-06,
-      "loss": 0.4639,
-      "step": 5900
-    },
-    {
-      "epoch": 0.3856610679040586,
-      "grad_norm": 0.5392407774925232,
-      "learning_rate": 9.700397508480852e-06,
-      "loss": 0.4082,
-      "step": 5901
-    },
-    {
-      "epoch": 0.3857264231096007,
-      "grad_norm": 0.5083760619163513,
-      "learning_rate": 9.70027843882756e-06,
-      "loss": 0.4325,
-      "step": 5902
-    },
-    {
-      "epoch": 0.3857917783151428,
-      "grad_norm": 0.45893141627311707,
-      "learning_rate": 9.700159346249383e-06,
-      "loss": 0.4033,
-      "step": 5903
-    },
-    {
-      "epoch": 0.3858571335206849,
-      "grad_norm": 0.4481857717037201,
-      "learning_rate": 9.7000402307469e-06,
-      "loss": 0.3696,
-      "step": 5904
-    },
-    {
-      "epoch": 0.38592248872622703,
-      "grad_norm": 0.5332913994789124,
-      "learning_rate": 9.69992109232069e-06,
-      "loss": 0.4616,
-      "step": 5905
-    },
-    {
-      "epoch": 0.38598784393176916,
-      "grad_norm": 0.474316269159317,
-      "learning_rate": 9.69980193097134e-06,
-      "loss": 0.3921,
-      "step": 5906
-    },
-    {
-      "epoch": 0.3860531991373113,
-      "grad_norm": 0.43687090277671814,
-      "learning_rate": 9.699682746699425e-06,
-      "loss": 0.3547,
-      "step": 5907
-    },
-    {
-      "epoch": 0.3861185543428534,
-      "grad_norm": 0.47035229206085205,
-      "learning_rate": 9.699563539505532e-06,
-      "loss": 0.4036,
-      "step": 5908
-    },
-    {
-      "epoch": 0.38618390954839554,
-      "grad_norm": 0.4641532599925995,
-      "learning_rate": 9.699444309390237e-06,
-      "loss": 0.3854,
-      "step": 5909
-    },
-    {
-      "epoch": 0.38624926475393767,
-      "grad_norm": 0.44381463527679443,
-      "learning_rate": 9.699325056354126e-06,
-      "loss": 0.3714,
-      "step": 5910
-    },
-    {
-      "epoch": 0.3863146199594798,
-      "grad_norm": 0.4531979262828827,
-      "learning_rate": 9.699205780397777e-06,
-      "loss": 0.3808,
-      "step": 5911
-    },
-    {
-      "epoch": 0.38637997516502187,
-      "grad_norm": 0.49356502294540405,
-      "learning_rate": 9.699086481521774e-06,
-      "loss": 0.3875,
-      "step": 5912
-    },
-    {
-      "epoch": 0.386445330370564,
-      "grad_norm": 0.42907342314720154,
-      "learning_rate": 9.698967159726698e-06,
-      "loss": 0.3766,
-      "step": 5913
-    },
-    {
-      "epoch": 0.3865106855761061,
-      "grad_norm": 0.43039849400520325,
-      "learning_rate": 9.698847815013133e-06,
-      "loss": 0.3299,
-      "step": 5914
-    },
-    {
-      "epoch": 0.38657604078164826,
-      "grad_norm": 0.4388934075832367,
-      "learning_rate": 9.698728447381658e-06,
-      "loss": 0.3504,
-      "step": 5915
-    },
-    {
-      "epoch": 0.3866413959871904,
-      "grad_norm": 0.47367024421691895,
-      "learning_rate": 9.698609056832857e-06,
-      "loss": 0.4229,
-      "step": 5916
-    },
-    {
-      "epoch": 0.3867067511927325,
-      "grad_norm": 0.46311619877815247,
-      "learning_rate": 9.698489643367311e-06,
-      "loss": 0.4141,
-      "step": 5917
-    },
-    {
-      "epoch": 0.38677210639827464,
-      "grad_norm": 0.46769559383392334,
-      "learning_rate": 9.698370206985604e-06,
-      "loss": 0.3497,
-      "step": 5918
-    },
-    {
-      "epoch": 0.38683746160381677,
-      "grad_norm": 0.4645219147205353,
-      "learning_rate": 9.698250747688319e-06,
-      "loss": 0.3759,
-      "step": 5919
-    },
-    {
-      "epoch": 0.38690281680935884,
-      "grad_norm": 0.4617730975151062,
-      "learning_rate": 9.698131265476036e-06,
-      "loss": 0.3543,
-      "step": 5920
-    },
-    {
-      "epoch": 0.38696817201490097,
-      "grad_norm": 0.4533134400844574,
-      "learning_rate": 9.69801176034934e-06,
-      "loss": 0.3967,
-      "step": 5921
-    },
-    {
-      "epoch": 0.3870335272204431,
-      "grad_norm": 0.512360692024231,
-      "learning_rate": 9.697892232308812e-06,
-      "loss": 0.4733,
-      "step": 5922
-    },
-    {
-      "epoch": 0.3870988824259852,
-      "grad_norm": 0.5169496536254883,
-      "learning_rate": 9.697772681355035e-06,
-      "loss": 0.4406,
-      "step": 5923
-    },
-    {
-      "epoch": 0.38716423763152735,
-      "grad_norm": 0.4784247875213623,
-      "learning_rate": 9.697653107488596e-06,
-      "loss": 0.3734,
-      "step": 5924
-    },
-    {
-      "epoch": 0.3872295928370695,
-      "grad_norm": 0.429458886384964,
-      "learning_rate": 9.697533510710074e-06,
-      "loss": 0.3416,
-      "step": 5925
-    },
-    {
-      "epoch": 0.3872949480426116,
-      "grad_norm": 0.4785601794719696,
-      "learning_rate": 9.697413891020053e-06,
-      "loss": 0.4273,
-      "step": 5926
-    },
-    {
-      "epoch": 0.38736030324815374,
-      "grad_norm": 0.46648111939430237,
-      "learning_rate": 9.697294248419118e-06,
-      "loss": 0.398,
-      "step": 5927
-    },
-    {
-      "epoch": 0.3874256584536958,
-      "grad_norm": 0.4867631196975708,
-      "learning_rate": 9.697174582907854e-06,
-      "loss": 0.4216,
-      "step": 5928
-    },
-    {
-      "epoch": 0.38749101365923794,
-      "grad_norm": 0.4751908481121063,
-      "learning_rate": 9.69705489448684e-06,
-      "loss": 0.3818,
-      "step": 5929
-    },
-    {
-      "epoch": 0.38755636886478007,
-      "grad_norm": 0.49188607931137085,
-      "learning_rate": 9.696935183156661e-06,
-      "loss": 0.3607,
-      "step": 5930
-    },
-    {
-      "epoch": 0.3876217240703222,
-      "grad_norm": 0.48401615023612976,
-      "learning_rate": 9.696815448917902e-06,
-      "loss": 0.3898,
-      "step": 5931
-    },
-    {
-      "epoch": 0.3876870792758643,
-      "grad_norm": 0.4528600871562958,
-      "learning_rate": 9.696695691771149e-06,
-      "loss": 0.4031,
-      "step": 5932
-    },
-    {
-      "epoch": 0.38775243448140645,
-      "grad_norm": 0.49441906809806824,
-      "learning_rate": 9.696575911716982e-06,
-      "loss": 0.4348,
-      "step": 5933
-    },
-    {
-      "epoch": 0.3878177896869486,
-      "grad_norm": 0.4886414706707001,
-      "learning_rate": 9.696456108755989e-06,
-      "loss": 0.423,
-      "step": 5934
-    },
-    {
-      "epoch": 0.3878831448924907,
-      "grad_norm": 0.48988276720046997,
-      "learning_rate": 9.696336282888751e-06,
-      "loss": 0.4128,
-      "step": 5935
-    },
-    {
-      "epoch": 0.38794850009803283,
-      "grad_norm": 0.4868811368942261,
-      "learning_rate": 9.696216434115855e-06,
-      "loss": 0.3689,
-      "step": 5936
-    },
-    {
-      "epoch": 0.3880138553035749,
-      "grad_norm": 0.5110340118408203,
-      "learning_rate": 9.696096562437884e-06,
-      "loss": 0.4753,
-      "step": 5937
-    },
-    {
-      "epoch": 0.38807921050911703,
-      "grad_norm": 0.4750353991985321,
-      "learning_rate": 9.695976667855424e-06,
-      "loss": 0.4023,
-      "step": 5938
-    },
-    {
-      "epoch": 0.38814456571465916,
-      "grad_norm": 0.46796661615371704,
-      "learning_rate": 9.695856750369057e-06,
-      "loss": 0.3937,
-      "step": 5939
-    },
-    {
-      "epoch": 0.3882099209202013,
-      "grad_norm": 0.5175364017486572,
-      "learning_rate": 9.695736809979372e-06,
-      "loss": 0.4367,
-      "step": 5940
-    },
-    {
-      "epoch": 0.3882752761257434,
-      "grad_norm": 0.4603165090084076,
-      "learning_rate": 9.69561684668695e-06,
-      "loss": 0.3692,
-      "step": 5941
-    },
-    {
-      "epoch": 0.38834063133128555,
-      "grad_norm": 0.4634871184825897,
-      "learning_rate": 9.69549686049238e-06,
-      "loss": 0.4013,
-      "step": 5942
-    },
-    {
-      "epoch": 0.3884059865368277,
-      "grad_norm": 0.45909276604652405,
-      "learning_rate": 9.695376851396242e-06,
-      "loss": 0.3399,
-      "step": 5943
-    },
-    {
-      "epoch": 0.3884713417423698,
-      "grad_norm": 0.43440115451812744,
-      "learning_rate": 9.695256819399127e-06,
-      "loss": 0.3666,
-      "step": 5944
-    },
-    {
-      "epoch": 0.3885366969479119,
-      "grad_norm": 0.46124815940856934,
-      "learning_rate": 9.695136764501618e-06,
-      "loss": 0.3866,
-      "step": 5945
-    },
-    {
-      "epoch": 0.388602052153454,
-      "grad_norm": 0.4922986924648285,
-      "learning_rate": 9.6950166867043e-06,
-      "loss": 0.4598,
-      "step": 5946
-    },
-    {
-      "epoch": 0.38866740735899613,
-      "grad_norm": 0.4456659257411957,
-      "learning_rate": 9.694896586007759e-06,
-      "loss": 0.3983,
-      "step": 5947
-    },
-    {
-      "epoch": 0.38873276256453826,
-      "grad_norm": 0.4693983793258667,
-      "learning_rate": 9.69477646241258e-06,
-      "loss": 0.3757,
-      "step": 5948
-    },
-    {
-      "epoch": 0.3887981177700804,
-      "grad_norm": 0.4834117591381073,
-      "learning_rate": 9.69465631591935e-06,
-      "loss": 0.3837,
-      "step": 5949
-    },
-    {
-      "epoch": 0.3888634729756225,
-      "grad_norm": 0.47310110926628113,
-      "learning_rate": 9.694536146528657e-06,
-      "loss": 0.3852,
-      "step": 5950
-    },
-    {
-      "epoch": 0.38892882818116464,
-      "grad_norm": 0.48990702629089355,
-      "learning_rate": 9.694415954241083e-06,
-      "loss": 0.3872,
-      "step": 5951
-    },
-    {
-      "epoch": 0.38899418338670677,
-      "grad_norm": 0.5139848589897156,
-      "learning_rate": 9.694295739057215e-06,
-      "loss": 0.4552,
-      "step": 5952
-    },
-    {
-      "epoch": 0.3890595385922489,
-      "grad_norm": 0.45801395177841187,
-      "learning_rate": 9.694175500977643e-06,
-      "loss": 0.3868,
-      "step": 5953
-    },
-    {
-      "epoch": 0.38912489379779097,
-      "grad_norm": 0.4636388421058655,
-      "learning_rate": 9.69405524000295e-06,
-      "loss": 0.374,
-      "step": 5954
-    },
-    {
-      "epoch": 0.3891902490033331,
-      "grad_norm": 0.43106546998023987,
-      "learning_rate": 9.693934956133723e-06,
-      "loss": 0.3507,
-      "step": 5955
-    },
-    {
-      "epoch": 0.3892556042088752,
-      "grad_norm": 0.43061211705207825,
-      "learning_rate": 9.693814649370548e-06,
-      "loss": 0.3327,
-      "step": 5956
-    },
-    {
-      "epoch": 0.38932095941441736,
-      "grad_norm": 0.45796898007392883,
-      "learning_rate": 9.693694319714016e-06,
-      "loss": 0.372,
-      "step": 5957
-    },
-    {
-      "epoch": 0.3893863146199595,
-      "grad_norm": 0.4221617579460144,
-      "learning_rate": 9.693573967164707e-06,
-      "loss": 0.3127,
-      "step": 5958
-    },
-    {
-      "epoch": 0.3894516698255016,
-      "grad_norm": 0.4860340356826782,
-      "learning_rate": 9.693453591723215e-06,
-      "loss": 0.4414,
-      "step": 5959
-    },
-    {
-      "epoch": 0.38951702503104374,
-      "grad_norm": 0.49887949228286743,
-      "learning_rate": 9.69333319339012e-06,
-      "loss": 0.4547,
-      "step": 5960
-    },
-    {
-      "epoch": 0.38958238023658587,
-      "grad_norm": 0.44215089082717896,
-      "learning_rate": 9.693212772166016e-06,
-      "loss": 0.3971,
-      "step": 5961
-    },
-    {
-      "epoch": 0.38964773544212794,
-      "grad_norm": 0.4157714247703552,
-      "learning_rate": 9.693092328051488e-06,
-      "loss": 0.3221,
-      "step": 5962
-    },
-    {
-      "epoch": 0.38971309064767007,
-      "grad_norm": 0.45220786333084106,
-      "learning_rate": 9.692971861047122e-06,
-      "loss": 0.3862,
-      "step": 5963
-    },
-    {
-      "epoch": 0.3897784458532122,
-      "grad_norm": 0.4925874173641205,
-      "learning_rate": 9.692851371153504e-06,
-      "loss": 0.4128,
-      "step": 5964
-    },
-    {
-      "epoch": 0.3898438010587543,
-      "grad_norm": 0.458907812833786,
-      "learning_rate": 9.692730858371227e-06,
-      "loss": 0.3982,
-      "step": 5965
-    },
-    {
-      "epoch": 0.38990915626429645,
-      "grad_norm": 0.4813711941242218,
-      "learning_rate": 9.692610322700874e-06,
-      "loss": 0.4372,
-      "step": 5966
-    },
-    {
-      "epoch": 0.3899745114698386,
-      "grad_norm": 0.42877089977264404,
-      "learning_rate": 9.692489764143039e-06,
-      "loss": 0.329,
-      "step": 5967
-    },
-    {
-      "epoch": 0.3900398666753807,
-      "grad_norm": 0.4834284782409668,
-      "learning_rate": 9.6923691826983e-06,
-      "loss": 0.423,
-      "step": 5968
-    },
-    {
-      "epoch": 0.39010522188092284,
-      "grad_norm": 0.5371841788291931,
-      "learning_rate": 9.692248578367256e-06,
-      "loss": 0.4302,
-      "step": 5969
-    },
-    {
-      "epoch": 0.3901705770864649,
-      "grad_norm": 0.48016226291656494,
-      "learning_rate": 9.692127951150487e-06,
-      "loss": 0.4248,
-      "step": 5970
-    },
-    {
-      "epoch": 0.39023593229200704,
-      "grad_norm": 0.4683961272239685,
-      "learning_rate": 9.692007301048585e-06,
-      "loss": 0.3935,
-      "step": 5971
-    },
-    {
-      "epoch": 0.39030128749754917,
-      "grad_norm": 0.4606408178806305,
-      "learning_rate": 9.691886628062136e-06,
-      "loss": 0.3907,
-      "step": 5972
-    },
-    {
-      "epoch": 0.3903666427030913,
-      "grad_norm": 0.4929756820201874,
-      "learning_rate": 9.691765932191734e-06,
-      "loss": 0.4469,
-      "step": 5973
-    },
-    {
-      "epoch": 0.3904319979086334,
-      "grad_norm": 0.5043468475341797,
-      "learning_rate": 9.691645213437963e-06,
-      "loss": 0.4196,
-      "step": 5974
-    },
-    {
-      "epoch": 0.39049735311417555,
-      "grad_norm": 0.4898582100868225,
-      "learning_rate": 9.691524471801412e-06,
-      "loss": 0.4567,
-      "step": 5975
-    },
-    {
-      "epoch": 0.3905627083197177,
-      "grad_norm": 0.4886986315250397,
-      "learning_rate": 9.691403707282672e-06,
-      "loss": 0.3763,
-      "step": 5976
-    },
-    {
-      "epoch": 0.3906280635252598,
-      "grad_norm": 0.49640101194381714,
-      "learning_rate": 9.691282919882332e-06,
-      "loss": 0.4357,
-      "step": 5977
-    },
-    {
-      "epoch": 0.39069341873080193,
-      "grad_norm": 0.505385160446167,
-      "learning_rate": 9.691162109600978e-06,
-      "loss": 0.41,
-      "step": 5978
-    },
-    {
-      "epoch": 0.390758773936344,
-      "grad_norm": 0.49238911271095276,
-      "learning_rate": 9.691041276439202e-06,
-      "loss": 0.4404,
-      "step": 5979
-    },
-    {
-      "epoch": 0.39082412914188613,
-      "grad_norm": 0.4722394049167633,
-      "learning_rate": 9.690920420397592e-06,
-      "loss": 0.4149,
-      "step": 5980
-    },
-    {
-      "epoch": 0.39088948434742826,
-      "grad_norm": 0.460263729095459,
-      "learning_rate": 9.690799541476738e-06,
-      "loss": 0.4246,
-      "step": 5981
-    },
-    {
-      "epoch": 0.3909548395529704,
-      "grad_norm": 0.47054922580718994,
-      "learning_rate": 9.690678639677229e-06,
-      "loss": 0.4027,
-      "step": 5982
-    },
-    {
-      "epoch": 0.3910201947585125,
-      "grad_norm": 0.47182849049568176,
-      "learning_rate": 9.690557714999656e-06,
-      "loss": 0.3996,
-      "step": 5983
-    },
-    {
-      "epoch": 0.39108554996405465,
-      "grad_norm": 0.4657513201236725,
-      "learning_rate": 9.690436767444608e-06,
-      "loss": 0.4072,
-      "step": 5984
-    },
-    {
-      "epoch": 0.3911509051695968,
-      "grad_norm": 0.44540852308273315,
-      "learning_rate": 9.690315797012676e-06,
-      "loss": 0.3894,
-      "step": 5985
-    },
-    {
-      "epoch": 0.3912162603751389,
-      "grad_norm": 0.5249620079994202,
-      "learning_rate": 9.690194803704447e-06,
-      "loss": 0.4839,
-      "step": 5986
-    },
-    {
-      "epoch": 0.391281615580681,
-      "grad_norm": 0.4964336156845093,
-      "learning_rate": 9.690073787520516e-06,
-      "loss": 0.3777,
-      "step": 5987
-    },
-    {
-      "epoch": 0.3913469707862231,
-      "grad_norm": 0.4647408723831177,
-      "learning_rate": 9.689952748461466e-06,
-      "loss": 0.4157,
-      "step": 5988
-    },
-    {
-      "epoch": 0.39141232599176523,
-      "grad_norm": 0.4532619118690491,
-      "learning_rate": 9.689831686527895e-06,
-      "loss": 0.3737,
-      "step": 5989
-    },
-    {
-      "epoch": 0.39147768119730736,
-      "grad_norm": 0.47622814774513245,
-      "learning_rate": 9.68971060172039e-06,
-      "loss": 0.3808,
-      "step": 5990
-    },
-    {
-      "epoch": 0.3915430364028495,
-      "grad_norm": 0.44839900732040405,
-      "learning_rate": 9.68958949403954e-06,
-      "loss": 0.3873,
-      "step": 5991
-    },
-    {
-      "epoch": 0.3916083916083916,
-      "grad_norm": 0.519347608089447,
-      "learning_rate": 9.68946836348594e-06,
-      "loss": 0.4935,
-      "step": 5992
-    },
-    {
-      "epoch": 0.39167374681393374,
-      "grad_norm": 0.5313746929168701,
-      "learning_rate": 9.689347210060175e-06,
-      "loss": 0.4754,
-      "step": 5993
-    },
-    {
-      "epoch": 0.39173910201947587,
-      "grad_norm": 0.4559462368488312,
-      "learning_rate": 9.689226033762841e-06,
-      "loss": 0.354,
-      "step": 5994
-    },
-    {
-      "epoch": 0.391804457225018,
-      "grad_norm": 0.4504825174808502,
-      "learning_rate": 9.689104834594527e-06,
-      "loss": 0.3945,
-      "step": 5995
-    },
-    {
-      "epoch": 0.39186981243056007,
-      "grad_norm": 0.5034160017967224,
-      "learning_rate": 9.688983612555825e-06,
-      "loss": 0.4352,
-      "step": 5996
-    },
-    {
-      "epoch": 0.3919351676361022,
-      "grad_norm": 0.5011710524559021,
-      "learning_rate": 9.688862367647325e-06,
-      "loss": 0.3854,
-      "step": 5997
-    },
-    {
-      "epoch": 0.3920005228416443,
-      "grad_norm": 0.48244917392730713,
-      "learning_rate": 9.68874109986962e-06,
-      "loss": 0.3953,
-      "step": 5998
-    },
-    {
-      "epoch": 0.39206587804718646,
-      "grad_norm": 0.4704824686050415,
-      "learning_rate": 9.688619809223297e-06,
-      "loss": 0.3756,
-      "step": 5999
-    },
-    {
-      "epoch": 0.3921312332527286,
-      "grad_norm": 0.42899468541145325,
-      "learning_rate": 9.688498495708953e-06,
-      "loss": 0.3553,
-      "step": 6000
-    },
-    {
-      "epoch": 0.3921965884582707,
-      "grad_norm": 0.4486721158027649,
-      "learning_rate": 9.688377159327178e-06,
-      "loss": 0.3659,
-      "step": 6001
-    },
-    {
-      "epoch": 0.39226194366381284,
-      "grad_norm": 0.47961145639419556,
-      "learning_rate": 9.688255800078562e-06,
-      "loss": 0.3606,
-      "step": 6002
-    },
-    {
-      "epoch": 0.39232729886935497,
-      "grad_norm": 0.4899479150772095,
-      "learning_rate": 9.6881344179637e-06,
-      "loss": 0.3824,
-      "step": 6003
-    },
-    {
-      "epoch": 0.39239265407489704,
-      "grad_norm": 0.4607725441455841,
-      "learning_rate": 9.68801301298318e-06,
-      "loss": 0.4032,
-      "step": 6004
-    },
-    {
-      "epoch": 0.39245800928043917,
-      "grad_norm": 0.5207237005233765,
-      "learning_rate": 9.687891585137598e-06,
-      "loss": 0.4236,
-      "step": 6005
-    },
-    {
-      "epoch": 0.3925233644859813,
-      "grad_norm": 0.4831946790218353,
-      "learning_rate": 9.687770134427544e-06,
-      "loss": 0.4222,
-      "step": 6006
-    },
-    {
-      "epoch": 0.3925887196915234,
-      "grad_norm": 0.45149165391921997,
-      "learning_rate": 9.687648660853613e-06,
-      "loss": 0.38,
-      "step": 6007
-    },
-    {
-      "epoch": 0.39265407489706555,
-      "grad_norm": 0.4608800411224365,
-      "learning_rate": 9.687527164416392e-06,
-      "loss": 0.4043,
-      "step": 6008
-    },
-    {
-      "epoch": 0.3927194301026077,
-      "grad_norm": 0.48860853910446167,
-      "learning_rate": 9.687405645116481e-06,
-      "loss": 0.3856,
-      "step": 6009
-    },
-    {
-      "epoch": 0.3927847853081498,
-      "grad_norm": 0.46203213930130005,
-      "learning_rate": 9.687284102954467e-06,
-      "loss": 0.3805,
-      "step": 6010
-    },
-    {
-      "epoch": 0.39285014051369194,
-      "grad_norm": 0.4599906802177429,
-      "learning_rate": 9.687162537930944e-06,
-      "loss": 0.3983,
-      "step": 6011
-    },
-    {
-      "epoch": 0.39291549571923406,
-      "grad_norm": 0.48271965980529785,
-      "learning_rate": 9.687040950046506e-06,
-      "loss": 0.4349,
-      "step": 6012
-    },
-    {
-      "epoch": 0.39298085092477614,
-      "grad_norm": 0.483013391494751,
-      "learning_rate": 9.686919339301747e-06,
-      "loss": 0.4362,
-      "step": 6013
-    },
-    {
-      "epoch": 0.39304620613031827,
-      "grad_norm": 0.43602317571640015,
-      "learning_rate": 9.686797705697255e-06,
-      "loss": 0.3632,
-      "step": 6014
-    },
-    {
-      "epoch": 0.3931115613358604,
-      "grad_norm": 0.5030431747436523,
-      "learning_rate": 9.68667604923363e-06,
-      "loss": 0.48,
-      "step": 6015
-    },
-    {
-      "epoch": 0.3931769165414025,
-      "grad_norm": 0.43381696939468384,
-      "learning_rate": 9.686554369911462e-06,
-      "loss": 0.3573,
-      "step": 6016
-    },
-    {
-      "epoch": 0.39324227174694465,
-      "grad_norm": 0.46310433745384216,
-      "learning_rate": 9.686432667731344e-06,
-      "loss": 0.3957,
-      "step": 6017
-    },
-    {
-      "epoch": 0.3933076269524868,
-      "grad_norm": 0.4989355504512787,
-      "learning_rate": 9.68631094269387e-06,
-      "loss": 0.4141,
-      "step": 6018
-    },
-    {
-      "epoch": 0.3933729821580289,
-      "grad_norm": 0.49543890357017517,
-      "learning_rate": 9.686189194799635e-06,
-      "loss": 0.4506,
-      "step": 6019
-    },
-    {
-      "epoch": 0.39343833736357103,
-      "grad_norm": 0.4531422555446625,
-      "learning_rate": 9.686067424049232e-06,
-      "loss": 0.3995,
-      "step": 6020
-    },
-    {
-      "epoch": 0.3935036925691131,
-      "grad_norm": 0.4357486069202423,
-      "learning_rate": 9.685945630443254e-06,
-      "loss": 0.3293,
-      "step": 6021
-    },
-    {
-      "epoch": 0.39356904777465523,
-      "grad_norm": 0.49692434072494507,
-      "learning_rate": 9.685823813982295e-06,
-      "loss": 0.4287,
-      "step": 6022
-    },
-    {
-      "epoch": 0.39363440298019736,
-      "grad_norm": 0.49291113018989563,
-      "learning_rate": 9.685701974666952e-06,
-      "loss": 0.4617,
-      "step": 6023
-    },
-    {
-      "epoch": 0.3936997581857395,
-      "grad_norm": 0.4602315425872803,
-      "learning_rate": 9.685580112497816e-06,
-      "loss": 0.3932,
-      "step": 6024
-    },
-    {
-      "epoch": 0.3937651133912816,
-      "grad_norm": 0.4561045467853546,
-      "learning_rate": 9.685458227475483e-06,
-      "loss": 0.3948,
-      "step": 6025
-    },
-    {
-      "epoch": 0.39383046859682375,
-      "grad_norm": 0.48737287521362305,
-      "learning_rate": 9.685336319600548e-06,
-      "loss": 0.3936,
-      "step": 6026
-    },
-    {
-      "epoch": 0.3938958238023659,
-      "grad_norm": 0.6739256381988525,
-      "learning_rate": 9.685214388873602e-06,
-      "loss": 0.4216,
-      "step": 6027
-    },
-    {
-      "epoch": 0.393961179007908,
-      "grad_norm": 0.4541867971420288,
-      "learning_rate": 9.685092435295244e-06,
-      "loss": 0.3767,
-      "step": 6028
-    },
-    {
-      "epoch": 0.3940265342134501,
-      "grad_norm": 0.4627193808555603,
-      "learning_rate": 9.684970458866066e-06,
-      "loss": 0.3892,
-      "step": 6029
-    },
-    {
-      "epoch": 0.3940918894189922,
-      "grad_norm": 0.46207600831985474,
-      "learning_rate": 9.684848459586666e-06,
-      "loss": 0.3508,
-      "step": 6030
-    },
-    {
-      "epoch": 0.39415724462453433,
-      "grad_norm": 0.4354219436645508,
-      "learning_rate": 9.684726437457635e-06,
-      "loss": 0.38,
-      "step": 6031
-    },
-    {
-      "epoch": 0.39422259983007646,
-      "grad_norm": 0.47896140813827515,
-      "learning_rate": 9.68460439247957e-06,
-      "loss": 0.4224,
-      "step": 6032
-    },
-    {
-      "epoch": 0.3942879550356186,
-      "grad_norm": 0.4910140931606293,
-      "learning_rate": 9.68448232465307e-06,
-      "loss": 0.3653,
-      "step": 6033
-    },
-    {
-      "epoch": 0.3943533102411607,
-      "grad_norm": 0.4564701020717621,
-      "learning_rate": 9.684360233978724e-06,
-      "loss": 0.4004,
-      "step": 6034
-    },
-    {
-      "epoch": 0.39441866544670284,
-      "grad_norm": 0.698441207408905,
-      "learning_rate": 9.68423812045713e-06,
-      "loss": 0.3669,
-      "step": 6035
-    },
-    {
-      "epoch": 0.39448402065224497,
-      "grad_norm": 0.4709613621234894,
-      "learning_rate": 9.684115984088884e-06,
-      "loss": 0.3879,
-      "step": 6036
-    },
-    {
-      "epoch": 0.3945493758577871,
-      "grad_norm": 0.44479063153266907,
-      "learning_rate": 9.683993824874584e-06,
-      "loss": 0.3785,
-      "step": 6037
-    },
-    {
-      "epoch": 0.39461473106332917,
-      "grad_norm": 0.4567413330078125,
-      "learning_rate": 9.683871642814821e-06,
-      "loss": 0.4103,
-      "step": 6038
-    },
-    {
-      "epoch": 0.3946800862688713,
-      "grad_norm": 0.47958114743232727,
-      "learning_rate": 9.683749437910195e-06,
-      "loss": 0.3866,
-      "step": 6039
-    },
-    {
-      "epoch": 0.3947454414744134,
-      "grad_norm": 0.5066713690757751,
-      "learning_rate": 9.683627210161299e-06,
-      "loss": 0.4795,
-      "step": 6040
-    },
-    {
-      "epoch": 0.39481079667995556,
-      "grad_norm": 0.4703404903411865,
-      "learning_rate": 9.68350495956873e-06,
-      "loss": 0.3812,
-      "step": 6041
-    },
-    {
-      "epoch": 0.3948761518854977,
-      "grad_norm": 0.4703965187072754,
-      "learning_rate": 9.683382686133086e-06,
-      "loss": 0.4418,
-      "step": 6042
-    },
-    {
-      "epoch": 0.3949415070910398,
-      "grad_norm": 0.47360336780548096,
-      "learning_rate": 9.683260389854962e-06,
-      "loss": 0.3947,
-      "step": 6043
-    },
-    {
-      "epoch": 0.39500686229658194,
-      "grad_norm": 0.44658640027046204,
-      "learning_rate": 9.683138070734953e-06,
-      "loss": 0.3537,
-      "step": 6044
-    },
-    {
-      "epoch": 0.39507221750212407,
-      "grad_norm": 0.45639845728874207,
-      "learning_rate": 9.683015728773661e-06,
-      "loss": 0.3795,
-      "step": 6045
-    },
-    {
-      "epoch": 0.39513757270766614,
-      "grad_norm": 0.4557878077030182,
-      "learning_rate": 9.682893363971678e-06,
-      "loss": 0.3839,
-      "step": 6046
-    },
-    {
-      "epoch": 0.39520292791320827,
-      "grad_norm": 0.4923136532306671,
-      "learning_rate": 9.682770976329599e-06,
-      "loss": 0.4318,
-      "step": 6047
-    },
-    {
-      "epoch": 0.3952682831187504,
-      "grad_norm": 0.4797445833683014,
-      "learning_rate": 9.682648565848025e-06,
-      "loss": 0.4008,
-      "step": 6048
-    },
-    {
-      "epoch": 0.3953336383242925,
-      "grad_norm": 0.44140514731407166,
-      "learning_rate": 9.682526132527553e-06,
-      "loss": 0.3682,
-      "step": 6049
-    },
-    {
-      "epoch": 0.39539899352983465,
-      "grad_norm": 0.47167378664016724,
-      "learning_rate": 9.682403676368777e-06,
-      "loss": 0.4334,
-      "step": 6050
-    },
-    {
-      "epoch": 0.3954643487353768,
-      "grad_norm": 0.49207475781440735,
-      "learning_rate": 9.682281197372297e-06,
-      "loss": 0.4322,
-      "step": 6051
-    },
-    {
-      "epoch": 0.3955297039409189,
-      "grad_norm": 0.5148537755012512,
-      "learning_rate": 9.68215869553871e-06,
-      "loss": 0.4675,
-      "step": 6052
-    },
-    {
-      "epoch": 0.39559505914646104,
-      "grad_norm": 0.4530176520347595,
-      "learning_rate": 9.682036170868612e-06,
-      "loss": 0.4124,
-      "step": 6053
-    },
-    {
-      "epoch": 0.39566041435200316,
-      "grad_norm": 0.4727493226528168,
-      "learning_rate": 9.681913623362602e-06,
-      "loss": 0.4255,
-      "step": 6054
-    },
-    {
-      "epoch": 0.39572576955754524,
-      "grad_norm": 0.46910151839256287,
-      "learning_rate": 9.681791053021277e-06,
-      "loss": 0.3803,
-      "step": 6055
-    },
-    {
-      "epoch": 0.39579112476308737,
-      "grad_norm": 0.4994412958621979,
-      "learning_rate": 9.681668459845236e-06,
-      "loss": 0.4584,
-      "step": 6056
-    },
-    {
-      "epoch": 0.3958564799686295,
-      "grad_norm": 0.4252305328845978,
-      "learning_rate": 9.681545843835074e-06,
-      "loss": 0.3308,
-      "step": 6057
-    },
-    {
-      "epoch": 0.3959218351741716,
-      "grad_norm": 0.4465518295764923,
-      "learning_rate": 9.681423204991394e-06,
-      "loss": 0.3798,
-      "step": 6058
-    },
-    {
-      "epoch": 0.39598719037971375,
-      "grad_norm": 0.4711715579032898,
-      "learning_rate": 9.68130054331479e-06,
-      "loss": 0.4227,
-      "step": 6059
-    },
-    {
-      "epoch": 0.3960525455852559,
-      "grad_norm": 0.4443996846675873,
-      "learning_rate": 9.681177858805858e-06,
-      "loss": 0.3306,
-      "step": 6060
-    },
-    {
-      "epoch": 0.396117900790798,
-      "grad_norm": 0.4859817624092102,
-      "learning_rate": 9.681055151465205e-06,
-      "loss": 0.3916,
-      "step": 6061
-    },
-    {
-      "epoch": 0.39618325599634013,
-      "grad_norm": 0.4672118127346039,
-      "learning_rate": 9.680932421293424e-06,
-      "loss": 0.4478,
-      "step": 6062
-    },
-    {
-      "epoch": 0.3962486112018822,
-      "grad_norm": 0.48797398805618286,
-      "learning_rate": 9.680809668291111e-06,
-      "loss": 0.4316,
-      "step": 6063
-    },
-    {
-      "epoch": 0.39631396640742433,
-      "grad_norm": 0.45298853516578674,
-      "learning_rate": 9.680686892458869e-06,
-      "loss": 0.3621,
-      "step": 6064
-    },
-    {
-      "epoch": 0.39637932161296646,
-      "grad_norm": 0.4821392595767975,
-      "learning_rate": 9.680564093797296e-06,
-      "loss": 0.4776,
-      "step": 6065
-    },
-    {
-      "epoch": 0.3964446768185086,
-      "grad_norm": 0.44541066884994507,
-      "learning_rate": 9.68044127230699e-06,
-      "loss": 0.3821,
-      "step": 6066
-    },
-    {
-      "epoch": 0.3965100320240507,
-      "grad_norm": 0.43608272075653076,
-      "learning_rate": 9.68031842798855e-06,
-      "loss": 0.3553,
-      "step": 6067
-    },
-    {
-      "epoch": 0.39657538722959285,
-      "grad_norm": 0.41650286316871643,
-      "learning_rate": 9.680195560842575e-06,
-      "loss": 0.3473,
-      "step": 6068
-    },
-    {
-      "epoch": 0.396640742435135,
-      "grad_norm": 0.4567144513130188,
-      "learning_rate": 9.680072670869667e-06,
-      "loss": 0.3902,
-      "step": 6069
-    },
-    {
-      "epoch": 0.3967060976406771,
-      "grad_norm": 0.4912777543067932,
-      "learning_rate": 9.679949758070421e-06,
-      "loss": 0.4242,
-      "step": 6070
-    },
-    {
-      "epoch": 0.3967714528462192,
-      "grad_norm": 0.4811214506626129,
-      "learning_rate": 9.67982682244544e-06,
-      "loss": 0.3831,
-      "step": 6071
-    },
-    {
-      "epoch": 0.3968368080517613,
-      "grad_norm": 0.45678627490997314,
-      "learning_rate": 9.679703863995322e-06,
-      "loss": 0.4118,
-      "step": 6072
-    },
-    {
-      "epoch": 0.39690216325730343,
-      "grad_norm": 0.4407769441604614,
-      "learning_rate": 9.679580882720668e-06,
-      "loss": 0.3817,
-      "step": 6073
-    },
-    {
-      "epoch": 0.39696751846284556,
-      "grad_norm": 0.4498608708381653,
-      "learning_rate": 9.679457878622076e-06,
-      "loss": 0.3685,
-      "step": 6074
-    },
-    {
-      "epoch": 0.3970328736683877,
-      "grad_norm": 0.47542741894721985,
-      "learning_rate": 9.679334851700147e-06,
-      "loss": 0.4311,
-      "step": 6075
-    },
-    {
-      "epoch": 0.3970982288739298,
-      "grad_norm": 0.45421308279037476,
-      "learning_rate": 9.679211801955482e-06,
-      "loss": 0.3979,
-      "step": 6076
-    },
-    {
-      "epoch": 0.39716358407947194,
-      "grad_norm": 0.44924843311309814,
-      "learning_rate": 9.679088729388677e-06,
-      "loss": 0.3691,
-      "step": 6077
-    },
-    {
-      "epoch": 0.39722893928501407,
-      "grad_norm": 0.4603331983089447,
-      "learning_rate": 9.678965634000338e-06,
-      "loss": 0.3961,
-      "step": 6078
-    },
-    {
-      "epoch": 0.3972942944905562,
-      "grad_norm": 0.4692233204841614,
-      "learning_rate": 9.678842515791062e-06,
-      "loss": 0.4175,
-      "step": 6079
-    },
-    {
-      "epoch": 0.39735964969609827,
-      "grad_norm": 0.4724222719669342,
-      "learning_rate": 9.67871937476145e-06,
-      "loss": 0.4059,
-      "step": 6080
-    },
-    {
-      "epoch": 0.3974250049016404,
-      "grad_norm": 0.476258248090744,
-      "learning_rate": 9.678596210912102e-06,
-      "loss": 0.4137,
-      "step": 6081
-    },
-    {
-      "epoch": 0.3974903601071825,
-      "grad_norm": 0.45830920338630676,
-      "learning_rate": 9.67847302424362e-06,
-      "loss": 0.3781,
-      "step": 6082
-    },
-    {
-      "epoch": 0.39755571531272466,
-      "grad_norm": 0.46827584505081177,
-      "learning_rate": 9.678349814756605e-06,
-      "loss": 0.4389,
-      "step": 6083
-    },
-    {
-      "epoch": 0.3976210705182668,
-      "grad_norm": 0.40678760409355164,
-      "learning_rate": 9.678226582451655e-06,
-      "loss": 0.3175,
-      "step": 6084
-    },
-    {
-      "epoch": 0.3976864257238089,
-      "grad_norm": 0.417957067489624,
-      "learning_rate": 9.678103327329375e-06,
-      "loss": 0.3262,
-      "step": 6085
-    },
-    {
-      "epoch": 0.39775178092935104,
-      "grad_norm": 0.45792457461357117,
-      "learning_rate": 9.677980049390366e-06,
-      "loss": 0.3649,
-      "step": 6086
-    },
-    {
-      "epoch": 0.39781713613489317,
-      "grad_norm": 0.4505128860473633,
-      "learning_rate": 9.677856748635224e-06,
-      "loss": 0.3688,
-      "step": 6087
-    },
-    {
-      "epoch": 0.39788249134043524,
-      "grad_norm": 0.4464775621891022,
-      "learning_rate": 9.677733425064558e-06,
-      "loss": 0.3526,
-      "step": 6088
-    },
-    {
-      "epoch": 0.39794784654597737,
-      "grad_norm": 0.4564993381500244,
-      "learning_rate": 9.677610078678964e-06,
-      "loss": 0.4147,
-      "step": 6089
-    },
-    {
-      "epoch": 0.3980132017515195,
-      "grad_norm": 0.45127493143081665,
-      "learning_rate": 9.677486709479042e-06,
-      "loss": 0.3747,
-      "step": 6090
-    },
-    {
-      "epoch": 0.3980785569570616,
-      "grad_norm": 0.4899660646915436,
-      "learning_rate": 9.677363317465401e-06,
-      "loss": 0.4295,
-      "step": 6091
-    },
-    {
-      "epoch": 0.39814391216260375,
-      "grad_norm": 0.44159093499183655,
-      "learning_rate": 9.677239902638637e-06,
-      "loss": 0.3664,
-      "step": 6092
-    },
-    {
-      "epoch": 0.3982092673681459,
-      "grad_norm": 0.43933340907096863,
-      "learning_rate": 9.677116464999355e-06,
-      "loss": 0.3427,
-      "step": 6093
-    },
-    {
-      "epoch": 0.398274622573688,
-      "grad_norm": 0.48111939430236816,
-      "learning_rate": 9.676993004548153e-06,
-      "loss": 0.3958,
-      "step": 6094
-    },
-    {
-      "epoch": 0.39833997777923014,
-      "grad_norm": 0.45526638627052307,
-      "learning_rate": 9.676869521285638e-06,
-      "loss": 0.4179,
-      "step": 6095
-    },
-    {
-      "epoch": 0.39840533298477226,
-      "grad_norm": 0.42688971757888794,
-      "learning_rate": 9.676746015212411e-06,
-      "loss": 0.3652,
-      "step": 6096
-    },
-    {
-      "epoch": 0.39847068819031434,
-      "grad_norm": 0.48857513070106506,
-      "learning_rate": 9.676622486329071e-06,
-      "loss": 0.3724,
-      "step": 6097
-    },
-    {
-      "epoch": 0.39853604339585647,
-      "grad_norm": 0.5034164190292358,
-      "learning_rate": 9.676498934636224e-06,
-      "loss": 0.4495,
-      "step": 6098
-    },
-    {
-      "epoch": 0.3986013986013986,
-      "grad_norm": 0.4916623830795288,
-      "learning_rate": 9.676375360134471e-06,
-      "loss": 0.4278,
-      "step": 6099
-    },
-    {
-      "epoch": 0.3986667538069407,
-      "grad_norm": 0.4503425657749176,
-      "learning_rate": 9.676251762824416e-06,
-      "loss": 0.4003,
-      "step": 6100
-    },
-    {
-      "epoch": 0.39873210901248285,
-      "grad_norm": 0.47600290179252625,
-      "learning_rate": 9.676128142706663e-06,
-      "loss": 0.3832,
-      "step": 6101
-    },
-    {
-      "epoch": 0.398797464218025,
-      "grad_norm": 0.5034260153770447,
-      "learning_rate": 9.67600449978181e-06,
-      "loss": 0.4491,
-      "step": 6102
-    },
-    {
-      "epoch": 0.3988628194235671,
-      "grad_norm": 0.4781205654144287,
-      "learning_rate": 9.675880834050465e-06,
-      "loss": 0.4199,
-      "step": 6103
-    },
-    {
-      "epoch": 0.39892817462910923,
-      "grad_norm": 0.4337165057659149,
-      "learning_rate": 9.675757145513229e-06,
-      "loss": 0.3518,
-      "step": 6104
-    },
-    {
-      "epoch": 0.3989935298346513,
-      "grad_norm": 0.45678630471229553,
-      "learning_rate": 9.675633434170704e-06,
-      "loss": 0.3712,
-      "step": 6105
-    },
-    {
-      "epoch": 0.39905888504019343,
-      "grad_norm": 0.47965630888938904,
-      "learning_rate": 9.675509700023498e-06,
-      "loss": 0.4467,
-      "step": 6106
-    },
-    {
-      "epoch": 0.39912424024573556,
-      "grad_norm": 0.42392995953559875,
-      "learning_rate": 9.675385943072209e-06,
-      "loss": 0.37,
-      "step": 6107
-    },
-    {
-      "epoch": 0.3991895954512777,
-      "grad_norm": 0.48698365688323975,
-      "learning_rate": 9.675262163317442e-06,
-      "loss": 0.4305,
-      "step": 6108
-    },
-    {
-      "epoch": 0.3992549506568198,
-      "grad_norm": 0.5198854207992554,
-      "learning_rate": 9.675138360759805e-06,
-      "loss": 0.457,
-      "step": 6109
-    },
-    {
-      "epoch": 0.39932030586236195,
-      "grad_norm": 0.5073537826538086,
-      "learning_rate": 9.675014535399897e-06,
-      "loss": 0.408,
-      "step": 6110
-    },
-    {
-      "epoch": 0.3993856610679041,
-      "grad_norm": 0.476301908493042,
-      "learning_rate": 9.674890687238324e-06,
-      "loss": 0.4619,
-      "step": 6111
-    },
-    {
-      "epoch": 0.3994510162734462,
-      "grad_norm": 0.4359187185764313,
-      "learning_rate": 9.67476681627569e-06,
-      "loss": 0.3556,
-      "step": 6112
-    },
-    {
-      "epoch": 0.3995163714789883,
-      "grad_norm": 0.42895931005477905,
-      "learning_rate": 9.674642922512596e-06,
-      "loss": 0.343,
-      "step": 6113
-    },
-    {
-      "epoch": 0.3995817266845304,
-      "grad_norm": 0.48938706517219543,
-      "learning_rate": 9.674519005949652e-06,
-      "loss": 0.3933,
-      "step": 6114
-    },
-    {
-      "epoch": 0.39964708189007253,
-      "grad_norm": 0.42167866230010986,
-      "learning_rate": 9.674395066587457e-06,
-      "loss": 0.3351,
-      "step": 6115
-    },
-    {
-      "epoch": 0.39971243709561466,
-      "grad_norm": 0.4954562187194824,
-      "learning_rate": 9.67427110442662e-06,
-      "loss": 0.4437,
-      "step": 6116
-    },
-    {
-      "epoch": 0.3997777923011568,
-      "grad_norm": 0.4689842760562897,
-      "learning_rate": 9.674147119467742e-06,
-      "loss": 0.4131,
-      "step": 6117
-    },
-    {
-      "epoch": 0.3998431475066989,
-      "grad_norm": 0.44719985127449036,
-      "learning_rate": 9.67402311171143e-06,
-      "loss": 0.3869,
-      "step": 6118
-    },
-    {
-      "epoch": 0.39990850271224104,
-      "grad_norm": 0.4831448495388031,
-      "learning_rate": 9.673899081158289e-06,
-      "loss": 0.4166,
-      "step": 6119
-    },
-    {
-      "epoch": 0.39997385791778317,
-      "grad_norm": 0.4672890901565552,
-      "learning_rate": 9.673775027808922e-06,
-      "loss": 0.371,
-      "step": 6120
-    },
-    {
-      "epoch": 0.4000392131233253,
-      "grad_norm": 0.47074273228645325,
-      "learning_rate": 9.673650951663934e-06,
-      "loss": 0.3713,
-      "step": 6121
-    },
-    {
-      "epoch": 0.40010456832886737,
-      "grad_norm": 0.43529555201530457,
-      "learning_rate": 9.673526852723934e-06,
-      "loss": 0.3629,
-      "step": 6122
-    },
-    {
-      "epoch": 0.4001699235344095,
-      "grad_norm": 0.45171019434928894,
-      "learning_rate": 9.67340273098952e-06,
-      "loss": 0.3958,
-      "step": 6123
-    },
-    {
-      "epoch": 0.4002352787399516,
-      "grad_norm": 0.4475041925907135,
-      "learning_rate": 9.673278586461305e-06,
-      "loss": 0.3904,
-      "step": 6124
-    },
-    {
-      "epoch": 0.40030063394549376,
-      "grad_norm": 0.45028144121170044,
-      "learning_rate": 9.673154419139892e-06,
-      "loss": 0.374,
-      "step": 6125
-    },
-    {
-      "epoch": 0.4003659891510359,
-      "grad_norm": 0.46634340286254883,
-      "learning_rate": 9.673030229025885e-06,
-      "loss": 0.4349,
-      "step": 6126
-    },
-    {
-      "epoch": 0.400431344356578,
-      "grad_norm": 0.4717932641506195,
-      "learning_rate": 9.672906016119889e-06,
-      "loss": 0.4063,
-      "step": 6127
-    },
-    {
-      "epoch": 0.40049669956212014,
-      "grad_norm": 0.4801435172557831,
-      "learning_rate": 9.672781780422515e-06,
-      "loss": 0.459,
-      "step": 6128
-    },
-    {
-      "epoch": 0.40056205476766227,
-      "grad_norm": 0.4910016357898712,
-      "learning_rate": 9.672657521934364e-06,
-      "loss": 0.3214,
-      "step": 6129
-    },
-    {
-      "epoch": 0.40062740997320434,
-      "grad_norm": 0.45259609818458557,
-      "learning_rate": 9.672533240656041e-06,
-      "loss": 0.3868,
-      "step": 6130
-    },
-    {
-      "epoch": 0.40069276517874647,
-      "grad_norm": 0.4314712882041931,
-      "learning_rate": 9.672408936588158e-06,
-      "loss": 0.3731,
-      "step": 6131
-    },
-    {
-      "epoch": 0.4007581203842886,
-      "grad_norm": 0.4598412811756134,
-      "learning_rate": 9.672284609731316e-06,
-      "loss": 0.3494,
-      "step": 6132
-    },
-    {
-      "epoch": 0.4008234755898307,
-      "grad_norm": 0.44612693786621094,
-      "learning_rate": 9.672160260086124e-06,
-      "loss": 0.4143,
-      "step": 6133
-    },
-    {
-      "epoch": 0.40088883079537285,
-      "grad_norm": 0.440768301486969,
-      "learning_rate": 9.672035887653189e-06,
-      "loss": 0.3627,
-      "step": 6134
-    },
-    {
-      "epoch": 0.400954186000915,
-      "grad_norm": 0.4287469685077667,
-      "learning_rate": 9.671911492433114e-06,
-      "loss": 0.3514,
-      "step": 6135
-    },
-    {
-      "epoch": 0.4010195412064571,
-      "grad_norm": 0.45346057415008545,
-      "learning_rate": 9.671787074426509e-06,
-      "loss": 0.4192,
-      "step": 6136
-    },
-    {
-      "epoch": 0.40108489641199924,
-      "grad_norm": 0.46000832319259644,
-      "learning_rate": 9.67166263363398e-06,
-      "loss": 0.3943,
-      "step": 6137
-    },
-    {
-      "epoch": 0.40115025161754136,
-      "grad_norm": 0.45692506432533264,
-      "learning_rate": 9.671538170056134e-06,
-      "loss": 0.3913,
-      "step": 6138
-    },
-    {
-      "epoch": 0.40121560682308344,
-      "grad_norm": 0.47234079241752625,
-      "learning_rate": 9.67141368369358e-06,
-      "loss": 0.4193,
-      "step": 6139
-    },
-    {
-      "epoch": 0.40128096202862557,
-      "grad_norm": 0.4784364402294159,
-      "learning_rate": 9.67128917454692e-06,
-      "loss": 0.4285,
-      "step": 6140
-    },
-    {
-      "epoch": 0.4013463172341677,
-      "grad_norm": 0.44267991185188293,
-      "learning_rate": 9.671164642616766e-06,
-      "loss": 0.3749,
-      "step": 6141
-    },
-    {
-      "epoch": 0.4014116724397098,
-      "grad_norm": 0.4557999074459076,
-      "learning_rate": 9.671040087903723e-06,
-      "loss": 0.4004,
-      "step": 6142
-    },
-    {
-      "epoch": 0.40147702764525195,
-      "grad_norm": 0.4551524817943573,
-      "learning_rate": 9.6709155104084e-06,
-      "loss": 0.3863,
-      "step": 6143
-    },
-    {
-      "epoch": 0.4015423828507941,
-      "grad_norm": 0.47498413920402527,
-      "learning_rate": 9.670790910131405e-06,
-      "loss": 0.3952,
-      "step": 6144
-    },
-    {
-      "epoch": 0.4016077380563362,
-      "grad_norm": 0.4864051342010498,
-      "learning_rate": 9.670666287073343e-06,
-      "loss": 0.4735,
-      "step": 6145
-    },
-    {
-      "epoch": 0.40167309326187833,
-      "grad_norm": 0.4366544187068939,
-      "learning_rate": 9.670541641234823e-06,
-      "loss": 0.3455,
-      "step": 6146
-    },
-    {
-      "epoch": 0.4017384484674204,
-      "grad_norm": 0.4424472749233246,
-      "learning_rate": 9.670416972616454e-06,
-      "loss": 0.38,
-      "step": 6147
-    },
-    {
-      "epoch": 0.40180380367296253,
-      "grad_norm": 0.4703499674797058,
-      "learning_rate": 9.670292281218844e-06,
-      "loss": 0.3753,
-      "step": 6148
-    },
-    {
-      "epoch": 0.40186915887850466,
-      "grad_norm": 0.42187562584877014,
-      "learning_rate": 9.6701675670426e-06,
-      "loss": 0.3769,
-      "step": 6149
-    },
-    {
-      "epoch": 0.4019345140840468,
-      "grad_norm": 0.44826239347457886,
-      "learning_rate": 9.670042830088331e-06,
-      "loss": 0.3742,
-      "step": 6150
-    },
-    {
-      "epoch": 0.4019998692895889,
-      "grad_norm": 0.47213315963745117,
-      "learning_rate": 9.669918070356645e-06,
-      "loss": 0.406,
-      "step": 6151
-    },
-    {
-      "epoch": 0.40206522449513105,
-      "grad_norm": 0.456643670797348,
-      "learning_rate": 9.669793287848151e-06,
-      "loss": 0.3881,
-      "step": 6152
-    },
-    {
-      "epoch": 0.4021305797006732,
-      "grad_norm": 0.46143490076065063,
-      "learning_rate": 9.669668482563456e-06,
-      "loss": 0.4419,
-      "step": 6153
-    },
-    {
-      "epoch": 0.4021959349062153,
-      "grad_norm": 0.5147846341133118,
-      "learning_rate": 9.669543654503174e-06,
-      "loss": 0.4547,
-      "step": 6154
-    },
-    {
-      "epoch": 0.4022612901117574,
-      "grad_norm": 0.5235462188720703,
-      "learning_rate": 9.669418803667906e-06,
-      "loss": 0.4945,
-      "step": 6155
-    },
-    {
-      "epoch": 0.4023266453172995,
-      "grad_norm": 0.4780176281929016,
-      "learning_rate": 9.669293930058266e-06,
-      "loss": 0.4002,
-      "step": 6156
-    },
-    {
-      "epoch": 0.40239200052284163,
-      "grad_norm": 0.4642355144023895,
-      "learning_rate": 9.669169033674863e-06,
-      "loss": 0.377,
-      "step": 6157
-    },
-    {
-      "epoch": 0.40245735572838376,
-      "grad_norm": 0.4488992691040039,
-      "learning_rate": 9.669044114518304e-06,
-      "loss": 0.3831,
-      "step": 6158
-    },
-    {
-      "epoch": 0.4025227109339259,
-      "grad_norm": 0.4399069845676422,
-      "learning_rate": 9.668919172589199e-06,
-      "loss": 0.3759,
-      "step": 6159
-    },
-    {
-      "epoch": 0.402588066139468,
-      "grad_norm": 0.4538023769855499,
-      "learning_rate": 9.668794207888158e-06,
-      "loss": 0.3739,
-      "step": 6160
-    },
-    {
-      "epoch": 0.40265342134501014,
-      "grad_norm": 0.47913607954978943,
-      "learning_rate": 9.668669220415792e-06,
-      "loss": 0.4149,
-      "step": 6161
-    },
-    {
-      "epoch": 0.40271877655055227,
-      "grad_norm": 0.47564640641212463,
-      "learning_rate": 9.668544210172707e-06,
-      "loss": 0.4273,
-      "step": 6162
-    },
-    {
-      "epoch": 0.4027841317560944,
-      "grad_norm": 0.4367576539516449,
-      "learning_rate": 9.668419177159515e-06,
-      "loss": 0.3508,
-      "step": 6163
-    },
-    {
-      "epoch": 0.40284948696163647,
-      "grad_norm": 0.4587484300136566,
-      "learning_rate": 9.668294121376825e-06,
-      "loss": 0.3809,
-      "step": 6164
-    },
-    {
-      "epoch": 0.4029148421671786,
-      "grad_norm": 0.47635382413864136,
-      "learning_rate": 9.668169042825249e-06,
-      "loss": 0.3786,
-      "step": 6165
-    },
-    {
-      "epoch": 0.4029801973727207,
-      "grad_norm": 0.48471590876579285,
-      "learning_rate": 9.668043941505392e-06,
-      "loss": 0.4324,
-      "step": 6166
-    },
-    {
-      "epoch": 0.40304555257826286,
-      "grad_norm": 0.5001281499862671,
-      "learning_rate": 9.66791881741787e-06,
-      "loss": 0.4585,
-      "step": 6167
-    },
-    {
-      "epoch": 0.403110907783805,
-      "grad_norm": 0.47650250792503357,
-      "learning_rate": 9.66779367056329e-06,
-      "loss": 0.4205,
-      "step": 6168
-    },
-    {
-      "epoch": 0.4031762629893471,
-      "grad_norm": 0.44792792201042175,
-      "learning_rate": 9.667668500942264e-06,
-      "loss": 0.368,
-      "step": 6169
-    },
-    {
-      "epoch": 0.40324161819488924,
-      "grad_norm": 0.4794873893260956,
-      "learning_rate": 9.6675433085554e-06,
-      "loss": 0.3765,
-      "step": 6170
-    },
-    {
-      "epoch": 0.40330697340043137,
-      "grad_norm": 0.4200485646724701,
-      "learning_rate": 9.667418093403312e-06,
-      "loss": 0.3524,
-      "step": 6171
-    },
-    {
-      "epoch": 0.40337232860597344,
-      "grad_norm": 0.48350051045417786,
-      "learning_rate": 9.66729285548661e-06,
-      "loss": 0.4464,
-      "step": 6172
-    },
-    {
-      "epoch": 0.40343768381151557,
-      "grad_norm": 0.48739489912986755,
-      "learning_rate": 9.667167594805901e-06,
-      "loss": 0.4104,
-      "step": 6173
-    },
-    {
-      "epoch": 0.4035030390170577,
-      "grad_norm": 0.5147889852523804,
-      "learning_rate": 9.6670423113618e-06,
-      "loss": 0.3991,
-      "step": 6174
-    },
-    {
-      "epoch": 0.4035683942225998,
-      "grad_norm": 0.5043358206748962,
-      "learning_rate": 9.666917005154918e-06,
-      "loss": 0.473,
-      "step": 6175
-    },
-    {
-      "epoch": 0.40363374942814195,
-      "grad_norm": 0.4084410071372986,
-      "learning_rate": 9.666791676185862e-06,
-      "loss": 0.339,
-      "step": 6176
-    },
-    {
-      "epoch": 0.4036991046336841,
-      "grad_norm": 0.4801981449127197,
-      "learning_rate": 9.66666632445525e-06,
-      "loss": 0.392,
-      "step": 6177
-    },
-    {
-      "epoch": 0.4037644598392262,
-      "grad_norm": 0.46107038855552673,
-      "learning_rate": 9.666540949963687e-06,
-      "loss": 0.3739,
-      "step": 6178
-    },
-    {
-      "epoch": 0.40382981504476834,
-      "grad_norm": 0.4498555064201355,
-      "learning_rate": 9.666415552711789e-06,
-      "loss": 0.3855,
-      "step": 6179
-    },
-    {
-      "epoch": 0.40389517025031046,
-      "grad_norm": 0.4480822682380676,
-      "learning_rate": 9.666290132700165e-06,
-      "loss": 0.3804,
-      "step": 6180
-    },
-    {
-      "epoch": 0.40396052545585254,
-      "grad_norm": 0.4630272388458252,
-      "learning_rate": 9.666164689929427e-06,
-      "loss": 0.4211,
-      "step": 6181
-    },
-    {
-      "epoch": 0.40402588066139467,
-      "grad_norm": 0.4566306173801422,
-      "learning_rate": 9.666039224400187e-06,
-      "loss": 0.4363,
-      "step": 6182
-    },
-    {
-      "epoch": 0.4040912358669368,
-      "grad_norm": 0.450600802898407,
-      "learning_rate": 9.665913736113057e-06,
-      "loss": 0.3984,
-      "step": 6183
-    },
-    {
-      "epoch": 0.4041565910724789,
-      "grad_norm": 0.449144184589386,
-      "learning_rate": 9.66578822506865e-06,
-      "loss": 0.3577,
-      "step": 6184
-    },
-    {
-      "epoch": 0.40422194627802105,
-      "grad_norm": 0.45714282989501953,
-      "learning_rate": 9.665662691267578e-06,
-      "loss": 0.4253,
-      "step": 6185
-    },
-    {
-      "epoch": 0.4042873014835632,
-      "grad_norm": 0.44167929887771606,
-      "learning_rate": 9.665537134710452e-06,
-      "loss": 0.3617,
-      "step": 6186
-    },
-    {
-      "epoch": 0.4043526566891053,
-      "grad_norm": 0.4349885880947113,
-      "learning_rate": 9.665411555397885e-06,
-      "loss": 0.3551,
-      "step": 6187
-    },
-    {
-      "epoch": 0.40441801189464743,
-      "grad_norm": 0.4807620942592621,
-      "learning_rate": 9.66528595333049e-06,
-      "loss": 0.3952,
-      "step": 6188
-    },
-    {
-      "epoch": 0.4044833671001895,
-      "grad_norm": 0.45490214228630066,
-      "learning_rate": 9.66516032850888e-06,
-      "loss": 0.4011,
-      "step": 6189
-    },
-    {
-      "epoch": 0.40454872230573163,
-      "grad_norm": 0.504051148891449,
-      "learning_rate": 9.665034680933665e-06,
-      "loss": 0.4593,
-      "step": 6190
-    },
-    {
-      "epoch": 0.40461407751127376,
-      "grad_norm": 0.4414256513118744,
-      "learning_rate": 9.664909010605462e-06,
-      "loss": 0.436,
-      "step": 6191
-    },
-    {
-      "epoch": 0.4046794327168159,
-      "grad_norm": 0.49999725818634033,
-      "learning_rate": 9.664783317524879e-06,
-      "loss": 0.4606,
-      "step": 6192
-    },
-    {
-      "epoch": 0.404744787922358,
-      "grad_norm": 0.44729292392730713,
-      "learning_rate": 9.664657601692532e-06,
-      "loss": 0.3958,
-      "step": 6193
-    },
-    {
-      "epoch": 0.40481014312790015,
-      "grad_norm": 0.45030468702316284,
-      "learning_rate": 9.664531863109036e-06,
-      "loss": 0.3377,
-      "step": 6194
-    },
-    {
-      "epoch": 0.4048754983334423,
-      "grad_norm": 0.4917182922363281,
-      "learning_rate": 9.664406101775001e-06,
-      "loss": 0.4469,
-      "step": 6195
-    },
-    {
-      "epoch": 0.4049408535389844,
-      "grad_norm": 0.4703455865383148,
-      "learning_rate": 9.664280317691042e-06,
-      "loss": 0.3949,
-      "step": 6196
-    },
-    {
-      "epoch": 0.4050062087445265,
-      "grad_norm": 0.4861055016517639,
-      "learning_rate": 9.664154510857772e-06,
-      "loss": 0.4385,
-      "step": 6197
-    },
-    {
-      "epoch": 0.4050715639500686,
-      "grad_norm": 0.47776445746421814,
-      "learning_rate": 9.664028681275804e-06,
-      "loss": 0.3932,
-      "step": 6198
-    },
-    {
-      "epoch": 0.40513691915561073,
-      "grad_norm": 0.5069558620452881,
-      "learning_rate": 9.663902828945753e-06,
-      "loss": 0.4466,
-      "step": 6199
-    },
-    {
-      "epoch": 0.40520227436115286,
-      "grad_norm": 0.4434361457824707,
-      "learning_rate": 9.663776953868232e-06,
-      "loss": 0.3606,
-      "step": 6200
-    },
-    {
-      "epoch": 0.405267629566695,
-      "grad_norm": 0.4312838912010193,
-      "learning_rate": 9.663651056043855e-06,
-      "loss": 0.3551,
-      "step": 6201
-    },
-    {
-      "epoch": 0.4053329847722371,
-      "grad_norm": 0.4512668251991272,
-      "learning_rate": 9.663525135473235e-06,
-      "loss": 0.4085,
-      "step": 6202
-    },
-    {
-      "epoch": 0.40539833997777924,
-      "grad_norm": 0.4664841294288635,
-      "learning_rate": 9.663399192156988e-06,
-      "loss": 0.3828,
-      "step": 6203
-    },
-    {
-      "epoch": 0.40546369518332137,
-      "grad_norm": 0.5049656629562378,
-      "learning_rate": 9.663273226095729e-06,
-      "loss": 0.444,
-      "step": 6204
-    },
-    {
-      "epoch": 0.4055290503888635,
-      "grad_norm": 0.51225346326828,
-      "learning_rate": 9.663147237290069e-06,
-      "loss": 0.4186,
-      "step": 6205
-    },
-    {
-      "epoch": 0.40559440559440557,
-      "grad_norm": 0.4588335156440735,
-      "learning_rate": 9.663021225740626e-06,
-      "loss": 0.361,
-      "step": 6206
-    },
-    {
-      "epoch": 0.4056597607999477,
-      "grad_norm": 0.46261194348335266,
-      "learning_rate": 9.662895191448013e-06,
-      "loss": 0.3684,
-      "step": 6207
-    },
-    {
-      "epoch": 0.4057251160054898,
-      "grad_norm": 0.4598039984703064,
-      "learning_rate": 9.662769134412843e-06,
-      "loss": 0.3666,
-      "step": 6208
-    },
-    {
-      "epoch": 0.40579047121103196,
-      "grad_norm": 0.5030199885368347,
-      "learning_rate": 9.662643054635735e-06,
-      "loss": 0.4112,
-      "step": 6209
-    },
-    {
-      "epoch": 0.4058558264165741,
-      "grad_norm": 0.46446874737739563,
-      "learning_rate": 9.6625169521173e-06,
-      "loss": 0.4004,
-      "step": 6210
-    },
-    {
-      "epoch": 0.4059211816221162,
-      "grad_norm": 0.5079129338264465,
-      "learning_rate": 9.662390826858154e-06,
-      "loss": 0.4407,
-      "step": 6211
-    },
-    {
-      "epoch": 0.40598653682765834,
-      "grad_norm": 0.4871063828468323,
-      "learning_rate": 9.662264678858915e-06,
-      "loss": 0.3826,
-      "step": 6212
-    },
-    {
-      "epoch": 0.40605189203320047,
-      "grad_norm": 0.5497263073921204,
-      "learning_rate": 9.662138508120195e-06,
-      "loss": 0.5118,
-      "step": 6213
-    },
-    {
-      "epoch": 0.40611724723874254,
-      "grad_norm": 0.48818713426589966,
-      "learning_rate": 9.662012314642609e-06,
-      "loss": 0.4587,
-      "step": 6214
-    },
-    {
-      "epoch": 0.40618260244428467,
-      "grad_norm": 0.44175398349761963,
-      "learning_rate": 9.661886098426777e-06,
-      "loss": 0.3615,
-      "step": 6215
-    },
-    {
-      "epoch": 0.4062479576498268,
-      "grad_norm": 0.49936994910240173,
-      "learning_rate": 9.661759859473307e-06,
-      "loss": 0.448,
-      "step": 6216
-    },
-    {
-      "epoch": 0.4063133128553689,
-      "grad_norm": 0.4806167185306549,
-      "learning_rate": 9.661633597782823e-06,
-      "loss": 0.3578,
-      "step": 6217
-    },
-    {
-      "epoch": 0.40637866806091105,
-      "grad_norm": 0.4918518662452698,
-      "learning_rate": 9.661507313355935e-06,
-      "loss": 0.398,
-      "step": 6218
-    },
-    {
-      "epoch": 0.4064440232664532,
-      "grad_norm": 0.4765973389148712,
-      "learning_rate": 9.661381006193261e-06,
-      "loss": 0.4001,
-      "step": 6219
-    },
-    {
-      "epoch": 0.4065093784719953,
-      "grad_norm": 0.47209399938583374,
-      "learning_rate": 9.661254676295418e-06,
-      "loss": 0.358,
-      "step": 6220
-    },
-    {
-      "epoch": 0.40657473367753744,
-      "grad_norm": 0.4761291444301605,
-      "learning_rate": 9.66112832366302e-06,
-      "loss": 0.4053,
-      "step": 6221
-    },
-    {
-      "epoch": 0.40664008888307956,
-      "grad_norm": 0.4611034691333771,
-      "learning_rate": 9.661001948296685e-06,
-      "loss": 0.3908,
-      "step": 6222
-    },
-    {
-      "epoch": 0.40670544408862164,
-      "grad_norm": 0.46636754274368286,
-      "learning_rate": 9.660875550197028e-06,
-      "loss": 0.4007,
-      "step": 6223
-    },
-    {
-      "epoch": 0.40677079929416377,
-      "grad_norm": 0.45153120160102844,
-      "learning_rate": 9.660749129364668e-06,
-      "loss": 0.3634,
-      "step": 6224
-    },
-    {
-      "epoch": 0.4068361544997059,
-      "grad_norm": 0.4640316069126129,
-      "learning_rate": 9.660622685800218e-06,
-      "loss": 0.4311,
-      "step": 6225
-    },
-    {
-      "epoch": 0.406901509705248,
-      "grad_norm": 0.48421111702919006,
-      "learning_rate": 9.660496219504298e-06,
-      "loss": 0.3934,
-      "step": 6226
-    },
-    {
-      "epoch": 0.40696686491079015,
-      "grad_norm": 0.44161295890808105,
-      "learning_rate": 9.660369730477524e-06,
-      "loss": 0.3405,
-      "step": 6227
-    },
-    {
-      "epoch": 0.4070322201163323,
-      "grad_norm": 0.5003371238708496,
-      "learning_rate": 9.660243218720511e-06,
-      "loss": 0.4521,
-      "step": 6228
-    },
-    {
-      "epoch": 0.4070975753218744,
-      "grad_norm": 0.413867712020874,
-      "learning_rate": 9.660116684233877e-06,
-      "loss": 0.3357,
-      "step": 6229
-    },
-    {
-      "epoch": 0.40716293052741653,
-      "grad_norm": 0.4682891368865967,
-      "learning_rate": 9.65999012701824e-06,
-      "loss": 0.3931,
-      "step": 6230
-    },
-    {
-      "epoch": 0.4072282857329586,
-      "grad_norm": 0.4446837902069092,
-      "learning_rate": 9.659863547074218e-06,
-      "loss": 0.3493,
-      "step": 6231
-    },
-    {
-      "epoch": 0.40729364093850073,
-      "grad_norm": 0.4651505947113037,
-      "learning_rate": 9.659736944402424e-06,
-      "loss": 0.3864,
-      "step": 6232
-    },
-    {
-      "epoch": 0.40735899614404286,
-      "grad_norm": 0.4548189342021942,
-      "learning_rate": 9.659610319003481e-06,
-      "loss": 0.398,
-      "step": 6233
-    },
-    {
-      "epoch": 0.407424351349585,
-      "grad_norm": 0.4308524429798126,
-      "learning_rate": 9.659483670878004e-06,
-      "loss": 0.3717,
-      "step": 6234
-    },
-    {
-      "epoch": 0.4074897065551271,
-      "grad_norm": 0.498067170381546,
-      "learning_rate": 9.65935700002661e-06,
-      "loss": 0.4998,
-      "step": 6235
-    },
-    {
-      "epoch": 0.40755506176066925,
-      "grad_norm": 0.47328585386276245,
-      "learning_rate": 9.65923030644992e-06,
-      "loss": 0.4105,
-      "step": 6236
-    },
-    {
-      "epoch": 0.4076204169662114,
-      "grad_norm": 0.48229318857192993,
-      "learning_rate": 9.659103590148548e-06,
-      "loss": 0.3991,
-      "step": 6237
-    },
-    {
-      "epoch": 0.4076857721717535,
-      "grad_norm": 0.4877176880836487,
-      "learning_rate": 9.658976851123113e-06,
-      "loss": 0.427,
-      "step": 6238
-    },
-    {
-      "epoch": 0.4077511273772956,
-      "grad_norm": 0.4455759823322296,
-      "learning_rate": 9.658850089374234e-06,
-      "loss": 0.3682,
-      "step": 6239
-    },
-    {
-      "epoch": 0.4078164825828377,
-      "grad_norm": 0.4522955119609833,
-      "learning_rate": 9.65872330490253e-06,
-      "loss": 0.4046,
-      "step": 6240
-    },
-    {
-      "epoch": 0.40788183778837983,
-      "grad_norm": 0.4727816879749298,
-      "learning_rate": 9.658596497708618e-06,
-      "loss": 0.3414,
-      "step": 6241
-    },
-    {
-      "epoch": 0.40794719299392196,
-      "grad_norm": 0.4037397503852844,
-      "learning_rate": 9.65846966779312e-06,
-      "loss": 0.317,
-      "step": 6242
-    },
-    {
-      "epoch": 0.4080125481994641,
-      "grad_norm": 0.45989301800727844,
-      "learning_rate": 9.658342815156646e-06,
-      "loss": 0.4004,
-      "step": 6243
-    },
-    {
-      "epoch": 0.4080779034050062,
-      "grad_norm": 0.5048174262046814,
-      "learning_rate": 9.658215939799824e-06,
-      "loss": 0.4635,
-      "step": 6244
-    },
-    {
-      "epoch": 0.40814325861054834,
-      "grad_norm": 0.4990854561328888,
-      "learning_rate": 9.658089041723269e-06,
-      "loss": 0.4536,
-      "step": 6245
-    },
-    {
-      "epoch": 0.40820861381609047,
-      "grad_norm": 0.43396371603012085,
-      "learning_rate": 9.6579621209276e-06,
-      "loss": 0.3521,
-      "step": 6246
-    },
-    {
-      "epoch": 0.4082739690216326,
-      "grad_norm": 0.43475762009620667,
-      "learning_rate": 9.657835177413434e-06,
-      "loss": 0.3556,
-      "step": 6247
-    },
-    {
-      "epoch": 0.40833932422717467,
-      "grad_norm": 0.48515042662620544,
-      "learning_rate": 9.657708211181395e-06,
-      "loss": 0.4212,
-      "step": 6248
-    },
-    {
-      "epoch": 0.4084046794327168,
-      "grad_norm": 0.47048693895339966,
-      "learning_rate": 9.657581222232098e-06,
-      "loss": 0.3557,
-      "step": 6249
-    },
-    {
-      "epoch": 0.4084700346382589,
-      "grad_norm": 0.4376108944416046,
-      "learning_rate": 9.657454210566164e-06,
-      "loss": 0.3954,
-      "step": 6250
-    },
-    {
-      "epoch": 0.40853538984380106,
-      "grad_norm": 0.47545501589775085,
-      "learning_rate": 9.657327176184212e-06,
-      "loss": 0.4319,
-      "step": 6251
-    },
-    {
-      "epoch": 0.4086007450493432,
-      "grad_norm": 0.4782899022102356,
-      "learning_rate": 9.657200119086862e-06,
-      "loss": 0.3818,
-      "step": 6252
-    },
-    {
-      "epoch": 0.4086661002548853,
-      "grad_norm": 0.44529226422309875,
-      "learning_rate": 9.657073039274733e-06,
-      "loss": 0.397,
-      "step": 6253
-    },
-    {
-      "epoch": 0.40873145546042744,
-      "grad_norm": 0.4592170715332031,
-      "learning_rate": 9.656945936748448e-06,
-      "loss": 0.39,
-      "step": 6254
-    },
-    {
-      "epoch": 0.40879681066596957,
-      "grad_norm": 0.449066698551178,
-      "learning_rate": 9.65681881150862e-06,
-      "loss": 0.3809,
-      "step": 6255
-    },
-    {
-      "epoch": 0.40886216587151164,
-      "grad_norm": 0.4741554856300354,
-      "learning_rate": 9.656691663555877e-06,
-      "loss": 0.4109,
-      "step": 6256
-    },
-    {
-      "epoch": 0.40892752107705377,
-      "grad_norm": 0.48184025287628174,
-      "learning_rate": 9.656564492890835e-06,
-      "loss": 0.4077,
-      "step": 6257
-    },
-    {
-      "epoch": 0.4089928762825959,
-      "grad_norm": 0.4373905062675476,
-      "learning_rate": 9.656437299514114e-06,
-      "loss": 0.3539,
-      "step": 6258
-    },
-    {
-      "epoch": 0.409058231488138,
-      "grad_norm": 0.4374079406261444,
-      "learning_rate": 9.656310083426337e-06,
-      "loss": 0.3625,
-      "step": 6259
-    },
-    {
-      "epoch": 0.40912358669368015,
-      "grad_norm": 0.41604936122894287,
-      "learning_rate": 9.65618284462812e-06,
-      "loss": 0.3324,
-      "step": 6260
-    },
-    {
-      "epoch": 0.4091889418992223,
-      "grad_norm": 0.4871421456336975,
-      "learning_rate": 9.656055583120087e-06,
-      "loss": 0.3604,
-      "step": 6261
-    },
-    {
-      "epoch": 0.4092542971047644,
-      "grad_norm": 0.5023635625839233,
-      "learning_rate": 9.655928298902858e-06,
-      "loss": 0.4469,
-      "step": 6262
-    },
-    {
-      "epoch": 0.40931965231030654,
-      "grad_norm": 0.5003955960273743,
-      "learning_rate": 9.655800991977054e-06,
-      "loss": 0.3828,
-      "step": 6263
-    },
-    {
-      "epoch": 0.40938500751584866,
-      "grad_norm": 0.4711710810661316,
-      "learning_rate": 9.655673662343296e-06,
-      "loss": 0.405,
-      "step": 6264
-    },
-    {
-      "epoch": 0.40945036272139074,
-      "grad_norm": 0.46596553921699524,
-      "learning_rate": 9.655546310002204e-06,
-      "loss": 0.3941,
-      "step": 6265
-    },
-    {
-      "epoch": 0.40951571792693287,
-      "grad_norm": 0.48585209250450134,
-      "learning_rate": 9.6554189349544e-06,
-      "loss": 0.4594,
-      "step": 6266
-    },
-    {
-      "epoch": 0.409581073132475,
-      "grad_norm": 0.4733290672302246,
-      "learning_rate": 9.655291537200505e-06,
-      "loss": 0.3973,
-      "step": 6267
-    },
-    {
-      "epoch": 0.4096464283380171,
-      "grad_norm": 0.4435085952281952,
-      "learning_rate": 9.65516411674114e-06,
-      "loss": 0.371,
-      "step": 6268
-    },
-    {
-      "epoch": 0.40971178354355925,
-      "grad_norm": 0.5237358212471008,
-      "learning_rate": 9.655036673576927e-06,
-      "loss": 0.4664,
-      "step": 6269
-    },
-    {
-      "epoch": 0.4097771387491014,
-      "grad_norm": 0.4727523922920227,
-      "learning_rate": 9.654909207708489e-06,
-      "loss": 0.4214,
-      "step": 6270
-    },
-    {
-      "epoch": 0.4098424939546435,
-      "grad_norm": 0.4445841908454895,
-      "learning_rate": 9.654781719136445e-06,
-      "loss": 0.3532,
-      "step": 6271
-    },
-    {
-      "epoch": 0.40990784916018563,
-      "grad_norm": 0.470441609621048,
-      "learning_rate": 9.654654207861418e-06,
-      "loss": 0.3739,
-      "step": 6272
-    },
-    {
-      "epoch": 0.4099732043657277,
-      "grad_norm": 0.4501136541366577,
-      "learning_rate": 9.65452667388403e-06,
-      "loss": 0.3555,
-      "step": 6273
-    },
-    {
-      "epoch": 0.41003855957126983,
-      "grad_norm": 0.49914395809173584,
-      "learning_rate": 9.654399117204902e-06,
-      "loss": 0.4225,
-      "step": 6274
-    },
-    {
-      "epoch": 0.41010391477681196,
-      "grad_norm": 0.4688049256801605,
-      "learning_rate": 9.654271537824658e-06,
-      "loss": 0.4027,
-      "step": 6275
-    },
-    {
-      "epoch": 0.4101692699823541,
-      "grad_norm": 0.4889581799507141,
-      "learning_rate": 9.65414393574392e-06,
-      "loss": 0.4532,
-      "step": 6276
-    },
-    {
-      "epoch": 0.4102346251878962,
-      "grad_norm": 0.470244437456131,
-      "learning_rate": 9.654016310963308e-06,
-      "loss": 0.3591,
-      "step": 6277
-    },
-    {
-      "epoch": 0.41029998039343835,
-      "grad_norm": 0.48283493518829346,
-      "learning_rate": 9.653888663483448e-06,
-      "loss": 0.4036,
-      "step": 6278
-    },
-    {
-      "epoch": 0.4103653355989805,
-      "grad_norm": 0.49872222542762756,
-      "learning_rate": 9.653760993304959e-06,
-      "loss": 0.4432,
-      "step": 6279
-    },
-    {
-      "epoch": 0.4104306908045226,
-      "grad_norm": 0.5243247151374817,
-      "learning_rate": 9.653633300428467e-06,
-      "loss": 0.4611,
-      "step": 6280
-    },
-    {
-      "epoch": 0.4104960460100647,
-      "grad_norm": 0.4761289060115814,
-      "learning_rate": 9.653505584854592e-06,
-      "loss": 0.439,
-      "step": 6281
-    },
-    {
-      "epoch": 0.4105614012156068,
-      "grad_norm": 0.4796181619167328,
-      "learning_rate": 9.653377846583957e-06,
-      "loss": 0.4189,
-      "step": 6282
-    },
-    {
-      "epoch": 0.41062675642114893,
-      "grad_norm": 0.5055959820747375,
-      "learning_rate": 9.65325008561719e-06,
-      "loss": 0.4411,
-      "step": 6283
-    },
-    {
-      "epoch": 0.41069211162669106,
-      "grad_norm": 0.49860087037086487,
-      "learning_rate": 9.653122301954907e-06,
-      "loss": 0.4269,
-      "step": 6284
-    },
-    {
-      "epoch": 0.4107574668322332,
-      "grad_norm": 0.4254683554172516,
-      "learning_rate": 9.652994495597737e-06,
-      "loss": 0.3599,
-      "step": 6285
-    },
-    {
-      "epoch": 0.4108228220377753,
-      "grad_norm": 0.4287961721420288,
-      "learning_rate": 9.652866666546298e-06,
-      "loss": 0.2996,
-      "step": 6286
-    },
-    {
-      "epoch": 0.41088817724331744,
-      "grad_norm": 0.4624471962451935,
-      "learning_rate": 9.652738814801218e-06,
-      "loss": 0.397,
-      "step": 6287
-    },
-    {
-      "epoch": 0.41095353244885957,
-      "grad_norm": 0.45612502098083496,
-      "learning_rate": 9.652610940363119e-06,
-      "loss": 0.375,
-      "step": 6288
-    },
-    {
-      "epoch": 0.4110188876544017,
-      "grad_norm": 0.4525100290775299,
-      "learning_rate": 9.652483043232624e-06,
-      "loss": 0.4226,
-      "step": 6289
-    },
-    {
-      "epoch": 0.41108424285994377,
-      "grad_norm": 0.5095105767250061,
-      "learning_rate": 9.652355123410357e-06,
-      "loss": 0.4664,
-      "step": 6290
-    },
-    {
-      "epoch": 0.4111495980654859,
-      "grad_norm": 0.46207407116889954,
-      "learning_rate": 9.652227180896943e-06,
-      "loss": 0.3348,
-      "step": 6291
-    },
-    {
-      "epoch": 0.411214953271028,
-      "grad_norm": 0.4590258300304413,
-      "learning_rate": 9.652099215693006e-06,
-      "loss": 0.3997,
-      "step": 6292
-    },
-    {
-      "epoch": 0.41128030847657016,
-      "grad_norm": 0.4889128804206848,
-      "learning_rate": 9.651971227799168e-06,
-      "loss": 0.4472,
-      "step": 6293
-    },
-    {
-      "epoch": 0.4113456636821123,
-      "grad_norm": 0.45457106828689575,
-      "learning_rate": 9.651843217216057e-06,
-      "loss": 0.3805,
-      "step": 6294
-    },
-    {
-      "epoch": 0.4114110188876544,
-      "grad_norm": 0.4759896993637085,
-      "learning_rate": 9.651715183944294e-06,
-      "loss": 0.3926,
-      "step": 6295
-    },
-    {
-      "epoch": 0.41147637409319654,
-      "grad_norm": 0.4567562937736511,
-      "learning_rate": 9.651587127984503e-06,
-      "loss": 0.338,
-      "step": 6296
-    },
-    {
-      "epoch": 0.41154172929873867,
-      "grad_norm": 0.49032914638519287,
-      "learning_rate": 9.651459049337313e-06,
-      "loss": 0.4234,
-      "step": 6297
-    },
-    {
-      "epoch": 0.41160708450428074,
-      "grad_norm": 0.4713863432407379,
-      "learning_rate": 9.651330948003343e-06,
-      "loss": 0.4053,
-      "step": 6298
-    },
-    {
-      "epoch": 0.41167243970982287,
-      "grad_norm": 0.46855491399765015,
-      "learning_rate": 9.651202823983223e-06,
-      "loss": 0.3953,
-      "step": 6299
-    },
-    {
-      "epoch": 0.411737794915365,
-      "grad_norm": 0.5075733065605164,
-      "learning_rate": 9.651074677277574e-06,
-      "loss": 0.4222,
-      "step": 6300
-    },
-    {
-      "epoch": 0.4118031501209071,
-      "grad_norm": 0.4374162256717682,
-      "learning_rate": 9.650946507887022e-06,
-      "loss": 0.3527,
-      "step": 6301
-    },
-    {
-      "epoch": 0.41186850532644925,
-      "grad_norm": 0.4779459238052368,
-      "learning_rate": 9.650818315812194e-06,
-      "loss": 0.4286,
-      "step": 6302
-    },
-    {
-      "epoch": 0.4119338605319914,
-      "grad_norm": 0.4661043882369995,
-      "learning_rate": 9.650690101053712e-06,
-      "loss": 0.3923,
-      "step": 6303
-    },
-    {
-      "epoch": 0.4119992157375335,
-      "grad_norm": 0.436037540435791,
-      "learning_rate": 9.650561863612206e-06,
-      "loss": 0.3697,
-      "step": 6304
-    },
-    {
-      "epoch": 0.41206457094307564,
-      "grad_norm": 0.5179614424705505,
-      "learning_rate": 9.650433603488297e-06,
-      "loss": 0.4282,
-      "step": 6305
-    },
-    {
-      "epoch": 0.41212992614861776,
-      "grad_norm": 0.49485376477241516,
-      "learning_rate": 9.650305320682612e-06,
-      "loss": 0.4296,
-      "step": 6306
-    },
-    {
-      "epoch": 0.41219528135415984,
-      "grad_norm": 0.4761447310447693,
-      "learning_rate": 9.650177015195778e-06,
-      "loss": 0.4351,
-      "step": 6307
-    },
-    {
-      "epoch": 0.41226063655970197,
-      "grad_norm": 0.44707223773002625,
-      "learning_rate": 9.650048687028419e-06,
-      "loss": 0.3594,
-      "step": 6308
-    },
-    {
-      "epoch": 0.4123259917652441,
-      "grad_norm": 0.46764659881591797,
-      "learning_rate": 9.649920336181162e-06,
-      "loss": 0.3742,
-      "step": 6309
-    },
-    {
-      "epoch": 0.4123913469707862,
-      "grad_norm": 0.46206626296043396,
-      "learning_rate": 9.649791962654631e-06,
-      "loss": 0.3575,
-      "step": 6310
-    },
-    {
-      "epoch": 0.41245670217632835,
-      "grad_norm": 0.46091535687446594,
-      "learning_rate": 9.649663566449457e-06,
-      "loss": 0.3992,
-      "step": 6311
-    },
-    {
-      "epoch": 0.4125220573818705,
-      "grad_norm": 0.5275874137878418,
-      "learning_rate": 9.64953514756626e-06,
-      "loss": 0.3964,
-      "step": 6312
-    },
-    {
-      "epoch": 0.4125874125874126,
-      "grad_norm": 0.45352184772491455,
-      "learning_rate": 9.64940670600567e-06,
-      "loss": 0.4023,
-      "step": 6313
-    },
-    {
-      "epoch": 0.41265276779295473,
-      "grad_norm": 0.41809654235839844,
-      "learning_rate": 9.649278241768313e-06,
-      "loss": 0.3301,
-      "step": 6314
-    },
-    {
-      "epoch": 0.4127181229984968,
-      "grad_norm": 0.44547584652900696,
-      "learning_rate": 9.649149754854817e-06,
-      "loss": 0.4001,
-      "step": 6315
-    },
-    {
-      "epoch": 0.41278347820403893,
-      "grad_norm": 0.4872359037399292,
-      "learning_rate": 9.649021245265804e-06,
-      "loss": 0.4315,
-      "step": 6316
-    },
-    {
-      "epoch": 0.41284883340958106,
-      "grad_norm": 0.4649243652820587,
-      "learning_rate": 9.648892713001906e-06,
-      "loss": 0.4069,
-      "step": 6317
-    },
-    {
-      "epoch": 0.4129141886151232,
-      "grad_norm": 0.44608205556869507,
-      "learning_rate": 9.648764158063747e-06,
-      "loss": 0.3972,
-      "step": 6318
-    },
-    {
-      "epoch": 0.4129795438206653,
-      "grad_norm": 0.4536326229572296,
-      "learning_rate": 9.648635580451955e-06,
-      "loss": 0.3966,
-      "step": 6319
-    },
-    {
-      "epoch": 0.41304489902620745,
-      "grad_norm": 0.4427248239517212,
-      "learning_rate": 9.648506980167156e-06,
-      "loss": 0.3812,
-      "step": 6320
-    },
-    {
-      "epoch": 0.4131102542317496,
-      "grad_norm": 0.45658716559410095,
-      "learning_rate": 9.64837835720998e-06,
-      "loss": 0.3798,
-      "step": 6321
-    },
-    {
-      "epoch": 0.4131756094372917,
-      "grad_norm": 0.4524311125278473,
-      "learning_rate": 9.64824971158105e-06,
-      "loss": 0.374,
-      "step": 6322
-    },
-    {
-      "epoch": 0.4132409646428338,
-      "grad_norm": 0.4253866970539093,
-      "learning_rate": 9.648121043280997e-06,
-      "loss": 0.3699,
-      "step": 6323
-    },
-    {
-      "epoch": 0.4133063198483759,
-      "grad_norm": 0.4903950095176697,
-      "learning_rate": 9.647992352310449e-06,
-      "loss": 0.4332,
-      "step": 6324
-    },
-    {
-      "epoch": 0.41337167505391803,
-      "grad_norm": 0.477671355009079,
-      "learning_rate": 9.64786363867003e-06,
-      "loss": 0.43,
-      "step": 6325
-    },
-    {
-      "epoch": 0.41343703025946016,
-      "grad_norm": 0.48784369230270386,
-      "learning_rate": 9.647734902360369e-06,
-      "loss": 0.4405,
-      "step": 6326
-    },
-    {
-      "epoch": 0.4135023854650023,
-      "grad_norm": 0.5050990581512451,
-      "learning_rate": 9.647606143382097e-06,
-      "loss": 0.4562,
-      "step": 6327
-    },
-    {
-      "epoch": 0.4135677406705444,
-      "grad_norm": 0.4481131434440613,
-      "learning_rate": 9.64747736173584e-06,
-      "loss": 0.3883,
-      "step": 6328
-    },
-    {
-      "epoch": 0.41363309587608654,
-      "grad_norm": 0.43018585443496704,
-      "learning_rate": 9.647348557422223e-06,
-      "loss": 0.3423,
-      "step": 6329
-    },
-    {
-      "epoch": 0.41369845108162867,
-      "grad_norm": 0.4817647337913513,
-      "learning_rate": 9.64721973044188e-06,
-      "loss": 0.4058,
-      "step": 6330
-    },
-    {
-      "epoch": 0.4137638062871708,
-      "grad_norm": 0.46122077107429504,
-      "learning_rate": 9.647090880795434e-06,
-      "loss": 0.4009,
-      "step": 6331
-    },
-    {
-      "epoch": 0.41382916149271287,
-      "grad_norm": 0.4703610837459564,
-      "learning_rate": 9.646962008483518e-06,
-      "loss": 0.4012,
-      "step": 6332
-    },
-    {
-      "epoch": 0.413894516698255,
-      "grad_norm": 0.4957987666130066,
-      "learning_rate": 9.646833113506758e-06,
-      "loss": 0.4215,
-      "step": 6333
-    },
-    {
-      "epoch": 0.4139598719037971,
-      "grad_norm": 0.4563731849193573,
-      "learning_rate": 9.64670419586578e-06,
-      "loss": 0.3844,
-      "step": 6334
-    },
-    {
-      "epoch": 0.41402522710933926,
-      "grad_norm": 0.4593028426170349,
-      "learning_rate": 9.64657525556122e-06,
-      "loss": 0.3897,
-      "step": 6335
-    },
-    {
-      "epoch": 0.4140905823148814,
-      "grad_norm": 0.47703373432159424,
-      "learning_rate": 9.6464462925937e-06,
-      "loss": 0.4261,
-      "step": 6336
-    },
-    {
-      "epoch": 0.4141559375204235,
-      "grad_norm": 0.45836660265922546,
-      "learning_rate": 9.646317306963853e-06,
-      "loss": 0.4094,
-      "step": 6337
-    },
-    {
-      "epoch": 0.41422129272596564,
-      "grad_norm": 0.47570812702178955,
-      "learning_rate": 9.646188298672308e-06,
-      "loss": 0.4156,
-      "step": 6338
-    },
-    {
-      "epoch": 0.41428664793150777,
-      "grad_norm": 0.5125575661659241,
-      "learning_rate": 9.646059267719691e-06,
-      "loss": 0.3903,
-      "step": 6339
-    },
-    {
-      "epoch": 0.41435200313704984,
-      "grad_norm": 0.42409124970436096,
-      "learning_rate": 9.645930214106635e-06,
-      "loss": 0.3551,
-      "step": 6340
-    },
-    {
-      "epoch": 0.41441735834259197,
-      "grad_norm": 0.48722681403160095,
-      "learning_rate": 9.645801137833766e-06,
-      "loss": 0.4434,
-      "step": 6341
-    },
-    {
-      "epoch": 0.4144827135481341,
-      "grad_norm": 0.45247524976730347,
-      "learning_rate": 9.645672038901716e-06,
-      "loss": 0.3257,
-      "step": 6342
-    },
-    {
-      "epoch": 0.4145480687536762,
-      "grad_norm": 0.4561573565006256,
-      "learning_rate": 9.645542917311116e-06,
-      "loss": 0.3802,
-      "step": 6343
-    },
-    {
-      "epoch": 0.41461342395921835,
-      "grad_norm": 0.4621433615684509,
-      "learning_rate": 9.645413773062593e-06,
-      "loss": 0.3769,
-      "step": 6344
-    },
-    {
-      "epoch": 0.4146787791647605,
-      "grad_norm": 0.44296616315841675,
-      "learning_rate": 9.645284606156776e-06,
-      "loss": 0.3653,
-      "step": 6345
-    },
-    {
-      "epoch": 0.4147441343703026,
-      "grad_norm": 0.46599406003952026,
-      "learning_rate": 9.645155416594299e-06,
-      "loss": 0.3809,
-      "step": 6346
-    },
-    {
-      "epoch": 0.41480948957584474,
-      "grad_norm": 0.4901500344276428,
-      "learning_rate": 9.645026204375787e-06,
-      "loss": 0.4152,
-      "step": 6347
-    },
-    {
-      "epoch": 0.41487484478138686,
-      "grad_norm": 0.5185682773590088,
-      "learning_rate": 9.644896969501876e-06,
-      "loss": 0.4723,
-      "step": 6348
-    },
-    {
-      "epoch": 0.41494019998692894,
-      "grad_norm": 0.469554603099823,
-      "learning_rate": 9.64476771197319e-06,
-      "loss": 0.3674,
-      "step": 6349
-    },
-    {
-      "epoch": 0.41500555519247107,
-      "grad_norm": 0.41888344287872314,
-      "learning_rate": 9.644638431790366e-06,
-      "loss": 0.3479,
-      "step": 6350
-    },
-    {
-      "epoch": 0.4150709103980132,
-      "grad_norm": 0.4631621837615967,
-      "learning_rate": 9.644509128954029e-06,
-      "loss": 0.3965,
-      "step": 6351
-    },
-    {
-      "epoch": 0.4151362656035553,
-      "grad_norm": 0.43988871574401855,
-      "learning_rate": 9.644379803464814e-06,
-      "loss": 0.3656,
-      "step": 6352
-    },
-    {
-      "epoch": 0.41520162080909745,
-      "grad_norm": 0.4931758642196655,
-      "learning_rate": 9.644250455323349e-06,
-      "loss": 0.4208,
-      "step": 6353
-    },
-    {
-      "epoch": 0.4152669760146396,
-      "grad_norm": 0.42915892601013184,
-      "learning_rate": 9.644121084530265e-06,
-      "loss": 0.3935,
-      "step": 6354
-    },
-    {
-      "epoch": 0.4153323312201817,
-      "grad_norm": 0.4514559209346771,
-      "learning_rate": 9.643991691086194e-06,
-      "loss": 0.3745,
-      "step": 6355
-    },
-    {
-      "epoch": 0.41539768642572383,
-      "grad_norm": 0.4412904381752014,
-      "learning_rate": 9.643862274991766e-06,
-      "loss": 0.4059,
-      "step": 6356
-    },
-    {
-      "epoch": 0.4154630416312659,
-      "grad_norm": 0.5144051313400269,
-      "learning_rate": 9.643732836247614e-06,
-      "loss": 0.4785,
-      "step": 6357
-    },
-    {
-      "epoch": 0.41552839683680803,
-      "grad_norm": 0.45030677318573,
-      "learning_rate": 9.643603374854367e-06,
-      "loss": 0.4072,
-      "step": 6358
-    },
-    {
-      "epoch": 0.41559375204235016,
-      "grad_norm": 0.4990215301513672,
-      "learning_rate": 9.643473890812658e-06,
-      "loss": 0.3493,
-      "step": 6359
-    },
-    {
-      "epoch": 0.4156591072478923,
-      "grad_norm": 0.45090121030807495,
-      "learning_rate": 9.64334438412312e-06,
-      "loss": 0.3531,
-      "step": 6360
-    },
-    {
-      "epoch": 0.4157244624534344,
-      "grad_norm": 0.4247357249259949,
-      "learning_rate": 9.64321485478638e-06,
-      "loss": 0.3396,
-      "step": 6361
-    },
-    {
-      "epoch": 0.41578981765897655,
-      "grad_norm": 0.4885372817516327,
-      "learning_rate": 9.643085302803074e-06,
-      "loss": 0.3966,
-      "step": 6362
-    },
-    {
-      "epoch": 0.4158551728645187,
-      "grad_norm": 0.4586491584777832,
-      "learning_rate": 9.642955728173833e-06,
-      "loss": 0.3746,
-      "step": 6363
-    },
-    {
-      "epoch": 0.4159205280700608,
-      "grad_norm": 0.47208213806152344,
-      "learning_rate": 9.642826130899287e-06,
-      "loss": 0.4177,
-      "step": 6364
-    },
-    {
-      "epoch": 0.4159858832756029,
-      "grad_norm": 0.4739512503147125,
-      "learning_rate": 9.642696510980072e-06,
-      "loss": 0.4394,
-      "step": 6365
-    },
-    {
-      "epoch": 0.416051238481145,
-      "grad_norm": 0.4629870653152466,
-      "learning_rate": 9.642566868416814e-06,
-      "loss": 0.4089,
-      "step": 6366
-    },
-    {
-      "epoch": 0.41611659368668713,
-      "grad_norm": 0.4545626938343048,
-      "learning_rate": 9.642437203210154e-06,
-      "loss": 0.4,
-      "step": 6367
-    },
-    {
-      "epoch": 0.41618194889222926,
-      "grad_norm": 0.4711894989013672,
-      "learning_rate": 9.642307515360715e-06,
-      "loss": 0.4555,
-      "step": 6368
-    },
-    {
-      "epoch": 0.4162473040977714,
-      "grad_norm": 0.45401862263679504,
-      "learning_rate": 9.642177804869136e-06,
-      "loss": 0.4058,
-      "step": 6369
-    },
-    {
-      "epoch": 0.4163126593033135,
-      "grad_norm": 0.4288789927959442,
-      "learning_rate": 9.642048071736047e-06,
-      "loss": 0.3452,
-      "step": 6370
-    },
-    {
-      "epoch": 0.41637801450885564,
-      "grad_norm": 0.4517304599285126,
-      "learning_rate": 9.641918315962082e-06,
-      "loss": 0.3917,
-      "step": 6371
-    },
-    {
-      "epoch": 0.41644336971439777,
-      "grad_norm": 0.4933716058731079,
-      "learning_rate": 9.641788537547873e-06,
-      "loss": 0.4526,
-      "step": 6372
-    },
-    {
-      "epoch": 0.4165087249199399,
-      "grad_norm": 0.47293055057525635,
-      "learning_rate": 9.641658736494053e-06,
-      "loss": 0.3568,
-      "step": 6373
-    },
-    {
-      "epoch": 0.41657408012548197,
-      "grad_norm": 0.46282824873924255,
-      "learning_rate": 9.641528912801255e-06,
-      "loss": 0.3822,
-      "step": 6374
-    },
-    {
-      "epoch": 0.4166394353310241,
-      "grad_norm": 0.4570055305957794,
-      "learning_rate": 9.641399066470112e-06,
-      "loss": 0.4075,
-      "step": 6375
-    },
-    {
-      "epoch": 0.4167047905365662,
-      "grad_norm": 0.43674564361572266,
-      "learning_rate": 9.64126919750126e-06,
-      "loss": 0.3788,
-      "step": 6376
-    },
-    {
-      "epoch": 0.41677014574210836,
-      "grad_norm": 0.5012200474739075,
-      "learning_rate": 9.641139305895329e-06,
-      "loss": 0.4448,
-      "step": 6377
-    },
-    {
-      "epoch": 0.4168355009476505,
-      "grad_norm": 0.476524293422699,
-      "learning_rate": 9.641009391652954e-06,
-      "loss": 0.3513,
-      "step": 6378
-    },
-    {
-      "epoch": 0.4169008561531926,
-      "grad_norm": 0.47984185814857483,
-      "learning_rate": 9.640879454774768e-06,
-      "loss": 0.3993,
-      "step": 6379
-    },
-    {
-      "epoch": 0.41696621135873474,
-      "grad_norm": 0.4569160044193268,
-      "learning_rate": 9.640749495261406e-06,
-      "loss": 0.3524,
-      "step": 6380
-    },
-    {
-      "epoch": 0.41703156656427687,
-      "grad_norm": 0.4451634883880615,
-      "learning_rate": 9.640619513113499e-06,
-      "loss": 0.384,
-      "step": 6381
-    },
-    {
-      "epoch": 0.41709692176981894,
-      "grad_norm": 0.47604092955589294,
-      "learning_rate": 9.640489508331684e-06,
-      "loss": 0.4264,
-      "step": 6382
-    },
-    {
-      "epoch": 0.41716227697536107,
-      "grad_norm": 0.45808592438697815,
-      "learning_rate": 9.640359480916594e-06,
-      "loss": 0.3728,
-      "step": 6383
-    },
-    {
-      "epoch": 0.4172276321809032,
-      "grad_norm": 0.4437313675880432,
-      "learning_rate": 9.640229430868864e-06,
-      "loss": 0.3991,
-      "step": 6384
-    },
-    {
-      "epoch": 0.4172929873864453,
-      "grad_norm": 0.4536198377609253,
-      "learning_rate": 9.640099358189127e-06,
-      "loss": 0.3717,
-      "step": 6385
-    },
-    {
-      "epoch": 0.41735834259198745,
-      "grad_norm": 0.46030333638191223,
-      "learning_rate": 9.639969262878018e-06,
-      "loss": 0.388,
-      "step": 6386
-    },
-    {
-      "epoch": 0.4174236977975296,
-      "grad_norm": 0.45528873801231384,
-      "learning_rate": 9.639839144936172e-06,
-      "loss": 0.4102,
-      "step": 6387
-    },
-    {
-      "epoch": 0.4174890530030717,
-      "grad_norm": 0.43292397260665894,
-      "learning_rate": 9.639709004364222e-06,
-      "loss": 0.3313,
-      "step": 6388
-    },
-    {
-      "epoch": 0.41755440820861384,
-      "grad_norm": 0.4031912684440613,
-      "learning_rate": 9.639578841162804e-06,
-      "loss": 0.3074,
-      "step": 6389
-    },
-    {
-      "epoch": 0.41761976341415596,
-      "grad_norm": 0.44478580355644226,
-      "learning_rate": 9.639448655332553e-06,
-      "loss": 0.4039,
-      "step": 6390
-    },
-    {
-      "epoch": 0.41768511861969804,
-      "grad_norm": 0.500061571598053,
-      "learning_rate": 9.639318446874104e-06,
-      "loss": 0.4263,
-      "step": 6391
-    },
-    {
-      "epoch": 0.41775047382524017,
-      "grad_norm": 0.4445696175098419,
-      "learning_rate": 9.639188215788092e-06,
-      "loss": 0.3838,
-      "step": 6392
-    },
-    {
-      "epoch": 0.4178158290307823,
-      "grad_norm": 0.47839704155921936,
-      "learning_rate": 9.63905796207515e-06,
-      "loss": 0.3883,
-      "step": 6393
-    },
-    {
-      "epoch": 0.4178811842363244,
-      "grad_norm": 0.4655381739139557,
-      "learning_rate": 9.638927685735918e-06,
-      "loss": 0.3777,
-      "step": 6394
-    },
-    {
-      "epoch": 0.41794653944186655,
-      "grad_norm": 0.4355512261390686,
-      "learning_rate": 9.638797386771029e-06,
-      "loss": 0.3431,
-      "step": 6395
-    },
-    {
-      "epoch": 0.4180118946474087,
-      "grad_norm": 0.4772740304470062,
-      "learning_rate": 9.638667065181116e-06,
-      "loss": 0.3893,
-      "step": 6396
-    },
-    {
-      "epoch": 0.4180772498529508,
-      "grad_norm": 0.440621942281723,
-      "learning_rate": 9.638536720966816e-06,
-      "loss": 0.3854,
-      "step": 6397
-    },
-    {
-      "epoch": 0.41814260505849293,
-      "grad_norm": 0.46311086416244507,
-      "learning_rate": 9.638406354128767e-06,
-      "loss": 0.4199,
-      "step": 6398
-    },
-    {
-      "epoch": 0.418207960264035,
-      "grad_norm": 0.4810176193714142,
-      "learning_rate": 9.638275964667603e-06,
-      "loss": 0.4053,
-      "step": 6399
-    },
-    {
-      "epoch": 0.41827331546957713,
-      "grad_norm": 0.48047590255737305,
-      "learning_rate": 9.638145552583959e-06,
-      "loss": 0.4594,
-      "step": 6400
-    },
-    {
-      "epoch": 0.41833867067511926,
-      "grad_norm": 0.4568217098712921,
-      "learning_rate": 9.638015117878474e-06,
-      "loss": 0.3654,
-      "step": 6401
-    },
-    {
-      "epoch": 0.4184040258806614,
-      "grad_norm": 0.46990150213241577,
-      "learning_rate": 9.637884660551782e-06,
-      "loss": 0.4181,
-      "step": 6402
-    },
-    {
-      "epoch": 0.4184693810862035,
-      "grad_norm": 0.4413808584213257,
-      "learning_rate": 9.63775418060452e-06,
-      "loss": 0.3778,
-      "step": 6403
-    },
-    {
-      "epoch": 0.41853473629174565,
-      "grad_norm": 0.44534632563591003,
-      "learning_rate": 9.637623678037323e-06,
-      "loss": 0.383,
-      "step": 6404
-    },
-    {
-      "epoch": 0.4186000914972878,
-      "grad_norm": 0.4657782018184662,
-      "learning_rate": 9.63749315285083e-06,
-      "loss": 0.3808,
-      "step": 6405
-    },
-    {
-      "epoch": 0.4186654467028299,
-      "grad_norm": 0.4660731554031372,
-      "learning_rate": 9.637362605045675e-06,
-      "loss": 0.4353,
-      "step": 6406
-    },
-    {
-      "epoch": 0.418730801908372,
-      "grad_norm": 0.42939215898513794,
-      "learning_rate": 9.637232034622496e-06,
-      "loss": 0.3732,
-      "step": 6407
-    },
-    {
-      "epoch": 0.4187961571139141,
-      "grad_norm": 0.48954418301582336,
-      "learning_rate": 9.637101441581934e-06,
-      "loss": 0.4305,
-      "step": 6408
-    },
-    {
-      "epoch": 0.41886151231945623,
-      "grad_norm": 0.43653449416160583,
-      "learning_rate": 9.636970825924616e-06,
-      "loss": 0.3532,
-      "step": 6409
-    },
-    {
-      "epoch": 0.41892686752499836,
-      "grad_norm": 0.822792649269104,
-      "learning_rate": 9.636840187651188e-06,
-      "loss": 0.4229,
-      "step": 6410
-    },
-    {
-      "epoch": 0.4189922227305405,
-      "grad_norm": 0.4509654939174652,
-      "learning_rate": 9.636709526762285e-06,
-      "loss": 0.3586,
-      "step": 6411
-    },
-    {
-      "epoch": 0.4190575779360826,
-      "grad_norm": 0.4449145793914795,
-      "learning_rate": 9.636578843258541e-06,
-      "loss": 0.3807,
-      "step": 6412
-    },
-    {
-      "epoch": 0.41912293314162474,
-      "grad_norm": 0.4820125102996826,
-      "learning_rate": 9.636448137140597e-06,
-      "loss": 0.4073,
-      "step": 6413
-    },
-    {
-      "epoch": 0.41918828834716687,
-      "grad_norm": 0.4795423746109009,
-      "learning_rate": 9.636317408409088e-06,
-      "loss": 0.4163,
-      "step": 6414
-    },
-    {
-      "epoch": 0.419253643552709,
-      "grad_norm": 0.5062985420227051,
-      "learning_rate": 9.636186657064654e-06,
-      "loss": 0.4213,
-      "step": 6415
-    },
-    {
-      "epoch": 0.41931899875825107,
-      "grad_norm": 0.4475199580192566,
-      "learning_rate": 9.63605588310793e-06,
-      "loss": 0.3542,
-      "step": 6416
-    },
-    {
-      "epoch": 0.4193843539637932,
-      "grad_norm": 0.4813227355480194,
-      "learning_rate": 9.635925086539558e-06,
-      "loss": 0.4107,
-      "step": 6417
-    },
-    {
-      "epoch": 0.4194497091693353,
-      "grad_norm": 0.4664250910282135,
-      "learning_rate": 9.635794267360172e-06,
-      "loss": 0.3884,
-      "step": 6418
-    },
-    {
-      "epoch": 0.41951506437487746,
-      "grad_norm": 0.4789632558822632,
-      "learning_rate": 9.635663425570412e-06,
-      "loss": 0.4021,
-      "step": 6419
-    },
-    {
-      "epoch": 0.4195804195804196,
-      "grad_norm": 0.5104120373725891,
-      "learning_rate": 9.635532561170914e-06,
-      "loss": 0.4383,
-      "step": 6420
-    },
-    {
-      "epoch": 0.4196457747859617,
-      "grad_norm": 0.47397351264953613,
-      "learning_rate": 9.635401674162319e-06,
-      "loss": 0.3893,
-      "step": 6421
-    },
-    {
-      "epoch": 0.41971112999150384,
-      "grad_norm": 0.46960514783859253,
-      "learning_rate": 9.635270764545264e-06,
-      "loss": 0.3942,
-      "step": 6422
-    },
-    {
-      "epoch": 0.41977648519704597,
-      "grad_norm": 0.46536019444465637,
-      "learning_rate": 9.635139832320387e-06,
-      "loss": 0.403,
-      "step": 6423
-    },
-    {
-      "epoch": 0.41984184040258804,
-      "grad_norm": 0.45115309953689575,
-      "learning_rate": 9.635008877488327e-06,
-      "loss": 0.3534,
-      "step": 6424
-    },
-    {
-      "epoch": 0.41990719560813017,
-      "grad_norm": 0.44649505615234375,
-      "learning_rate": 9.634877900049724e-06,
-      "loss": 0.3768,
-      "step": 6425
-    },
-    {
-      "epoch": 0.4199725508136723,
-      "grad_norm": 0.4636431336402893,
-      "learning_rate": 9.634746900005216e-06,
-      "loss": 0.4057,
-      "step": 6426
-    },
-    {
-      "epoch": 0.4200379060192144,
-      "grad_norm": 0.47141778469085693,
-      "learning_rate": 9.634615877355441e-06,
-      "loss": 0.4041,
-      "step": 6427
-    },
-    {
-      "epoch": 0.42010326122475655,
-      "grad_norm": 0.4881165027618408,
-      "learning_rate": 9.634484832101039e-06,
-      "loss": 0.4366,
-      "step": 6428
-    },
-    {
-      "epoch": 0.4201686164302987,
-      "grad_norm": 0.4366433024406433,
-      "learning_rate": 9.63435376424265e-06,
-      "loss": 0.3677,
-      "step": 6429
-    },
-    {
-      "epoch": 0.4202339716358408,
-      "grad_norm": 0.47002291679382324,
-      "learning_rate": 9.63422267378091e-06,
-      "loss": 0.429,
-      "step": 6430
-    },
-    {
-      "epoch": 0.42029932684138294,
-      "grad_norm": 0.4882446527481079,
-      "learning_rate": 9.634091560716462e-06,
-      "loss": 0.4305,
-      "step": 6431
-    },
-    {
-      "epoch": 0.42036468204692506,
-      "grad_norm": 0.46278485655784607,
-      "learning_rate": 9.633960425049944e-06,
-      "loss": 0.4103,
-      "step": 6432
-    },
-    {
-      "epoch": 0.42043003725246714,
-      "grad_norm": 0.4388725459575653,
-      "learning_rate": 9.633829266781995e-06,
-      "loss": 0.3563,
-      "step": 6433
-    },
-    {
-      "epoch": 0.42049539245800927,
-      "grad_norm": 0.42902904748916626,
-      "learning_rate": 9.633698085913256e-06,
-      "loss": 0.3515,
-      "step": 6434
-    },
-    {
-      "epoch": 0.4205607476635514,
-      "grad_norm": 0.4717724621295929,
-      "learning_rate": 9.633566882444365e-06,
-      "loss": 0.3895,
-      "step": 6435
-    },
-    {
-      "epoch": 0.4206261028690935,
-      "grad_norm": 0.4701840579509735,
-      "learning_rate": 9.633435656375964e-06,
-      "loss": 0.4316,
-      "step": 6436
-    },
-    {
-      "epoch": 0.42069145807463565,
-      "grad_norm": 0.4383528530597687,
-      "learning_rate": 9.633304407708693e-06,
-      "loss": 0.4148,
-      "step": 6437
-    },
-    {
-      "epoch": 0.4207568132801778,
-      "grad_norm": 0.46183767914772034,
-      "learning_rate": 9.63317313644319e-06,
-      "loss": 0.4315,
-      "step": 6438
-    },
-    {
-      "epoch": 0.4208221684857199,
-      "grad_norm": 0.41978150606155396,
-      "learning_rate": 9.633041842580098e-06,
-      "loss": 0.3545,
-      "step": 6439
-    },
-    {
-      "epoch": 0.42088752369126203,
-      "grad_norm": 0.4599344730377197,
-      "learning_rate": 9.632910526120054e-06,
-      "loss": 0.3699,
-      "step": 6440
-    },
-    {
-      "epoch": 0.4209528788968041,
-      "grad_norm": 0.4426717162132263,
-      "learning_rate": 9.6327791870637e-06,
-      "loss": 0.3995,
-      "step": 6441
-    },
-    {
-      "epoch": 0.42101823410234623,
-      "grad_norm": 0.4519651234149933,
-      "learning_rate": 9.63264782541168e-06,
-      "loss": 0.4049,
-      "step": 6442
-    },
-    {
-      "epoch": 0.42108358930788836,
-      "grad_norm": 0.4410443603992462,
-      "learning_rate": 9.632516441164629e-06,
-      "loss": 0.3757,
-      "step": 6443
-    },
-    {
-      "epoch": 0.4211489445134305,
-      "grad_norm": 0.4512501060962677,
-      "learning_rate": 9.632385034323191e-06,
-      "loss": 0.3869,
-      "step": 6444
-    },
-    {
-      "epoch": 0.4212142997189726,
-      "grad_norm": 0.47190433740615845,
-      "learning_rate": 9.632253604888007e-06,
-      "loss": 0.4084,
-      "step": 6445
-    },
-    {
-      "epoch": 0.42127965492451475,
-      "grad_norm": 0.5133803486824036,
-      "learning_rate": 9.632122152859717e-06,
-      "loss": 0.4533,
-      "step": 6446
-    },
-    {
-      "epoch": 0.4213450101300569,
-      "grad_norm": 0.9352949261665344,
-      "learning_rate": 9.631990678238962e-06,
-      "loss": 0.408,
-      "step": 6447
-    },
-    {
-      "epoch": 0.421410365335599,
-      "grad_norm": 0.4503645896911621,
-      "learning_rate": 9.631859181026385e-06,
-      "loss": 0.3935,
-      "step": 6448
-    },
-    {
-      "epoch": 0.4214757205411411,
-      "grad_norm": 0.44001659750938416,
-      "learning_rate": 9.631727661222625e-06,
-      "loss": 0.354,
-      "step": 6449
-    },
-    {
-      "epoch": 0.4215410757466832,
-      "grad_norm": 0.4678772985935211,
-      "learning_rate": 9.631596118828326e-06,
-      "loss": 0.3696,
-      "step": 6450
-    },
-    {
-      "epoch": 0.42160643095222533,
-      "grad_norm": 0.4532861113548279,
-      "learning_rate": 9.631464553844128e-06,
-      "loss": 0.3874,
-      "step": 6451
-    },
-    {
-      "epoch": 0.42167178615776746,
-      "grad_norm": 0.4305182099342346,
-      "learning_rate": 9.631332966270671e-06,
-      "loss": 0.3309,
-      "step": 6452
-    },
-    {
-      "epoch": 0.4217371413633096,
-      "grad_norm": 0.44606882333755493,
-      "learning_rate": 9.6312013561086e-06,
-      "loss": 0.3164,
-      "step": 6453
-    },
-    {
-      "epoch": 0.4218024965688517,
-      "grad_norm": 0.4718857407569885,
-      "learning_rate": 9.631069723358555e-06,
-      "loss": 0.4263,
-      "step": 6454
-    },
-    {
-      "epoch": 0.42186785177439384,
-      "grad_norm": 0.48390716314315796,
-      "learning_rate": 9.630938068021181e-06,
-      "loss": 0.4291,
-      "step": 6455
-    },
-    {
-      "epoch": 0.42193320697993597,
-      "grad_norm": 0.42989227175712585,
-      "learning_rate": 9.630806390097113e-06,
-      "loss": 0.3709,
-      "step": 6456
-    },
-    {
-      "epoch": 0.4219985621854781,
-      "grad_norm": 0.4639935791492462,
-      "learning_rate": 9.630674689587003e-06,
-      "loss": 0.3991,
-      "step": 6457
-    },
-    {
-      "epoch": 0.42206391739102017,
-      "grad_norm": 0.472952663898468,
-      "learning_rate": 9.630542966491485e-06,
-      "loss": 0.4536,
-      "step": 6458
-    },
-    {
-      "epoch": 0.4221292725965623,
-      "grad_norm": 0.45817291736602783,
-      "learning_rate": 9.630411220811207e-06,
-      "loss": 0.3825,
-      "step": 6459
-    },
-    {
-      "epoch": 0.4221946278021044,
-      "grad_norm": 0.4815440773963928,
-      "learning_rate": 9.630279452546808e-06,
-      "loss": 0.43,
-      "step": 6460
-    },
-    {
-      "epoch": 0.42225998300764656,
-      "grad_norm": 0.45036017894744873,
-      "learning_rate": 9.630147661698931e-06,
-      "loss": 0.3647,
-      "step": 6461
-    },
-    {
-      "epoch": 0.4223253382131887,
-      "grad_norm": 0.4516957700252533,
-      "learning_rate": 9.630015848268221e-06,
-      "loss": 0.388,
-      "step": 6462
-    },
-    {
-      "epoch": 0.4223906934187308,
-      "grad_norm": 0.44138821959495544,
-      "learning_rate": 9.62988401225532e-06,
-      "loss": 0.384,
-      "step": 6463
-    },
-    {
-      "epoch": 0.42245604862427294,
-      "grad_norm": 0.4491787850856781,
-      "learning_rate": 9.629752153660871e-06,
-      "loss": 0.3638,
-      "step": 6464
-    },
-    {
-      "epoch": 0.42252140382981507,
-      "grad_norm": 0.48579147458076477,
-      "learning_rate": 9.629620272485517e-06,
-      "loss": 0.3969,
-      "step": 6465
-    },
-    {
-      "epoch": 0.42258675903535714,
-      "grad_norm": 0.4487808346748352,
-      "learning_rate": 9.6294883687299e-06,
-      "loss": 0.3842,
-      "step": 6466
-    },
-    {
-      "epoch": 0.42265211424089927,
-      "grad_norm": 0.4325678050518036,
-      "learning_rate": 9.629356442394666e-06,
-      "loss": 0.352,
-      "step": 6467
-    },
-    {
-      "epoch": 0.4227174694464414,
-      "grad_norm": 0.41130340099334717,
-      "learning_rate": 9.629224493480455e-06,
-      "loss": 0.3225,
-      "step": 6468
-    },
-    {
-      "epoch": 0.4227828246519835,
-      "grad_norm": 0.4503832757472992,
-      "learning_rate": 9.629092521987913e-06,
-      "loss": 0.3908,
-      "step": 6469
-    },
-    {
-      "epoch": 0.42284817985752565,
-      "grad_norm": 0.45339593291282654,
-      "learning_rate": 9.628960527917683e-06,
-      "loss": 0.3658,
-      "step": 6470
-    },
-    {
-      "epoch": 0.4229135350630678,
-      "grad_norm": 0.4623308777809143,
-      "learning_rate": 9.62882851127041e-06,
-      "loss": 0.4133,
-      "step": 6471
-    },
-    {
-      "epoch": 0.4229788902686099,
-      "grad_norm": 0.44871044158935547,
-      "learning_rate": 9.628696472046734e-06,
-      "loss": 0.3715,
-      "step": 6472
-    },
-    {
-      "epoch": 0.42304424547415204,
-      "grad_norm": 0.4574100077152252,
-      "learning_rate": 9.628564410247306e-06,
-      "loss": 0.4189,
-      "step": 6473
-    },
-    {
-      "epoch": 0.42310960067969416,
-      "grad_norm": 0.48398715257644653,
-      "learning_rate": 9.628432325872764e-06,
-      "loss": 0.406,
-      "step": 6474
-    },
-    {
-      "epoch": 0.42317495588523624,
-      "grad_norm": 0.44641536474227905,
-      "learning_rate": 9.628300218923752e-06,
-      "loss": 0.3963,
-      "step": 6475
-    },
-    {
-      "epoch": 0.42324031109077836,
-      "grad_norm": 0.47431057691574097,
-      "learning_rate": 9.628168089400917e-06,
-      "loss": 0.3926,
-      "step": 6476
-    },
-    {
-      "epoch": 0.4233056662963205,
-      "grad_norm": 0.48836711049079895,
-      "learning_rate": 9.628035937304905e-06,
-      "loss": 0.3975,
-      "step": 6477
-    },
-    {
-      "epoch": 0.4233710215018626,
-      "grad_norm": 0.46272969245910645,
-      "learning_rate": 9.627903762636358e-06,
-      "loss": 0.4067,
-      "step": 6478
-    },
-    {
-      "epoch": 0.42343637670740475,
-      "grad_norm": 0.47614696621894836,
-      "learning_rate": 9.62777156539592e-06,
-      "loss": 0.4382,
-      "step": 6479
-    },
-    {
-      "epoch": 0.4235017319129469,
-      "grad_norm": 0.4734412133693695,
-      "learning_rate": 9.627639345584236e-06,
-      "loss": 0.4084,
-      "step": 6480
-    },
-    {
-      "epoch": 0.423567087118489,
-      "grad_norm": 0.4440581500530243,
-      "learning_rate": 9.627507103201954e-06,
-      "loss": 0.3692,
-      "step": 6481
-    },
-    {
-      "epoch": 0.42363244232403113,
-      "grad_norm": 0.4822497069835663,
-      "learning_rate": 9.627374838249716e-06,
-      "loss": 0.4218,
-      "step": 6482
-    },
-    {
-      "epoch": 0.4236977975295732,
-      "grad_norm": 0.42719003558158875,
-      "learning_rate": 9.627242550728167e-06,
-      "loss": 0.3567,
-      "step": 6483
-    },
-    {
-      "epoch": 0.42376315273511533,
-      "grad_norm": 0.4925616979598999,
-      "learning_rate": 9.627110240637954e-06,
-      "loss": 0.4256,
-      "step": 6484
-    },
-    {
-      "epoch": 0.42382850794065746,
-      "grad_norm": 0.49054938554763794,
-      "learning_rate": 9.626977907979722e-06,
-      "loss": 0.4291,
-      "step": 6485
-    },
-    {
-      "epoch": 0.4238938631461996,
-      "grad_norm": 0.4381217062473297,
-      "learning_rate": 9.626845552754113e-06,
-      "loss": 0.3521,
-      "step": 6486
-    },
-    {
-      "epoch": 0.4239592183517417,
-      "grad_norm": 0.4951472282409668,
-      "learning_rate": 9.62671317496178e-06,
-      "loss": 0.4498,
-      "step": 6487
-    },
-    {
-      "epoch": 0.42402457355728385,
-      "grad_norm": 0.45642176270484924,
-      "learning_rate": 9.62658077460336e-06,
-      "loss": 0.3895,
-      "step": 6488
-    },
-    {
-      "epoch": 0.424089928762826,
-      "grad_norm": 0.4530963599681854,
-      "learning_rate": 9.626448351679504e-06,
-      "loss": 0.3609,
-      "step": 6489
-    },
-    {
-      "epoch": 0.4241552839683681,
-      "grad_norm": 0.4927248954772949,
-      "learning_rate": 9.626315906190856e-06,
-      "loss": 0.4507,
-      "step": 6490
-    },
-    {
-      "epoch": 0.4242206391739102,
-      "grad_norm": 0.45638951659202576,
-      "learning_rate": 9.626183438138064e-06,
-      "loss": 0.386,
-      "step": 6491
-    },
-    {
-      "epoch": 0.4242859943794523,
-      "grad_norm": 0.4506241977214813,
-      "learning_rate": 9.626050947521772e-06,
-      "loss": 0.4048,
-      "step": 6492
-    },
-    {
-      "epoch": 0.42435134958499443,
-      "grad_norm": 0.4654195010662079,
-      "learning_rate": 9.625918434342627e-06,
-      "loss": 0.3924,
-      "step": 6493
-    },
-    {
-      "epoch": 0.42441670479053656,
-      "grad_norm": 0.4722681939601898,
-      "learning_rate": 9.625785898601274e-06,
-      "loss": 0.3935,
-      "step": 6494
-    },
-    {
-      "epoch": 0.4244820599960787,
-      "grad_norm": 0.47420647740364075,
-      "learning_rate": 9.625653340298363e-06,
-      "loss": 0.4008,
-      "step": 6495
-    },
-    {
-      "epoch": 0.4245474152016208,
-      "grad_norm": 0.44712671637535095,
-      "learning_rate": 9.625520759434537e-06,
-      "loss": 0.3915,
-      "step": 6496
-    },
-    {
-      "epoch": 0.42461277040716294,
-      "grad_norm": 0.5193232297897339,
-      "learning_rate": 9.625388156010443e-06,
-      "loss": 0.4885,
-      "step": 6497
-    },
-    {
-      "epoch": 0.42467812561270507,
-      "grad_norm": 0.4673554301261902,
-      "learning_rate": 9.62525553002673e-06,
-      "loss": 0.3768,
-      "step": 6498
-    },
-    {
-      "epoch": 0.4247434808182472,
-      "grad_norm": 0.462789922952652,
-      "learning_rate": 9.625122881484041e-06,
-      "loss": 0.4034,
-      "step": 6499
-    },
-    {
-      "epoch": 0.42480883602378927,
-      "grad_norm": 0.4847203493118286,
-      "learning_rate": 9.624990210383027e-06,
-      "loss": 0.387,
-      "step": 6500
-    },
-    {
-      "epoch": 0.4248741912293314,
-      "grad_norm": 0.44709813594818115,
-      "learning_rate": 9.624857516724336e-06,
-      "loss": 0.3402,
-      "step": 6501
-    },
-    {
-      "epoch": 0.4249395464348735,
-      "grad_norm": 0.48443934321403503,
-      "learning_rate": 9.624724800508609e-06,
-      "loss": 0.4141,
-      "step": 6502
-    },
-    {
-      "epoch": 0.42500490164041566,
-      "grad_norm": 0.48059728741645813,
-      "learning_rate": 9.624592061736499e-06,
-      "loss": 0.4162,
-      "step": 6503
-    },
-    {
-      "epoch": 0.4250702568459578,
-      "grad_norm": 0.43630126118659973,
-      "learning_rate": 9.624459300408651e-06,
-      "loss": 0.4015,
-      "step": 6504
-    },
-    {
-      "epoch": 0.4251356120514999,
-      "grad_norm": 0.44657135009765625,
-      "learning_rate": 9.624326516525712e-06,
-      "loss": 0.3795,
-      "step": 6505
-    },
-    {
-      "epoch": 0.42520096725704204,
-      "grad_norm": 0.42472267150878906,
-      "learning_rate": 9.62419371008833e-06,
-      "loss": 0.3547,
-      "step": 6506
-    },
-    {
-      "epoch": 0.42526632246258417,
-      "grad_norm": 0.4591834843158722,
-      "learning_rate": 9.624060881097155e-06,
-      "loss": 0.3791,
-      "step": 6507
-    },
-    {
-      "epoch": 0.42533167766812624,
-      "grad_norm": 0.4807187020778656,
-      "learning_rate": 9.623928029552833e-06,
-      "loss": 0.4424,
-      "step": 6508
-    },
-    {
-      "epoch": 0.42539703287366837,
-      "grad_norm": 0.4762464463710785,
-      "learning_rate": 9.623795155456013e-06,
-      "loss": 0.398,
-      "step": 6509
-    },
-    {
-      "epoch": 0.4254623880792105,
-      "grad_norm": 0.4546484053134918,
-      "learning_rate": 9.623662258807341e-06,
-      "loss": 0.4143,
-      "step": 6510
-    },
-    {
-      "epoch": 0.4255277432847526,
-      "grad_norm": 0.49070531129837036,
-      "learning_rate": 9.623529339607465e-06,
-      "loss": 0.4133,
-      "step": 6511
-    },
-    {
-      "epoch": 0.42559309849029475,
-      "grad_norm": 0.4635466933250427,
-      "learning_rate": 9.623396397857036e-06,
-      "loss": 0.3814,
-      "step": 6512
-    },
-    {
-      "epoch": 0.4256584536958369,
-      "grad_norm": 0.476260781288147,
-      "learning_rate": 9.623263433556701e-06,
-      "loss": 0.393,
-      "step": 6513
-    },
-    {
-      "epoch": 0.425723808901379,
-      "grad_norm": 0.4596383571624756,
-      "learning_rate": 9.623130446707109e-06,
-      "loss": 0.3824,
-      "step": 6514
-    },
-    {
-      "epoch": 0.42578916410692114,
-      "grad_norm": 0.5997282266616821,
-      "learning_rate": 9.622997437308907e-06,
-      "loss": 0.3895,
-      "step": 6515
-    },
-    {
-      "epoch": 0.42585451931246326,
-      "grad_norm": 0.4949215352535248,
-      "learning_rate": 9.622864405362745e-06,
-      "loss": 0.4091,
-      "step": 6516
-    },
-    {
-      "epoch": 0.42591987451800534,
-      "grad_norm": 0.45569026470184326,
-      "learning_rate": 9.622731350869275e-06,
-      "loss": 0.4015,
-      "step": 6517
-    },
-    {
-      "epoch": 0.42598522972354746,
-      "grad_norm": 0.48326602578163147,
-      "learning_rate": 9.62259827382914e-06,
-      "loss": 0.4152,
-      "step": 6518
-    },
-    {
-      "epoch": 0.4260505849290896,
-      "grad_norm": 0.48029467463493347,
-      "learning_rate": 9.622465174242992e-06,
-      "loss": 0.42,
-      "step": 6519
-    },
-    {
-      "epoch": 0.4261159401346317,
-      "grad_norm": 0.4485394060611725,
-      "learning_rate": 9.62233205211148e-06,
-      "loss": 0.3772,
-      "step": 6520
-    },
-    {
-      "epoch": 0.42618129534017385,
-      "grad_norm": 0.4546124041080475,
-      "learning_rate": 9.622198907435253e-06,
-      "loss": 0.3682,
-      "step": 6521
-    },
-    {
-      "epoch": 0.426246650545716,
-      "grad_norm": 0.45446571707725525,
-      "learning_rate": 9.62206574021496e-06,
-      "loss": 0.4168,
-      "step": 6522
-    },
-    {
-      "epoch": 0.4263120057512581,
-      "grad_norm": 0.4523228704929352,
-      "learning_rate": 9.621932550451253e-06,
-      "loss": 0.3528,
-      "step": 6523
-    },
-    {
-      "epoch": 0.42637736095680023,
-      "grad_norm": 0.48179081082344055,
-      "learning_rate": 9.621799338144779e-06,
-      "loss": 0.4355,
-      "step": 6524
-    },
-    {
-      "epoch": 0.4264427161623423,
-      "grad_norm": 0.48248958587646484,
-      "learning_rate": 9.621666103296188e-06,
-      "loss": 0.4306,
-      "step": 6525
-    },
-    {
-      "epoch": 0.42650807136788443,
-      "grad_norm": 0.48483365774154663,
-      "learning_rate": 9.621532845906133e-06,
-      "loss": 0.3844,
-      "step": 6526
-    },
-    {
-      "epoch": 0.42657342657342656,
-      "grad_norm": 0.46387338638305664,
-      "learning_rate": 9.621399565975258e-06,
-      "loss": 0.405,
-      "step": 6527
-    },
-    {
-      "epoch": 0.4266387817789687,
-      "grad_norm": 0.4772443473339081,
-      "learning_rate": 9.621266263504216e-06,
-      "loss": 0.3901,
-      "step": 6528
-    },
-    {
-      "epoch": 0.4267041369845108,
-      "grad_norm": 0.4691782295703888,
-      "learning_rate": 9.621132938493658e-06,
-      "loss": 0.4061,
-      "step": 6529
-    },
-    {
-      "epoch": 0.42676949219005295,
-      "grad_norm": 0.4504014551639557,
-      "learning_rate": 9.620999590944235e-06,
-      "loss": 0.3875,
-      "step": 6530
-    },
-    {
-      "epoch": 0.4268348473955951,
-      "grad_norm": 0.5112310647964478,
-      "learning_rate": 9.620866220856595e-06,
-      "loss": 0.4634,
-      "step": 6531
-    },
-    {
-      "epoch": 0.4269002026011372,
-      "grad_norm": 0.49836134910583496,
-      "learning_rate": 9.620732828231391e-06,
-      "loss": 0.4101,
-      "step": 6532
-    },
-    {
-      "epoch": 0.4269655578066793,
-      "grad_norm": 0.446916788816452,
-      "learning_rate": 9.620599413069272e-06,
-      "loss": 0.4071,
-      "step": 6533
-    },
-    {
-      "epoch": 0.4270309130122214,
-      "grad_norm": 0.4787539541721344,
-      "learning_rate": 9.620465975370888e-06,
-      "loss": 0.4323,
-      "step": 6534
-    },
-    {
-      "epoch": 0.42709626821776353,
-      "grad_norm": 0.4808615744113922,
-      "learning_rate": 9.620332515136893e-06,
-      "loss": 0.4084,
-      "step": 6535
-    },
-    {
-      "epoch": 0.42716162342330566,
-      "grad_norm": 0.4856660068035126,
-      "learning_rate": 9.620199032367931e-06,
-      "loss": 0.4295,
-      "step": 6536
-    },
-    {
-      "epoch": 0.4272269786288478,
-      "grad_norm": 0.44857537746429443,
-      "learning_rate": 9.620065527064661e-06,
-      "loss": 0.3565,
-      "step": 6537
-    },
-    {
-      "epoch": 0.4272923338343899,
-      "grad_norm": 0.4612455368041992,
-      "learning_rate": 9.619931999227731e-06,
-      "loss": 0.4084,
-      "step": 6538
-    },
-    {
-      "epoch": 0.42735768903993204,
-      "grad_norm": 0.4732638895511627,
-      "learning_rate": 9.61979844885779e-06,
-      "loss": 0.4076,
-      "step": 6539
-    },
-    {
-      "epoch": 0.42742304424547417,
-      "grad_norm": 0.4432867169380188,
-      "learning_rate": 9.619664875955494e-06,
-      "loss": 0.3668,
-      "step": 6540
-    },
-    {
-      "epoch": 0.4274883994510163,
-      "grad_norm": 0.46726810932159424,
-      "learning_rate": 9.619531280521493e-06,
-      "loss": 0.3699,
-      "step": 6541
-    },
-    {
-      "epoch": 0.42755375465655837,
-      "grad_norm": 0.47225064039230347,
-      "learning_rate": 9.619397662556434e-06,
-      "loss": 0.3946,
-      "step": 6542
-    },
-    {
-      "epoch": 0.4276191098621005,
-      "grad_norm": 0.44583478569984436,
-      "learning_rate": 9.619264022060974e-06,
-      "loss": 0.396,
-      "step": 6543
-    },
-    {
-      "epoch": 0.4276844650676426,
-      "grad_norm": 0.48455506563186646,
-      "learning_rate": 9.619130359035765e-06,
-      "loss": 0.4289,
-      "step": 6544
-    },
-    {
-      "epoch": 0.42774982027318476,
-      "grad_norm": 0.4884521961212158,
-      "learning_rate": 9.618996673481453e-06,
-      "loss": 0.3857,
-      "step": 6545
-    },
-    {
-      "epoch": 0.4278151754787269,
-      "grad_norm": 0.5418428778648376,
-      "learning_rate": 9.618862965398696e-06,
-      "loss": 0.4713,
-      "step": 6546
-    },
-    {
-      "epoch": 0.427880530684269,
-      "grad_norm": 0.46160823106765747,
-      "learning_rate": 9.618729234788144e-06,
-      "loss": 0.3885,
-      "step": 6547
-    },
-    {
-      "epoch": 0.42794588588981114,
-      "grad_norm": 0.45779356360435486,
-      "learning_rate": 9.61859548165045e-06,
-      "loss": 0.3828,
-      "step": 6548
-    },
-    {
-      "epoch": 0.42801124109535327,
-      "grad_norm": 0.4436955749988556,
-      "learning_rate": 9.618461705986265e-06,
-      "loss": 0.3879,
-      "step": 6549
-    },
-    {
-      "epoch": 0.42807659630089534,
-      "grad_norm": 0.48656827211380005,
-      "learning_rate": 9.618327907796244e-06,
-      "loss": 0.3839,
-      "step": 6550
-    },
-    {
-      "epoch": 0.42814195150643747,
-      "grad_norm": 0.4393969476222992,
-      "learning_rate": 9.618194087081037e-06,
-      "loss": 0.3826,
-      "step": 6551
-    },
-    {
-      "epoch": 0.4282073067119796,
-      "grad_norm": 0.4598098695278168,
-      "learning_rate": 9.618060243841298e-06,
-      "loss": 0.4324,
-      "step": 6552
-    },
-    {
-      "epoch": 0.4282726619175217,
-      "grad_norm": 0.45974496006965637,
-      "learning_rate": 9.617926378077678e-06,
-      "loss": 0.3863,
-      "step": 6553
-    },
-    {
-      "epoch": 0.42833801712306385,
-      "grad_norm": 0.5933836102485657,
-      "learning_rate": 9.617792489790831e-06,
-      "loss": 0.3986,
-      "step": 6554
-    },
-    {
-      "epoch": 0.428403372328606,
-      "grad_norm": 0.473839670419693,
-      "learning_rate": 9.617658578981412e-06,
-      "loss": 0.4184,
-      "step": 6555
-    },
-    {
-      "epoch": 0.4284687275341481,
-      "grad_norm": 0.47087037563323975,
-      "learning_rate": 9.617524645650071e-06,
-      "loss": 0.4181,
-      "step": 6556
-    },
-    {
-      "epoch": 0.42853408273969024,
-      "grad_norm": 0.4623969793319702,
-      "learning_rate": 9.617390689797464e-06,
-      "loss": 0.3599,
-      "step": 6557
-    },
-    {
-      "epoch": 0.42859943794523236,
-      "grad_norm": 0.4589996933937073,
-      "learning_rate": 9.617256711424241e-06,
-      "loss": 0.3902,
-      "step": 6558
-    },
-    {
-      "epoch": 0.42866479315077444,
-      "grad_norm": 0.5123886466026306,
-      "learning_rate": 9.61712271053106e-06,
-      "loss": 0.4765,
-      "step": 6559
-    },
-    {
-      "epoch": 0.42873014835631656,
-      "grad_norm": 0.4635045528411865,
-      "learning_rate": 9.61698868711857e-06,
-      "loss": 0.3764,
-      "step": 6560
-    },
-    {
-      "epoch": 0.4287955035618587,
-      "grad_norm": 0.441690593957901,
-      "learning_rate": 9.616854641187426e-06,
-      "loss": 0.3511,
-      "step": 6561
-    },
-    {
-      "epoch": 0.4288608587674008,
-      "grad_norm": 0.44375911355018616,
-      "learning_rate": 9.616720572738285e-06,
-      "loss": 0.3672,
-      "step": 6562
-    },
-    {
-      "epoch": 0.42892621397294295,
-      "grad_norm": 0.46911972761154175,
-      "learning_rate": 9.616586481771797e-06,
-      "loss": 0.4233,
-      "step": 6563
-    },
-    {
-      "epoch": 0.4289915691784851,
-      "grad_norm": 0.4201802611351013,
-      "learning_rate": 9.616452368288616e-06,
-      "loss": 0.3447,
-      "step": 6564
-    },
-    {
-      "epoch": 0.4290569243840272,
-      "grad_norm": 0.4446719288825989,
-      "learning_rate": 9.6163182322894e-06,
-      "loss": 0.3708,
-      "step": 6565
-    },
-    {
-      "epoch": 0.42912227958956933,
-      "grad_norm": 0.4703514575958252,
-      "learning_rate": 9.616184073774798e-06,
-      "loss": 0.4014,
-      "step": 6566
-    },
-    {
-      "epoch": 0.4291876347951114,
-      "grad_norm": 0.5087060332298279,
-      "learning_rate": 9.61604989274547e-06,
-      "loss": 0.4018,
-      "step": 6567
-    },
-    {
-      "epoch": 0.42925299000065353,
-      "grad_norm": 0.4773683547973633,
-      "learning_rate": 9.615915689202066e-06,
-      "loss": 0.4229,
-      "step": 6568
-    },
-    {
-      "epoch": 0.42931834520619566,
-      "grad_norm": 0.48420435190200806,
-      "learning_rate": 9.615781463145244e-06,
-      "loss": 0.4031,
-      "step": 6569
-    },
-    {
-      "epoch": 0.4293837004117378,
-      "grad_norm": 0.4794633090496063,
-      "learning_rate": 9.615647214575655e-06,
-      "loss": 0.3943,
-      "step": 6570
-    },
-    {
-      "epoch": 0.4294490556172799,
-      "grad_norm": 0.4491889178752899,
-      "learning_rate": 9.615512943493955e-06,
-      "loss": 0.3662,
-      "step": 6571
-    },
-    {
-      "epoch": 0.42951441082282205,
-      "grad_norm": 0.4783543348312378,
-      "learning_rate": 9.6153786499008e-06,
-      "loss": 0.4304,
-      "step": 6572
-    },
-    {
-      "epoch": 0.4295797660283642,
-      "grad_norm": 0.43879467248916626,
-      "learning_rate": 9.615244333796844e-06,
-      "loss": 0.3744,
-      "step": 6573
-    },
-    {
-      "epoch": 0.4296451212339063,
-      "grad_norm": 0.44893524050712585,
-      "learning_rate": 9.615109995182744e-06,
-      "loss": 0.3888,
-      "step": 6574
-    },
-    {
-      "epoch": 0.4297104764394484,
-      "grad_norm": 0.4799850285053253,
-      "learning_rate": 9.614975634059152e-06,
-      "loss": 0.4486,
-      "step": 6575
-    },
-    {
-      "epoch": 0.4297758316449905,
-      "grad_norm": 0.4692569077014923,
-      "learning_rate": 9.614841250426726e-06,
-      "loss": 0.4038,
-      "step": 6576
-    },
-    {
-      "epoch": 0.42984118685053263,
-      "grad_norm": 0.4473058879375458,
-      "learning_rate": 9.614706844286122e-06,
-      "loss": 0.3885,
-      "step": 6577
-    },
-    {
-      "epoch": 0.42990654205607476,
-      "grad_norm": 0.469270259141922,
-      "learning_rate": 9.614572415637991e-06,
-      "loss": 0.4171,
-      "step": 6578
-    },
-    {
-      "epoch": 0.4299718972616169,
-      "grad_norm": 0.43836092948913574,
-      "learning_rate": 9.614437964482993e-06,
-      "loss": 0.3768,
-      "step": 6579
-    },
-    {
-      "epoch": 0.430037252467159,
-      "grad_norm": 0.4554644525051117,
-      "learning_rate": 9.614303490821783e-06,
-      "loss": 0.352,
-      "step": 6580
-    },
-    {
-      "epoch": 0.43010260767270114,
-      "grad_norm": 0.4886828064918518,
-      "learning_rate": 9.614168994655016e-06,
-      "loss": 0.4283,
-      "step": 6581
-    },
-    {
-      "epoch": 0.43016796287824327,
-      "grad_norm": 0.46111127734184265,
-      "learning_rate": 9.614034475983347e-06,
-      "loss": 0.3735,
-      "step": 6582
-    },
-    {
-      "epoch": 0.4302333180837854,
-      "grad_norm": 0.4684242904186249,
-      "learning_rate": 9.613899934807435e-06,
-      "loss": 0.4032,
-      "step": 6583
-    },
-    {
-      "epoch": 0.43029867328932747,
-      "grad_norm": 0.47861090302467346,
-      "learning_rate": 9.613765371127935e-06,
-      "loss": 0.3815,
-      "step": 6584
-    },
-    {
-      "epoch": 0.4303640284948696,
-      "grad_norm": 0.4744029939174652,
-      "learning_rate": 9.613630784945501e-06,
-      "loss": 0.3965,
-      "step": 6585
-    },
-    {
-      "epoch": 0.4304293837004117,
-      "grad_norm": 0.5068906545639038,
-      "learning_rate": 9.613496176260793e-06,
-      "loss": 0.4415,
-      "step": 6586
-    },
-    {
-      "epoch": 0.43049473890595386,
-      "grad_norm": 0.449878066778183,
-      "learning_rate": 9.613361545074465e-06,
-      "loss": 0.3999,
-      "step": 6587
-    },
-    {
-      "epoch": 0.430560094111496,
-      "grad_norm": 0.44311073422431946,
-      "learning_rate": 9.613226891387174e-06,
-      "loss": 0.384,
-      "step": 6588
-    },
-    {
-      "epoch": 0.4306254493170381,
-      "grad_norm": 0.44952529668807983,
-      "learning_rate": 9.613092215199577e-06,
-      "loss": 0.3478,
-      "step": 6589
-    },
-    {
-      "epoch": 0.43069080452258024,
-      "grad_norm": 0.454725056886673,
-      "learning_rate": 9.612957516512333e-06,
-      "loss": 0.3628,
-      "step": 6590
-    },
-    {
-      "epoch": 0.43075615972812237,
-      "grad_norm": 0.4151926338672638,
-      "learning_rate": 9.612822795326096e-06,
-      "loss": 0.3249,
-      "step": 6591
-    },
-    {
-      "epoch": 0.43082151493366444,
-      "grad_norm": 0.4689428210258484,
-      "learning_rate": 9.612688051641526e-06,
-      "loss": 0.4147,
-      "step": 6592
-    },
-    {
-      "epoch": 0.43088687013920657,
-      "grad_norm": 0.48327550292015076,
-      "learning_rate": 9.612553285459276e-06,
-      "loss": 0.3895,
-      "step": 6593
-    },
-    {
-      "epoch": 0.4309522253447487,
-      "grad_norm": 0.4606989622116089,
-      "learning_rate": 9.612418496780008e-06,
-      "loss": 0.4329,
-      "step": 6594
-    },
-    {
-      "epoch": 0.4310175805502908,
-      "grad_norm": 0.4373370409011841,
-      "learning_rate": 9.612283685604374e-06,
-      "loss": 0.3607,
-      "step": 6595
-    },
-    {
-      "epoch": 0.43108293575583295,
-      "grad_norm": 0.4986790716648102,
-      "learning_rate": 9.612148851933037e-06,
-      "loss": 0.467,
-      "step": 6596
-    },
-    {
-      "epoch": 0.4311482909613751,
-      "grad_norm": 0.45467954874038696,
-      "learning_rate": 9.612013995766652e-06,
-      "loss": 0.3826,
-      "step": 6597
-    },
-    {
-      "epoch": 0.4312136461669172,
-      "grad_norm": 0.4827488958835602,
-      "learning_rate": 9.611879117105876e-06,
-      "loss": 0.4219,
-      "step": 6598
-    },
-    {
-      "epoch": 0.43127900137245934,
-      "grad_norm": 0.4785784184932709,
-      "learning_rate": 9.611744215951369e-06,
-      "loss": 0.4181,
-      "step": 6599
-    },
-    {
-      "epoch": 0.43134435657800146,
-      "grad_norm": 0.4393242597579956,
-      "learning_rate": 9.611609292303787e-06,
-      "loss": 0.3263,
-      "step": 6600
-    },
-    {
-      "epoch": 0.43140971178354354,
-      "grad_norm": 0.45603376626968384,
-      "learning_rate": 9.611474346163788e-06,
-      "loss": 0.4086,
-      "step": 6601
-    },
-    {
-      "epoch": 0.43147506698908566,
-      "grad_norm": 0.4390104413032532,
-      "learning_rate": 9.611339377532031e-06,
-      "loss": 0.3695,
-      "step": 6602
-    },
-    {
-      "epoch": 0.4315404221946278,
-      "grad_norm": 0.41211044788360596,
-      "learning_rate": 9.611204386409176e-06,
-      "loss": 0.3459,
-      "step": 6603
-    },
-    {
-      "epoch": 0.4316057774001699,
-      "grad_norm": 0.4459226429462433,
-      "learning_rate": 9.61106937279588e-06,
-      "loss": 0.3845,
-      "step": 6604
-    },
-    {
-      "epoch": 0.43167113260571205,
-      "grad_norm": 0.45860961079597473,
-      "learning_rate": 9.610934336692799e-06,
-      "loss": 0.4194,
-      "step": 6605
-    },
-    {
-      "epoch": 0.4317364878112542,
-      "grad_norm": 0.47973817586898804,
-      "learning_rate": 9.610799278100595e-06,
-      "loss": 0.4071,
-      "step": 6606
-    },
-    {
-      "epoch": 0.4318018430167963,
-      "grad_norm": 0.5148317217826843,
-      "learning_rate": 9.610664197019927e-06,
-      "loss": 0.4215,
-      "step": 6607
-    },
-    {
-      "epoch": 0.43186719822233843,
-      "grad_norm": 0.4817494750022888,
-      "learning_rate": 9.610529093451451e-06,
-      "loss": 0.368,
-      "step": 6608
-    },
-    {
-      "epoch": 0.4319325534278805,
-      "grad_norm": 0.5141207575798035,
-      "learning_rate": 9.610393967395827e-06,
-      "loss": 0.4433,
-      "step": 6609
-    },
-    {
-      "epoch": 0.43199790863342263,
-      "grad_norm": 0.4598926305770874,
-      "learning_rate": 9.610258818853716e-06,
-      "loss": 0.4034,
-      "step": 6610
-    },
-    {
-      "epoch": 0.43206326383896476,
-      "grad_norm": 0.48298901319503784,
-      "learning_rate": 9.610123647825775e-06,
-      "loss": 0.4176,
-      "step": 6611
-    },
-    {
-      "epoch": 0.4321286190445069,
-      "grad_norm": 0.4944208860397339,
-      "learning_rate": 9.609988454312664e-06,
-      "loss": 0.4159,
-      "step": 6612
-    },
-    {
-      "epoch": 0.432193974250049,
-      "grad_norm": 0.5217379331588745,
-      "learning_rate": 9.609853238315041e-06,
-      "loss": 0.4539,
-      "step": 6613
-    },
-    {
-      "epoch": 0.43225932945559115,
-      "grad_norm": 0.4784821569919586,
-      "learning_rate": 9.609717999833568e-06,
-      "loss": 0.4028,
-      "step": 6614
-    },
-    {
-      "epoch": 0.4323246846611333,
-      "grad_norm": 0.49810606241226196,
-      "learning_rate": 9.609582738868903e-06,
-      "loss": 0.4603,
-      "step": 6615
-    },
-    {
-      "epoch": 0.4323900398666754,
-      "grad_norm": 0.49869388341903687,
-      "learning_rate": 9.609447455421706e-06,
-      "loss": 0.4198,
-      "step": 6616
-    },
-    {
-      "epoch": 0.4324553950722175,
-      "grad_norm": 0.4355716407299042,
-      "learning_rate": 9.609312149492636e-06,
-      "loss": 0.3627,
-      "step": 6617
-    },
-    {
-      "epoch": 0.4325207502777596,
-      "grad_norm": 0.434512734413147,
-      "learning_rate": 9.609176821082354e-06,
-      "loss": 0.3362,
-      "step": 6618
-    },
-    {
-      "epoch": 0.43258610548330173,
-      "grad_norm": 0.46171969175338745,
-      "learning_rate": 9.60904147019152e-06,
-      "loss": 0.3941,
-      "step": 6619
-    },
-    {
-      "epoch": 0.43265146068884386,
-      "grad_norm": 0.45991650223731995,
-      "learning_rate": 9.608906096820796e-06,
-      "loss": 0.3781,
-      "step": 6620
-    },
-    {
-      "epoch": 0.432716815894386,
-      "grad_norm": 0.9240472912788391,
-      "learning_rate": 9.608770700970838e-06,
-      "loss": 0.4062,
-      "step": 6621
-    },
-    {
-      "epoch": 0.4327821710999281,
-      "grad_norm": 0.48443368077278137,
-      "learning_rate": 9.60863528264231e-06,
-      "loss": 0.4296,
-      "step": 6622
-    },
-    {
-      "epoch": 0.43284752630547024,
-      "grad_norm": 0.48755085468292236,
-      "learning_rate": 9.60849984183587e-06,
-      "loss": 0.3904,
-      "step": 6623
-    },
-    {
-      "epoch": 0.43291288151101237,
-      "grad_norm": 0.47976285219192505,
-      "learning_rate": 9.608364378552181e-06,
-      "loss": 0.4182,
-      "step": 6624
-    },
-    {
-      "epoch": 0.4329782367165545,
-      "grad_norm": 0.4211183786392212,
-      "learning_rate": 9.608228892791902e-06,
-      "loss": 0.3518,
-      "step": 6625
-    },
-    {
-      "epoch": 0.43304359192209657,
-      "grad_norm": 0.45069074630737305,
-      "learning_rate": 9.608093384555695e-06,
-      "loss": 0.3975,
-      "step": 6626
-    },
-    {
-      "epoch": 0.4331089471276387,
-      "grad_norm": 0.5069361329078674,
-      "learning_rate": 9.607957853844218e-06,
-      "loss": 0.4385,
-      "step": 6627
-    },
-    {
-      "epoch": 0.4331743023331808,
-      "grad_norm": 0.5243623852729797,
-      "learning_rate": 9.607822300658136e-06,
-      "loss": 0.4152,
-      "step": 6628
-    },
-    {
-      "epoch": 0.43323965753872296,
-      "grad_norm": 0.456946462392807,
-      "learning_rate": 9.607686724998106e-06,
-      "loss": 0.4219,
-      "step": 6629
-    },
-    {
-      "epoch": 0.4333050127442651,
-      "grad_norm": 0.4868513345718384,
-      "learning_rate": 9.607551126864794e-06,
-      "loss": 0.4491,
-      "step": 6630
-    },
-    {
-      "epoch": 0.4333703679498072,
-      "grad_norm": 0.4848618805408478,
-      "learning_rate": 9.60741550625886e-06,
-      "loss": 0.4436,
-      "step": 6631
-    },
-    {
-      "epoch": 0.43343572315534934,
-      "grad_norm": 0.4648330807685852,
-      "learning_rate": 9.60727986318096e-06,
-      "loss": 0.4258,
-      "step": 6632
-    },
-    {
-      "epoch": 0.43350107836089147,
-      "grad_norm": 0.4540043771266937,
-      "learning_rate": 9.607144197631764e-06,
-      "loss": 0.3961,
-      "step": 6633
-    },
-    {
-      "epoch": 0.43356643356643354,
-      "grad_norm": 0.46273645758628845,
-      "learning_rate": 9.607008509611928e-06,
-      "loss": 0.3933,
-      "step": 6634
-    },
-    {
-      "epoch": 0.43363178877197567,
-      "grad_norm": 0.499675989151001,
-      "learning_rate": 9.606872799122115e-06,
-      "loss": 0.4195,
-      "step": 6635
-    },
-    {
-      "epoch": 0.4336971439775178,
-      "grad_norm": 0.45402762293815613,
-      "learning_rate": 9.60673706616299e-06,
-      "loss": 0.3823,
-      "step": 6636
-    },
-    {
-      "epoch": 0.4337624991830599,
-      "grad_norm": 0.45103034377098083,
-      "learning_rate": 9.60660131073521e-06,
-      "loss": 0.3487,
-      "step": 6637
-    },
-    {
-      "epoch": 0.43382785438860205,
-      "grad_norm": 0.4304608702659607,
-      "learning_rate": 9.60646553283944e-06,
-      "loss": 0.3297,
-      "step": 6638
-    },
-    {
-      "epoch": 0.4338932095941442,
-      "grad_norm": 0.46604615449905396,
-      "learning_rate": 9.606329732476343e-06,
-      "loss": 0.4074,
-      "step": 6639
-    },
-    {
-      "epoch": 0.4339585647996863,
-      "grad_norm": 0.42506149411201477,
-      "learning_rate": 9.606193909646579e-06,
-      "loss": 0.3117,
-      "step": 6640
-    },
-    {
-      "epoch": 0.43402392000522844,
-      "grad_norm": 0.473295122385025,
-      "learning_rate": 9.60605806435081e-06,
-      "loss": 0.4023,
-      "step": 6641
-    },
-    {
-      "epoch": 0.43408927521077056,
-      "grad_norm": 0.4828851521015167,
-      "learning_rate": 9.605922196589704e-06,
-      "loss": 0.4079,
-      "step": 6642
-    },
-    {
-      "epoch": 0.43415463041631264,
-      "grad_norm": 0.4616178572177887,
-      "learning_rate": 9.605786306363916e-06,
-      "loss": 0.3928,
-      "step": 6643
-    },
-    {
-      "epoch": 0.43421998562185476,
-      "grad_norm": 0.4404052197933197,
-      "learning_rate": 9.605650393674114e-06,
-      "loss": 0.381,
-      "step": 6644
-    },
-    {
-      "epoch": 0.4342853408273969,
-      "grad_norm": 0.4153364896774292,
-      "learning_rate": 9.605514458520959e-06,
-      "loss": 0.347,
-      "step": 6645
-    },
-    {
-      "epoch": 0.434350696032939,
-      "grad_norm": 0.4468926787376404,
-      "learning_rate": 9.605378500905116e-06,
-      "loss": 0.3646,
-      "step": 6646
-    },
-    {
-      "epoch": 0.43441605123848115,
-      "grad_norm": 0.48296794295310974,
-      "learning_rate": 9.605242520827245e-06,
-      "loss": 0.4171,
-      "step": 6647
-    },
-    {
-      "epoch": 0.4344814064440233,
-      "grad_norm": 0.4960406422615051,
-      "learning_rate": 9.605106518288012e-06,
-      "loss": 0.4635,
-      "step": 6648
-    },
-    {
-      "epoch": 0.4345467616495654,
-      "grad_norm": 0.45715218782424927,
-      "learning_rate": 9.604970493288078e-06,
-      "loss": 0.4043,
-      "step": 6649
-    },
-    {
-      "epoch": 0.43461211685510753,
-      "grad_norm": 0.47133275866508484,
-      "learning_rate": 9.604834445828109e-06,
-      "loss": 0.3652,
-      "step": 6650
-    },
-    {
-      "epoch": 0.4346774720606496,
-      "grad_norm": 0.43846094608306885,
-      "learning_rate": 9.604698375908766e-06,
-      "loss": 0.3651,
-      "step": 6651
-    },
-    {
-      "epoch": 0.43474282726619173,
-      "grad_norm": 0.4737315773963928,
-      "learning_rate": 9.604562283530714e-06,
-      "loss": 0.3809,
-      "step": 6652
-    },
-    {
-      "epoch": 0.43480818247173386,
-      "grad_norm": 0.4771938920021057,
-      "learning_rate": 9.604426168694618e-06,
-      "loss": 0.4058,
-      "step": 6653
-    },
-    {
-      "epoch": 0.434873537677276,
-      "grad_norm": 0.48013123869895935,
-      "learning_rate": 9.604290031401137e-06,
-      "loss": 0.4381,
-      "step": 6654
-    },
-    {
-      "epoch": 0.4349388928828181,
-      "grad_norm": 0.4409782886505127,
-      "learning_rate": 9.604153871650942e-06,
-      "loss": 0.3935,
-      "step": 6655
-    },
-    {
-      "epoch": 0.43500424808836025,
-      "grad_norm": 0.4736935496330261,
-      "learning_rate": 9.604017689444691e-06,
-      "loss": 0.3832,
-      "step": 6656
-    },
-    {
-      "epoch": 0.4350696032939024,
-      "grad_norm": 0.451788067817688,
-      "learning_rate": 9.603881484783054e-06,
-      "loss": 0.3746,
-      "step": 6657
-    },
-    {
-      "epoch": 0.4351349584994445,
-      "grad_norm": 0.45443475246429443,
-      "learning_rate": 9.60374525766669e-06,
-      "loss": 0.4336,
-      "step": 6658
-    },
-    {
-      "epoch": 0.43520031370498663,
-      "grad_norm": 0.44025322794914246,
-      "learning_rate": 9.603609008096265e-06,
-      "loss": 0.3863,
-      "step": 6659
-    },
-    {
-      "epoch": 0.4352656689105287,
-      "grad_norm": 0.5253258943557739,
-      "learning_rate": 9.603472736072443e-06,
-      "loss": 0.4156,
-      "step": 6660
-    },
-    {
-      "epoch": 0.43533102411607083,
-      "grad_norm": 0.4346178472042084,
-      "learning_rate": 9.603336441595892e-06,
-      "loss": 0.3726,
-      "step": 6661
-    },
-    {
-      "epoch": 0.43539637932161296,
-      "grad_norm": 0.48458564281463623,
-      "learning_rate": 9.603200124667273e-06,
-      "loss": 0.4368,
-      "step": 6662
-    },
-    {
-      "epoch": 0.4354617345271551,
-      "grad_norm": 0.4136316180229187,
-      "learning_rate": 9.603063785287252e-06,
-      "loss": 0.3479,
-      "step": 6663
-    },
-    {
-      "epoch": 0.4355270897326972,
-      "grad_norm": 0.4404667615890503,
-      "learning_rate": 9.602927423456497e-06,
-      "loss": 0.3883,
-      "step": 6664
-    },
-    {
-      "epoch": 0.43559244493823934,
-      "grad_norm": 0.44658559560775757,
-      "learning_rate": 9.602791039175668e-06,
-      "loss": 0.377,
-      "step": 6665
-    },
-    {
-      "epoch": 0.43565780014378147,
-      "grad_norm": 0.4379778802394867,
-      "learning_rate": 9.602654632445434e-06,
-      "loss": 0.3947,
-      "step": 6666
-    },
-    {
-      "epoch": 0.4357231553493236,
-      "grad_norm": 0.4784342348575592,
-      "learning_rate": 9.602518203266456e-06,
-      "loss": 0.4134,
-      "step": 6667
-    },
-    {
-      "epoch": 0.43578851055486567,
-      "grad_norm": 0.40214741230010986,
-      "learning_rate": 9.602381751639405e-06,
-      "loss": 0.3257,
-      "step": 6668
-    },
-    {
-      "epoch": 0.4358538657604078,
-      "grad_norm": 0.47082242369651794,
-      "learning_rate": 9.602245277564944e-06,
-      "loss": 0.4101,
-      "step": 6669
-    },
-    {
-      "epoch": 0.4359192209659499,
-      "grad_norm": 0.47592031955718994,
-      "learning_rate": 9.602108781043735e-06,
-      "loss": 0.4164,
-      "step": 6670
-    },
-    {
-      "epoch": 0.43598457617149206,
-      "grad_norm": 0.43255478143692017,
-      "learning_rate": 9.601972262076452e-06,
-      "loss": 0.3636,
-      "step": 6671
-    },
-    {
-      "epoch": 0.4360499313770342,
-      "grad_norm": 0.43682658672332764,
-      "learning_rate": 9.601835720663752e-06,
-      "loss": 0.3748,
-      "step": 6672
-    },
-    {
-      "epoch": 0.4361152865825763,
-      "grad_norm": 0.43344199657440186,
-      "learning_rate": 9.601699156806306e-06,
-      "loss": 0.3813,
-      "step": 6673
-    },
-    {
-      "epoch": 0.43618064178811844,
-      "grad_norm": 0.4494476318359375,
-      "learning_rate": 9.60156257050478e-06,
-      "loss": 0.372,
-      "step": 6674
-    },
-    {
-      "epoch": 0.43624599699366057,
-      "grad_norm": 0.44902560114860535,
-      "learning_rate": 9.601425961759837e-06,
-      "loss": 0.3584,
-      "step": 6675
-    },
-    {
-      "epoch": 0.43631135219920264,
-      "grad_norm": 0.4559073746204376,
-      "learning_rate": 9.601289330572149e-06,
-      "loss": 0.3785,
-      "step": 6676
-    },
-    {
-      "epoch": 0.43637670740474477,
-      "grad_norm": 0.4348633289337158,
-      "learning_rate": 9.601152676942376e-06,
-      "loss": 0.3499,
-      "step": 6677
-    },
-    {
-      "epoch": 0.4364420626102869,
-      "grad_norm": 0.4531196355819702,
-      "learning_rate": 9.601016000871189e-06,
-      "loss": 0.4095,
-      "step": 6678
-    },
-    {
-      "epoch": 0.436507417815829,
-      "grad_norm": 0.4794025719165802,
-      "learning_rate": 9.600879302359253e-06,
-      "loss": 0.4136,
-      "step": 6679
-    },
-    {
-      "epoch": 0.43657277302137115,
-      "grad_norm": 0.46024563908576965,
-      "learning_rate": 9.600742581407234e-06,
-      "loss": 0.3687,
-      "step": 6680
-    },
-    {
-      "epoch": 0.4366381282269133,
-      "grad_norm": 0.4755650758743286,
-      "learning_rate": 9.6006058380158e-06,
-      "loss": 0.4086,
-      "step": 6681
-    },
-    {
-      "epoch": 0.4367034834324554,
-      "grad_norm": 0.4605053663253784,
-      "learning_rate": 9.600469072185616e-06,
-      "loss": 0.3909,
-      "step": 6682
-    },
-    {
-      "epoch": 0.43676883863799754,
-      "grad_norm": 0.4819033145904541,
-      "learning_rate": 9.600332283917352e-06,
-      "loss": 0.4149,
-      "step": 6683
-    },
-    {
-      "epoch": 0.43683419384353966,
-      "grad_norm": 0.42655113339424133,
-      "learning_rate": 9.600195473211676e-06,
-      "loss": 0.3483,
-      "step": 6684
-    },
-    {
-      "epoch": 0.43689954904908174,
-      "grad_norm": 0.4536789059638977,
-      "learning_rate": 9.600058640069249e-06,
-      "loss": 0.4081,
-      "step": 6685
-    },
-    {
-      "epoch": 0.43696490425462386,
-      "grad_norm": 0.48175111413002014,
-      "learning_rate": 9.599921784490745e-06,
-      "loss": 0.4044,
-      "step": 6686
-    },
-    {
-      "epoch": 0.437030259460166,
-      "grad_norm": 0.4394722878932953,
-      "learning_rate": 9.599784906476827e-06,
-      "loss": 0.369,
-      "step": 6687
-    },
-    {
-      "epoch": 0.4370956146657081,
-      "grad_norm": 0.45626410841941833,
-      "learning_rate": 9.599648006028166e-06,
-      "loss": 0.3925,
-      "step": 6688
-    },
-    {
-      "epoch": 0.43716096987125025,
-      "grad_norm": 0.47347065806388855,
-      "learning_rate": 9.599511083145427e-06,
-      "loss": 0.4162,
-      "step": 6689
-    },
-    {
-      "epoch": 0.4372263250767924,
-      "grad_norm": 0.4870966076850891,
-      "learning_rate": 9.59937413782928e-06,
-      "loss": 0.4259,
-      "step": 6690
-    },
-    {
-      "epoch": 0.4372916802823345,
-      "grad_norm": 0.4528791308403015,
-      "learning_rate": 9.599237170080391e-06,
-      "loss": 0.3816,
-      "step": 6691
-    },
-    {
-      "epoch": 0.43735703548787663,
-      "grad_norm": 0.4730829894542694,
-      "learning_rate": 9.59910017989943e-06,
-      "loss": 0.4202,
-      "step": 6692
-    },
-    {
-      "epoch": 0.4374223906934187,
-      "grad_norm": 0.4537069797515869,
-      "learning_rate": 9.598963167287064e-06,
-      "loss": 0.4044,
-      "step": 6693
-    },
-    {
-      "epoch": 0.43748774589896083,
-      "grad_norm": 0.4376910626888275,
-      "learning_rate": 9.59882613224396e-06,
-      "loss": 0.3798,
-      "step": 6694
-    },
-    {
-      "epoch": 0.43755310110450296,
-      "grad_norm": 0.4318070113658905,
-      "learning_rate": 9.59868907477079e-06,
-      "loss": 0.3898,
-      "step": 6695
-    },
-    {
-      "epoch": 0.4376184563100451,
-      "grad_norm": 0.43683746457099915,
-      "learning_rate": 9.598551994868219e-06,
-      "loss": 0.3551,
-      "step": 6696
-    },
-    {
-      "epoch": 0.4376838115155872,
-      "grad_norm": 0.4378679692745209,
-      "learning_rate": 9.598414892536917e-06,
-      "loss": 0.391,
-      "step": 6697
-    },
-    {
-      "epoch": 0.43774916672112935,
-      "grad_norm": 0.46658703684806824,
-      "learning_rate": 9.598277767777553e-06,
-      "loss": 0.4143,
-      "step": 6698
-    },
-    {
-      "epoch": 0.4378145219266715,
-      "grad_norm": 0.4765806198120117,
-      "learning_rate": 9.598140620590794e-06,
-      "loss": 0.3998,
-      "step": 6699
-    },
-    {
-      "epoch": 0.4378798771322136,
-      "grad_norm": 0.48689937591552734,
-      "learning_rate": 9.59800345097731e-06,
-      "loss": 0.4645,
-      "step": 6700
-    },
-    {
-      "epoch": 0.43794523233775573,
-      "grad_norm": 0.43289056420326233,
-      "learning_rate": 9.597866258937774e-06,
-      "loss": 0.3401,
-      "step": 6701
-    },
-    {
-      "epoch": 0.4380105875432978,
-      "grad_norm": 0.4396793842315674,
-      "learning_rate": 9.597729044472847e-06,
-      "loss": 0.3612,
-      "step": 6702
-    },
-    {
-      "epoch": 0.43807594274883993,
-      "grad_norm": 0.4856289327144623,
-      "learning_rate": 9.597591807583206e-06,
-      "loss": 0.4165,
-      "step": 6703
-    },
-    {
-      "epoch": 0.43814129795438206,
-      "grad_norm": 0.4623776376247406,
-      "learning_rate": 9.597454548269514e-06,
-      "loss": 0.3631,
-      "step": 6704
-    },
-    {
-      "epoch": 0.4382066531599242,
-      "grad_norm": 0.44471532106399536,
-      "learning_rate": 9.597317266532446e-06,
-      "loss": 0.3867,
-      "step": 6705
-    },
-    {
-      "epoch": 0.4382720083654663,
-      "grad_norm": 0.47742483019828796,
-      "learning_rate": 9.597179962372668e-06,
-      "loss": 0.3764,
-      "step": 6706
-    },
-    {
-      "epoch": 0.43833736357100844,
-      "grad_norm": 0.4707384407520294,
-      "learning_rate": 9.59704263579085e-06,
-      "loss": 0.4285,
-      "step": 6707
-    },
-    {
-      "epoch": 0.43840271877655057,
-      "grad_norm": 0.4595271348953247,
-      "learning_rate": 9.596905286787663e-06,
-      "loss": 0.4254,
-      "step": 6708
-    },
-    {
-      "epoch": 0.4384680739820927,
-      "grad_norm": 0.4742881953716278,
-      "learning_rate": 9.596767915363777e-06,
-      "loss": 0.45,
-      "step": 6709
-    },
-    {
-      "epoch": 0.43853342918763477,
-      "grad_norm": 0.42788028717041016,
-      "learning_rate": 9.59663052151986e-06,
-      "loss": 0.3547,
-      "step": 6710
-    },
-    {
-      "epoch": 0.4385987843931769,
-      "grad_norm": 0.4922092854976654,
-      "learning_rate": 9.596493105256584e-06,
-      "loss": 0.4076,
-      "step": 6711
-    },
-    {
-      "epoch": 0.438664139598719,
-      "grad_norm": 0.480979859828949,
-      "learning_rate": 9.59635566657462e-06,
-      "loss": 0.4404,
-      "step": 6712
-    },
-    {
-      "epoch": 0.43872949480426116,
-      "grad_norm": 0.5063999891281128,
-      "learning_rate": 9.596218205474637e-06,
-      "loss": 0.4622,
-      "step": 6713
-    },
-    {
-      "epoch": 0.4387948500098033,
-      "grad_norm": 0.42097312211990356,
-      "learning_rate": 9.596080721957304e-06,
-      "loss": 0.3682,
-      "step": 6714
-    },
-    {
-      "epoch": 0.4388602052153454,
-      "grad_norm": 0.4473300278186798,
-      "learning_rate": 9.595943216023293e-06,
-      "loss": 0.3855,
-      "step": 6715
-    },
-    {
-      "epoch": 0.43892556042088754,
-      "grad_norm": 0.4737445116043091,
-      "learning_rate": 9.595805687673276e-06,
-      "loss": 0.387,
-      "step": 6716
-    },
-    {
-      "epoch": 0.43899091562642967,
-      "grad_norm": 0.4513935148715973,
-      "learning_rate": 9.595668136907924e-06,
-      "loss": 0.3955,
-      "step": 6717
-    },
-    {
-      "epoch": 0.43905627083197174,
-      "grad_norm": 0.4501236379146576,
-      "learning_rate": 9.595530563727904e-06,
-      "loss": 0.3791,
-      "step": 6718
-    },
-    {
-      "epoch": 0.43912162603751387,
-      "grad_norm": 0.46202975511550903,
-      "learning_rate": 9.59539296813389e-06,
-      "loss": 0.3753,
-      "step": 6719
-    },
-    {
-      "epoch": 0.439186981243056,
-      "grad_norm": 0.44744008779525757,
-      "learning_rate": 9.59525535012655e-06,
-      "loss": 0.3958,
-      "step": 6720
-    },
-    {
-      "epoch": 0.4392523364485981,
-      "grad_norm": 0.4539359211921692,
-      "learning_rate": 9.595117709706562e-06,
-      "loss": 0.3914,
-      "step": 6721
-    },
-    {
-      "epoch": 0.43931769165414025,
-      "grad_norm": 0.48395809531211853,
-      "learning_rate": 9.59498004687459e-06,
-      "loss": 0.4611,
-      "step": 6722
-    },
-    {
-      "epoch": 0.4393830468596824,
-      "grad_norm": 0.4520648419857025,
-      "learning_rate": 9.59484236163131e-06,
-      "loss": 0.3771,
-      "step": 6723
-    },
-    {
-      "epoch": 0.4394484020652245,
-      "grad_norm": 0.42805686593055725,
-      "learning_rate": 9.594704653977392e-06,
-      "loss": 0.3749,
-      "step": 6724
-    },
-    {
-      "epoch": 0.43951375727076664,
-      "grad_norm": 0.48044294118881226,
-      "learning_rate": 9.594566923913506e-06,
-      "loss": 0.4229,
-      "step": 6725
-    },
-    {
-      "epoch": 0.43957911247630876,
-      "grad_norm": 0.43216997385025024,
-      "learning_rate": 9.594429171440328e-06,
-      "loss": 0.3678,
-      "step": 6726
-    },
-    {
-      "epoch": 0.43964446768185084,
-      "grad_norm": 0.44116532802581787,
-      "learning_rate": 9.594291396558526e-06,
-      "loss": 0.3945,
-      "step": 6727
-    },
-    {
-      "epoch": 0.43970982288739296,
-      "grad_norm": 0.4396519362926483,
-      "learning_rate": 9.594153599268773e-06,
-      "loss": 0.3574,
-      "step": 6728
-    },
-    {
-      "epoch": 0.4397751780929351,
-      "grad_norm": 0.4698348641395569,
-      "learning_rate": 9.594015779571741e-06,
-      "loss": 0.4113,
-      "step": 6729
-    },
-    {
-      "epoch": 0.4398405332984772,
-      "grad_norm": 0.44042375683784485,
-      "learning_rate": 9.593877937468104e-06,
-      "loss": 0.3623,
-      "step": 6730
-    },
-    {
-      "epoch": 0.43990588850401935,
-      "grad_norm": 0.4320286810398102,
-      "learning_rate": 9.593740072958531e-06,
-      "loss": 0.3746,
-      "step": 6731
-    },
-    {
-      "epoch": 0.4399712437095615,
-      "grad_norm": 0.463545560836792,
-      "learning_rate": 9.593602186043698e-06,
-      "loss": 0.412,
-      "step": 6732
-    },
-    {
-      "epoch": 0.4400365989151036,
-      "grad_norm": 0.42265942692756653,
-      "learning_rate": 9.593464276724273e-06,
-      "loss": 0.3622,
-      "step": 6733
-    },
-    {
-      "epoch": 0.44010195412064573,
-      "grad_norm": 0.4521913230419159,
-      "learning_rate": 9.593326345000935e-06,
-      "loss": 0.3623,
-      "step": 6734
-    },
-    {
-      "epoch": 0.4401673093261878,
-      "grad_norm": 0.4236926734447479,
-      "learning_rate": 9.59318839087435e-06,
-      "loss": 0.3446,
-      "step": 6735
-    },
-    {
-      "epoch": 0.44023266453172993,
-      "grad_norm": 0.4713384509086609,
-      "learning_rate": 9.593050414345197e-06,
-      "loss": 0.4078,
-      "step": 6736
-    },
-    {
-      "epoch": 0.44029801973727206,
-      "grad_norm": 0.43656012415885925,
-      "learning_rate": 9.592912415414145e-06,
-      "loss": 0.3451,
-      "step": 6737
-    },
-    {
-      "epoch": 0.4403633749428142,
-      "grad_norm": 0.44159042835235596,
-      "learning_rate": 9.592774394081867e-06,
-      "loss": 0.374,
-      "step": 6738
-    },
-    {
-      "epoch": 0.4404287301483563,
-      "grad_norm": 0.47376754879951477,
-      "learning_rate": 9.592636350349036e-06,
-      "loss": 0.4055,
-      "step": 6739
-    },
-    {
-      "epoch": 0.44049408535389845,
-      "grad_norm": 0.48617663979530334,
-      "learning_rate": 9.592498284216328e-06,
-      "loss": 0.4265,
-      "step": 6740
-    },
-    {
-      "epoch": 0.4405594405594406,
-      "grad_norm": 0.47439900040626526,
-      "learning_rate": 9.592360195684417e-06,
-      "loss": 0.4048,
-      "step": 6741
-    },
-    {
-      "epoch": 0.4406247957649827,
-      "grad_norm": 0.4841964542865753,
-      "learning_rate": 9.59222208475397e-06,
-      "loss": 0.4097,
-      "step": 6742
-    },
-    {
-      "epoch": 0.44069015097052483,
-      "grad_norm": 0.4714210033416748,
-      "learning_rate": 9.592083951425668e-06,
-      "loss": 0.4126,
-      "step": 6743
-    },
-    {
-      "epoch": 0.4407555061760669,
-      "grad_norm": 0.46400946378707886,
-      "learning_rate": 9.591945795700181e-06,
-      "loss": 0.3896,
-      "step": 6744
-    },
-    {
-      "epoch": 0.44082086138160903,
-      "grad_norm": 0.47542211413383484,
-      "learning_rate": 9.591807617578184e-06,
-      "loss": 0.3663,
-      "step": 6745
-    },
-    {
-      "epoch": 0.44088621658715116,
-      "grad_norm": 0.4832114577293396,
-      "learning_rate": 9.59166941706035e-06,
-      "loss": 0.3592,
-      "step": 6746
-    },
-    {
-      "epoch": 0.4409515717926933,
-      "grad_norm": 0.4695150852203369,
-      "learning_rate": 9.591531194147352e-06,
-      "loss": 0.3695,
-      "step": 6747
-    },
-    {
-      "epoch": 0.4410169269982354,
-      "grad_norm": 0.4999988377094269,
-      "learning_rate": 9.591392948839867e-06,
-      "loss": 0.4373,
-      "step": 6748
-    },
-    {
-      "epoch": 0.44108228220377754,
-      "grad_norm": 0.45000162720680237,
-      "learning_rate": 9.591254681138568e-06,
-      "loss": 0.365,
-      "step": 6749
-    },
-    {
-      "epoch": 0.44114763740931967,
-      "grad_norm": 0.46372920274734497,
-      "learning_rate": 9.591116391044131e-06,
-      "loss": 0.3743,
-      "step": 6750
-    },
-    {
-      "epoch": 0.4412129926148618,
-      "grad_norm": 0.46051591634750366,
-      "learning_rate": 9.590978078557227e-06,
-      "loss": 0.3858,
-      "step": 6751
-    },
-    {
-      "epoch": 0.44127834782040387,
-      "grad_norm": 0.4539625942707062,
-      "learning_rate": 9.590839743678532e-06,
-      "loss": 0.3847,
-      "step": 6752
-    },
-    {
-      "epoch": 0.441343703025946,
-      "grad_norm": 0.4396856725215912,
-      "learning_rate": 9.590701386408723e-06,
-      "loss": 0.3595,
-      "step": 6753
-    },
-    {
-      "epoch": 0.4414090582314881,
-      "grad_norm": 0.4734814465045929,
-      "learning_rate": 9.590563006748472e-06,
-      "loss": 0.4586,
-      "step": 6754
-    },
-    {
-      "epoch": 0.44147441343703026,
-      "grad_norm": 0.4724050760269165,
-      "learning_rate": 9.590424604698455e-06,
-      "loss": 0.4052,
-      "step": 6755
-    },
-    {
-      "epoch": 0.4415397686425724,
-      "grad_norm": 0.44638800621032715,
-      "learning_rate": 9.590286180259347e-06,
-      "loss": 0.3669,
-      "step": 6756
-    },
-    {
-      "epoch": 0.4416051238481145,
-      "grad_norm": 0.44825056195259094,
-      "learning_rate": 9.590147733431823e-06,
-      "loss": 0.3405,
-      "step": 6757
-    },
-    {
-      "epoch": 0.44167047905365664,
-      "grad_norm": 0.46173352003097534,
-      "learning_rate": 9.590009264216557e-06,
-      "loss": 0.4144,
-      "step": 6758
-    },
-    {
-      "epoch": 0.44173583425919877,
-      "grad_norm": 0.4688032567501068,
-      "learning_rate": 9.589870772614227e-06,
-      "loss": 0.3929,
-      "step": 6759
-    },
-    {
-      "epoch": 0.44180118946474084,
-      "grad_norm": 0.4624279737472534,
-      "learning_rate": 9.589732258625508e-06,
-      "loss": 0.4148,
-      "step": 6760
-    },
-    {
-      "epoch": 0.44186654467028297,
-      "grad_norm": 0.5510392189025879,
-      "learning_rate": 9.589593722251074e-06,
-      "loss": 0.4373,
-      "step": 6761
-    },
-    {
-      "epoch": 0.4419318998758251,
-      "grad_norm": 0.46463918685913086,
-      "learning_rate": 9.589455163491601e-06,
-      "loss": 0.3678,
-      "step": 6762
-    },
-    {
-      "epoch": 0.4419972550813672,
-      "grad_norm": 0.47901713848114014,
-      "learning_rate": 9.589316582347766e-06,
-      "loss": 0.3773,
-      "step": 6763
-    },
-    {
-      "epoch": 0.44206261028690935,
-      "grad_norm": 0.43875572085380554,
-      "learning_rate": 9.589177978820244e-06,
-      "loss": 0.3904,
-      "step": 6764
-    },
-    {
-      "epoch": 0.4421279654924515,
-      "grad_norm": 0.4612356424331665,
-      "learning_rate": 9.58903935290971e-06,
-      "loss": 0.4024,
-      "step": 6765
-    },
-    {
-      "epoch": 0.4421933206979936,
-      "grad_norm": 0.4210243821144104,
-      "learning_rate": 9.588900704616841e-06,
-      "loss": 0.3601,
-      "step": 6766
-    },
-    {
-      "epoch": 0.44225867590353574,
-      "grad_norm": 0.46239984035491943,
-      "learning_rate": 9.588762033942316e-06,
-      "loss": 0.384,
-      "step": 6767
-    },
-    {
-      "epoch": 0.44232403110907786,
-      "grad_norm": 0.48965543508529663,
-      "learning_rate": 9.588623340886807e-06,
-      "loss": 0.4171,
-      "step": 6768
-    },
-    {
-      "epoch": 0.44238938631461994,
-      "grad_norm": 0.4603058099746704,
-      "learning_rate": 9.588484625450993e-06,
-      "loss": 0.4124,
-      "step": 6769
-    },
-    {
-      "epoch": 0.44245474152016206,
-      "grad_norm": 0.47310999035835266,
-      "learning_rate": 9.588345887635549e-06,
-      "loss": 0.4351,
-      "step": 6770
-    },
-    {
-      "epoch": 0.4425200967257042,
-      "grad_norm": 0.4399702847003937,
-      "learning_rate": 9.588207127441153e-06,
-      "loss": 0.342,
-      "step": 6771
-    },
-    {
-      "epoch": 0.4425854519312463,
-      "grad_norm": 0.4237600266933441,
-      "learning_rate": 9.588068344868482e-06,
-      "loss": 0.3306,
-      "step": 6772
-    },
-    {
-      "epoch": 0.44265080713678845,
-      "grad_norm": 0.4667479693889618,
-      "learning_rate": 9.587929539918212e-06,
-      "loss": 0.4316,
-      "step": 6773
-    },
-    {
-      "epoch": 0.4427161623423306,
-      "grad_norm": 0.42612531781196594,
-      "learning_rate": 9.587790712591018e-06,
-      "loss": 0.3685,
-      "step": 6774
-    },
-    {
-      "epoch": 0.4427815175478727,
-      "grad_norm": 0.46188637614250183,
-      "learning_rate": 9.587651862887582e-06,
-      "loss": 0.3896,
-      "step": 6775
-    },
-    {
-      "epoch": 0.44284687275341483,
-      "grad_norm": 0.43306100368499756,
-      "learning_rate": 9.587512990808578e-06,
-      "loss": 0.3615,
-      "step": 6776
-    },
-    {
-      "epoch": 0.4429122279589569,
-      "grad_norm": 0.4422586262226105,
-      "learning_rate": 9.587374096354685e-06,
-      "loss": 0.3931,
-      "step": 6777
-    },
-    {
-      "epoch": 0.44297758316449903,
-      "grad_norm": 0.47793295979499817,
-      "learning_rate": 9.587235179526578e-06,
-      "loss": 0.3986,
-      "step": 6778
-    },
-    {
-      "epoch": 0.44304293837004116,
-      "grad_norm": 0.4292960464954376,
-      "learning_rate": 9.587096240324935e-06,
-      "loss": 0.3387,
-      "step": 6779
-    },
-    {
-      "epoch": 0.4431082935755833,
-      "grad_norm": 0.4719371795654297,
-      "learning_rate": 9.586957278750436e-06,
-      "loss": 0.3838,
-      "step": 6780
-    },
-    {
-      "epoch": 0.4431736487811254,
-      "grad_norm": 0.45420241355895996,
-      "learning_rate": 9.586818294803756e-06,
-      "loss": 0.3659,
-      "step": 6781
-    },
-    {
-      "epoch": 0.44323900398666755,
-      "grad_norm": 0.42631033062934875,
-      "learning_rate": 9.586679288485575e-06,
-      "loss": 0.3518,
-      "step": 6782
-    },
-    {
-      "epoch": 0.4433043591922097,
-      "grad_norm": 0.5003683567047119,
-      "learning_rate": 9.58654025979657e-06,
-      "loss": 0.4471,
-      "step": 6783
-    },
-    {
-      "epoch": 0.4433697143977518,
-      "grad_norm": 0.4721042811870575,
-      "learning_rate": 9.586401208737419e-06,
-      "loss": 0.3909,
-      "step": 6784
-    },
-    {
-      "epoch": 0.44343506960329393,
-      "grad_norm": 0.4883996844291687,
-      "learning_rate": 9.5862621353088e-06,
-      "loss": 0.4348,
-      "step": 6785
-    },
-    {
-      "epoch": 0.443500424808836,
-      "grad_norm": 0.4878160059452057,
-      "learning_rate": 9.586123039511393e-06,
-      "loss": 0.437,
-      "step": 6786
-    },
-    {
-      "epoch": 0.44356578001437813,
-      "grad_norm": 0.47435104846954346,
-      "learning_rate": 9.585983921345875e-06,
-      "loss": 0.3554,
-      "step": 6787
-    },
-    {
-      "epoch": 0.44363113521992026,
-      "grad_norm": 0.4903028905391693,
-      "learning_rate": 9.585844780812922e-06,
-      "loss": 0.4556,
-      "step": 6788
-    },
-    {
-      "epoch": 0.4436964904254624,
-      "grad_norm": 0.44504037499427795,
-      "learning_rate": 9.58570561791322e-06,
-      "loss": 0.3913,
-      "step": 6789
-    },
-    {
-      "epoch": 0.4437618456310045,
-      "grad_norm": 0.4510614275932312,
-      "learning_rate": 9.58556643264744e-06,
-      "loss": 0.3747,
-      "step": 6790
-    },
-    {
-      "epoch": 0.44382720083654664,
-      "grad_norm": 0.48686614632606506,
-      "learning_rate": 9.585427225016264e-06,
-      "loss": 0.4173,
-      "step": 6791
-    },
-    {
-      "epoch": 0.44389255604208877,
-      "grad_norm": 0.4253038167953491,
-      "learning_rate": 9.585287995020371e-06,
-      "loss": 0.3401,
-      "step": 6792
-    },
-    {
-      "epoch": 0.4439579112476309,
-      "grad_norm": 0.4288231432437897,
-      "learning_rate": 9.58514874266044e-06,
-      "loss": 0.3617,
-      "step": 6793
-    },
-    {
-      "epoch": 0.44402326645317297,
-      "grad_norm": 0.4962444603443146,
-      "learning_rate": 9.58500946793715e-06,
-      "loss": 0.4344,
-      "step": 6794
-    },
-    {
-      "epoch": 0.4440886216587151,
-      "grad_norm": 0.3991558849811554,
-      "learning_rate": 9.584870170851182e-06,
-      "loss": 0.333,
-      "step": 6795
-    },
-    {
-      "epoch": 0.4441539768642572,
-      "grad_norm": 0.4364485740661621,
-      "learning_rate": 9.584730851403212e-06,
-      "loss": 0.3452,
-      "step": 6796
-    },
-    {
-      "epoch": 0.44421933206979936,
-      "grad_norm": 0.46209946274757385,
-      "learning_rate": 9.584591509593922e-06,
-      "loss": 0.4015,
-      "step": 6797
-    },
-    {
-      "epoch": 0.4442846872753415,
-      "grad_norm": 0.48259955644607544,
-      "learning_rate": 9.58445214542399e-06,
-      "loss": 0.3962,
-      "step": 6798
-    },
-    {
-      "epoch": 0.4443500424808836,
-      "grad_norm": 0.4641874432563782,
-      "learning_rate": 9.584312758894099e-06,
-      "loss": 0.3965,
-      "step": 6799
-    },
-    {
-      "epoch": 0.44441539768642574,
-      "grad_norm": 0.45316869020462036,
-      "learning_rate": 9.584173350004924e-06,
-      "loss": 0.4277,
-      "step": 6800
-    },
-    {
-      "epoch": 0.44448075289196787,
-      "grad_norm": 0.5300980806350708,
-      "learning_rate": 9.58403391875715e-06,
-      "loss": 0.4546,
-      "step": 6801
-    },
-    {
-      "epoch": 0.44454610809750994,
-      "grad_norm": 0.4747301936149597,
-      "learning_rate": 9.583894465151452e-06,
-      "loss": 0.3843,
-      "step": 6802
-    },
-    {
-      "epoch": 0.44461146330305207,
-      "grad_norm": 0.45713356137275696,
-      "learning_rate": 9.583754989188514e-06,
-      "loss": 0.3652,
-      "step": 6803
-    },
-    {
-      "epoch": 0.4446768185085942,
-      "grad_norm": 0.442258358001709,
-      "learning_rate": 9.583615490869014e-06,
-      "loss": 0.3867,
-      "step": 6804
-    },
-    {
-      "epoch": 0.4447421737141363,
-      "grad_norm": 0.488505095243454,
-      "learning_rate": 9.583475970193634e-06,
-      "loss": 0.434,
-      "step": 6805
-    },
-    {
-      "epoch": 0.44480752891967845,
-      "grad_norm": 0.4674488604068756,
-      "learning_rate": 9.583336427163054e-06,
-      "loss": 0.3893,
-      "step": 6806
-    },
-    {
-      "epoch": 0.4448728841252206,
-      "grad_norm": 0.4637359082698822,
-      "learning_rate": 9.583196861777955e-06,
-      "loss": 0.3814,
-      "step": 6807
-    },
-    {
-      "epoch": 0.4449382393307627,
-      "grad_norm": 0.42842769622802734,
-      "learning_rate": 9.583057274039016e-06,
-      "loss": 0.3652,
-      "step": 6808
-    },
-    {
-      "epoch": 0.44500359453630484,
-      "grad_norm": 0.4319152235984802,
-      "learning_rate": 9.58291766394692e-06,
-      "loss": 0.3533,
-      "step": 6809
-    },
-    {
-      "epoch": 0.44506894974184696,
-      "grad_norm": 0.530834436416626,
-      "learning_rate": 9.582778031502347e-06,
-      "loss": 0.4572,
-      "step": 6810
-    },
-    {
-      "epoch": 0.44513430494738904,
-      "grad_norm": 0.4620935022830963,
-      "learning_rate": 9.582638376705976e-06,
-      "loss": 0.3921,
-      "step": 6811
-    },
-    {
-      "epoch": 0.44519966015293116,
-      "grad_norm": 0.43870776891708374,
-      "learning_rate": 9.582498699558492e-06,
-      "loss": 0.3411,
-      "step": 6812
-    },
-    {
-      "epoch": 0.4452650153584733,
-      "grad_norm": 0.46446239948272705,
-      "learning_rate": 9.582359000060572e-06,
-      "loss": 0.4065,
-      "step": 6813
-    },
-    {
-      "epoch": 0.4453303705640154,
-      "grad_norm": 0.46167463064193726,
-      "learning_rate": 9.582219278212903e-06,
-      "loss": 0.4311,
-      "step": 6814
-    },
-    {
-      "epoch": 0.44539572576955755,
-      "grad_norm": 0.4565676152706146,
-      "learning_rate": 9.58207953401616e-06,
-      "loss": 0.4029,
-      "step": 6815
-    },
-    {
-      "epoch": 0.4454610809750997,
-      "grad_norm": 0.4476493000984192,
-      "learning_rate": 9.58193976747103e-06,
-      "loss": 0.402,
-      "step": 6816
-    },
-    {
-      "epoch": 0.4455264361806418,
-      "grad_norm": 0.4766305983066559,
-      "learning_rate": 9.581799978578191e-06,
-      "loss": 0.4243,
-      "step": 6817
-    },
-    {
-      "epoch": 0.44559179138618393,
-      "grad_norm": 0.5118014216423035,
-      "learning_rate": 9.581660167338327e-06,
-      "loss": 0.4625,
-      "step": 6818
-    },
-    {
-      "epoch": 0.445657146591726,
-      "grad_norm": 0.4302530884742737,
-      "learning_rate": 9.581520333752119e-06,
-      "loss": 0.3676,
-      "step": 6819
-    },
-    {
-      "epoch": 0.44572250179726813,
-      "grad_norm": 0.4763137698173523,
-      "learning_rate": 9.581380477820249e-06,
-      "loss": 0.4104,
-      "step": 6820
-    },
-    {
-      "epoch": 0.44578785700281026,
-      "grad_norm": 0.43492022156715393,
-      "learning_rate": 9.581240599543398e-06,
-      "loss": 0.3454,
-      "step": 6821
-    },
-    {
-      "epoch": 0.4458532122083524,
-      "grad_norm": 0.44110429286956787,
-      "learning_rate": 9.581100698922252e-06,
-      "loss": 0.3757,
-      "step": 6822
-    },
-    {
-      "epoch": 0.4459185674138945,
-      "grad_norm": 0.44938743114471436,
-      "learning_rate": 9.58096077595749e-06,
-      "loss": 0.3557,
-      "step": 6823
-    },
-    {
-      "epoch": 0.44598392261943665,
-      "grad_norm": 0.4719734489917755,
-      "learning_rate": 9.580820830649795e-06,
-      "loss": 0.4106,
-      "step": 6824
-    },
-    {
-      "epoch": 0.4460492778249788,
-      "grad_norm": 0.4498710334300995,
-      "learning_rate": 9.580680862999849e-06,
-      "loss": 0.4167,
-      "step": 6825
-    },
-    {
-      "epoch": 0.4461146330305209,
-      "grad_norm": 0.4458502531051636,
-      "learning_rate": 9.580540873008338e-06,
-      "loss": 0.3593,
-      "step": 6826
-    },
-    {
-      "epoch": 0.44617998823606303,
-      "grad_norm": 0.41527310013771057,
-      "learning_rate": 9.58040086067594e-06,
-      "loss": 0.3222,
-      "step": 6827
-    },
-    {
-      "epoch": 0.4462453434416051,
-      "grad_norm": 0.44922083616256714,
-      "learning_rate": 9.580260826003341e-06,
-      "loss": 0.3871,
-      "step": 6828
-    },
-    {
-      "epoch": 0.44631069864714723,
-      "grad_norm": 0.46734818816185,
-      "learning_rate": 9.58012076899122e-06,
-      "loss": 0.3843,
-      "step": 6829
-    },
-    {
-      "epoch": 0.44637605385268936,
-      "grad_norm": 0.4546533226966858,
-      "learning_rate": 9.579980689640268e-06,
-      "loss": 0.3803,
-      "step": 6830
-    },
-    {
-      "epoch": 0.4464414090582315,
-      "grad_norm": 0.4769834578037262,
-      "learning_rate": 9.579840587951161e-06,
-      "loss": 0.4091,
-      "step": 6831
-    },
-    {
-      "epoch": 0.4465067642637736,
-      "grad_norm": 0.4315880537033081,
-      "learning_rate": 9.579700463924586e-06,
-      "loss": 0.3267,
-      "step": 6832
-    },
-    {
-      "epoch": 0.44657211946931574,
-      "grad_norm": 0.44697728753089905,
-      "learning_rate": 9.579560317561225e-06,
-      "loss": 0.3975,
-      "step": 6833
-    },
-    {
-      "epoch": 0.44663747467485787,
-      "grad_norm": 0.45439666509628296,
-      "learning_rate": 9.579420148861763e-06,
-      "loss": 0.4024,
-      "step": 6834
-    },
-    {
-      "epoch": 0.4467028298804,
-      "grad_norm": 0.4275607168674469,
-      "learning_rate": 9.579279957826882e-06,
-      "loss": 0.3422,
-      "step": 6835
-    },
-    {
-      "epoch": 0.44676818508594207,
-      "grad_norm": 0.437131404876709,
-      "learning_rate": 9.579139744457264e-06,
-      "loss": 0.3626,
-      "step": 6836
-    },
-    {
-      "epoch": 0.4468335402914842,
-      "grad_norm": 0.42850175499916077,
-      "learning_rate": 9.578999508753597e-06,
-      "loss": 0.3785,
-      "step": 6837
-    },
-    {
-      "epoch": 0.4468988954970263,
-      "grad_norm": 0.47736141085624695,
-      "learning_rate": 9.578859250716562e-06,
-      "loss": 0.4251,
-      "step": 6838
-    },
-    {
-      "epoch": 0.44696425070256846,
-      "grad_norm": 0.41900646686553955,
-      "learning_rate": 9.578718970346846e-06,
-      "loss": 0.3303,
-      "step": 6839
-    },
-    {
-      "epoch": 0.4470296059081106,
-      "grad_norm": 0.4415460526943207,
-      "learning_rate": 9.578578667645128e-06,
-      "loss": 0.3121,
-      "step": 6840
-    },
-    {
-      "epoch": 0.4470949611136527,
-      "grad_norm": 0.44675517082214355,
-      "learning_rate": 9.5784383426121e-06,
-      "loss": 0.3917,
-      "step": 6841
-    },
-    {
-      "epoch": 0.44716031631919484,
-      "grad_norm": 0.43669185042381287,
-      "learning_rate": 9.57829799524844e-06,
-      "loss": 0.3253,
-      "step": 6842
-    },
-    {
-      "epoch": 0.44722567152473697,
-      "grad_norm": 0.45259734988212585,
-      "learning_rate": 9.578157625554833e-06,
-      "loss": 0.3789,
-      "step": 6843
-    },
-    {
-      "epoch": 0.44729102673027904,
-      "grad_norm": 0.44392073154449463,
-      "learning_rate": 9.578017233531968e-06,
-      "loss": 0.3581,
-      "step": 6844
-    },
-    {
-      "epoch": 0.44735638193582117,
-      "grad_norm": 0.49673157930374146,
-      "learning_rate": 9.577876819180525e-06,
-      "loss": 0.4376,
-      "step": 6845
-    },
-    {
-      "epoch": 0.4474217371413633,
-      "grad_norm": 0.43708527088165283,
-      "learning_rate": 9.577736382501192e-06,
-      "loss": 0.3545,
-      "step": 6846
-    },
-    {
-      "epoch": 0.4474870923469054,
-      "grad_norm": 0.45776787400245667,
-      "learning_rate": 9.577595923494651e-06,
-      "loss": 0.4155,
-      "step": 6847
-    },
-    {
-      "epoch": 0.44755244755244755,
-      "grad_norm": 0.47815778851509094,
-      "learning_rate": 9.577455442161591e-06,
-      "loss": 0.4418,
-      "step": 6848
-    },
-    {
-      "epoch": 0.4476178027579897,
-      "grad_norm": 0.45162200927734375,
-      "learning_rate": 9.577314938502696e-06,
-      "loss": 0.3803,
-      "step": 6849
-    },
-    {
-      "epoch": 0.4476831579635318,
-      "grad_norm": 0.4717872738838196,
-      "learning_rate": 9.577174412518648e-06,
-      "loss": 0.4355,
-      "step": 6850
-    },
-    {
-      "epoch": 0.44774851316907394,
-      "grad_norm": 0.45953986048698425,
-      "learning_rate": 9.577033864210135e-06,
-      "loss": 0.4222,
-      "step": 6851
-    },
-    {
-      "epoch": 0.44781386837461606,
-      "grad_norm": 0.4623877704143524,
-      "learning_rate": 9.576893293577842e-06,
-      "loss": 0.3616,
-      "step": 6852
-    },
-    {
-      "epoch": 0.44787922358015814,
-      "grad_norm": 0.5005372762680054,
-      "learning_rate": 9.576752700622455e-06,
-      "loss": 0.4262,
-      "step": 6853
-    },
-    {
-      "epoch": 0.44794457878570026,
-      "grad_norm": 0.4469951093196869,
-      "learning_rate": 9.57661208534466e-06,
-      "loss": 0.369,
-      "step": 6854
-    },
-    {
-      "epoch": 0.4480099339912424,
-      "grad_norm": 0.5558889508247375,
-      "learning_rate": 9.57647144774514e-06,
-      "loss": 0.503,
-      "step": 6855
-    },
-    {
-      "epoch": 0.4480752891967845,
-      "grad_norm": 0.47243532538414,
-      "learning_rate": 9.576330787824588e-06,
-      "loss": 0.4,
-      "step": 6856
-    },
-    {
-      "epoch": 0.44814064440232665,
-      "grad_norm": 0.4681210219860077,
-      "learning_rate": 9.576190105583683e-06,
-      "loss": 0.4047,
-      "step": 6857
-    },
-    {
-      "epoch": 0.4482059996078688,
-      "grad_norm": 0.48239651322364807,
-      "learning_rate": 9.576049401023112e-06,
-      "loss": 0.4473,
-      "step": 6858
-    },
-    {
-      "epoch": 0.4482713548134109,
-      "grad_norm": 0.44161128997802734,
-      "learning_rate": 9.575908674143564e-06,
-      "loss": 0.3642,
-      "step": 6859
-    },
-    {
-      "epoch": 0.44833671001895303,
-      "grad_norm": 0.5230162739753723,
-      "learning_rate": 9.575767924945725e-06,
-      "loss": 0.4412,
-      "step": 6860
-    },
-    {
-      "epoch": 0.4484020652244951,
-      "grad_norm": 0.4514910876750946,
-      "learning_rate": 9.57562715343028e-06,
-      "loss": 0.3841,
-      "step": 6861
-    },
-    {
-      "epoch": 0.44846742043003723,
-      "grad_norm": 0.44265827536582947,
-      "learning_rate": 9.575486359597916e-06,
-      "loss": 0.3373,
-      "step": 6862
-    },
-    {
-      "epoch": 0.44853277563557936,
-      "grad_norm": 0.4110846519470215,
-      "learning_rate": 9.57534554344932e-06,
-      "loss": 0.3182,
-      "step": 6863
-    },
-    {
-      "epoch": 0.4485981308411215,
-      "grad_norm": 0.43720105290412903,
-      "learning_rate": 9.575204704985178e-06,
-      "loss": 0.3838,
-      "step": 6864
-    },
-    {
-      "epoch": 0.4486634860466636,
-      "grad_norm": 0.44153308868408203,
-      "learning_rate": 9.57506384420618e-06,
-      "loss": 0.3737,
-      "step": 6865
-    },
-    {
-      "epoch": 0.44872884125220575,
-      "grad_norm": 0.44844532012939453,
-      "learning_rate": 9.574922961113009e-06,
-      "loss": 0.3858,
-      "step": 6866
-    },
-    {
-      "epoch": 0.4487941964577479,
-      "grad_norm": 0.4577677547931671,
-      "learning_rate": 9.574782055706353e-06,
-      "loss": 0.4042,
-      "step": 6867
-    },
-    {
-      "epoch": 0.44885955166329,
-      "grad_norm": 0.4581975042819977,
-      "learning_rate": 9.5746411279869e-06,
-      "loss": 0.3989,
-      "step": 6868
-    },
-    {
-      "epoch": 0.44892490686883213,
-      "grad_norm": 0.515884280204773,
-      "learning_rate": 9.574500177955338e-06,
-      "loss": 0.4982,
-      "step": 6869
-    },
-    {
-      "epoch": 0.4489902620743742,
-      "grad_norm": 0.43185150623321533,
-      "learning_rate": 9.574359205612356e-06,
-      "loss": 0.3413,
-      "step": 6870
-    },
-    {
-      "epoch": 0.44905561727991633,
-      "grad_norm": 0.4548414945602417,
-      "learning_rate": 9.574218210958638e-06,
-      "loss": 0.3875,
-      "step": 6871
-    },
-    {
-      "epoch": 0.44912097248545846,
-      "grad_norm": 0.45214909315109253,
-      "learning_rate": 9.574077193994873e-06,
-      "loss": 0.3822,
-      "step": 6872
-    },
-    {
-      "epoch": 0.4491863276910006,
-      "grad_norm": 0.481594443321228,
-      "learning_rate": 9.573936154721749e-06,
-      "loss": 0.3916,
-      "step": 6873
-    },
-    {
-      "epoch": 0.4492516828965427,
-      "grad_norm": 0.4700080156326294,
-      "learning_rate": 9.573795093139952e-06,
-      "loss": 0.4066,
-      "step": 6874
-    },
-    {
-      "epoch": 0.44931703810208484,
-      "grad_norm": 0.4320732057094574,
-      "learning_rate": 9.573654009250174e-06,
-      "loss": 0.3726,
-      "step": 6875
-    },
-    {
-      "epoch": 0.44938239330762697,
-      "grad_norm": 0.4457520544528961,
-      "learning_rate": 9.573512903053101e-06,
-      "loss": 0.407,
-      "step": 6876
-    },
-    {
-      "epoch": 0.4494477485131691,
-      "grad_norm": 0.531183660030365,
-      "learning_rate": 9.57337177454942e-06,
-      "loss": 0.4639,
-      "step": 6877
-    },
-    {
-      "epoch": 0.44951310371871117,
-      "grad_norm": 0.4703523814678192,
-      "learning_rate": 9.573230623739821e-06,
-      "loss": 0.3691,
-      "step": 6878
-    },
-    {
-      "epoch": 0.4495784589242533,
-      "grad_norm": 0.47748491168022156,
-      "learning_rate": 9.573089450624992e-06,
-      "loss": 0.4312,
-      "step": 6879
-    },
-    {
-      "epoch": 0.4496438141297954,
-      "grad_norm": 0.4636233150959015,
-      "learning_rate": 9.57294825520562e-06,
-      "loss": 0.3896,
-      "step": 6880
-    },
-    {
-      "epoch": 0.44970916933533756,
-      "grad_norm": 0.46522000432014465,
-      "learning_rate": 9.572807037482397e-06,
-      "loss": 0.4053,
-      "step": 6881
-    },
-    {
-      "epoch": 0.4497745245408797,
-      "grad_norm": 0.4766260087490082,
-      "learning_rate": 9.57266579745601e-06,
-      "loss": 0.4767,
-      "step": 6882
-    },
-    {
-      "epoch": 0.4498398797464218,
-      "grad_norm": 0.47308024764060974,
-      "learning_rate": 9.572524535127148e-06,
-      "loss": 0.4555,
-      "step": 6883
-    },
-    {
-      "epoch": 0.44990523495196394,
-      "grad_norm": 0.475248783826828,
-      "learning_rate": 9.572383250496498e-06,
-      "loss": 0.411,
-      "step": 6884
-    },
-    {
-      "epoch": 0.44997059015750607,
-      "grad_norm": 0.45154842734336853,
-      "learning_rate": 9.572241943564752e-06,
-      "loss": 0.3848,
-      "step": 6885
-    },
-    {
-      "epoch": 0.45003594536304814,
-      "grad_norm": 0.45776087045669556,
-      "learning_rate": 9.572100614332598e-06,
-      "loss": 0.4069,
-      "step": 6886
-    },
-    {
-      "epoch": 0.45010130056859027,
-      "grad_norm": 0.4290597140789032,
-      "learning_rate": 9.571959262800725e-06,
-      "loss": 0.3443,
-      "step": 6887
-    },
-    {
-      "epoch": 0.4501666557741324,
-      "grad_norm": 0.4693318009376526,
-      "learning_rate": 9.571817888969823e-06,
-      "loss": 0.4365,
-      "step": 6888
-    },
-    {
-      "epoch": 0.4502320109796745,
-      "grad_norm": 0.46669575572013855,
-      "learning_rate": 9.571676492840582e-06,
-      "loss": 0.3688,
-      "step": 6889
-    },
-    {
-      "epoch": 0.45029736618521665,
-      "grad_norm": 0.4279143214225769,
-      "learning_rate": 9.57153507441369e-06,
-      "loss": 0.3747,
-      "step": 6890
-    },
-    {
-      "epoch": 0.4503627213907588,
-      "grad_norm": 0.46470198035240173,
-      "learning_rate": 9.571393633689838e-06,
-      "loss": 0.3674,
-      "step": 6891
-    },
-    {
-      "epoch": 0.4504280765963009,
-      "grad_norm": 0.44927316904067993,
-      "learning_rate": 9.571252170669715e-06,
-      "loss": 0.3816,
-      "step": 6892
-    },
-    {
-      "epoch": 0.45049343180184304,
-      "grad_norm": 0.4965997636318207,
-      "learning_rate": 9.571110685354012e-06,
-      "loss": 0.3928,
-      "step": 6893
-    },
-    {
-      "epoch": 0.45055878700738516,
-      "grad_norm": 0.4474985897541046,
-      "learning_rate": 9.570969177743419e-06,
-      "loss": 0.3898,
-      "step": 6894
-    },
-    {
-      "epoch": 0.45062414221292724,
-      "grad_norm": 0.4462169110774994,
-      "learning_rate": 9.570827647838625e-06,
-      "loss": 0.3908,
-      "step": 6895
-    },
-    {
-      "epoch": 0.45068949741846936,
-      "grad_norm": 0.47301992774009705,
-      "learning_rate": 9.570686095640323e-06,
-      "loss": 0.4078,
-      "step": 6896
-    },
-    {
-      "epoch": 0.4507548526240115,
-      "grad_norm": 0.45238858461380005,
-      "learning_rate": 9.570544521149199e-06,
-      "loss": 0.3705,
-      "step": 6897
-    },
-    {
-      "epoch": 0.4508202078295536,
-      "grad_norm": 0.5131879448890686,
-      "learning_rate": 9.570402924365949e-06,
-      "loss": 0.4599,
-      "step": 6898
-    },
-    {
-      "epoch": 0.45088556303509575,
-      "grad_norm": 0.46553248167037964,
-      "learning_rate": 9.570261305291258e-06,
-      "loss": 0.4098,
-      "step": 6899
-    },
-    {
-      "epoch": 0.4509509182406379,
-      "grad_norm": 0.43932557106018066,
-      "learning_rate": 9.570119663925819e-06,
-      "loss": 0.3626,
-      "step": 6900
-    },
-    {
-      "epoch": 0.45101627344618,
-      "grad_norm": 0.45153963565826416,
-      "learning_rate": 9.569978000270325e-06,
-      "loss": 0.3823,
-      "step": 6901
-    },
-    {
-      "epoch": 0.45108162865172213,
-      "grad_norm": 1.0369046926498413,
-      "learning_rate": 9.569836314325463e-06,
-      "loss": 0.4856,
-      "step": 6902
-    },
-    {
-      "epoch": 0.4511469838572642,
-      "grad_norm": 0.4859563708305359,
-      "learning_rate": 9.569694606091928e-06,
-      "loss": 0.4273,
-      "step": 6903
-    },
-    {
-      "epoch": 0.45121233906280633,
-      "grad_norm": 0.41873952746391296,
-      "learning_rate": 9.56955287557041e-06,
-      "loss": 0.3501,
-      "step": 6904
-    },
-    {
-      "epoch": 0.45127769426834846,
-      "grad_norm": 0.42056483030319214,
-      "learning_rate": 9.569411122761597e-06,
-      "loss": 0.3593,
-      "step": 6905
-    },
-    {
-      "epoch": 0.4513430494738906,
-      "grad_norm": 0.45343002676963806,
-      "learning_rate": 9.569269347666185e-06,
-      "loss": 0.3415,
-      "step": 6906
-    },
-    {
-      "epoch": 0.4514084046794327,
-      "grad_norm": 0.49917858839035034,
-      "learning_rate": 9.569127550284863e-06,
-      "loss": 0.4187,
-      "step": 6907
-    },
-    {
-      "epoch": 0.45147375988497485,
-      "grad_norm": 0.43981149792671204,
-      "learning_rate": 9.568985730618321e-06,
-      "loss": 0.3795,
-      "step": 6908
-    },
-    {
-      "epoch": 0.451539115090517,
-      "grad_norm": 0.5186979174613953,
-      "learning_rate": 9.568843888667256e-06,
-      "loss": 0.3623,
-      "step": 6909
-    },
-    {
-      "epoch": 0.4516044702960591,
-      "grad_norm": 0.43448546528816223,
-      "learning_rate": 9.568702024432355e-06,
-      "loss": 0.36,
-      "step": 6910
-    },
-    {
-      "epoch": 0.45166982550160123,
-      "grad_norm": 0.45746171474456787,
-      "learning_rate": 9.568560137914312e-06,
-      "loss": 0.3569,
-      "step": 6911
-    },
-    {
-      "epoch": 0.4517351807071433,
-      "grad_norm": 0.5021383762359619,
-      "learning_rate": 9.568418229113816e-06,
-      "loss": 0.4158,
-      "step": 6912
-    },
-    {
-      "epoch": 0.45180053591268543,
-      "grad_norm": 0.4388960599899292,
-      "learning_rate": 9.568276298031565e-06,
-      "loss": 0.361,
-      "step": 6913
-    },
-    {
-      "epoch": 0.45186589111822756,
-      "grad_norm": 0.48843470215797424,
-      "learning_rate": 9.568134344668245e-06,
-      "loss": 0.3759,
-      "step": 6914
-    },
-    {
-      "epoch": 0.4519312463237697,
-      "grad_norm": 0.443772554397583,
-      "learning_rate": 9.567992369024554e-06,
-      "loss": 0.3917,
-      "step": 6915
-    },
-    {
-      "epoch": 0.4519966015293118,
-      "grad_norm": 0.48165953159332275,
-      "learning_rate": 9.56785037110118e-06,
-      "loss": 0.395,
-      "step": 6916
-    },
-    {
-      "epoch": 0.45206195673485394,
-      "grad_norm": 0.5079911351203918,
-      "learning_rate": 9.56770835089882e-06,
-      "loss": 0.4319,
-      "step": 6917
-    },
-    {
-      "epoch": 0.45212731194039607,
-      "grad_norm": 0.46010375022888184,
-      "learning_rate": 9.567566308418159e-06,
-      "loss": 0.4077,
-      "step": 6918
-    },
-    {
-      "epoch": 0.4521926671459382,
-      "grad_norm": 0.4574672281742096,
-      "learning_rate": 9.567424243659898e-06,
-      "loss": 0.4111,
-      "step": 6919
-    },
-    {
-      "epoch": 0.45225802235148027,
-      "grad_norm": 0.4674014449119568,
-      "learning_rate": 9.567282156624727e-06,
-      "loss": 0.3923,
-      "step": 6920
-    },
-    {
-      "epoch": 0.4523233775570224,
-      "grad_norm": 0.48392459750175476,
-      "learning_rate": 9.567140047313337e-06,
-      "loss": 0.446,
-      "step": 6921
-    },
-    {
-      "epoch": 0.4523887327625645,
-      "grad_norm": 0.4962601959705353,
-      "learning_rate": 9.566997915726423e-06,
-      "loss": 0.4486,
-      "step": 6922
-    },
-    {
-      "epoch": 0.45245408796810666,
-      "grad_norm": 0.4572274088859558,
-      "learning_rate": 9.566855761864679e-06,
-      "loss": 0.4068,
-      "step": 6923
-    },
-    {
-      "epoch": 0.4525194431736488,
-      "grad_norm": 0.4558427333831787,
-      "learning_rate": 9.566713585728797e-06,
-      "loss": 0.3974,
-      "step": 6924
-    },
-    {
-      "epoch": 0.4525847983791909,
-      "grad_norm": 0.47658196091651917,
-      "learning_rate": 9.566571387319473e-06,
-      "loss": 0.4158,
-      "step": 6925
-    },
-    {
-      "epoch": 0.45265015358473304,
-      "grad_norm": 0.47146621346473694,
-      "learning_rate": 9.566429166637395e-06,
-      "loss": 0.3965,
-      "step": 6926
-    },
-    {
-      "epoch": 0.45271550879027517,
-      "grad_norm": 0.44216328859329224,
-      "learning_rate": 9.566286923683261e-06,
-      "loss": 0.3713,
-      "step": 6927
-    },
-    {
-      "epoch": 0.45278086399581724,
-      "grad_norm": 0.4713079631328583,
-      "learning_rate": 9.566144658457763e-06,
-      "loss": 0.4071,
-      "step": 6928
-    },
-    {
-      "epoch": 0.45284621920135937,
-      "grad_norm": 0.489364355802536,
-      "learning_rate": 9.566002370961596e-06,
-      "loss": 0.4098,
-      "step": 6929
-    },
-    {
-      "epoch": 0.4529115744069015,
-      "grad_norm": 0.445444256067276,
-      "learning_rate": 9.565860061195455e-06,
-      "loss": 0.3858,
-      "step": 6930
-    },
-    {
-      "epoch": 0.4529769296124436,
-      "grad_norm": 0.46047458052635193,
-      "learning_rate": 9.56571772916003e-06,
-      "loss": 0.3969,
-      "step": 6931
-    },
-    {
-      "epoch": 0.45304228481798575,
-      "grad_norm": 0.4774382412433624,
-      "learning_rate": 9.565575374856022e-06,
-      "loss": 0.4024,
-      "step": 6932
-    },
-    {
-      "epoch": 0.4531076400235279,
-      "grad_norm": 0.463263601064682,
-      "learning_rate": 9.565432998284118e-06,
-      "loss": 0.4047,
-      "step": 6933
-    },
-    {
-      "epoch": 0.45317299522907,
-      "grad_norm": 0.46228039264678955,
-      "learning_rate": 9.565290599445016e-06,
-      "loss": 0.4126,
-      "step": 6934
-    },
-    {
-      "epoch": 0.45323835043461214,
-      "grad_norm": 0.476277232170105,
-      "learning_rate": 9.565148178339411e-06,
-      "loss": 0.4198,
-      "step": 6935
-    },
-    {
-      "epoch": 0.45330370564015426,
-      "grad_norm": 0.43697965145111084,
-      "learning_rate": 9.565005734967997e-06,
-      "loss": 0.3851,
-      "step": 6936
-    },
-    {
-      "epoch": 0.45336906084569634,
-      "grad_norm": 0.48467156291007996,
-      "learning_rate": 9.564863269331469e-06,
-      "loss": 0.4153,
-      "step": 6937
-    },
-    {
-      "epoch": 0.45343441605123846,
-      "grad_norm": 0.4783487021923065,
-      "learning_rate": 9.56472078143052e-06,
-      "loss": 0.3704,
-      "step": 6938
-    },
-    {
-      "epoch": 0.4534997712567806,
-      "grad_norm": 0.4710427224636078,
-      "learning_rate": 9.564578271265847e-06,
-      "loss": 0.4071,
-      "step": 6939
-    },
-    {
-      "epoch": 0.4535651264623227,
-      "grad_norm": 0.458793580532074,
-      "learning_rate": 9.564435738838144e-06,
-      "loss": 0.4282,
-      "step": 6940
-    },
-    {
-      "epoch": 0.45363048166786485,
-      "grad_norm": 0.4751000702381134,
-      "learning_rate": 9.564293184148109e-06,
-      "loss": 0.4182,
-      "step": 6941
-    },
-    {
-      "epoch": 0.453695836873407,
-      "grad_norm": 0.5245192646980286,
-      "learning_rate": 9.564150607196431e-06,
-      "loss": 0.4666,
-      "step": 6942
-    },
-    {
-      "epoch": 0.4537611920789491,
-      "grad_norm": 0.44270089268684387,
-      "learning_rate": 9.564008007983811e-06,
-      "loss": 0.3558,
-      "step": 6943
-    },
-    {
-      "epoch": 0.45382654728449123,
-      "grad_norm": 0.5179519057273865,
-      "learning_rate": 9.563865386510943e-06,
-      "loss": 0.4648,
-      "step": 6944
-    },
-    {
-      "epoch": 0.4538919024900333,
-      "grad_norm": 0.4951685070991516,
-      "learning_rate": 9.563722742778523e-06,
-      "loss": 0.4332,
-      "step": 6945
-    },
-    {
-      "epoch": 0.45395725769557543,
-      "grad_norm": 0.5189263224601746,
-      "learning_rate": 9.563580076787246e-06,
-      "loss": 0.4989,
-      "step": 6946
-    },
-    {
-      "epoch": 0.45402261290111756,
-      "grad_norm": 0.46859562397003174,
-      "learning_rate": 9.563437388537808e-06,
-      "loss": 0.4128,
-      "step": 6947
-    },
-    {
-      "epoch": 0.4540879681066597,
-      "grad_norm": 0.4332079589366913,
-      "learning_rate": 9.563294678030903e-06,
-      "loss": 0.3603,
-      "step": 6948
-    },
-    {
-      "epoch": 0.4541533233122018,
-      "grad_norm": 0.46565601229667664,
-      "learning_rate": 9.563151945267232e-06,
-      "loss": 0.4352,
-      "step": 6949
-    },
-    {
-      "epoch": 0.45421867851774395,
-      "grad_norm": 0.48951128125190735,
-      "learning_rate": 9.563009190247487e-06,
-      "loss": 0.4578,
-      "step": 6950
-    },
-    {
-      "epoch": 0.4542840337232861,
-      "grad_norm": 0.47698163986206055,
-      "learning_rate": 9.562866412972365e-06,
-      "loss": 0.3924,
-      "step": 6951
-    },
-    {
-      "epoch": 0.4543493889288282,
-      "grad_norm": 0.46763136982917786,
-      "learning_rate": 9.562723613442562e-06,
-      "loss": 0.4206,
-      "step": 6952
-    },
-    {
-      "epoch": 0.45441474413437033,
-      "grad_norm": 0.43223679065704346,
-      "learning_rate": 9.562580791658776e-06,
-      "loss": 0.372,
-      "step": 6953
-    },
-    {
-      "epoch": 0.4544800993399124,
-      "grad_norm": 0.4109695255756378,
-      "learning_rate": 9.562437947621703e-06,
-      "loss": 0.3387,
-      "step": 6954
-    },
-    {
-      "epoch": 0.45454545454545453,
-      "grad_norm": 0.4674420654773712,
-      "learning_rate": 9.56229508133204e-06,
-      "loss": 0.4166,
-      "step": 6955
-    },
-    {
-      "epoch": 0.45461080975099666,
-      "grad_norm": 0.4164307713508606,
-      "learning_rate": 9.562152192790482e-06,
-      "loss": 0.3337,
-      "step": 6956
-    },
-    {
-      "epoch": 0.4546761649565388,
-      "grad_norm": 0.45005887746810913,
-      "learning_rate": 9.562009281997728e-06,
-      "loss": 0.368,
-      "step": 6957
-    },
-    {
-      "epoch": 0.4547415201620809,
-      "grad_norm": 0.47211381793022156,
-      "learning_rate": 9.561866348954477e-06,
-      "loss": 0.3919,
-      "step": 6958
-    },
-    {
-      "epoch": 0.45480687536762304,
-      "grad_norm": 0.449897438287735,
-      "learning_rate": 9.56172339366142e-06,
-      "loss": 0.3737,
-      "step": 6959
-    },
-    {
-      "epoch": 0.45487223057316517,
-      "grad_norm": 0.46587154269218445,
-      "learning_rate": 9.561580416119259e-06,
-      "loss": 0.4175,
-      "step": 6960
-    },
-    {
-      "epoch": 0.4549375857787073,
-      "grad_norm": 0.479175865650177,
-      "learning_rate": 9.561437416328688e-06,
-      "loss": 0.4255,
-      "step": 6961
-    },
-    {
-      "epoch": 0.45500294098424937,
-      "grad_norm": 0.4773583710193634,
-      "learning_rate": 9.561294394290408e-06,
-      "loss": 0.3959,
-      "step": 6962
-    },
-    {
-      "epoch": 0.4550682961897915,
-      "grad_norm": 0.5353477001190186,
-      "learning_rate": 9.561151350005115e-06,
-      "loss": 0.507,
-      "step": 6963
-    },
-    {
-      "epoch": 0.4551336513953336,
-      "grad_norm": 0.4761268198490143,
-      "learning_rate": 9.561008283473507e-06,
-      "loss": 0.4126,
-      "step": 6964
-    },
-    {
-      "epoch": 0.45519900660087576,
-      "grad_norm": 0.43995970487594604,
-      "learning_rate": 9.560865194696282e-06,
-      "loss": 0.3706,
-      "step": 6965
-    },
-    {
-      "epoch": 0.4552643618064179,
-      "grad_norm": 0.49021244049072266,
-      "learning_rate": 9.560722083674136e-06,
-      "loss": 0.4448,
-      "step": 6966
-    },
-    {
-      "epoch": 0.45532971701196,
-      "grad_norm": 0.5180768370628357,
-      "learning_rate": 9.56057895040777e-06,
-      "loss": 0.4792,
-      "step": 6967
-    },
-    {
-      "epoch": 0.45539507221750214,
-      "grad_norm": 0.4723127484321594,
-      "learning_rate": 9.56043579489788e-06,
-      "loss": 0.4008,
-      "step": 6968
-    },
-    {
-      "epoch": 0.45546042742304427,
-      "grad_norm": 0.4562893509864807,
-      "learning_rate": 9.560292617145163e-06,
-      "loss": 0.4033,
-      "step": 6969
-    },
-    {
-      "epoch": 0.45552578262858634,
-      "grad_norm": 0.48076966404914856,
-      "learning_rate": 9.560149417150322e-06,
-      "loss": 0.4177,
-      "step": 6970
-    },
-    {
-      "epoch": 0.45559113783412847,
-      "grad_norm": 0.4708404541015625,
-      "learning_rate": 9.560006194914051e-06,
-      "loss": 0.4152,
-      "step": 6971
-    },
-    {
-      "epoch": 0.4556564930396706,
-      "grad_norm": 0.481901615858078,
-      "learning_rate": 9.55986295043705e-06,
-      "loss": 0.4467,
-      "step": 6972
-    },
-    {
-      "epoch": 0.4557218482452127,
-      "grad_norm": 0.47024330496788025,
-      "learning_rate": 9.559719683720017e-06,
-      "loss": 0.3817,
-      "step": 6973
-    },
-    {
-      "epoch": 0.45578720345075485,
-      "grad_norm": 0.46320924162864685,
-      "learning_rate": 9.559576394763652e-06,
-      "loss": 0.3843,
-      "step": 6974
-    },
-    {
-      "epoch": 0.455852558656297,
-      "grad_norm": 0.5211080312728882,
-      "learning_rate": 9.559433083568654e-06,
-      "loss": 0.4468,
-      "step": 6975
-    },
-    {
-      "epoch": 0.4559179138618391,
-      "grad_norm": 0.4338943660259247,
-      "learning_rate": 9.55928975013572e-06,
-      "loss": 0.3722,
-      "step": 6976
-    },
-    {
-      "epoch": 0.45598326906738124,
-      "grad_norm": 0.44695672392845154,
-      "learning_rate": 9.559146394465553e-06,
-      "loss": 0.3984,
-      "step": 6977
-    },
-    {
-      "epoch": 0.45604862427292336,
-      "grad_norm": 0.44805628061294556,
-      "learning_rate": 9.559003016558848e-06,
-      "loss": 0.3668,
-      "step": 6978
-    },
-    {
-      "epoch": 0.45611397947846544,
-      "grad_norm": 0.45200228691101074,
-      "learning_rate": 9.558859616416305e-06,
-      "loss": 0.3769,
-      "step": 6979
-    },
-    {
-      "epoch": 0.45617933468400756,
-      "grad_norm": 0.4438546299934387,
-      "learning_rate": 9.558716194038625e-06,
-      "loss": 0.387,
-      "step": 6980
-    },
-    {
-      "epoch": 0.4562446898895497,
-      "grad_norm": 0.4782806932926178,
-      "learning_rate": 9.558572749426507e-06,
-      "loss": 0.405,
-      "step": 6981
-    },
-    {
-      "epoch": 0.4563100450950918,
-      "grad_norm": 0.5054789185523987,
-      "learning_rate": 9.55842928258065e-06,
-      "loss": 0.4615,
-      "step": 6982
-    },
-    {
-      "epoch": 0.45637540030063395,
-      "grad_norm": 0.4915611147880554,
-      "learning_rate": 9.558285793501756e-06,
-      "loss": 0.4339,
-      "step": 6983
-    },
-    {
-      "epoch": 0.4564407555061761,
-      "grad_norm": 0.5091811418533325,
-      "learning_rate": 9.558142282190521e-06,
-      "loss": 0.4851,
-      "step": 6984
-    },
-    {
-      "epoch": 0.4565061107117182,
-      "grad_norm": 0.48367637395858765,
-      "learning_rate": 9.557998748647648e-06,
-      "loss": 0.3858,
-      "step": 6985
-    },
-    {
-      "epoch": 0.45657146591726033,
-      "grad_norm": 0.4622894525527954,
-      "learning_rate": 9.557855192873834e-06,
-      "loss": 0.4186,
-      "step": 6986
-    },
-    {
-      "epoch": 0.4566368211228024,
-      "grad_norm": 0.5128580331802368,
-      "learning_rate": 9.557711614869785e-06,
-      "loss": 0.4223,
-      "step": 6987
-    },
-    {
-      "epoch": 0.45670217632834453,
-      "grad_norm": 0.4722349941730499,
-      "learning_rate": 9.557568014636195e-06,
-      "loss": 0.4225,
-      "step": 6988
-    },
-    {
-      "epoch": 0.45676753153388666,
-      "grad_norm": 0.4965941905975342,
-      "learning_rate": 9.557424392173766e-06,
-      "loss": 0.3799,
-      "step": 6989
-    },
-    {
-      "epoch": 0.4568328867394288,
-      "grad_norm": 0.5437185168266296,
-      "learning_rate": 9.557280747483202e-06,
-      "loss": 0.3664,
-      "step": 6990
-    },
-    {
-      "epoch": 0.4568982419449709,
-      "grad_norm": 0.45322853326797485,
-      "learning_rate": 9.5571370805652e-06,
-      "loss": 0.4068,
-      "step": 6991
-    },
-    {
-      "epoch": 0.45696359715051305,
-      "grad_norm": 0.46104517579078674,
-      "learning_rate": 9.556993391420462e-06,
-      "loss": 0.3672,
-      "step": 6992
-    },
-    {
-      "epoch": 0.4570289523560552,
-      "grad_norm": 0.4782438278198242,
-      "learning_rate": 9.556849680049687e-06,
-      "loss": 0.3992,
-      "step": 6993
-    },
-    {
-      "epoch": 0.4570943075615973,
-      "grad_norm": 0.4755147397518158,
-      "learning_rate": 9.556705946453578e-06,
-      "loss": 0.3868,
-      "step": 6994
-    },
-    {
-      "epoch": 0.45715966276713943,
-      "grad_norm": 0.47630441188812256,
-      "learning_rate": 9.556562190632837e-06,
-      "loss": 0.4016,
-      "step": 6995
-    },
-    {
-      "epoch": 0.4572250179726815,
-      "grad_norm": 0.4549620747566223,
-      "learning_rate": 9.556418412588163e-06,
-      "loss": 0.378,
-      "step": 6996
-    },
-    {
-      "epoch": 0.45729037317822363,
-      "grad_norm": 0.4581489861011505,
-      "learning_rate": 9.556274612320257e-06,
-      "loss": 0.3912,
-      "step": 6997
-    },
-    {
-      "epoch": 0.45735572838376576,
-      "grad_norm": 0.4601094424724579,
-      "learning_rate": 9.556130789829821e-06,
-      "loss": 0.4044,
-      "step": 6998
-    },
-    {
-      "epoch": 0.4574210835893079,
-      "grad_norm": 0.448824942111969,
-      "learning_rate": 9.555986945117558e-06,
-      "loss": 0.3811,
-      "step": 6999
-    },
-    {
-      "epoch": 0.45748643879485,
-      "grad_norm": 0.484931617975235,
-      "learning_rate": 9.555843078184169e-06,
-      "loss": 0.3791,
-      "step": 7000
-    },
-    {
-      "epoch": 0.45755179400039214,
-      "grad_norm": 0.41543468832969666,
-      "learning_rate": 9.555699189030354e-06,
-      "loss": 0.3233,
-      "step": 7001
-    },
-    {
-      "epoch": 0.45761714920593427,
-      "grad_norm": 0.47216349840164185,
-      "learning_rate": 9.555555277656815e-06,
-      "loss": 0.4125,
-      "step": 7002
-    },
-    {
-      "epoch": 0.4576825044114764,
-      "grad_norm": 0.45830056071281433,
-      "learning_rate": 9.555411344064255e-06,
-      "loss": 0.3899,
-      "step": 7003
-    },
-    {
-      "epoch": 0.45774785961701847,
-      "grad_norm": 0.45786866545677185,
-      "learning_rate": 9.555267388253375e-06,
-      "loss": 0.3367,
-      "step": 7004
-    },
-    {
-      "epoch": 0.4578132148225606,
-      "grad_norm": 0.4865111708641052,
-      "learning_rate": 9.55512341022488e-06,
-      "loss": 0.3759,
-      "step": 7005
-    },
-    {
-      "epoch": 0.4578785700281027,
-      "grad_norm": 0.47519540786743164,
-      "learning_rate": 9.55497940997947e-06,
-      "loss": 0.3779,
-      "step": 7006
-    },
-    {
-      "epoch": 0.45794392523364486,
-      "grad_norm": 0.45519840717315674,
-      "learning_rate": 9.554835387517844e-06,
-      "loss": 0.4354,
-      "step": 7007
-    },
-    {
-      "epoch": 0.458009280439187,
-      "grad_norm": 0.47062450647354126,
-      "learning_rate": 9.554691342840711e-06,
-      "loss": 0.4141,
-      "step": 7008
-    },
-    {
-      "epoch": 0.4580746356447291,
-      "grad_norm": 0.4759996235370636,
-      "learning_rate": 9.554547275948772e-06,
-      "loss": 0.4266,
-      "step": 7009
-    },
-    {
-      "epoch": 0.45813999085027124,
-      "grad_norm": 0.4584622383117676,
-      "learning_rate": 9.554403186842725e-06,
-      "loss": 0.4054,
-      "step": 7010
-    },
-    {
-      "epoch": 0.45820534605581337,
-      "grad_norm": 0.46390944719314575,
-      "learning_rate": 9.554259075523276e-06,
-      "loss": 0.4432,
-      "step": 7011
-    },
-    {
-      "epoch": 0.45827070126135544,
-      "grad_norm": 0.4170962870121002,
-      "learning_rate": 9.554114941991128e-06,
-      "loss": 0.3316,
-      "step": 7012
-    },
-    {
-      "epoch": 0.45833605646689757,
-      "grad_norm": 0.4917965531349182,
-      "learning_rate": 9.553970786246985e-06,
-      "loss": 0.422,
-      "step": 7013
-    },
-    {
-      "epoch": 0.4584014116724397,
-      "grad_norm": 0.47754913568496704,
-      "learning_rate": 9.553826608291547e-06,
-      "loss": 0.3928,
-      "step": 7014
-    },
-    {
-      "epoch": 0.4584667668779818,
-      "grad_norm": 0.45876345038414,
-      "learning_rate": 9.553682408125521e-06,
-      "loss": 0.4065,
-      "step": 7015
-    },
-    {
-      "epoch": 0.45853212208352395,
-      "grad_norm": 0.47021669149398804,
-      "learning_rate": 9.553538185749607e-06,
-      "loss": 0.4149,
-      "step": 7016
-    },
-    {
-      "epoch": 0.4585974772890661,
-      "grad_norm": 0.47361770272254944,
-      "learning_rate": 9.55339394116451e-06,
-      "loss": 0.4048,
-      "step": 7017
-    },
-    {
-      "epoch": 0.4586628324946082,
-      "grad_norm": 0.4924563467502594,
-      "learning_rate": 9.553249674370935e-06,
-      "loss": 0.457,
-      "step": 7018
-    },
-    {
-      "epoch": 0.45872818770015034,
-      "grad_norm": 0.47624871134757996,
-      "learning_rate": 9.553105385369581e-06,
-      "loss": 0.4158,
-      "step": 7019
-    },
-    {
-      "epoch": 0.45879354290569246,
-      "grad_norm": 0.4325328469276428,
-      "learning_rate": 9.552961074161156e-06,
-      "loss": 0.3604,
-      "step": 7020
-    },
-    {
-      "epoch": 0.45885889811123454,
-      "grad_norm": 0.4883722960948944,
-      "learning_rate": 9.552816740746363e-06,
-      "loss": 0.4353,
-      "step": 7021
-    },
-    {
-      "epoch": 0.45892425331677666,
-      "grad_norm": 0.45907896757125854,
-      "learning_rate": 9.552672385125906e-06,
-      "loss": 0.3444,
-      "step": 7022
-    },
-    {
-      "epoch": 0.4589896085223188,
-      "grad_norm": 0.4705089330673218,
-      "learning_rate": 9.552528007300488e-06,
-      "loss": 0.4002,
-      "step": 7023
-    },
-    {
-      "epoch": 0.4590549637278609,
-      "grad_norm": 0.47056421637535095,
-      "learning_rate": 9.552383607270812e-06,
-      "loss": 0.3944,
-      "step": 7024
-    },
-    {
-      "epoch": 0.45912031893340305,
-      "grad_norm": 0.44806599617004395,
-      "learning_rate": 9.552239185037586e-06,
-      "loss": 0.3808,
-      "step": 7025
-    },
-    {
-      "epoch": 0.4591856741389452,
-      "grad_norm": 0.4405684769153595,
-      "learning_rate": 9.552094740601512e-06,
-      "loss": 0.3718,
-      "step": 7026
-    },
-    {
-      "epoch": 0.4592510293444873,
-      "grad_norm": 0.4540782570838928,
-      "learning_rate": 9.551950273963296e-06,
-      "loss": 0.3708,
-      "step": 7027
-    },
-    {
-      "epoch": 0.45931638455002943,
-      "grad_norm": 0.47009095549583435,
-      "learning_rate": 9.55180578512364e-06,
-      "loss": 0.4093,
-      "step": 7028
-    },
-    {
-      "epoch": 0.4593817397555715,
-      "grad_norm": 0.48366907238960266,
-      "learning_rate": 9.55166127408325e-06,
-      "loss": 0.3835,
-      "step": 7029
-    },
-    {
-      "epoch": 0.45944709496111363,
-      "grad_norm": 0.46774783730506897,
-      "learning_rate": 9.551516740842833e-06,
-      "loss": 0.3985,
-      "step": 7030
-    },
-    {
-      "epoch": 0.45951245016665576,
-      "grad_norm": 0.45185360312461853,
-      "learning_rate": 9.551372185403091e-06,
-      "loss": 0.3879,
-      "step": 7031
-    },
-    {
-      "epoch": 0.4595778053721979,
-      "grad_norm": 0.4647712707519531,
-      "learning_rate": 9.551227607764728e-06,
-      "loss": 0.4021,
-      "step": 7032
-    },
-    {
-      "epoch": 0.45964316057774,
-      "grad_norm": 0.4509907066822052,
-      "learning_rate": 9.551083007928455e-06,
-      "loss": 0.3703,
-      "step": 7033
-    },
-    {
-      "epoch": 0.45970851578328215,
-      "grad_norm": 0.4928545653820038,
-      "learning_rate": 9.550938385894973e-06,
-      "loss": 0.4177,
-      "step": 7034
-    },
-    {
-      "epoch": 0.4597738709888243,
-      "grad_norm": 0.6041052937507629,
-      "learning_rate": 9.550793741664985e-06,
-      "loss": 0.4794,
-      "step": 7035
-    },
-    {
-      "epoch": 0.4598392261943664,
-      "grad_norm": 0.4401216506958008,
-      "learning_rate": 9.550649075239203e-06,
-      "loss": 0.3415,
-      "step": 7036
-    },
-    {
-      "epoch": 0.45990458139990853,
-      "grad_norm": 0.47015464305877686,
-      "learning_rate": 9.550504386618326e-06,
-      "loss": 0.382,
-      "step": 7037
-    },
-    {
-      "epoch": 0.4599699366054506,
-      "grad_norm": 0.5325417518615723,
-      "learning_rate": 9.550359675803064e-06,
-      "loss": 0.4573,
-      "step": 7038
-    },
-    {
-      "epoch": 0.46003529181099273,
-      "grad_norm": 0.4384169578552246,
-      "learning_rate": 9.55021494279412e-06,
-      "loss": 0.3669,
-      "step": 7039
-    },
-    {
-      "epoch": 0.46010064701653486,
-      "grad_norm": 0.4498749077320099,
-      "learning_rate": 9.550070187592204e-06,
-      "loss": 0.3737,
-      "step": 7040
-    },
-    {
-      "epoch": 0.460166002222077,
-      "grad_norm": 0.49636998772621155,
-      "learning_rate": 9.549925410198017e-06,
-      "loss": 0.4166,
-      "step": 7041
-    },
-    {
-      "epoch": 0.4602313574276191,
-      "grad_norm": 0.4357587397098541,
-      "learning_rate": 9.549780610612269e-06,
-      "loss": 0.3362,
-      "step": 7042
-    },
-    {
-      "epoch": 0.46029671263316124,
-      "grad_norm": 0.4485979676246643,
-      "learning_rate": 9.549635788835665e-06,
-      "loss": 0.2888,
-      "step": 7043
-    },
-    {
-      "epoch": 0.46036206783870337,
-      "grad_norm": 0.4663568139076233,
-      "learning_rate": 9.549490944868908e-06,
-      "loss": 0.4285,
-      "step": 7044
-    },
-    {
-      "epoch": 0.4604274230442455,
-      "grad_norm": 0.4595623016357422,
-      "learning_rate": 9.549346078712711e-06,
-      "loss": 0.3949,
-      "step": 7045
-    },
-    {
-      "epoch": 0.46049277824978757,
-      "grad_norm": 0.45535972714424133,
-      "learning_rate": 9.549201190367776e-06,
-      "loss": 0.4061,
-      "step": 7046
-    },
-    {
-      "epoch": 0.4605581334553297,
-      "grad_norm": 0.4724358320236206,
-      "learning_rate": 9.549056279834811e-06,
-      "loss": 0.415,
-      "step": 7047
-    },
-    {
-      "epoch": 0.4606234886608718,
-      "grad_norm": 0.47900664806365967,
-      "learning_rate": 9.548911347114523e-06,
-      "loss": 0.416,
-      "step": 7048
-    },
-    {
-      "epoch": 0.46068884386641396,
-      "grad_norm": 0.45628440380096436,
-      "learning_rate": 9.548766392207618e-06,
-      "loss": 0.428,
-      "step": 7049
-    },
-    {
-      "epoch": 0.4607541990719561,
-      "grad_norm": 0.4388725459575653,
-      "learning_rate": 9.548621415114802e-06,
-      "loss": 0.3626,
-      "step": 7050
-    },
-    {
-      "epoch": 0.4608195542774982,
-      "grad_norm": 0.4458193778991699,
-      "learning_rate": 9.548476415836788e-06,
-      "loss": 0.3527,
-      "step": 7051
-    },
-    {
-      "epoch": 0.46088490948304034,
-      "grad_norm": 0.4481854736804962,
-      "learning_rate": 9.548331394374276e-06,
-      "loss": 0.3717,
-      "step": 7052
-    },
-    {
-      "epoch": 0.46095026468858247,
-      "grad_norm": 0.4708429276943207,
-      "learning_rate": 9.548186350727974e-06,
-      "loss": 0.3935,
-      "step": 7053
-    },
-    {
-      "epoch": 0.46101561989412454,
-      "grad_norm": 0.4286001920700073,
-      "learning_rate": 9.548041284898595e-06,
-      "loss": 0.3441,
-      "step": 7054
-    },
-    {
-      "epoch": 0.46108097509966667,
-      "grad_norm": 0.48979654908180237,
-      "learning_rate": 9.54789619688684e-06,
-      "loss": 0.419,
-      "step": 7055
-    },
-    {
-      "epoch": 0.4611463303052088,
-      "grad_norm": 0.4655849039554596,
-      "learning_rate": 9.547751086693422e-06,
-      "loss": 0.4043,
-      "step": 7056
-    },
-    {
-      "epoch": 0.4612116855107509,
-      "grad_norm": 0.4343613088130951,
-      "learning_rate": 9.547605954319045e-06,
-      "loss": 0.3885,
-      "step": 7057
-    },
-    {
-      "epoch": 0.46127704071629305,
-      "grad_norm": 0.46460336446762085,
-      "learning_rate": 9.547460799764418e-06,
-      "loss": 0.4341,
-      "step": 7058
-    },
-    {
-      "epoch": 0.4613423959218352,
-      "grad_norm": 0.451212078332901,
-      "learning_rate": 9.547315623030251e-06,
-      "loss": 0.3995,
-      "step": 7059
-    },
-    {
-      "epoch": 0.4614077511273773,
-      "grad_norm": 0.4431312382221222,
-      "learning_rate": 9.547170424117249e-06,
-      "loss": 0.3942,
-      "step": 7060
-    },
-    {
-      "epoch": 0.46147310633291944,
-      "grad_norm": 0.4502315819263458,
-      "learning_rate": 9.547025203026122e-06,
-      "loss": 0.4037,
-      "step": 7061
-    },
-    {
-      "epoch": 0.46153846153846156,
-      "grad_norm": 0.47694632411003113,
-      "learning_rate": 9.546879959757578e-06,
-      "loss": 0.3787,
-      "step": 7062
-    },
-    {
-      "epoch": 0.46160381674400364,
-      "grad_norm": 0.44212377071380615,
-      "learning_rate": 9.546734694312325e-06,
-      "loss": 0.4185,
-      "step": 7063
-    },
-    {
-      "epoch": 0.46166917194954576,
-      "grad_norm": 0.4314897656440735,
-      "learning_rate": 9.546589406691073e-06,
-      "loss": 0.3314,
-      "step": 7064
-    },
-    {
-      "epoch": 0.4617345271550879,
-      "grad_norm": 0.4295647442340851,
-      "learning_rate": 9.546444096894527e-06,
-      "loss": 0.396,
-      "step": 7065
-    },
-    {
-      "epoch": 0.46179988236063,
-      "grad_norm": 0.44813117384910583,
-      "learning_rate": 9.546298764923401e-06,
-      "loss": 0.3634,
-      "step": 7066
-    },
-    {
-      "epoch": 0.46186523756617215,
-      "grad_norm": 0.46085724234580994,
-      "learning_rate": 9.546153410778397e-06,
-      "loss": 0.393,
-      "step": 7067
-    },
-    {
-      "epoch": 0.4619305927717143,
-      "grad_norm": 0.4404080808162689,
-      "learning_rate": 9.546008034460233e-06,
-      "loss": 0.393,
-      "step": 7068
-    },
-    {
-      "epoch": 0.4619959479772564,
-      "grad_norm": 0.4515850841999054,
-      "learning_rate": 9.54586263596961e-06,
-      "loss": 0.3802,
-      "step": 7069
-    },
-    {
-      "epoch": 0.46206130318279853,
-      "grad_norm": 0.48453107476234436,
-      "learning_rate": 9.54571721530724e-06,
-      "loss": 0.4598,
-      "step": 7070
-    },
-    {
-      "epoch": 0.4621266583883406,
-      "grad_norm": 0.44939786195755005,
-      "learning_rate": 9.545571772473832e-06,
-      "loss": 0.4033,
-      "step": 7071
-    },
-    {
-      "epoch": 0.46219201359388273,
-      "grad_norm": 0.46518874168395996,
-      "learning_rate": 9.545426307470096e-06,
-      "loss": 0.4157,
-      "step": 7072
-    },
-    {
-      "epoch": 0.46225736879942486,
-      "grad_norm": 0.48767927289009094,
-      "learning_rate": 9.545280820296742e-06,
-      "loss": 0.4307,
-      "step": 7073
-    },
-    {
-      "epoch": 0.462322724004967,
-      "grad_norm": 0.4950529932975769,
-      "learning_rate": 9.545135310954479e-06,
-      "loss": 0.4363,
-      "step": 7074
-    },
-    {
-      "epoch": 0.4623880792105091,
-      "grad_norm": 0.45484399795532227,
-      "learning_rate": 9.544989779444017e-06,
-      "loss": 0.4088,
-      "step": 7075
-    },
-    {
-      "epoch": 0.46245343441605125,
-      "grad_norm": 0.4709393382072449,
-      "learning_rate": 9.544844225766064e-06,
-      "loss": 0.4355,
-      "step": 7076
-    },
-    {
-      "epoch": 0.4625187896215934,
-      "grad_norm": 0.46647417545318604,
-      "learning_rate": 9.54469864992133e-06,
-      "loss": 0.3886,
-      "step": 7077
-    },
-    {
-      "epoch": 0.4625841448271355,
-      "grad_norm": 0.4877423048019409,
-      "learning_rate": 9.544553051910527e-06,
-      "loss": 0.4058,
-      "step": 7078
-    },
-    {
-      "epoch": 0.46264950003267763,
-      "grad_norm": 0.44288796186447144,
-      "learning_rate": 9.544407431734366e-06,
-      "loss": 0.3706,
-      "step": 7079
-    },
-    {
-      "epoch": 0.4627148552382197,
-      "grad_norm": 0.41956257820129395,
-      "learning_rate": 9.544261789393554e-06,
-      "loss": 0.3172,
-      "step": 7080
-    },
-    {
-      "epoch": 0.46278021044376183,
-      "grad_norm": 0.46104365587234497,
-      "learning_rate": 9.544116124888806e-06,
-      "loss": 0.4235,
-      "step": 7081
-    },
-    {
-      "epoch": 0.46284556564930396,
-      "grad_norm": 0.4715181589126587,
-      "learning_rate": 9.543970438220825e-06,
-      "loss": 0.418,
-      "step": 7082
-    },
-    {
-      "epoch": 0.4629109208548461,
-      "grad_norm": 0.47676488757133484,
-      "learning_rate": 9.543824729390329e-06,
-      "loss": 0.3958,
-      "step": 7083
-    },
-    {
-      "epoch": 0.4629762760603882,
-      "grad_norm": 0.4482547640800476,
-      "learning_rate": 9.543678998398024e-06,
-      "loss": 0.4033,
-      "step": 7084
-    },
-    {
-      "epoch": 0.46304163126593034,
-      "grad_norm": 0.4512160122394562,
-      "learning_rate": 9.543533245244624e-06,
-      "loss": 0.4038,
-      "step": 7085
-    },
-    {
-      "epoch": 0.46310698647147247,
-      "grad_norm": 0.4391566514968872,
-      "learning_rate": 9.543387469930835e-06,
-      "loss": 0.3776,
-      "step": 7086
-    },
-    {
-      "epoch": 0.4631723416770146,
-      "grad_norm": 0.43100181221961975,
-      "learning_rate": 9.543241672457376e-06,
-      "loss": 0.3627,
-      "step": 7087
-    },
-    {
-      "epoch": 0.46323769688255667,
-      "grad_norm": 0.48517122864723206,
-      "learning_rate": 9.54309585282495e-06,
-      "loss": 0.4069,
-      "step": 7088
-    },
-    {
-      "epoch": 0.4633030520880988,
-      "grad_norm": 0.5059811472892761,
-      "learning_rate": 9.542950011034273e-06,
-      "loss": 0.4317,
-      "step": 7089
-    },
-    {
-      "epoch": 0.4633684072936409,
-      "grad_norm": 0.4587441384792328,
-      "learning_rate": 9.542804147086055e-06,
-      "loss": 0.4202,
-      "step": 7090
-    },
-    {
-      "epoch": 0.46343376249918306,
-      "grad_norm": 0.4466225802898407,
-      "learning_rate": 9.542658260981008e-06,
-      "loss": 0.4004,
-      "step": 7091
-    },
-    {
-      "epoch": 0.4634991177047252,
-      "grad_norm": 0.4705326557159424,
-      "learning_rate": 9.542512352719842e-06,
-      "loss": 0.3731,
-      "step": 7092
-    },
-    {
-      "epoch": 0.4635644729102673,
-      "grad_norm": 0.4367637634277344,
-      "learning_rate": 9.542366422303269e-06,
-      "loss": 0.3374,
-      "step": 7093
-    },
-    {
-      "epoch": 0.46362982811580944,
-      "grad_norm": 0.4710226058959961,
-      "learning_rate": 9.542220469732004e-06,
-      "loss": 0.4031,
-      "step": 7094
-    },
-    {
-      "epoch": 0.46369518332135157,
-      "grad_norm": 0.4866265654563904,
-      "learning_rate": 9.542074495006754e-06,
-      "loss": 0.4594,
-      "step": 7095
-    },
-    {
-      "epoch": 0.46376053852689364,
-      "grad_norm": 0.47064992785453796,
-      "learning_rate": 9.541928498128232e-06,
-      "loss": 0.3811,
-      "step": 7096
-    },
-    {
-      "epoch": 0.46382589373243577,
-      "grad_norm": 0.4414912462234497,
-      "learning_rate": 9.541782479097152e-06,
-      "loss": 0.3543,
-      "step": 7097
-    },
-    {
-      "epoch": 0.4638912489379779,
-      "grad_norm": 0.4506942331790924,
-      "learning_rate": 9.541636437914228e-06,
-      "loss": 0.3733,
-      "step": 7098
-    },
-    {
-      "epoch": 0.46395660414352,
-      "grad_norm": 0.5158799886703491,
-      "learning_rate": 9.541490374580166e-06,
-      "loss": 0.4275,
-      "step": 7099
-    },
-    {
-      "epoch": 0.46402195934906215,
-      "grad_norm": 0.48787835240364075,
-      "learning_rate": 9.541344289095686e-06,
-      "loss": 0.4212,
-      "step": 7100
-    },
-    {
-      "epoch": 0.4640873145546043,
-      "grad_norm": 0.43957990407943726,
-      "learning_rate": 9.541198181461493e-06,
-      "loss": 0.405,
-      "step": 7101
-    },
-    {
-      "epoch": 0.4641526697601464,
-      "grad_norm": 0.48659685254096985,
-      "learning_rate": 9.541052051678306e-06,
-      "loss": 0.4616,
-      "step": 7102
-    },
-    {
-      "epoch": 0.46421802496568854,
-      "grad_norm": 0.5065358877182007,
-      "learning_rate": 9.540905899746832e-06,
-      "loss": 0.4145,
-      "step": 7103
-    },
-    {
-      "epoch": 0.46428338017123066,
-      "grad_norm": 0.45447084307670593,
-      "learning_rate": 9.54075972566779e-06,
-      "loss": 0.3754,
-      "step": 7104
-    },
-    {
-      "epoch": 0.46434873537677274,
-      "grad_norm": 0.46564382314682007,
-      "learning_rate": 9.54061352944189e-06,
-      "loss": 0.3861,
-      "step": 7105
-    },
-    {
-      "epoch": 0.46441409058231486,
-      "grad_norm": 0.47878503799438477,
-      "learning_rate": 9.54046731106984e-06,
-      "loss": 0.3877,
-      "step": 7106
-    },
-    {
-      "epoch": 0.464479445787857,
-      "grad_norm": 0.4690788984298706,
-      "learning_rate": 9.540321070552362e-06,
-      "loss": 0.404,
-      "step": 7107
-    },
-    {
-      "epoch": 0.4645448009933991,
-      "grad_norm": 0.4505467712879181,
-      "learning_rate": 9.540174807890165e-06,
-      "loss": 0.3374,
-      "step": 7108
-    },
-    {
-      "epoch": 0.46461015619894125,
-      "grad_norm": 0.5293868184089661,
-      "learning_rate": 9.540028523083962e-06,
-      "loss": 0.3617,
-      "step": 7109
-    },
-    {
-      "epoch": 0.4646755114044834,
-      "grad_norm": 0.4439355432987213,
-      "learning_rate": 9.539882216134467e-06,
-      "loss": 0.3704,
-      "step": 7110
-    },
-    {
-      "epoch": 0.4647408666100255,
-      "grad_norm": 0.47333186864852905,
-      "learning_rate": 9.539735887042395e-06,
-      "loss": 0.398,
-      "step": 7111
-    },
-    {
-      "epoch": 0.46480622181556763,
-      "grad_norm": 0.4669288396835327,
-      "learning_rate": 9.539589535808456e-06,
-      "loss": 0.4264,
-      "step": 7112
-    },
-    {
-      "epoch": 0.4648715770211097,
-      "grad_norm": 0.4460848271846771,
-      "learning_rate": 9.539443162433367e-06,
-      "loss": 0.385,
-      "step": 7113
-    },
-    {
-      "epoch": 0.46493693222665183,
-      "grad_norm": 0.4974602460861206,
-      "learning_rate": 9.539296766917841e-06,
-      "loss": 0.4145,
-      "step": 7114
-    },
-    {
-      "epoch": 0.46500228743219396,
-      "grad_norm": 0.4577859044075012,
-      "learning_rate": 9.539150349262592e-06,
-      "loss": 0.4027,
-      "step": 7115
-    },
-    {
-      "epoch": 0.4650676426377361,
-      "grad_norm": 0.48755016922950745,
-      "learning_rate": 9.539003909468335e-06,
-      "loss": 0.4269,
-      "step": 7116
-    },
-    {
-      "epoch": 0.4651329978432782,
-      "grad_norm": 0.4554654657840729,
-      "learning_rate": 9.538857447535784e-06,
-      "loss": 0.3876,
-      "step": 7117
-    },
-    {
-      "epoch": 0.46519835304882035,
-      "grad_norm": 0.49087807536125183,
-      "learning_rate": 9.538710963465652e-06,
-      "loss": 0.4259,
-      "step": 7118
-    },
-    {
-      "epoch": 0.4652637082543625,
-      "grad_norm": 0.4510446786880493,
-      "learning_rate": 9.538564457258653e-06,
-      "loss": 0.3821,
-      "step": 7119
-    },
-    {
-      "epoch": 0.4653290634599046,
-      "grad_norm": 0.5753329992294312,
-      "learning_rate": 9.538417928915504e-06,
-      "loss": 0.4437,
-      "step": 7120
-    },
-    {
-      "epoch": 0.46539441866544673,
-      "grad_norm": 0.4266151487827301,
-      "learning_rate": 9.538271378436918e-06,
-      "loss": 0.3637,
-      "step": 7121
-    },
-    {
-      "epoch": 0.4654597738709888,
-      "grad_norm": 0.44344064593315125,
-      "learning_rate": 9.538124805823612e-06,
-      "loss": 0.3848,
-      "step": 7122
-    },
-    {
-      "epoch": 0.46552512907653093,
-      "grad_norm": 0.46759217977523804,
-      "learning_rate": 9.537978211076298e-06,
-      "loss": 0.4182,
-      "step": 7123
-    },
-    {
-      "epoch": 0.46559048428207306,
-      "grad_norm": 0.44066500663757324,
-      "learning_rate": 9.537831594195693e-06,
-      "loss": 0.3965,
-      "step": 7124
-    },
-    {
-      "epoch": 0.4656558394876152,
-      "grad_norm": 0.48578962683677673,
-      "learning_rate": 9.537684955182508e-06,
-      "loss": 0.3876,
-      "step": 7125
-    },
-    {
-      "epoch": 0.4657211946931573,
-      "grad_norm": 0.47319337725639343,
-      "learning_rate": 9.537538294037464e-06,
-      "loss": 0.404,
-      "step": 7126
-    },
-    {
-      "epoch": 0.46578654989869944,
-      "grad_norm": 0.4367903470993042,
-      "learning_rate": 9.537391610761275e-06,
-      "loss": 0.3791,
-      "step": 7127
-    },
-    {
-      "epoch": 0.46585190510424157,
-      "grad_norm": 0.4693010151386261,
-      "learning_rate": 9.537244905354655e-06,
-      "loss": 0.4292,
-      "step": 7128
-    },
-    {
-      "epoch": 0.4659172603097837,
-      "grad_norm": 0.44172173738479614,
-      "learning_rate": 9.537098177818318e-06,
-      "loss": 0.3718,
-      "step": 7129
-    },
-    {
-      "epoch": 0.46598261551532577,
-      "grad_norm": 0.47351235151290894,
-      "learning_rate": 9.53695142815298e-06,
-      "loss": 0.4046,
-      "step": 7130
-    },
-    {
-      "epoch": 0.4660479707208679,
-      "grad_norm": 0.4396744966506958,
-      "learning_rate": 9.536804656359362e-06,
-      "loss": 0.3659,
-      "step": 7131
-    },
-    {
-      "epoch": 0.46611332592641,
-      "grad_norm": 0.44549560546875,
-      "learning_rate": 9.536657862438173e-06,
-      "loss": 0.3389,
-      "step": 7132
-    },
-    {
-      "epoch": 0.46617868113195216,
-      "grad_norm": 0.4338420033454895,
-      "learning_rate": 9.536511046390134e-06,
-      "loss": 0.3673,
-      "step": 7133
-    },
-    {
-      "epoch": 0.4662440363374943,
-      "grad_norm": 0.46013087034225464,
-      "learning_rate": 9.536364208215957e-06,
-      "loss": 0.3877,
-      "step": 7134
-    },
-    {
-      "epoch": 0.4663093915430364,
-      "grad_norm": 0.47175872325897217,
-      "learning_rate": 9.536217347916362e-06,
-      "loss": 0.4194,
-      "step": 7135
-    },
-    {
-      "epoch": 0.46637474674857854,
-      "grad_norm": 0.4527713358402252,
-      "learning_rate": 9.536070465492062e-06,
-      "loss": 0.3947,
-      "step": 7136
-    },
-    {
-      "epoch": 0.46644010195412067,
-      "grad_norm": 0.44613954424858093,
-      "learning_rate": 9.535923560943776e-06,
-      "loss": 0.3834,
-      "step": 7137
-    },
-    {
-      "epoch": 0.46650545715966274,
-      "grad_norm": 0.4772361218929291,
-      "learning_rate": 9.53577663427222e-06,
-      "loss": 0.3975,
-      "step": 7138
-    },
-    {
-      "epoch": 0.46657081236520487,
-      "grad_norm": 0.5101822018623352,
-      "learning_rate": 9.53562968547811e-06,
-      "loss": 0.4177,
-      "step": 7139
-    },
-    {
-      "epoch": 0.466636167570747,
-      "grad_norm": 0.4716101586818695,
-      "learning_rate": 9.53548271456216e-06,
-      "loss": 0.3683,
-      "step": 7140
-    },
-    {
-      "epoch": 0.4667015227762891,
-      "grad_norm": 0.4602915346622467,
-      "learning_rate": 9.535335721525091e-06,
-      "loss": 0.38,
-      "step": 7141
-    },
-    {
-      "epoch": 0.46676687798183125,
-      "grad_norm": 0.5293527841567993,
-      "learning_rate": 9.53518870636762e-06,
-      "loss": 0.4298,
-      "step": 7142
-    },
-    {
-      "epoch": 0.4668322331873734,
-      "grad_norm": 0.4298951029777527,
-      "learning_rate": 9.535041669090461e-06,
-      "loss": 0.3779,
-      "step": 7143
-    },
-    {
-      "epoch": 0.4668975883929155,
-      "grad_norm": 0.47229844331741333,
-      "learning_rate": 9.534894609694333e-06,
-      "loss": 0.4069,
-      "step": 7144
-    },
-    {
-      "epoch": 0.46696294359845764,
-      "grad_norm": 0.4897076189517975,
-      "learning_rate": 9.534747528179953e-06,
-      "loss": 0.4297,
-      "step": 7145
-    },
-    {
-      "epoch": 0.46702829880399976,
-      "grad_norm": 0.4641999900341034,
-      "learning_rate": 9.53460042454804e-06,
-      "loss": 0.3812,
-      "step": 7146
-    },
-    {
-      "epoch": 0.46709365400954184,
-      "grad_norm": 0.45778489112854004,
-      "learning_rate": 9.534453298799307e-06,
-      "loss": 0.387,
-      "step": 7147
-    },
-    {
-      "epoch": 0.46715900921508396,
-      "grad_norm": 0.45132094621658325,
-      "learning_rate": 9.534306150934476e-06,
-      "loss": 0.4103,
-      "step": 7148
-    },
-    {
-      "epoch": 0.4672243644206261,
-      "grad_norm": 0.5260260701179504,
-      "learning_rate": 9.534158980954263e-06,
-      "loss": 0.403,
-      "step": 7149
-    },
-    {
-      "epoch": 0.4672897196261682,
-      "grad_norm": 0.46466195583343506,
-      "learning_rate": 9.534011788859386e-06,
-      "loss": 0.3899,
-      "step": 7150
-    },
-    {
-      "epoch": 0.46735507483171035,
-      "grad_norm": 0.4723871648311615,
-      "learning_rate": 9.53386457465056e-06,
-      "loss": 0.4237,
-      "step": 7151
-    },
-    {
-      "epoch": 0.4674204300372525,
-      "grad_norm": 0.42926061153411865,
-      "learning_rate": 9.533717338328508e-06,
-      "loss": 0.3933,
-      "step": 7152
-    },
-    {
-      "epoch": 0.4674857852427946,
-      "grad_norm": 0.4415152370929718,
-      "learning_rate": 9.533570079893946e-06,
-      "loss": 0.3496,
-      "step": 7153
-    },
-    {
-      "epoch": 0.46755114044833673,
-      "grad_norm": 0.5533658862113953,
-      "learning_rate": 9.533422799347594e-06,
-      "loss": 0.3923,
-      "step": 7154
-    },
-    {
-      "epoch": 0.4676164956538788,
-      "grad_norm": 0.4387933611869812,
-      "learning_rate": 9.533275496690165e-06,
-      "loss": 0.3481,
-      "step": 7155
-    },
-    {
-      "epoch": 0.46768185085942093,
-      "grad_norm": 0.4454059302806854,
-      "learning_rate": 9.533128171922384e-06,
-      "loss": 0.378,
-      "step": 7156
-    },
-    {
-      "epoch": 0.46774720606496306,
-      "grad_norm": 0.45458248257637024,
-      "learning_rate": 9.532980825044963e-06,
-      "loss": 0.4273,
-      "step": 7157
-    },
-    {
-      "epoch": 0.4678125612705052,
-      "grad_norm": 0.4645242393016815,
-      "learning_rate": 9.532833456058627e-06,
-      "loss": 0.4408,
-      "step": 7158
-    },
-    {
-      "epoch": 0.4678779164760473,
-      "grad_norm": 0.5080562829971313,
-      "learning_rate": 9.532686064964093e-06,
-      "loss": 0.4961,
-      "step": 7159
-    },
-    {
-      "epoch": 0.46794327168158945,
-      "grad_norm": 0.45649388432502747,
-      "learning_rate": 9.532538651762076e-06,
-      "loss": 0.3736,
-      "step": 7160
-    },
-    {
-      "epoch": 0.4680086268871316,
-      "grad_norm": 0.4337565302848816,
-      "learning_rate": 9.532391216453299e-06,
-      "loss": 0.3504,
-      "step": 7161
-    },
-    {
-      "epoch": 0.4680739820926737,
-      "grad_norm": 0.4585168659687042,
-      "learning_rate": 9.532243759038478e-06,
-      "loss": 0.3586,
-      "step": 7162
-    },
-    {
-      "epoch": 0.46813933729821583,
-      "grad_norm": 0.4669223725795746,
-      "learning_rate": 9.532096279518335e-06,
-      "loss": 0.4139,
-      "step": 7163
-    },
-    {
-      "epoch": 0.4682046925037579,
-      "grad_norm": 0.4383951723575592,
-      "learning_rate": 9.531948777893589e-06,
-      "loss": 0.3546,
-      "step": 7164
-    },
-    {
-      "epoch": 0.46827004770930003,
-      "grad_norm": 0.4296281635761261,
-      "learning_rate": 9.531801254164958e-06,
-      "loss": 0.3568,
-      "step": 7165
-    },
-    {
-      "epoch": 0.46833540291484216,
-      "grad_norm": 0.4933410882949829,
-      "learning_rate": 9.531653708333164e-06,
-      "loss": 0.4256,
-      "step": 7166
-    },
-    {
-      "epoch": 0.4684007581203843,
-      "grad_norm": 0.441978394985199,
-      "learning_rate": 9.531506140398925e-06,
-      "loss": 0.3638,
-      "step": 7167
-    },
-    {
-      "epoch": 0.4684661133259264,
-      "grad_norm": 0.4891020655632019,
-      "learning_rate": 9.53135855036296e-06,
-      "loss": 0.4353,
-      "step": 7168
-    },
-    {
-      "epoch": 0.46853146853146854,
-      "grad_norm": 0.44048675894737244,
-      "learning_rate": 9.531210938225988e-06,
-      "loss": 0.3646,
-      "step": 7169
-    },
-    {
-      "epoch": 0.46859682373701067,
-      "grad_norm": 0.4505417048931122,
-      "learning_rate": 9.531063303988732e-06,
-      "loss": 0.3837,
-      "step": 7170
-    },
-    {
-      "epoch": 0.4686621789425528,
-      "grad_norm": 0.42214491963386536,
-      "learning_rate": 9.53091564765191e-06,
-      "loss": 0.3387,
-      "step": 7171
-    },
-    {
-      "epoch": 0.46872753414809487,
-      "grad_norm": 0.456562876701355,
-      "learning_rate": 9.530767969216244e-06,
-      "loss": 0.3912,
-      "step": 7172
-    },
-    {
-      "epoch": 0.468792889353637,
-      "grad_norm": 0.43945470452308655,
-      "learning_rate": 9.53062026868245e-06,
-      "loss": 0.3756,
-      "step": 7173
-    },
-    {
-      "epoch": 0.4688582445591791,
-      "grad_norm": 0.44226452708244324,
-      "learning_rate": 9.530472546051255e-06,
-      "loss": 0.3674,
-      "step": 7174
-    },
-    {
-      "epoch": 0.46892359976472126,
-      "grad_norm": 0.4294166564941406,
-      "learning_rate": 9.530324801323375e-06,
-      "loss": 0.3487,
-      "step": 7175
-    },
-    {
-      "epoch": 0.4689889549702634,
-      "grad_norm": 0.4668973982334137,
-      "learning_rate": 9.53017703449953e-06,
-      "loss": 0.3798,
-      "step": 7176
-    },
-    {
-      "epoch": 0.4690543101758055,
-      "grad_norm": 0.5044274926185608,
-      "learning_rate": 9.530029245580442e-06,
-      "loss": 0.4084,
-      "step": 7177
-    },
-    {
-      "epoch": 0.46911966538134764,
-      "grad_norm": 0.4264613687992096,
-      "learning_rate": 9.529881434566833e-06,
-      "loss": 0.3728,
-      "step": 7178
-    },
-    {
-      "epoch": 0.46918502058688977,
-      "grad_norm": 0.4415418803691864,
-      "learning_rate": 9.529733601459424e-06,
-      "loss": 0.3912,
-      "step": 7179
-    },
-    {
-      "epoch": 0.46925037579243184,
-      "grad_norm": 0.4888932406902313,
-      "learning_rate": 9.529585746258934e-06,
-      "loss": 0.4292,
-      "step": 7180
-    },
-    {
-      "epoch": 0.46931573099797397,
-      "grad_norm": 0.4589439332485199,
-      "learning_rate": 9.529437868966085e-06,
-      "loss": 0.4056,
-      "step": 7181
-    },
-    {
-      "epoch": 0.4693810862035161,
-      "grad_norm": 0.47460922598838806,
-      "learning_rate": 9.529289969581596e-06,
-      "loss": 0.4395,
-      "step": 7182
-    },
-    {
-      "epoch": 0.4694464414090582,
-      "grad_norm": 0.43631982803344727,
-      "learning_rate": 9.529142048106194e-06,
-      "loss": 0.3347,
-      "step": 7183
-    },
-    {
-      "epoch": 0.46951179661460035,
-      "grad_norm": 0.44571515917778015,
-      "learning_rate": 9.528994104540596e-06,
-      "loss": 0.379,
-      "step": 7184
-    },
-    {
-      "epoch": 0.4695771518201425,
-      "grad_norm": 0.5230749249458313,
-      "learning_rate": 9.528846138885526e-06,
-      "loss": 0.493,
-      "step": 7185
-    },
-    {
-      "epoch": 0.4696425070256846,
-      "grad_norm": 0.5007441639900208,
-      "learning_rate": 9.528698151141702e-06,
-      "loss": 0.4655,
-      "step": 7186
-    },
-    {
-      "epoch": 0.46970786223122674,
-      "grad_norm": 0.4592949450016022,
-      "learning_rate": 9.52855014130985e-06,
-      "loss": 0.4223,
-      "step": 7187
-    },
-    {
-      "epoch": 0.46977321743676886,
-      "grad_norm": 0.40913125872612,
-      "learning_rate": 9.52840210939069e-06,
-      "loss": 0.32,
-      "step": 7188
-    },
-    {
-      "epoch": 0.46983857264231094,
-      "grad_norm": 0.46526312828063965,
-      "learning_rate": 9.528254055384944e-06,
-      "loss": 0.3993,
-      "step": 7189
-    },
-    {
-      "epoch": 0.46990392784785306,
-      "grad_norm": 0.4625626802444458,
-      "learning_rate": 9.528105979293334e-06,
-      "loss": 0.4317,
-      "step": 7190
-    },
-    {
-      "epoch": 0.4699692830533952,
-      "grad_norm": 0.46927690505981445,
-      "learning_rate": 9.527957881116582e-06,
-      "loss": 0.4157,
-      "step": 7191
-    },
-    {
-      "epoch": 0.4700346382589373,
-      "grad_norm": 0.4961572289466858,
-      "learning_rate": 9.527809760855412e-06,
-      "loss": 0.4453,
-      "step": 7192
-    },
-    {
-      "epoch": 0.47009999346447945,
-      "grad_norm": 0.43030011653900146,
-      "learning_rate": 9.527661618510545e-06,
-      "loss": 0.3681,
-      "step": 7193
-    },
-    {
-      "epoch": 0.4701653486700216,
-      "grad_norm": 0.461275190114975,
-      "learning_rate": 9.527513454082705e-06,
-      "loss": 0.3916,
-      "step": 7194
-    },
-    {
-      "epoch": 0.4702307038755637,
-      "grad_norm": 0.48548248410224915,
-      "learning_rate": 9.52736526757261e-06,
-      "loss": 0.4444,
-      "step": 7195
-    },
-    {
-      "epoch": 0.47029605908110583,
-      "grad_norm": 0.4657756984233856,
-      "learning_rate": 9.527217058980989e-06,
-      "loss": 0.4206,
-      "step": 7196
-    },
-    {
-      "epoch": 0.4703614142866479,
-      "grad_norm": 0.48155540227890015,
-      "learning_rate": 9.52706882830856e-06,
-      "loss": 0.421,
-      "step": 7197
-    },
-    {
-      "epoch": 0.47042676949219003,
-      "grad_norm": 0.4589233696460724,
-      "learning_rate": 9.526920575556048e-06,
-      "loss": 0.362,
-      "step": 7198
-    },
-    {
-      "epoch": 0.47049212469773216,
-      "grad_norm": 0.4520863890647888,
-      "learning_rate": 9.52677230072418e-06,
-      "loss": 0.4167,
-      "step": 7199
-    },
-    {
-      "epoch": 0.4705574799032743,
-      "grad_norm": 0.46362555027008057,
-      "learning_rate": 9.526624003813671e-06,
-      "loss": 0.3834,
-      "step": 7200
-    },
-    {
-      "epoch": 0.4706228351088164,
-      "grad_norm": 0.4626002907752991,
-      "learning_rate": 9.52647568482525e-06,
-      "loss": 0.4045,
-      "step": 7201
-    },
-    {
-      "epoch": 0.47068819031435855,
-      "grad_norm": 0.4694306254386902,
-      "learning_rate": 9.526327343759639e-06,
-      "loss": 0.4315,
-      "step": 7202
-    },
-    {
-      "epoch": 0.4707535455199007,
-      "grad_norm": 0.49735140800476074,
-      "learning_rate": 9.52617898061756e-06,
-      "loss": 0.4278,
-      "step": 7203
-    },
-    {
-      "epoch": 0.4708189007254428,
-      "grad_norm": 0.5209844708442688,
-      "learning_rate": 9.52603059539974e-06,
-      "loss": 0.484,
-      "step": 7204
-    },
-    {
-      "epoch": 0.47088425593098493,
-      "grad_norm": 0.4144379794597626,
-      "learning_rate": 9.525882188106899e-06,
-      "loss": 0.3293,
-      "step": 7205
-    },
-    {
-      "epoch": 0.470949611136527,
-      "grad_norm": 0.4277053475379944,
-      "learning_rate": 9.525733758739766e-06,
-      "loss": 0.3742,
-      "step": 7206
-    },
-    {
-      "epoch": 0.47101496634206913,
-      "grad_norm": 0.42117348313331604,
-      "learning_rate": 9.525585307299059e-06,
-      "loss": 0.3091,
-      "step": 7207
-    },
-    {
-      "epoch": 0.47108032154761126,
-      "grad_norm": 0.4527733325958252,
-      "learning_rate": 9.525436833785505e-06,
-      "loss": 0.3608,
-      "step": 7208
-    },
-    {
-      "epoch": 0.4711456767531534,
-      "grad_norm": 0.45344555377960205,
-      "learning_rate": 9.525288338199828e-06,
-      "loss": 0.3909,
-      "step": 7209
-    },
-    {
-      "epoch": 0.4712110319586955,
-      "grad_norm": 0.4333653151988983,
-      "learning_rate": 9.52513982054275e-06,
-      "loss": 0.3615,
-      "step": 7210
-    },
-    {
-      "epoch": 0.47127638716423764,
-      "grad_norm": 0.43686145544052124,
-      "learning_rate": 9.524991280815e-06,
-      "loss": 0.3294,
-      "step": 7211
-    },
-    {
-      "epoch": 0.47134174236977977,
-      "grad_norm": 0.46286964416503906,
-      "learning_rate": 9.5248427190173e-06,
-      "loss": 0.3977,
-      "step": 7212
-    },
-    {
-      "epoch": 0.4714070975753219,
-      "grad_norm": 0.5068244934082031,
-      "learning_rate": 9.524694135150374e-06,
-      "loss": 0.3957,
-      "step": 7213
-    },
-    {
-      "epoch": 0.47147245278086397,
-      "grad_norm": 0.5215017199516296,
-      "learning_rate": 9.524545529214945e-06,
-      "loss": 0.4687,
-      "step": 7214
-    },
-    {
-      "epoch": 0.4715378079864061,
-      "grad_norm": 0.4662645161151886,
-      "learning_rate": 9.524396901211743e-06,
-      "loss": 0.4009,
-      "step": 7215
-    },
-    {
-      "epoch": 0.4716031631919482,
-      "grad_norm": 0.4635929763317108,
-      "learning_rate": 9.524248251141488e-06,
-      "loss": 0.4474,
-      "step": 7216
-    },
-    {
-      "epoch": 0.47166851839749036,
-      "grad_norm": 0.45604032278060913,
-      "learning_rate": 9.524099579004909e-06,
-      "loss": 0.3905,
-      "step": 7217
-    },
-    {
-      "epoch": 0.4717338736030325,
-      "grad_norm": 0.4709457755088806,
-      "learning_rate": 9.523950884802728e-06,
-      "loss": 0.3947,
-      "step": 7218
-    },
-    {
-      "epoch": 0.4717992288085746,
-      "grad_norm": 0.4669475853443146,
-      "learning_rate": 9.523802168535673e-06,
-      "loss": 0.4133,
-      "step": 7219
-    },
-    {
-      "epoch": 0.47186458401411674,
-      "grad_norm": 0.4406854510307312,
-      "learning_rate": 9.523653430204464e-06,
-      "loss": 0.397,
-      "step": 7220
-    },
-    {
-      "epoch": 0.47192993921965887,
-      "grad_norm": 0.4447558522224426,
-      "learning_rate": 9.523504669809832e-06,
-      "loss": 0.361,
-      "step": 7221
-    },
-    {
-      "epoch": 0.47199529442520094,
-      "grad_norm": 0.43887776136398315,
-      "learning_rate": 9.523355887352501e-06,
-      "loss": 0.3803,
-      "step": 7222
-    },
-    {
-      "epoch": 0.47206064963074307,
-      "grad_norm": 0.427418053150177,
-      "learning_rate": 9.523207082833195e-06,
-      "loss": 0.3869,
-      "step": 7223
-    },
-    {
-      "epoch": 0.4721260048362852,
-      "grad_norm": 0.48189249634742737,
-      "learning_rate": 9.523058256252643e-06,
-      "loss": 0.4352,
-      "step": 7224
-    },
-    {
-      "epoch": 0.4721913600418273,
-      "grad_norm": 0.4351761043071747,
-      "learning_rate": 9.522909407611566e-06,
-      "loss": 0.3769,
-      "step": 7225
-    },
-    {
-      "epoch": 0.47225671524736945,
-      "grad_norm": 0.44468432664871216,
-      "learning_rate": 9.522760536910696e-06,
-      "loss": 0.3805,
-      "step": 7226
-    },
-    {
-      "epoch": 0.4723220704529116,
-      "grad_norm": 0.48574814200401306,
-      "learning_rate": 9.522611644150754e-06,
-      "loss": 0.4481,
-      "step": 7227
-    },
-    {
-      "epoch": 0.4723874256584537,
-      "grad_norm": 0.4702390730381012,
-      "learning_rate": 9.52246272933247e-06,
-      "loss": 0.3978,
-      "step": 7228
-    },
-    {
-      "epoch": 0.47245278086399584,
-      "grad_norm": 0.4346328377723694,
-      "learning_rate": 9.522313792456567e-06,
-      "loss": 0.3601,
-      "step": 7229
-    },
-    {
-      "epoch": 0.47251813606953796,
-      "grad_norm": 0.454008013010025,
-      "learning_rate": 9.522164833523775e-06,
-      "loss": 0.394,
-      "step": 7230
-    },
-    {
-      "epoch": 0.47258349127508004,
-      "grad_norm": 0.42442142963409424,
-      "learning_rate": 9.522015852534817e-06,
-      "loss": 0.303,
-      "step": 7231
-    },
-    {
-      "epoch": 0.47264884648062216,
-      "grad_norm": 0.4440680146217346,
-      "learning_rate": 9.52186684949042e-06,
-      "loss": 0.4029,
-      "step": 7232
-    },
-    {
-      "epoch": 0.4727142016861643,
-      "grad_norm": 0.42702242732048035,
-      "learning_rate": 9.521717824391312e-06,
-      "loss": 0.3779,
-      "step": 7233
-    },
-    {
-      "epoch": 0.4727795568917064,
-      "grad_norm": 0.4719691872596741,
-      "learning_rate": 9.521568777238221e-06,
-      "loss": 0.4281,
-      "step": 7234
-    },
-    {
-      "epoch": 0.47284491209724855,
-      "grad_norm": 0.47126439213752747,
-      "learning_rate": 9.521419708031873e-06,
-      "loss": 0.4204,
-      "step": 7235
-    },
-    {
-      "epoch": 0.4729102673027907,
-      "grad_norm": 0.43783560395240784,
-      "learning_rate": 9.521270616772993e-06,
-      "loss": 0.3696,
-      "step": 7236
-    },
-    {
-      "epoch": 0.4729756225083328,
-      "grad_norm": 0.43278399109840393,
-      "learning_rate": 9.521121503462311e-06,
-      "loss": 0.3602,
-      "step": 7237
-    },
-    {
-      "epoch": 0.47304097771387493,
-      "grad_norm": 0.47101151943206787,
-      "learning_rate": 9.520972368100554e-06,
-      "loss": 0.4192,
-      "step": 7238
-    },
-    {
-      "epoch": 0.473106332919417,
-      "grad_norm": 0.5181102752685547,
-      "learning_rate": 9.520823210688446e-06,
-      "loss": 0.4488,
-      "step": 7239
-    },
-    {
-      "epoch": 0.47317168812495913,
-      "grad_norm": 0.4515284299850464,
-      "learning_rate": 9.520674031226719e-06,
-      "loss": 0.388,
-      "step": 7240
-    },
-    {
-      "epoch": 0.47323704333050126,
-      "grad_norm": 0.4675734043121338,
-      "learning_rate": 9.5205248297161e-06,
-      "loss": 0.4155,
-      "step": 7241
-    },
-    {
-      "epoch": 0.4733023985360434,
-      "grad_norm": 0.46302253007888794,
-      "learning_rate": 9.520375606157312e-06,
-      "loss": 0.4102,
-      "step": 7242
-    },
-    {
-      "epoch": 0.4733677537415855,
-      "grad_norm": 0.4747518301010132,
-      "learning_rate": 9.520226360551089e-06,
-      "loss": 0.4373,
-      "step": 7243
-    },
-    {
-      "epoch": 0.47343310894712765,
-      "grad_norm": 0.4323144555091858,
-      "learning_rate": 9.520077092898155e-06,
-      "loss": 0.3076,
-      "step": 7244
-    },
-    {
-      "epoch": 0.4734984641526698,
-      "grad_norm": 0.44021075963974,
-      "learning_rate": 9.519927803199239e-06,
-      "loss": 0.334,
-      "step": 7245
-    },
-    {
-      "epoch": 0.4735638193582119,
-      "grad_norm": 0.4736710786819458,
-      "learning_rate": 9.519778491455068e-06,
-      "loss": 0.4281,
-      "step": 7246
-    },
-    {
-      "epoch": 0.47362917456375403,
-      "grad_norm": 0.48081645369529724,
-      "learning_rate": 9.519629157666373e-06,
-      "loss": 0.3862,
-      "step": 7247
-    },
-    {
-      "epoch": 0.4736945297692961,
-      "grad_norm": 0.4680396020412445,
-      "learning_rate": 9.51947980183388e-06,
-      "loss": 0.3978,
-      "step": 7248
-    },
-    {
-      "epoch": 0.47375988497483823,
-      "grad_norm": 0.4743956923484802,
-      "learning_rate": 9.519330423958319e-06,
-      "loss": 0.4567,
-      "step": 7249
-    },
-    {
-      "epoch": 0.47382524018038036,
-      "grad_norm": 0.46481868624687195,
-      "learning_rate": 9.519181024040418e-06,
-      "loss": 0.3635,
-      "step": 7250
-    },
-    {
-      "epoch": 0.4738905953859225,
-      "grad_norm": 0.4730256199836731,
-      "learning_rate": 9.519031602080905e-06,
-      "loss": 0.4023,
-      "step": 7251
-    },
-    {
-      "epoch": 0.4739559505914646,
-      "grad_norm": 0.46956583857536316,
-      "learning_rate": 9.51888215808051e-06,
-      "loss": 0.4263,
-      "step": 7252
-    },
-    {
-      "epoch": 0.47402130579700674,
-      "grad_norm": 0.4671914875507355,
-      "learning_rate": 9.518732692039962e-06,
-      "loss": 0.4106,
-      "step": 7253
-    },
-    {
-      "epoch": 0.47408666100254887,
-      "grad_norm": 0.4920426309108734,
-      "learning_rate": 9.518583203959988e-06,
-      "loss": 0.3802,
-      "step": 7254
-    },
-    {
-      "epoch": 0.474152016208091,
-      "grad_norm": 0.45804592967033386,
-      "learning_rate": 9.518433693841318e-06,
-      "loss": 0.405,
-      "step": 7255
-    },
-    {
-      "epoch": 0.47421737141363307,
-      "grad_norm": 0.4789954125881195,
-      "learning_rate": 9.518284161684681e-06,
-      "loss": 0.4155,
-      "step": 7256
-    },
-    {
-      "epoch": 0.4742827266191752,
-      "grad_norm": 0.4323576092720032,
-      "learning_rate": 9.518134607490808e-06,
-      "loss": 0.3587,
-      "step": 7257
-    },
-    {
-      "epoch": 0.4743480818247173,
-      "grad_norm": 0.4509740173816681,
-      "learning_rate": 9.517985031260429e-06,
-      "loss": 0.3557,
-      "step": 7258
-    },
-    {
-      "epoch": 0.47441343703025945,
-      "grad_norm": 0.4612359404563904,
-      "learning_rate": 9.517835432994269e-06,
-      "loss": 0.4052,
-      "step": 7259
-    },
-    {
-      "epoch": 0.4744787922358016,
-      "grad_norm": 0.4386858344078064,
-      "learning_rate": 9.51768581269306e-06,
-      "loss": 0.3728,
-      "step": 7260
-    },
-    {
-      "epoch": 0.4745441474413437,
-      "grad_norm": 0.47079646587371826,
-      "learning_rate": 9.517536170357535e-06,
-      "loss": 0.4406,
-      "step": 7261
-    },
-    {
-      "epoch": 0.47460950264688584,
-      "grad_norm": 0.46120864152908325,
-      "learning_rate": 9.517386505988419e-06,
-      "loss": 0.3856,
-      "step": 7262
-    },
-    {
-      "epoch": 0.47467485785242797,
-      "grad_norm": 0.4666607975959778,
-      "learning_rate": 9.517236819586445e-06,
-      "loss": 0.4161,
-      "step": 7263
-    },
-    {
-      "epoch": 0.47474021305797004,
-      "grad_norm": 0.4508492946624756,
-      "learning_rate": 9.51708711115234e-06,
-      "loss": 0.3842,
-      "step": 7264
-    },
-    {
-      "epoch": 0.47480556826351217,
-      "grad_norm": 0.4076070189476013,
-      "learning_rate": 9.516937380686839e-06,
-      "loss": 0.3313,
-      "step": 7265
-    },
-    {
-      "epoch": 0.4748709234690543,
-      "grad_norm": 0.4719148874282837,
-      "learning_rate": 9.516787628190668e-06,
-      "loss": 0.4112,
-      "step": 7266
-    },
-    {
-      "epoch": 0.4749362786745964,
-      "grad_norm": 0.41744285821914673,
-      "learning_rate": 9.51663785366456e-06,
-      "loss": 0.3318,
-      "step": 7267
-    },
-    {
-      "epoch": 0.47500163388013855,
-      "grad_norm": 0.44334912300109863,
-      "learning_rate": 9.516488057109245e-06,
-      "loss": 0.3724,
-      "step": 7268
-    },
-    {
-      "epoch": 0.4750669890856807,
-      "grad_norm": 0.4568454921245575,
-      "learning_rate": 9.516338238525453e-06,
-      "loss": 0.3515,
-      "step": 7269
-    },
-    {
-      "epoch": 0.4751323442912228,
-      "grad_norm": 0.4562966227531433,
-      "learning_rate": 9.516188397913913e-06,
-      "loss": 0.3878,
-      "step": 7270
-    },
-    {
-      "epoch": 0.47519769949676494,
-      "grad_norm": 0.4447886049747467,
-      "learning_rate": 9.51603853527536e-06,
-      "loss": 0.4111,
-      "step": 7271
-    },
-    {
-      "epoch": 0.47526305470230706,
-      "grad_norm": 0.47418299317359924,
-      "learning_rate": 9.51588865061052e-06,
-      "loss": 0.4049,
-      "step": 7272
-    },
-    {
-      "epoch": 0.47532840990784914,
-      "grad_norm": 0.44725048542022705,
-      "learning_rate": 9.515738743920128e-06,
-      "loss": 0.3943,
-      "step": 7273
-    },
-    {
-      "epoch": 0.47539376511339126,
-      "grad_norm": 0.4517868161201477,
-      "learning_rate": 9.515588815204913e-06,
-      "loss": 0.4052,
-      "step": 7274
-    },
-    {
-      "epoch": 0.4754591203189334,
-      "grad_norm": 0.4453171491622925,
-      "learning_rate": 9.515438864465608e-06,
-      "loss": 0.3633,
-      "step": 7275
-    },
-    {
-      "epoch": 0.4755244755244755,
-      "grad_norm": 0.48039841651916504,
-      "learning_rate": 9.515288891702943e-06,
-      "loss": 0.441,
-      "step": 7276
-    },
-    {
-      "epoch": 0.47558983073001765,
-      "grad_norm": 0.43807274103164673,
-      "learning_rate": 9.51513889691765e-06,
-      "loss": 0.3898,
-      "step": 7277
-    },
-    {
-      "epoch": 0.4756551859355598,
-      "grad_norm": 0.4547906816005707,
-      "learning_rate": 9.514988880110461e-06,
-      "loss": 0.4063,
-      "step": 7278
-    },
-    {
-      "epoch": 0.4757205411411019,
-      "grad_norm": 0.48315906524658203,
-      "learning_rate": 9.514838841282107e-06,
-      "loss": 0.4325,
-      "step": 7279
-    },
-    {
-      "epoch": 0.47578589634664403,
-      "grad_norm": 0.46735313534736633,
-      "learning_rate": 9.514688780433316e-06,
-      "loss": 0.4105,
-      "step": 7280
-    },
-    {
-      "epoch": 0.4758512515521861,
-      "grad_norm": 0.4949561059474945,
-      "learning_rate": 9.514538697564828e-06,
-      "loss": 0.4434,
-      "step": 7281
-    },
-    {
-      "epoch": 0.47591660675772823,
-      "grad_norm": 0.44944673776626587,
-      "learning_rate": 9.51438859267737e-06,
-      "loss": 0.3767,
-      "step": 7282
-    },
-    {
-      "epoch": 0.47598196196327036,
-      "grad_norm": 0.46788227558135986,
-      "learning_rate": 9.514238465771675e-06,
-      "loss": 0.4028,
-      "step": 7283
-    },
-    {
-      "epoch": 0.4760473171688125,
-      "grad_norm": 0.4889834523200989,
-      "learning_rate": 9.514088316848474e-06,
-      "loss": 0.4382,
-      "step": 7284
-    },
-    {
-      "epoch": 0.4761126723743546,
-      "grad_norm": 0.4904220998287201,
-      "learning_rate": 9.5139381459085e-06,
-      "loss": 0.4392,
-      "step": 7285
-    },
-    {
-      "epoch": 0.47617802757989675,
-      "grad_norm": 0.45664700865745544,
-      "learning_rate": 9.51378795295249e-06,
-      "loss": 0.4045,
-      "step": 7286
-    },
-    {
-      "epoch": 0.4762433827854389,
-      "grad_norm": 0.5146574378013611,
-      "learning_rate": 9.513637737981168e-06,
-      "loss": 0.4579,
-      "step": 7287
-    },
-    {
-      "epoch": 0.476308737990981,
-      "grad_norm": 0.4887774586677551,
-      "learning_rate": 9.513487500995273e-06,
-      "loss": 0.4011,
-      "step": 7288
-    },
-    {
-      "epoch": 0.47637409319652313,
-      "grad_norm": 0.46555158495903015,
-      "learning_rate": 9.513337241995535e-06,
-      "loss": 0.3898,
-      "step": 7289
-    },
-    {
-      "epoch": 0.4764394484020652,
-      "grad_norm": 0.4263874292373657,
-      "learning_rate": 9.513186960982688e-06,
-      "loss": 0.3445,
-      "step": 7290
-    },
-    {
-      "epoch": 0.47650480360760733,
-      "grad_norm": 0.4768146276473999,
-      "learning_rate": 9.513036657957464e-06,
-      "loss": 0.4141,
-      "step": 7291
-    },
-    {
-      "epoch": 0.47657015881314946,
-      "grad_norm": 0.5001056790351868,
-      "learning_rate": 9.512886332920598e-06,
-      "loss": 0.4119,
-      "step": 7292
-    },
-    {
-      "epoch": 0.4766355140186916,
-      "grad_norm": 0.47101029753685,
-      "learning_rate": 9.51273598587282e-06,
-      "loss": 0.3835,
-      "step": 7293
-    },
-    {
-      "epoch": 0.4767008692242337,
-      "grad_norm": 0.4715198874473572,
-      "learning_rate": 9.512585616814869e-06,
-      "loss": 0.4164,
-      "step": 7294
-    },
-    {
-      "epoch": 0.47676622442977584,
-      "grad_norm": 0.4667539596557617,
-      "learning_rate": 9.512435225747472e-06,
-      "loss": 0.3791,
-      "step": 7295
-    },
-    {
-      "epoch": 0.47683157963531797,
-      "grad_norm": 0.4642941355705261,
-      "learning_rate": 9.512284812671365e-06,
-      "loss": 0.3675,
-      "step": 7296
-    },
-    {
-      "epoch": 0.4768969348408601,
-      "grad_norm": 0.4901065230369568,
-      "learning_rate": 9.512134377587282e-06,
-      "loss": 0.3923,
-      "step": 7297
-    },
-    {
-      "epoch": 0.47696229004640217,
-      "grad_norm": 0.4683338701725006,
-      "learning_rate": 9.511983920495957e-06,
-      "loss": 0.3923,
-      "step": 7298
-    },
-    {
-      "epoch": 0.4770276452519443,
-      "grad_norm": 0.43025875091552734,
-      "learning_rate": 9.511833441398123e-06,
-      "loss": 0.3393,
-      "step": 7299
-    },
-    {
-      "epoch": 0.4770930004574864,
-      "grad_norm": 0.5275658369064331,
-      "learning_rate": 9.511682940294514e-06,
-      "loss": 0.4693,
-      "step": 7300
-    },
-    {
-      "epoch": 0.47715835566302855,
-      "grad_norm": 0.5425550937652588,
-      "learning_rate": 9.511532417185864e-06,
-      "loss": 0.4165,
-      "step": 7301
-    },
-    {
-      "epoch": 0.4772237108685707,
-      "grad_norm": 0.4689454138278961,
-      "learning_rate": 9.51138187207291e-06,
-      "loss": 0.395,
-      "step": 7302
-    },
-    {
-      "epoch": 0.4772890660741128,
-      "grad_norm": 0.4754444360733032,
-      "learning_rate": 9.51123130495638e-06,
-      "loss": 0.393,
-      "step": 7303
-    },
-    {
-      "epoch": 0.47735442127965494,
-      "grad_norm": 0.447358638048172,
-      "learning_rate": 9.511080715837015e-06,
-      "loss": 0.3535,
-      "step": 7304
-    },
-    {
-      "epoch": 0.47741977648519707,
-      "grad_norm": 0.4476067125797272,
-      "learning_rate": 9.510930104715545e-06,
-      "loss": 0.3602,
-      "step": 7305
-    },
-    {
-      "epoch": 0.47748513169073914,
-      "grad_norm": 0.48552852869033813,
-      "learning_rate": 9.510779471592706e-06,
-      "loss": 0.448,
-      "step": 7306
-    },
-    {
-      "epoch": 0.47755048689628127,
-      "grad_norm": 0.4522158205509186,
-      "learning_rate": 9.510628816469234e-06,
-      "loss": 0.3395,
-      "step": 7307
-    },
-    {
-      "epoch": 0.4776158421018234,
-      "grad_norm": 0.49904000759124756,
-      "learning_rate": 9.510478139345862e-06,
-      "loss": 0.4331,
-      "step": 7308
-    },
-    {
-      "epoch": 0.4776811973073655,
-      "grad_norm": 0.448441743850708,
-      "learning_rate": 9.510327440223324e-06,
-      "loss": 0.3785,
-      "step": 7309
-    },
-    {
-      "epoch": 0.47774655251290765,
-      "grad_norm": 0.4391007125377655,
-      "learning_rate": 9.51017671910236e-06,
-      "loss": 0.3522,
-      "step": 7310
-    },
-    {
-      "epoch": 0.4778119077184498,
-      "grad_norm": 0.5057657957077026,
-      "learning_rate": 9.510025975983699e-06,
-      "loss": 0.4436,
-      "step": 7311
-    },
-    {
-      "epoch": 0.4778772629239919,
-      "grad_norm": 0.43368351459503174,
-      "learning_rate": 9.50987521086808e-06,
-      "loss": 0.36,
-      "step": 7312
-    },
-    {
-      "epoch": 0.47794261812953404,
-      "grad_norm": 0.4711674451828003,
-      "learning_rate": 9.509724423756237e-06,
-      "loss": 0.4364,
-      "step": 7313
-    },
-    {
-      "epoch": 0.47800797333507616,
-      "grad_norm": 0.4643467664718628,
-      "learning_rate": 9.509573614648905e-06,
-      "loss": 0.4411,
-      "step": 7314
-    },
-    {
-      "epoch": 0.47807332854061824,
-      "grad_norm": 0.4575525224208832,
-      "learning_rate": 9.50942278354682e-06,
-      "loss": 0.3946,
-      "step": 7315
-    },
-    {
-      "epoch": 0.47813868374616036,
-      "grad_norm": 0.4377342462539673,
-      "learning_rate": 9.509271930450718e-06,
-      "loss": 0.3586,
-      "step": 7316
-    },
-    {
-      "epoch": 0.4782040389517025,
-      "grad_norm": 0.4354041814804077,
-      "learning_rate": 9.509121055361337e-06,
-      "loss": 0.3639,
-      "step": 7317
-    },
-    {
-      "epoch": 0.4782693941572446,
-      "grad_norm": 0.5191472172737122,
-      "learning_rate": 9.508970158279409e-06,
-      "loss": 0.4548,
-      "step": 7318
-    },
-    {
-      "epoch": 0.47833474936278675,
-      "grad_norm": 0.4496753215789795,
-      "learning_rate": 9.508819239205672e-06,
-      "loss": 0.383,
-      "step": 7319
-    },
-    {
-      "epoch": 0.4784001045683289,
-      "grad_norm": 0.4612496793270111,
-      "learning_rate": 9.508668298140859e-06,
-      "loss": 0.4039,
-      "step": 7320
-    },
-    {
-      "epoch": 0.478465459773871,
-      "grad_norm": 0.4505055546760559,
-      "learning_rate": 9.50851733508571e-06,
-      "loss": 0.3934,
-      "step": 7321
-    },
-    {
-      "epoch": 0.47853081497941313,
-      "grad_norm": 0.43406152725219727,
-      "learning_rate": 9.50836635004096e-06,
-      "loss": 0.3405,
-      "step": 7322
-    },
-    {
-      "epoch": 0.4785961701849552,
-      "grad_norm": 0.4261730909347534,
-      "learning_rate": 9.508215343007348e-06,
-      "loss": 0.3902,
-      "step": 7323
-    },
-    {
-      "epoch": 0.47866152539049733,
-      "grad_norm": 0.43612366914749146,
-      "learning_rate": 9.508064313985606e-06,
-      "loss": 0.3778,
-      "step": 7324
-    },
-    {
-      "epoch": 0.47872688059603946,
-      "grad_norm": 0.4492633640766144,
-      "learning_rate": 9.507913262976472e-06,
-      "loss": 0.3798,
-      "step": 7325
-    },
-    {
-      "epoch": 0.4787922358015816,
-      "grad_norm": 0.4606937766075134,
-      "learning_rate": 9.507762189980684e-06,
-      "loss": 0.3851,
-      "step": 7326
-    },
-    {
-      "epoch": 0.4788575910071237,
-      "grad_norm": 0.44337230920791626,
-      "learning_rate": 9.507611094998979e-06,
-      "loss": 0.3714,
-      "step": 7327
-    },
-    {
-      "epoch": 0.47892294621266585,
-      "grad_norm": 0.4208022654056549,
-      "learning_rate": 9.507459978032093e-06,
-      "loss": 0.3463,
-      "step": 7328
-    },
-    {
-      "epoch": 0.478988301418208,
-      "grad_norm": 0.4231928288936615,
-      "learning_rate": 9.507308839080761e-06,
-      "loss": 0.3846,
-      "step": 7329
-    },
-    {
-      "epoch": 0.4790536566237501,
-      "grad_norm": 0.46564728021621704,
-      "learning_rate": 9.507157678145725e-06,
-      "loss": 0.3758,
-      "step": 7330
-    },
-    {
-      "epoch": 0.47911901182929223,
-      "grad_norm": 0.45815080404281616,
-      "learning_rate": 9.507006495227718e-06,
-      "loss": 0.4189,
-      "step": 7331
-    },
-    {
-      "epoch": 0.4791843670348343,
-      "grad_norm": 0.4404975473880768,
-      "learning_rate": 9.506855290327479e-06,
-      "loss": 0.3809,
-      "step": 7332
-    },
-    {
-      "epoch": 0.47924972224037643,
-      "grad_norm": 0.4769819378852844,
-      "learning_rate": 9.506704063445743e-06,
-      "loss": 0.4282,
-      "step": 7333
-    },
-    {
-      "epoch": 0.47931507744591856,
-      "grad_norm": 0.4443832039833069,
-      "learning_rate": 9.506552814583254e-06,
-      "loss": 0.371,
-      "step": 7334
-    },
-    {
-      "epoch": 0.4793804326514607,
-      "grad_norm": 0.46927258372306824,
-      "learning_rate": 9.506401543740746e-06,
-      "loss": 0.3837,
-      "step": 7335
-    },
-    {
-      "epoch": 0.4794457878570028,
-      "grad_norm": 0.4402283728122711,
-      "learning_rate": 9.506250250918954e-06,
-      "loss": 0.3758,
-      "step": 7336
-    },
-    {
-      "epoch": 0.47951114306254494,
-      "grad_norm": 0.4432384669780731,
-      "learning_rate": 9.50609893611862e-06,
-      "loss": 0.3726,
-      "step": 7337
-    },
-    {
-      "epoch": 0.47957649826808707,
-      "grad_norm": 0.4765337109565735,
-      "learning_rate": 9.50594759934048e-06,
-      "loss": 0.3942,
-      "step": 7338
-    },
-    {
-      "epoch": 0.4796418534736292,
-      "grad_norm": 0.4251444935798645,
-      "learning_rate": 9.50579624058527e-06,
-      "loss": 0.3684,
-      "step": 7339
-    },
-    {
-      "epoch": 0.47970720867917127,
-      "grad_norm": 0.4497198462486267,
-      "learning_rate": 9.505644859853734e-06,
-      "loss": 0.3869,
-      "step": 7340
-    },
-    {
-      "epoch": 0.4797725638847134,
-      "grad_norm": 0.4948360025882721,
-      "learning_rate": 9.505493457146605e-06,
-      "loss": 0.4207,
-      "step": 7341
-    },
-    {
-      "epoch": 0.4798379190902555,
-      "grad_norm": 0.45769166946411133,
-      "learning_rate": 9.505342032464627e-06,
-      "loss": 0.3838,
-      "step": 7342
-    },
-    {
-      "epoch": 0.47990327429579765,
-      "grad_norm": 0.48405545949935913,
-      "learning_rate": 9.505190585808533e-06,
-      "loss": 0.456,
-      "step": 7343
-    },
-    {
-      "epoch": 0.4799686295013398,
-      "grad_norm": 0.49579980969429016,
-      "learning_rate": 9.505039117179063e-06,
-      "loss": 0.4385,
-      "step": 7344
-    },
-    {
-      "epoch": 0.4800339847068819,
-      "grad_norm": 0.6010561585426331,
-      "learning_rate": 9.504887626576958e-06,
-      "loss": 0.4404,
-      "step": 7345
-    },
-    {
-      "epoch": 0.48009933991242404,
-      "grad_norm": 0.4512021839618683,
-      "learning_rate": 9.504736114002953e-06,
-      "loss": 0.3693,
-      "step": 7346
-    },
-    {
-      "epoch": 0.48016469511796617,
-      "grad_norm": 0.4892708361148834,
-      "learning_rate": 9.50458457945779e-06,
-      "loss": 0.4398,
-      "step": 7347
-    },
-    {
-      "epoch": 0.4802300503235083,
-      "grad_norm": 0.494975209236145,
-      "learning_rate": 9.50443302294221e-06,
-      "loss": 0.3536,
-      "step": 7348
-    },
-    {
-      "epoch": 0.48029540552905037,
-      "grad_norm": 0.4584241509437561,
-      "learning_rate": 9.504281444456947e-06,
-      "loss": 0.4125,
-      "step": 7349
-    },
-    {
-      "epoch": 0.4803607607345925,
-      "grad_norm": 0.4221155345439911,
-      "learning_rate": 9.504129844002745e-06,
-      "loss": 0.3738,
-      "step": 7350
-    },
-    {
-      "epoch": 0.4804261159401346,
-      "grad_norm": 0.48600223660469055,
-      "learning_rate": 9.503978221580338e-06,
-      "loss": 0.4373,
-      "step": 7351
-    },
-    {
-      "epoch": 0.48049147114567675,
-      "grad_norm": 0.41237521171569824,
-      "learning_rate": 9.503826577190473e-06,
-      "loss": 0.3448,
-      "step": 7352
-    },
-    {
-      "epoch": 0.4805568263512189,
-      "grad_norm": 0.46173515915870667,
-      "learning_rate": 9.503674910833884e-06,
-      "loss": 0.3958,
-      "step": 7353
-    },
-    {
-      "epoch": 0.480622181556761,
-      "grad_norm": 0.47616085410118103,
-      "learning_rate": 9.503523222511311e-06,
-      "loss": 0.4173,
-      "step": 7354
-    },
-    {
-      "epoch": 0.48068753676230314,
-      "grad_norm": 0.4457026422023773,
-      "learning_rate": 9.503371512223497e-06,
-      "loss": 0.3822,
-      "step": 7355
-    },
-    {
-      "epoch": 0.48075289196784526,
-      "grad_norm": 0.46840551495552063,
-      "learning_rate": 9.503219779971178e-06,
-      "loss": 0.3957,
-      "step": 7356
-    },
-    {
-      "epoch": 0.48081824717338734,
-      "grad_norm": 0.43921956419944763,
-      "learning_rate": 9.503068025755099e-06,
-      "loss": 0.3762,
-      "step": 7357
-    },
-    {
-      "epoch": 0.48088360237892946,
-      "grad_norm": 0.47099873423576355,
-      "learning_rate": 9.502916249575993e-06,
-      "loss": 0.419,
-      "step": 7358
-    },
-    {
-      "epoch": 0.4809489575844716,
-      "grad_norm": 0.4435424506664276,
-      "learning_rate": 9.502764451434608e-06,
-      "loss": 0.3604,
-      "step": 7359
-    },
-    {
-      "epoch": 0.4810143127900137,
-      "grad_norm": 0.463450163602829,
-      "learning_rate": 9.50261263133168e-06,
-      "loss": 0.3797,
-      "step": 7360
-    },
-    {
-      "epoch": 0.48107966799555585,
-      "grad_norm": 0.5260501503944397,
-      "learning_rate": 9.50246078926795e-06,
-      "loss": 0.4759,
-      "step": 7361
-    },
-    {
-      "epoch": 0.481145023201098,
-      "grad_norm": 0.4505915641784668,
-      "learning_rate": 9.502308925244159e-06,
-      "loss": 0.4012,
-      "step": 7362
-    },
-    {
-      "epoch": 0.4812103784066401,
-      "grad_norm": 0.4749990403652191,
-      "learning_rate": 9.502157039261047e-06,
-      "loss": 0.3791,
-      "step": 7363
-    },
-    {
-      "epoch": 0.48127573361218223,
-      "grad_norm": 0.44820836186408997,
-      "learning_rate": 9.502005131319357e-06,
-      "loss": 0.4018,
-      "step": 7364
-    },
-    {
-      "epoch": 0.4813410888177243,
-      "grad_norm": 0.44692501425743103,
-      "learning_rate": 9.501853201419826e-06,
-      "loss": 0.3693,
-      "step": 7365
-    },
-    {
-      "epoch": 0.48140644402326643,
-      "grad_norm": 0.4645141065120697,
-      "learning_rate": 9.5017012495632e-06,
-      "loss": 0.3833,
-      "step": 7366
-    },
-    {
-      "epoch": 0.48147179922880856,
-      "grad_norm": 0.7762148380279541,
-      "learning_rate": 9.501549275750215e-06,
-      "loss": 0.3834,
-      "step": 7367
-    },
-    {
-      "epoch": 0.4815371544343507,
-      "grad_norm": 0.48271432518959045,
-      "learning_rate": 9.501397279981616e-06,
-      "loss": 0.4235,
-      "step": 7368
-    },
-    {
-      "epoch": 0.4816025096398928,
-      "grad_norm": 0.4508444666862488,
-      "learning_rate": 9.501245262258142e-06,
-      "loss": 0.3982,
-      "step": 7369
-    },
-    {
-      "epoch": 0.48166786484543495,
-      "grad_norm": 0.48299744725227356,
-      "learning_rate": 9.501093222580537e-06,
-      "loss": 0.394,
-      "step": 7370
-    },
-    {
-      "epoch": 0.4817332200509771,
-      "grad_norm": 0.43675726652145386,
-      "learning_rate": 9.500941160949541e-06,
-      "loss": 0.3428,
-      "step": 7371
-    },
-    {
-      "epoch": 0.4817985752565192,
-      "grad_norm": 0.44021663069725037,
-      "learning_rate": 9.500789077365894e-06,
-      "loss": 0.4024,
-      "step": 7372
-    },
-    {
-      "epoch": 0.48186393046206133,
-      "grad_norm": 0.4571131467819214,
-      "learning_rate": 9.500636971830342e-06,
-      "loss": 0.3693,
-      "step": 7373
-    },
-    {
-      "epoch": 0.4819292856676034,
-      "grad_norm": 0.44831281900405884,
-      "learning_rate": 9.500484844343622e-06,
-      "loss": 0.3586,
-      "step": 7374
-    },
-    {
-      "epoch": 0.48199464087314553,
-      "grad_norm": 0.4352133572101593,
-      "learning_rate": 9.50033269490648e-06,
-      "loss": 0.3445,
-      "step": 7375
-    },
-    {
-      "epoch": 0.48205999607868766,
-      "grad_norm": 0.4469340443611145,
-      "learning_rate": 9.500180523519655e-06,
-      "loss": 0.3669,
-      "step": 7376
-    },
-    {
-      "epoch": 0.4821253512842298,
-      "grad_norm": 0.4346008598804474,
-      "learning_rate": 9.500028330183892e-06,
-      "loss": 0.369,
-      "step": 7377
-    },
-    {
-      "epoch": 0.4821907064897719,
-      "grad_norm": 0.4379720091819763,
-      "learning_rate": 9.49987611489993e-06,
-      "loss": 0.3932,
-      "step": 7378
-    },
-    {
-      "epoch": 0.48225606169531404,
-      "grad_norm": 0.43308043479919434,
-      "learning_rate": 9.499723877668514e-06,
-      "loss": 0.3591,
-      "step": 7379
-    },
-    {
-      "epoch": 0.48232141690085617,
-      "grad_norm": 0.4782882630825043,
-      "learning_rate": 9.499571618490387e-06,
-      "loss": 0.4213,
-      "step": 7380
-    },
-    {
-      "epoch": 0.4823867721063983,
-      "grad_norm": 0.47381147742271423,
-      "learning_rate": 9.499419337366289e-06,
-      "loss": 0.4241,
-      "step": 7381
-    },
-    {
-      "epoch": 0.48245212731194037,
-      "grad_norm": 0.44306692481040955,
-      "learning_rate": 9.499267034296966e-06,
-      "loss": 0.3411,
-      "step": 7382
-    },
-    {
-      "epoch": 0.4825174825174825,
-      "grad_norm": 0.44983986020088196,
-      "learning_rate": 9.499114709283157e-06,
-      "loss": 0.3958,
-      "step": 7383
-    },
-    {
-      "epoch": 0.4825828377230246,
-      "grad_norm": 0.4825235605239868,
-      "learning_rate": 9.498962362325608e-06,
-      "loss": 0.4092,
-      "step": 7384
-    },
-    {
-      "epoch": 0.48264819292856675,
-      "grad_norm": 0.42523130774497986,
-      "learning_rate": 9.49880999342506e-06,
-      "loss": 0.3681,
-      "step": 7385
-    },
-    {
-      "epoch": 0.4827135481341089,
-      "grad_norm": 0.4949938654899597,
-      "learning_rate": 9.498657602582258e-06,
-      "loss": 0.4341,
-      "step": 7386
-    },
-    {
-      "epoch": 0.482778903339651,
-      "grad_norm": 0.49087655544281006,
-      "learning_rate": 9.498505189797945e-06,
-      "loss": 0.4445,
-      "step": 7387
-    },
-    {
-      "epoch": 0.48284425854519314,
-      "grad_norm": 0.4897131621837616,
-      "learning_rate": 9.498352755072861e-06,
-      "loss": 0.4489,
-      "step": 7388
-    },
-    {
-      "epoch": 0.48290961375073527,
-      "grad_norm": 0.47331148386001587,
-      "learning_rate": 9.498200298407754e-06,
-      "loss": 0.3556,
-      "step": 7389
-    },
-    {
-      "epoch": 0.4829749689562774,
-      "grad_norm": 0.47523513436317444,
-      "learning_rate": 9.498047819803367e-06,
-      "loss": 0.375,
-      "step": 7390
-    },
-    {
-      "epoch": 0.48304032416181947,
-      "grad_norm": 0.5040974020957947,
-      "learning_rate": 9.497895319260439e-06,
-      "loss": 0.4224,
-      "step": 7391
-    },
-    {
-      "epoch": 0.4831056793673616,
-      "grad_norm": 0.4312072694301605,
-      "learning_rate": 9.497742796779721e-06,
-      "loss": 0.3427,
-      "step": 7392
-    },
-    {
-      "epoch": 0.4831710345729037,
-      "grad_norm": 0.469063401222229,
-      "learning_rate": 9.497590252361952e-06,
-      "loss": 0.3891,
-      "step": 7393
-    },
-    {
-      "epoch": 0.48323638977844585,
-      "grad_norm": 0.4873962104320526,
-      "learning_rate": 9.497437686007877e-06,
-      "loss": 0.4093,
-      "step": 7394
-    },
-    {
-      "epoch": 0.483301744983988,
-      "grad_norm": 0.4571136236190796,
-      "learning_rate": 9.497285097718241e-06,
-      "loss": 0.3556,
-      "step": 7395
-    },
-    {
-      "epoch": 0.4833671001895301,
-      "grad_norm": 0.43147221207618713,
-      "learning_rate": 9.497132487493785e-06,
-      "loss": 0.3768,
-      "step": 7396
-    },
-    {
-      "epoch": 0.48343245539507224,
-      "grad_norm": 0.43853816390037537,
-      "learning_rate": 9.49697985533526e-06,
-      "loss": 0.3771,
-      "step": 7397
-    },
-    {
-      "epoch": 0.48349781060061436,
-      "grad_norm": 0.49010199308395386,
-      "learning_rate": 9.496827201243404e-06,
-      "loss": 0.4498,
-      "step": 7398
-    },
-    {
-      "epoch": 0.48356316580615644,
-      "grad_norm": 0.4329914450645447,
-      "learning_rate": 9.496674525218963e-06,
-      "loss": 0.3719,
-      "step": 7399
-    },
-    {
-      "epoch": 0.48362852101169856,
-      "grad_norm": 0.4928269386291504,
-      "learning_rate": 9.496521827262684e-06,
-      "loss": 0.4307,
-      "step": 7400
-    },
-    {
-      "epoch": 0.4836938762172407,
-      "grad_norm": 0.5182434320449829,
-      "learning_rate": 9.49636910737531e-06,
-      "loss": 0.3826,
-      "step": 7401
-    },
-    {
-      "epoch": 0.4837592314227828,
-      "grad_norm": 0.46465063095092773,
-      "learning_rate": 9.496216365557586e-06,
-      "loss": 0.393,
-      "step": 7402
-    },
-    {
-      "epoch": 0.48382458662832495,
-      "grad_norm": 0.4669272005558014,
-      "learning_rate": 9.496063601810257e-06,
-      "loss": 0.3953,
-      "step": 7403
-    },
-    {
-      "epoch": 0.4838899418338671,
-      "grad_norm": 0.4677024483680725,
-      "learning_rate": 9.495910816134069e-06,
-      "loss": 0.3773,
-      "step": 7404
-    },
-    {
-      "epoch": 0.4839552970394092,
-      "grad_norm": 0.4466164708137512,
-      "learning_rate": 9.495758008529766e-06,
-      "loss": 0.376,
-      "step": 7405
-    },
-    {
-      "epoch": 0.48402065224495133,
-      "grad_norm": 0.48743775486946106,
-      "learning_rate": 9.495605178998094e-06,
-      "loss": 0.3743,
-      "step": 7406
-    },
-    {
-      "epoch": 0.4840860074504934,
-      "grad_norm": 0.4799644351005554,
-      "learning_rate": 9.495452327539797e-06,
-      "loss": 0.4082,
-      "step": 7407
-    },
-    {
-      "epoch": 0.48415136265603553,
-      "grad_norm": 0.4941820800304413,
-      "learning_rate": 9.495299454155621e-06,
-      "loss": 0.4203,
-      "step": 7408
-    },
-    {
-      "epoch": 0.48421671786157766,
-      "grad_norm": 0.5379561185836792,
-      "learning_rate": 9.495146558846315e-06,
-      "loss": 0.3769,
-      "step": 7409
-    },
-    {
-      "epoch": 0.4842820730671198,
-      "grad_norm": 0.4449411630630493,
-      "learning_rate": 9.49499364161262e-06,
-      "loss": 0.3813,
-      "step": 7410
-    },
-    {
-      "epoch": 0.4843474282726619,
-      "grad_norm": 0.46515390276908875,
-      "learning_rate": 9.494840702455284e-06,
-      "loss": 0.393,
-      "step": 7411
-    },
-    {
-      "epoch": 0.48441278347820405,
-      "grad_norm": 0.5021597146987915,
-      "learning_rate": 9.494687741375054e-06,
-      "loss": 0.4439,
-      "step": 7412
-    },
-    {
-      "epoch": 0.4844781386837462,
-      "grad_norm": 0.5706249475479126,
-      "learning_rate": 9.494534758372674e-06,
-      "loss": 0.5165,
-      "step": 7413
-    },
-    {
-      "epoch": 0.4845434938892883,
-      "grad_norm": 0.46963274478912354,
-      "learning_rate": 9.494381753448891e-06,
-      "loss": 0.4729,
-      "step": 7414
-    },
-    {
-      "epoch": 0.48460884909483043,
-      "grad_norm": 0.4755740761756897,
-      "learning_rate": 9.49422872660445e-06,
-      "loss": 0.3955,
-      "step": 7415
-    },
-    {
-      "epoch": 0.4846742043003725,
-      "grad_norm": 0.4645135700702667,
-      "learning_rate": 9.4940756778401e-06,
-      "loss": 0.3823,
-      "step": 7416
-    },
-    {
-      "epoch": 0.48473955950591463,
-      "grad_norm": 0.4323805868625641,
-      "learning_rate": 9.493922607156585e-06,
-      "loss": 0.335,
-      "step": 7417
-    },
-    {
-      "epoch": 0.48480491471145676,
-      "grad_norm": 0.42189720273017883,
-      "learning_rate": 9.493769514554654e-06,
-      "loss": 0.3291,
-      "step": 7418
-    },
-    {
-      "epoch": 0.4848702699169989,
-      "grad_norm": 0.48483920097351074,
-      "learning_rate": 9.49361640003505e-06,
-      "loss": 0.4195,
-      "step": 7419
-    },
-    {
-      "epoch": 0.484935625122541,
-      "grad_norm": 0.4638504087924957,
-      "learning_rate": 9.493463263598525e-06,
-      "loss": 0.3907,
-      "step": 7420
-    },
-    {
-      "epoch": 0.48500098032808314,
-      "grad_norm": 0.5052520036697388,
-      "learning_rate": 9.49331010524582e-06,
-      "loss": 0.4968,
-      "step": 7421
-    },
-    {
-      "epoch": 0.48506633553362527,
-      "grad_norm": 0.4519069194793701,
-      "learning_rate": 9.493156924977689e-06,
-      "loss": 0.3484,
-      "step": 7422
-    },
-    {
-      "epoch": 0.4851316907391674,
-      "grad_norm": 0.4787712097167969,
-      "learning_rate": 9.493003722794871e-06,
-      "loss": 0.4144,
-      "step": 7423
-    },
-    {
-      "epoch": 0.48519704594470947,
-      "grad_norm": 0.45541858673095703,
-      "learning_rate": 9.492850498698119e-06,
-      "loss": 0.3542,
-      "step": 7424
-    },
-    {
-      "epoch": 0.4852624011502516,
-      "grad_norm": 0.4757027328014374,
-      "learning_rate": 9.492697252688178e-06,
-      "loss": 0.4123,
-      "step": 7425
-    },
-    {
-      "epoch": 0.4853277563557937,
-      "grad_norm": 0.5296993255615234,
-      "learning_rate": 9.492543984765797e-06,
-      "loss": 0.4714,
-      "step": 7426
-    },
-    {
-      "epoch": 0.48539311156133585,
-      "grad_norm": 0.43457597494125366,
-      "learning_rate": 9.492390694931722e-06,
-      "loss": 0.3275,
-      "step": 7427
-    },
-    {
-      "epoch": 0.485458466766878,
-      "grad_norm": 0.49319249391555786,
-      "learning_rate": 9.492237383186702e-06,
-      "loss": 0.4424,
-      "step": 7428
-    },
-    {
-      "epoch": 0.4855238219724201,
-      "grad_norm": 0.45508813858032227,
-      "learning_rate": 9.492084049531483e-06,
-      "loss": 0.3829,
-      "step": 7429
-    },
-    {
-      "epoch": 0.48558917717796224,
-      "grad_norm": 0.5218712687492371,
-      "learning_rate": 9.491930693966816e-06,
-      "loss": 0.4896,
-      "step": 7430
-    },
-    {
-      "epoch": 0.48565453238350437,
-      "grad_norm": 0.43735891580581665,
-      "learning_rate": 9.491777316493444e-06,
-      "loss": 0.3774,
-      "step": 7431
-    },
-    {
-      "epoch": 0.4857198875890465,
-      "grad_norm": 0.4714779257774353,
-      "learning_rate": 9.49162391711212e-06,
-      "loss": 0.3966,
-      "step": 7432
-    },
-    {
-      "epoch": 0.48578524279458857,
-      "grad_norm": 0.4681144654750824,
-      "learning_rate": 9.491470495823589e-06,
-      "loss": 0.417,
-      "step": 7433
-    },
-    {
-      "epoch": 0.4858505980001307,
-      "grad_norm": 0.439555823802948,
-      "learning_rate": 9.491317052628601e-06,
-      "loss": 0.372,
-      "step": 7434
-    },
-    {
-      "epoch": 0.4859159532056728,
-      "grad_norm": 0.46620187163352966,
-      "learning_rate": 9.491163587527902e-06,
-      "loss": 0.381,
-      "step": 7435
-    },
-    {
-      "epoch": 0.48598130841121495,
-      "grad_norm": 0.4731757342815399,
-      "learning_rate": 9.491010100522245e-06,
-      "loss": 0.4169,
-      "step": 7436
-    },
-    {
-      "epoch": 0.4860466636167571,
-      "grad_norm": 0.4192809462547302,
-      "learning_rate": 9.490856591612373e-06,
-      "loss": 0.3447,
-      "step": 7437
-    },
-    {
-      "epoch": 0.4861120188222992,
-      "grad_norm": 0.41696199774742126,
-      "learning_rate": 9.490703060799041e-06,
-      "loss": 0.3449,
-      "step": 7438
-    },
-    {
-      "epoch": 0.48617737402784134,
-      "grad_norm": 0.4259806275367737,
-      "learning_rate": 9.490549508082994e-06,
-      "loss": 0.3837,
-      "step": 7439
-    },
-    {
-      "epoch": 0.48624272923338346,
-      "grad_norm": 0.4262341856956482,
-      "learning_rate": 9.490395933464978e-06,
-      "loss": 0.3328,
-      "step": 7440
-    },
-    {
-      "epoch": 0.48630808443892554,
-      "grad_norm": 0.44839173555374146,
-      "learning_rate": 9.490242336945748e-06,
-      "loss": 0.4062,
-      "step": 7441
-    },
-    {
-      "epoch": 0.48637343964446766,
-      "grad_norm": 0.45787325501441956,
-      "learning_rate": 9.49008871852605e-06,
-      "loss": 0.4065,
-      "step": 7442
-    },
-    {
-      "epoch": 0.4864387948500098,
-      "grad_norm": 0.46419623494148254,
-      "learning_rate": 9.489935078206634e-06,
-      "loss": 0.4104,
-      "step": 7443
-    },
-    {
-      "epoch": 0.4865041500555519,
-      "grad_norm": 0.4577346742153168,
-      "learning_rate": 9.48978141598825e-06,
-      "loss": 0.3915,
-      "step": 7444
-    },
-    {
-      "epoch": 0.48656950526109405,
-      "grad_norm": 0.4654706120491028,
-      "learning_rate": 9.489627731871644e-06,
-      "loss": 0.3935,
-      "step": 7445
-    },
-    {
-      "epoch": 0.4866348604666362,
-      "grad_norm": 0.45058754086494446,
-      "learning_rate": 9.48947402585757e-06,
-      "loss": 0.3947,
-      "step": 7446
-    },
-    {
-      "epoch": 0.4867002156721783,
-      "grad_norm": 0.4263327419757843,
-      "learning_rate": 9.489320297946777e-06,
-      "loss": 0.3359,
-      "step": 7447
-    },
-    {
-      "epoch": 0.48676557087772043,
-      "grad_norm": 0.4461744427680969,
-      "learning_rate": 9.489166548140012e-06,
-      "loss": 0.3928,
-      "step": 7448
-    },
-    {
-      "epoch": 0.4868309260832625,
-      "grad_norm": 0.4722278416156769,
-      "learning_rate": 9.489012776438028e-06,
-      "loss": 0.4061,
-      "step": 7449
-    },
-    {
-      "epoch": 0.48689628128880463,
-      "grad_norm": 0.4542923867702484,
-      "learning_rate": 9.488858982841572e-06,
-      "loss": 0.3547,
-      "step": 7450
-    },
-    {
-      "epoch": 0.48696163649434676,
-      "grad_norm": 0.4683986008167267,
-      "learning_rate": 9.488705167351396e-06,
-      "loss": 0.4105,
-      "step": 7451
-    },
-    {
-      "epoch": 0.4870269916998889,
-      "grad_norm": 0.44002366065979004,
-      "learning_rate": 9.48855132996825e-06,
-      "loss": 0.3653,
-      "step": 7452
-    },
-    {
-      "epoch": 0.487092346905431,
-      "grad_norm": 0.500486433506012,
-      "learning_rate": 9.488397470692884e-06,
-      "loss": 0.4377,
-      "step": 7453
-    },
-    {
-      "epoch": 0.48715770211097315,
-      "grad_norm": 0.4667614996433258,
-      "learning_rate": 9.488243589526049e-06,
-      "loss": 0.3596,
-      "step": 7454
-    },
-    {
-      "epoch": 0.4872230573165153,
-      "grad_norm": 0.4702526032924652,
-      "learning_rate": 9.488089686468494e-06,
-      "loss": 0.412,
-      "step": 7455
-    },
-    {
-      "epoch": 0.4872884125220574,
-      "grad_norm": 0.4730609059333801,
-      "learning_rate": 9.487935761520972e-06,
-      "loss": 0.4101,
-      "step": 7456
-    },
-    {
-      "epoch": 0.48735376772759953,
-      "grad_norm": 0.4924863278865814,
-      "learning_rate": 9.487781814684233e-06,
-      "loss": 0.4295,
-      "step": 7457
-    },
-    {
-      "epoch": 0.4874191229331416,
-      "grad_norm": 0.4361676573753357,
-      "learning_rate": 9.487627845959027e-06,
-      "loss": 0.3684,
-      "step": 7458
-    },
-    {
-      "epoch": 0.48748447813868373,
-      "grad_norm": 0.4205648601055145,
-      "learning_rate": 9.487473855346104e-06,
-      "loss": 0.3522,
-      "step": 7459
-    },
-    {
-      "epoch": 0.48754983334422586,
-      "grad_norm": 0.4551865756511688,
-      "learning_rate": 9.487319842846218e-06,
-      "loss": 0.3701,
-      "step": 7460
-    },
-    {
-      "epoch": 0.487615188549768,
-      "grad_norm": 0.4413236379623413,
-      "learning_rate": 9.487165808460118e-06,
-      "loss": 0.3592,
-      "step": 7461
-    },
-    {
-      "epoch": 0.4876805437553101,
-      "grad_norm": 0.4950585961341858,
-      "learning_rate": 9.487011752188555e-06,
-      "loss": 0.4188,
-      "step": 7462
-    },
-    {
-      "epoch": 0.48774589896085224,
-      "grad_norm": 0.46799200773239136,
-      "learning_rate": 9.486857674032281e-06,
-      "loss": 0.411,
-      "step": 7463
-    },
-    {
-      "epoch": 0.48781125416639437,
-      "grad_norm": 0.49813807010650635,
-      "learning_rate": 9.48670357399205e-06,
-      "loss": 0.4423,
-      "step": 7464
-    },
-    {
-      "epoch": 0.4878766093719365,
-      "grad_norm": 0.42745304107666016,
-      "learning_rate": 9.48654945206861e-06,
-      "loss": 0.3688,
-      "step": 7465
-    },
-    {
-      "epoch": 0.48794196457747857,
-      "grad_norm": 0.4313168227672577,
-      "learning_rate": 9.486395308262714e-06,
-      "loss": 0.3619,
-      "step": 7466
-    },
-    {
-      "epoch": 0.4880073197830207,
-      "grad_norm": 0.45661038160324097,
-      "learning_rate": 9.486241142575114e-06,
-      "loss": 0.4127,
-      "step": 7467
-    },
-    {
-      "epoch": 0.4880726749885628,
-      "grad_norm": 0.46303310990333557,
-      "learning_rate": 9.486086955006562e-06,
-      "loss": 0.3923,
-      "step": 7468
-    },
-    {
-      "epoch": 0.48813803019410495,
-      "grad_norm": 0.4744683504104614,
-      "learning_rate": 9.485932745557808e-06,
-      "loss": 0.3959,
-      "step": 7469
-    },
-    {
-      "epoch": 0.4882033853996471,
-      "grad_norm": 0.43088585138320923,
-      "learning_rate": 9.485778514229609e-06,
-      "loss": 0.3495,
-      "step": 7470
-    },
-    {
-      "epoch": 0.4882687406051892,
-      "grad_norm": 0.4546554684638977,
-      "learning_rate": 9.485624261022712e-06,
-      "loss": 0.3717,
-      "step": 7471
-    },
-    {
-      "epoch": 0.48833409581073134,
-      "grad_norm": 0.4435812532901764,
-      "learning_rate": 9.485469985937871e-06,
-      "loss": 0.3625,
-      "step": 7472
-    },
-    {
-      "epoch": 0.48839945101627347,
-      "grad_norm": 0.4912225902080536,
-      "learning_rate": 9.485315688975842e-06,
-      "loss": 0.4572,
-      "step": 7473
-    },
-    {
-      "epoch": 0.4884648062218156,
-      "grad_norm": 0.5011767745018005,
-      "learning_rate": 9.485161370137372e-06,
-      "loss": 0.4345,
-      "step": 7474
-    },
-    {
-      "epoch": 0.48853016142735767,
-      "grad_norm": 0.49139299988746643,
-      "learning_rate": 9.485007029423216e-06,
-      "loss": 0.3975,
-      "step": 7475
-    },
-    {
-      "epoch": 0.4885955166328998,
-      "grad_norm": 0.47588053345680237,
-      "learning_rate": 9.484852666834128e-06,
-      "loss": 0.4218,
-      "step": 7476
-    },
-    {
-      "epoch": 0.4886608718384419,
-      "grad_norm": 0.43201300501823425,
-      "learning_rate": 9.48469828237086e-06,
-      "loss": 0.3376,
-      "step": 7477
-    },
-    {
-      "epoch": 0.48872622704398405,
-      "grad_norm": 0.48201489448547363,
-      "learning_rate": 9.484543876034164e-06,
-      "loss": 0.4284,
-      "step": 7478
-    },
-    {
-      "epoch": 0.4887915822495262,
-      "grad_norm": 0.45606234669685364,
-      "learning_rate": 9.484389447824795e-06,
-      "loss": 0.4013,
-      "step": 7479
-    },
-    {
-      "epoch": 0.4888569374550683,
-      "grad_norm": 0.4488910138607025,
-      "learning_rate": 9.484234997743505e-06,
-      "loss": 0.4096,
-      "step": 7480
-    },
-    {
-      "epoch": 0.48892229266061044,
-      "grad_norm": 0.48581674695014954,
-      "learning_rate": 9.484080525791047e-06,
-      "loss": 0.421,
-      "step": 7481
-    },
-    {
-      "epoch": 0.48898764786615256,
-      "grad_norm": 0.4653454124927521,
-      "learning_rate": 9.483926031968174e-06,
-      "loss": 0.3806,
-      "step": 7482
-    },
-    {
-      "epoch": 0.48905300307169464,
-      "grad_norm": 0.4340418577194214,
-      "learning_rate": 9.48377151627564e-06,
-      "loss": 0.3481,
-      "step": 7483
-    },
-    {
-      "epoch": 0.48911835827723676,
-      "grad_norm": 0.47066178917884827,
-      "learning_rate": 9.483616978714201e-06,
-      "loss": 0.4243,
-      "step": 7484
-    },
-    {
-      "epoch": 0.4891837134827789,
-      "grad_norm": 0.5573233962059021,
-      "learning_rate": 9.483462419284607e-06,
-      "loss": 0.5169,
-      "step": 7485
-    },
-    {
-      "epoch": 0.489249068688321,
-      "grad_norm": 0.4547271430492401,
-      "learning_rate": 9.483307837987615e-06,
-      "loss": 0.3952,
-      "step": 7486
-    },
-    {
-      "epoch": 0.48931442389386315,
-      "grad_norm": 0.4757891893386841,
-      "learning_rate": 9.483153234823976e-06,
-      "loss": 0.3764,
-      "step": 7487
-    },
-    {
-      "epoch": 0.4893797790994053,
-      "grad_norm": 0.4782652258872986,
-      "learning_rate": 9.482998609794447e-06,
-      "loss": 0.4111,
-      "step": 7488
-    },
-    {
-      "epoch": 0.4894451343049474,
-      "grad_norm": 0.4628230035305023,
-      "learning_rate": 9.48284396289978e-06,
-      "loss": 0.3851,
-      "step": 7489
-    },
-    {
-      "epoch": 0.48951048951048953,
-      "grad_norm": 0.4211656153202057,
-      "learning_rate": 9.48268929414073e-06,
-      "loss": 0.3503,
-      "step": 7490
-    },
-    {
-      "epoch": 0.4895758447160316,
-      "grad_norm": 0.4701022505760193,
-      "learning_rate": 9.482534603518052e-06,
-      "loss": 0.4122,
-      "step": 7491
-    },
-    {
-      "epoch": 0.48964119992157373,
-      "grad_norm": 0.45558977127075195,
-      "learning_rate": 9.482379891032499e-06,
-      "loss": 0.377,
-      "step": 7492
-    },
-    {
-      "epoch": 0.48970655512711586,
-      "grad_norm": 0.46655386686325073,
-      "learning_rate": 9.482225156684826e-06,
-      "loss": 0.4386,
-      "step": 7493
-    },
-    {
-      "epoch": 0.489771910332658,
-      "grad_norm": 0.4944005310535431,
-      "learning_rate": 9.48207040047579e-06,
-      "loss": 0.4416,
-      "step": 7494
-    },
-    {
-      "epoch": 0.4898372655382001,
-      "grad_norm": 0.44188159704208374,
-      "learning_rate": 9.481915622406141e-06,
-      "loss": 0.3857,
-      "step": 7495
-    },
-    {
-      "epoch": 0.48990262074374225,
-      "grad_norm": 0.42931339144706726,
-      "learning_rate": 9.48176082247664e-06,
-      "loss": 0.3474,
-      "step": 7496
-    },
-    {
-      "epoch": 0.4899679759492844,
-      "grad_norm": 0.44169536232948303,
-      "learning_rate": 9.481606000688037e-06,
-      "loss": 0.3868,
-      "step": 7497
-    },
-    {
-      "epoch": 0.4900333311548265,
-      "grad_norm": 0.4467918276786804,
-      "learning_rate": 9.481451157041089e-06,
-      "loss": 0.4069,
-      "step": 7498
-    },
-    {
-      "epoch": 0.49009868636036863,
-      "grad_norm": 0.44880470633506775,
-      "learning_rate": 9.481296291536553e-06,
-      "loss": 0.3887,
-      "step": 7499
-    },
-    {
-      "epoch": 0.4901640415659107,
-      "grad_norm": 0.4240676760673523,
-      "learning_rate": 9.48114140417518e-06,
-      "loss": 0.3818,
-      "step": 7500
-    },
-    {
-      "epoch": 0.49022939677145283,
-      "grad_norm": 0.4574185311794281,
-      "learning_rate": 9.480986494957729e-06,
-      "loss": 0.4024,
-      "step": 7501
-    },
-    {
-      "epoch": 0.49029475197699496,
-      "grad_norm": 0.4692107141017914,
-      "learning_rate": 9.480831563884955e-06,
-      "loss": 0.4067,
-      "step": 7502
-    },
-    {
-      "epoch": 0.4903601071825371,
-      "grad_norm": 0.4708959758281708,
-      "learning_rate": 9.480676610957612e-06,
-      "loss": 0.4044,
-      "step": 7503
-    },
-    {
-      "epoch": 0.4904254623880792,
-      "grad_norm": 0.4837050437927246,
-      "learning_rate": 9.48052163617646e-06,
-      "loss": 0.4194,
-      "step": 7504
-    },
-    {
-      "epoch": 0.49049081759362134,
-      "grad_norm": 0.44029760360717773,
-      "learning_rate": 9.480366639542247e-06,
-      "loss": 0.3344,
-      "step": 7505
-    },
-    {
-      "epoch": 0.49055617279916347,
-      "grad_norm": 0.46228793263435364,
-      "learning_rate": 9.480211621055737e-06,
-      "loss": 0.3911,
-      "step": 7506
-    },
-    {
-      "epoch": 0.4906215280047056,
-      "grad_norm": 0.47261396050453186,
-      "learning_rate": 9.480056580717681e-06,
-      "loss": 0.4063,
-      "step": 7507
-    },
-    {
-      "epoch": 0.49068688321024767,
-      "grad_norm": 0.47378870844841003,
-      "learning_rate": 9.479901518528839e-06,
-      "loss": 0.4487,
-      "step": 7508
-    },
-    {
-      "epoch": 0.4907522384157898,
-      "grad_norm": 0.4447380602359772,
-      "learning_rate": 9.479746434489964e-06,
-      "loss": 0.3941,
-      "step": 7509
-    },
-    {
-      "epoch": 0.4908175936213319,
-      "grad_norm": 0.4341123700141907,
-      "learning_rate": 9.479591328601814e-06,
-      "loss": 0.3261,
-      "step": 7510
-    },
-    {
-      "epoch": 0.49088294882687405,
-      "grad_norm": 0.47509121894836426,
-      "learning_rate": 9.479436200865144e-06,
-      "loss": 0.3712,
-      "step": 7511
-    },
-    {
-      "epoch": 0.4909483040324162,
-      "grad_norm": 0.4571378827095032,
-      "learning_rate": 9.479281051280713e-06,
-      "loss": 0.3586,
-      "step": 7512
-    },
-    {
-      "epoch": 0.4910136592379583,
-      "grad_norm": 0.4923804998397827,
-      "learning_rate": 9.479125879849278e-06,
-      "loss": 0.4517,
-      "step": 7513
-    },
-    {
-      "epoch": 0.49107901444350044,
-      "grad_norm": 0.4679352939128876,
-      "learning_rate": 9.478970686571593e-06,
-      "loss": 0.4071,
-      "step": 7514
-    },
-    {
-      "epoch": 0.49114436964904257,
-      "grad_norm": 0.4228973686695099,
-      "learning_rate": 9.478815471448416e-06,
-      "loss": 0.3366,
-      "step": 7515
-    },
-    {
-      "epoch": 0.4912097248545847,
-      "grad_norm": 0.4530769884586334,
-      "learning_rate": 9.478660234480504e-06,
-      "loss": 0.3564,
-      "step": 7516
-    },
-    {
-      "epoch": 0.49127508006012677,
-      "grad_norm": 0.5609104037284851,
-      "learning_rate": 9.478504975668616e-06,
-      "loss": 0.4835,
-      "step": 7517
-    },
-    {
-      "epoch": 0.4913404352656689,
-      "grad_norm": 0.46037769317626953,
-      "learning_rate": 9.478349695013506e-06,
-      "loss": 0.3851,
-      "step": 7518
-    },
-    {
-      "epoch": 0.491405790471211,
-      "grad_norm": 0.4529840350151062,
-      "learning_rate": 9.478194392515934e-06,
-      "loss": 0.385,
-      "step": 7519
-    },
-    {
-      "epoch": 0.49147114567675315,
-      "grad_norm": 0.4782061278820038,
-      "learning_rate": 9.478039068176657e-06,
-      "loss": 0.4092,
-      "step": 7520
-    },
-    {
-      "epoch": 0.4915365008822953,
-      "grad_norm": 0.4545944929122925,
-      "learning_rate": 9.47788372199643e-06,
-      "loss": 0.402,
-      "step": 7521
-    },
-    {
-      "epoch": 0.4916018560878374,
-      "grad_norm": 0.4744200110435486,
-      "learning_rate": 9.477728353976015e-06,
-      "loss": 0.4138,
-      "step": 7522
-    },
-    {
-      "epoch": 0.49166721129337954,
-      "grad_norm": 0.4526703357696533,
-      "learning_rate": 9.477572964116166e-06,
-      "loss": 0.3796,
-      "step": 7523
-    },
-    {
-      "epoch": 0.49173256649892166,
-      "grad_norm": 0.4858260154724121,
-      "learning_rate": 9.477417552417644e-06,
-      "loss": 0.4297,
-      "step": 7524
-    },
-    {
-      "epoch": 0.49179792170446374,
-      "grad_norm": 0.4929284453392029,
-      "learning_rate": 9.477262118881206e-06,
-      "loss": 0.455,
-      "step": 7525
-    },
-    {
-      "epoch": 0.49186327691000586,
-      "grad_norm": 0.4835423231124878,
-      "learning_rate": 9.477106663507607e-06,
-      "loss": 0.4435,
-      "step": 7526
-    },
-    {
-      "epoch": 0.491928632115548,
-      "grad_norm": 0.4589439630508423,
-      "learning_rate": 9.476951186297609e-06,
-      "loss": 0.3727,
-      "step": 7527
-    },
-    {
-      "epoch": 0.4919939873210901,
-      "grad_norm": 0.43981215357780457,
-      "learning_rate": 9.47679568725197e-06,
-      "loss": 0.3579,
-      "step": 7528
-    },
-    {
-      "epoch": 0.49205934252663225,
-      "grad_norm": 0.4801551103591919,
-      "learning_rate": 9.476640166371446e-06,
-      "loss": 0.4125,
-      "step": 7529
-    },
-    {
-      "epoch": 0.4921246977321744,
-      "grad_norm": 0.4690263271331787,
-      "learning_rate": 9.476484623656799e-06,
-      "loss": 0.4297,
-      "step": 7530
-    },
-    {
-      "epoch": 0.4921900529377165,
-      "grad_norm": 0.4995599687099457,
-      "learning_rate": 9.476329059108783e-06,
-      "loss": 0.431,
-      "step": 7531
-    },
-    {
-      "epoch": 0.49225540814325863,
-      "grad_norm": 0.4623638093471527,
-      "learning_rate": 9.476173472728163e-06,
-      "loss": 0.3899,
-      "step": 7532
-    },
-    {
-      "epoch": 0.4923207633488007,
-      "grad_norm": 0.5120333433151245,
-      "learning_rate": 9.476017864515692e-06,
-      "loss": 0.4663,
-      "step": 7533
-    },
-    {
-      "epoch": 0.49238611855434283,
-      "grad_norm": 0.47190403938293457,
-      "learning_rate": 9.475862234472131e-06,
-      "loss": 0.4761,
-      "step": 7534
-    },
-    {
-      "epoch": 0.49245147375988496,
-      "grad_norm": 0.4381028711795807,
-      "learning_rate": 9.475706582598241e-06,
-      "loss": 0.3473,
-      "step": 7535
-    },
-    {
-      "epoch": 0.4925168289654271,
-      "grad_norm": 0.4686459004878998,
-      "learning_rate": 9.475550908894777e-06,
-      "loss": 0.413,
-      "step": 7536
-    },
-    {
-      "epoch": 0.4925821841709692,
-      "grad_norm": 0.4188326597213745,
-      "learning_rate": 9.475395213362502e-06,
-      "loss": 0.3327,
-      "step": 7537
-    },
-    {
-      "epoch": 0.49264753937651135,
-      "grad_norm": 0.43074822425842285,
-      "learning_rate": 9.475239496002174e-06,
-      "loss": 0.3377,
-      "step": 7538
-    },
-    {
-      "epoch": 0.4927128945820535,
-      "grad_norm": 0.460660457611084,
-      "learning_rate": 9.475083756814554e-06,
-      "loss": 0.3871,
-      "step": 7539
-    },
-    {
-      "epoch": 0.4927782497875956,
-      "grad_norm": 0.4765304923057556,
-      "learning_rate": 9.474927995800398e-06,
-      "loss": 0.449,
-      "step": 7540
-    },
-    {
-      "epoch": 0.49284360499313773,
-      "grad_norm": 0.4728533625602722,
-      "learning_rate": 9.47477221296047e-06,
-      "loss": 0.3679,
-      "step": 7541
-    },
-    {
-      "epoch": 0.4929089601986798,
-      "grad_norm": 0.41927409172058105,
-      "learning_rate": 9.474616408295525e-06,
-      "loss": 0.3451,
-      "step": 7542
-    },
-    {
-      "epoch": 0.49297431540422193,
-      "grad_norm": 0.4750388264656067,
-      "learning_rate": 9.474460581806328e-06,
-      "loss": 0.4336,
-      "step": 7543
-    },
-    {
-      "epoch": 0.49303967060976406,
-      "grad_norm": 0.46211883425712585,
-      "learning_rate": 9.474304733493635e-06,
-      "loss": 0.3752,
-      "step": 7544
-    },
-    {
-      "epoch": 0.4931050258153062,
-      "grad_norm": 0.427681028842926,
-      "learning_rate": 9.474148863358209e-06,
-      "loss": 0.3273,
-      "step": 7545
-    },
-    {
-      "epoch": 0.4931703810208483,
-      "grad_norm": 0.48589053750038147,
-      "learning_rate": 9.473992971400809e-06,
-      "loss": 0.4333,
-      "step": 7546
-    },
-    {
-      "epoch": 0.49323573622639044,
-      "grad_norm": 0.459023118019104,
-      "learning_rate": 9.473837057622195e-06,
-      "loss": 0.4447,
-      "step": 7547
-    },
-    {
-      "epoch": 0.49330109143193257,
-      "grad_norm": 0.44562047719955444,
-      "learning_rate": 9.473681122023128e-06,
-      "loss": 0.3549,
-      "step": 7548
-    },
-    {
-      "epoch": 0.4933664466374747,
-      "grad_norm": 0.45815688371658325,
-      "learning_rate": 9.473525164604367e-06,
-      "loss": 0.3903,
-      "step": 7549
-    },
-    {
-      "epoch": 0.49343180184301677,
-      "grad_norm": 0.4764108955860138,
-      "learning_rate": 9.473369185366676e-06,
-      "loss": 0.4393,
-      "step": 7550
-    },
-    {
-      "epoch": 0.4934971570485589,
-      "grad_norm": 0.46385669708251953,
-      "learning_rate": 9.473213184310812e-06,
-      "loss": 0.3712,
-      "step": 7551
-    },
-    {
-      "epoch": 0.493562512254101,
-      "grad_norm": 0.4341394603252411,
-      "learning_rate": 9.473057161437539e-06,
-      "loss": 0.3423,
-      "step": 7552
-    },
-    {
-      "epoch": 0.49362786745964315,
-      "grad_norm": 0.4430205821990967,
-      "learning_rate": 9.472901116747616e-06,
-      "loss": 0.3706,
-      "step": 7553
-    },
-    {
-      "epoch": 0.4936932226651853,
-      "grad_norm": 0.4221726655960083,
-      "learning_rate": 9.472745050241805e-06,
-      "loss": 0.3316,
-      "step": 7554
-    },
-    {
-      "epoch": 0.4937585778707274,
-      "grad_norm": 0.45707279443740845,
-      "learning_rate": 9.472588961920866e-06,
-      "loss": 0.3918,
-      "step": 7555
-    },
-    {
-      "epoch": 0.49382393307626954,
-      "grad_norm": 0.43674278259277344,
-      "learning_rate": 9.472432851785563e-06,
-      "loss": 0.3556,
-      "step": 7556
-    },
-    {
-      "epoch": 0.49388928828181167,
-      "grad_norm": 0.4903871417045593,
-      "learning_rate": 9.472276719836652e-06,
-      "loss": 0.4391,
-      "step": 7557
-    },
-    {
-      "epoch": 0.4939546434873538,
-      "grad_norm": 0.4739069640636444,
-      "learning_rate": 9.472120566074902e-06,
-      "loss": 0.3787,
-      "step": 7558
-    },
-    {
-      "epoch": 0.49401999869289587,
-      "grad_norm": 0.39869147539138794,
-      "learning_rate": 9.471964390501069e-06,
-      "loss": 0.3024,
-      "step": 7559
-    },
-    {
-      "epoch": 0.494085353898438,
-      "grad_norm": 0.45095956325531006,
-      "learning_rate": 9.471808193115917e-06,
-      "loss": 0.3789,
-      "step": 7560
-    },
-    {
-      "epoch": 0.4941507091039801,
-      "grad_norm": 0.41862669587135315,
-      "learning_rate": 9.471651973920206e-06,
-      "loss": 0.3523,
-      "step": 7561
-    },
-    {
-      "epoch": 0.49421606430952225,
-      "grad_norm": 0.46903014183044434,
-      "learning_rate": 9.471495732914699e-06,
-      "loss": 0.4093,
-      "step": 7562
-    },
-    {
-      "epoch": 0.4942814195150644,
-      "grad_norm": 0.4792992174625397,
-      "learning_rate": 9.47133947010016e-06,
-      "loss": 0.4428,
-      "step": 7563
-    },
-    {
-      "epoch": 0.4943467747206065,
-      "grad_norm": 0.46021440625190735,
-      "learning_rate": 9.47118318547735e-06,
-      "loss": 0.3637,
-      "step": 7564
-    },
-    {
-      "epoch": 0.49441212992614864,
-      "grad_norm": 0.4644688069820404,
-      "learning_rate": 9.47102687904703e-06,
-      "loss": 0.4203,
-      "step": 7565
-    },
-    {
-      "epoch": 0.49447748513169076,
-      "grad_norm": 0.42194512486457825,
-      "learning_rate": 9.470870550809961e-06,
-      "loss": 0.314,
-      "step": 7566
-    },
-    {
-      "epoch": 0.49454284033723284,
-      "grad_norm": 0.41313961148262024,
-      "learning_rate": 9.47071420076691e-06,
-      "loss": 0.3436,
-      "step": 7567
-    },
-    {
-      "epoch": 0.49460819554277496,
-      "grad_norm": 0.4607747495174408,
-      "learning_rate": 9.470557828918634e-06,
-      "loss": 0.4399,
-      "step": 7568
-    },
-    {
-      "epoch": 0.4946735507483171,
-      "grad_norm": 0.39260971546173096,
-      "learning_rate": 9.4704014352659e-06,
-      "loss": 0.3153,
-      "step": 7569
-    },
-    {
-      "epoch": 0.4947389059538592,
-      "grad_norm": 0.442340612411499,
-      "learning_rate": 9.47024501980947e-06,
-      "loss": 0.3602,
-      "step": 7570
-    },
-    {
-      "epoch": 0.49480426115940135,
-      "grad_norm": 0.43838855624198914,
-      "learning_rate": 9.470088582550108e-06,
-      "loss": 0.3537,
-      "step": 7571
-    },
-    {
-      "epoch": 0.4948696163649435,
-      "grad_norm": 0.41797998547554016,
-      "learning_rate": 9.469932123488574e-06,
-      "loss": 0.3481,
-      "step": 7572
-    },
-    {
-      "epoch": 0.4949349715704856,
-      "grad_norm": 0.45995160937309265,
-      "learning_rate": 9.469775642625632e-06,
-      "loss": 0.3902,
-      "step": 7573
-    },
-    {
-      "epoch": 0.49500032677602773,
-      "grad_norm": 0.4653254449367523,
-      "learning_rate": 9.469619139962046e-06,
-      "loss": 0.4177,
-      "step": 7574
-    },
-    {
-      "epoch": 0.4950656819815698,
-      "grad_norm": 0.41050922870635986,
-      "learning_rate": 9.469462615498579e-06,
-      "loss": 0.3026,
-      "step": 7575
-    },
-    {
-      "epoch": 0.49513103718711193,
-      "grad_norm": 0.46875399351119995,
-      "learning_rate": 9.469306069235994e-06,
-      "loss": 0.4102,
-      "step": 7576
-    },
-    {
-      "epoch": 0.49519639239265406,
-      "grad_norm": 0.4898216128349304,
-      "learning_rate": 9.469149501175056e-06,
-      "loss": 0.4231,
-      "step": 7577
-    },
-    {
-      "epoch": 0.4952617475981962,
-      "grad_norm": 0.46499302983283997,
-      "learning_rate": 9.468992911316527e-06,
-      "loss": 0.4477,
-      "step": 7578
-    },
-    {
-      "epoch": 0.4953271028037383,
-      "grad_norm": 0.44518306851387024,
-      "learning_rate": 9.468836299661171e-06,
-      "loss": 0.4132,
-      "step": 7579
-    },
-    {
-      "epoch": 0.49539245800928045,
-      "grad_norm": 0.4644605219364166,
-      "learning_rate": 9.468679666209752e-06,
-      "loss": 0.4504,
-      "step": 7580
-    },
-    {
-      "epoch": 0.4954578132148226,
-      "grad_norm": 0.4470239281654358,
-      "learning_rate": 9.468523010963036e-06,
-      "loss": 0.3617,
-      "step": 7581
-    },
-    {
-      "epoch": 0.4955231684203647,
-      "grad_norm": 0.43484559655189514,
-      "learning_rate": 9.468366333921783e-06,
-      "loss": 0.3422,
-      "step": 7582
-    },
-    {
-      "epoch": 0.49558852362590683,
-      "grad_norm": 0.4375055134296417,
-      "learning_rate": 9.468209635086762e-06,
-      "loss": 0.3793,
-      "step": 7583
-    },
-    {
-      "epoch": 0.4956538788314489,
-      "grad_norm": 0.4536997079849243,
-      "learning_rate": 9.468052914458732e-06,
-      "loss": 0.4377,
-      "step": 7584
-    },
-    {
-      "epoch": 0.49571923403699103,
-      "grad_norm": 0.4456861615180969,
-      "learning_rate": 9.467896172038462e-06,
-      "loss": 0.3526,
-      "step": 7585
-    },
-    {
-      "epoch": 0.49578458924253316,
-      "grad_norm": 0.4149416387081146,
-      "learning_rate": 9.467739407826714e-06,
-      "loss": 0.3493,
-      "step": 7586
-    },
-    {
-      "epoch": 0.4958499444480753,
-      "grad_norm": 0.47360849380493164,
-      "learning_rate": 9.467582621824252e-06,
-      "loss": 0.447,
-      "step": 7587
-    },
-    {
-      "epoch": 0.4959152996536174,
-      "grad_norm": 0.44779253005981445,
-      "learning_rate": 9.467425814031843e-06,
-      "loss": 0.3444,
-      "step": 7588
-    },
-    {
-      "epoch": 0.49598065485915954,
-      "grad_norm": 0.46606341004371643,
-      "learning_rate": 9.46726898445025e-06,
-      "loss": 0.3895,
-      "step": 7589
-    },
-    {
-      "epoch": 0.49604601006470167,
-      "grad_norm": 0.49216240644454956,
-      "learning_rate": 9.467112133080239e-06,
-      "loss": 0.4247,
-      "step": 7590
-    },
-    {
-      "epoch": 0.4961113652702438,
-      "grad_norm": 0.4630853831768036,
-      "learning_rate": 9.466955259922574e-06,
-      "loss": 0.3872,
-      "step": 7591
-    },
-    {
-      "epoch": 0.49617672047578587,
-      "grad_norm": 0.465962290763855,
-      "learning_rate": 9.466798364978023e-06,
-      "loss": 0.4334,
-      "step": 7592
-    },
-    {
-      "epoch": 0.496242075681328,
-      "grad_norm": 0.4391988515853882,
-      "learning_rate": 9.466641448247346e-06,
-      "loss": 0.3757,
-      "step": 7593
-    },
-    {
-      "epoch": 0.4963074308868701,
-      "grad_norm": 0.4161827564239502,
-      "learning_rate": 9.466484509731313e-06,
-      "loss": 0.3297,
-      "step": 7594
-    },
-    {
-      "epoch": 0.49637278609241225,
-      "grad_norm": 0.4143718481063843,
-      "learning_rate": 9.466327549430688e-06,
-      "loss": 0.3426,
-      "step": 7595
-    },
-    {
-      "epoch": 0.4964381412979544,
-      "grad_norm": 0.5041487216949463,
-      "learning_rate": 9.466170567346233e-06,
-      "loss": 0.4603,
-      "step": 7596
-    },
-    {
-      "epoch": 0.4965034965034965,
-      "grad_norm": 0.44594258069992065,
-      "learning_rate": 9.46601356347872e-06,
-      "loss": 0.3618,
-      "step": 7597
-    },
-    {
-      "epoch": 0.49656885170903864,
-      "grad_norm": 0.6060497760772705,
-      "learning_rate": 9.465856537828911e-06,
-      "loss": 0.3925,
-      "step": 7598
-    },
-    {
-      "epoch": 0.49663420691458077,
-      "grad_norm": 0.4989464282989502,
-      "learning_rate": 9.465699490397572e-06,
-      "loss": 0.4287,
-      "step": 7599
-    },
-    {
-      "epoch": 0.4966995621201229,
-      "grad_norm": 0.4288352131843567,
-      "learning_rate": 9.46554242118547e-06,
-      "loss": 0.3779,
-      "step": 7600
-    },
-    {
-      "epoch": 0.49676491732566497,
-      "grad_norm": 0.4628756046295166,
-      "learning_rate": 9.46538533019337e-06,
-      "loss": 0.3878,
-      "step": 7601
-    },
-    {
-      "epoch": 0.4968302725312071,
-      "grad_norm": 0.5000662803649902,
-      "learning_rate": 9.465228217422042e-06,
-      "loss": 0.4319,
-      "step": 7602
-    },
-    {
-      "epoch": 0.4968956277367492,
-      "grad_norm": 0.4614022672176361,
-      "learning_rate": 9.465071082872246e-06,
-      "loss": 0.3907,
-      "step": 7603
-    },
-    {
-      "epoch": 0.49696098294229135,
-      "grad_norm": 0.4405010938644409,
-      "learning_rate": 9.46491392654475e-06,
-      "loss": 0.3847,
-      "step": 7604
-    },
-    {
-      "epoch": 0.4970263381478335,
-      "grad_norm": 0.45698612928390503,
-      "learning_rate": 9.464756748440324e-06,
-      "loss": 0.3885,
-      "step": 7605
-    },
-    {
-      "epoch": 0.4970916933533756,
-      "grad_norm": 0.4648208022117615,
-      "learning_rate": 9.464599548559734e-06,
-      "loss": 0.4246,
-      "step": 7606
-    },
-    {
-      "epoch": 0.49715704855891774,
-      "grad_norm": 0.49123719334602356,
-      "learning_rate": 9.464442326903745e-06,
-      "loss": 0.4749,
-      "step": 7607
-    },
-    {
-      "epoch": 0.49722240376445986,
-      "grad_norm": 0.5127215385437012,
-      "learning_rate": 9.46428508347312e-06,
-      "loss": 0.4446,
-      "step": 7608
-    },
-    {
-      "epoch": 0.49728775897000194,
-      "grad_norm": 0.42929407954216003,
-      "learning_rate": 9.464127818268635e-06,
-      "loss": 0.3494,
-      "step": 7609
-    },
-    {
-      "epoch": 0.49735311417554406,
-      "grad_norm": 0.476552277803421,
-      "learning_rate": 9.46397053129105e-06,
-      "loss": 0.4076,
-      "step": 7610
-    },
-    {
-      "epoch": 0.4974184693810862,
-      "grad_norm": 0.41644465923309326,
-      "learning_rate": 9.463813222541134e-06,
-      "loss": 0.3119,
-      "step": 7611
-    },
-    {
-      "epoch": 0.4974838245866283,
-      "grad_norm": 0.4688579738140106,
-      "learning_rate": 9.463655892019656e-06,
-      "loss": 0.4709,
-      "step": 7612
-    },
-    {
-      "epoch": 0.49754917979217045,
-      "grad_norm": 0.4837631583213806,
-      "learning_rate": 9.463498539727381e-06,
-      "loss": 0.4388,
-      "step": 7613
-    },
-    {
-      "epoch": 0.4976145349977126,
-      "grad_norm": 0.44618090987205505,
-      "learning_rate": 9.463341165665076e-06,
-      "loss": 0.3478,
-      "step": 7614
-    },
-    {
-      "epoch": 0.4976798902032547,
-      "grad_norm": 0.4475180208683014,
-      "learning_rate": 9.46318376983351e-06,
-      "loss": 0.3427,
-      "step": 7615
-    },
-    {
-      "epoch": 0.49774524540879683,
-      "grad_norm": 0.45974820852279663,
-      "learning_rate": 9.463026352233454e-06,
-      "loss": 0.3651,
-      "step": 7616
-    },
-    {
-      "epoch": 0.4978106006143389,
-      "grad_norm": 0.4433004558086395,
-      "learning_rate": 9.462868912865669e-06,
-      "loss": 0.3682,
-      "step": 7617
-    },
-    {
-      "epoch": 0.49787595581988103,
-      "grad_norm": 0.4583650529384613,
-      "learning_rate": 9.462711451730926e-06,
-      "loss": 0.4335,
-      "step": 7618
-    },
-    {
-      "epoch": 0.49794131102542316,
-      "grad_norm": 0.48022961616516113,
-      "learning_rate": 9.462553968829995e-06,
-      "loss": 0.4164,
-      "step": 7619
-    },
-    {
-      "epoch": 0.4980066662309653,
-      "grad_norm": 0.440767765045166,
-      "learning_rate": 9.462396464163642e-06,
-      "loss": 0.3813,
-      "step": 7620
-    },
-    {
-      "epoch": 0.4980720214365074,
-      "grad_norm": 0.5964440703392029,
-      "learning_rate": 9.462238937732635e-06,
-      "loss": 0.3971,
-      "step": 7621
-    },
-    {
-      "epoch": 0.49813737664204955,
-      "grad_norm": 0.43554821610450745,
-      "learning_rate": 9.462081389537742e-06,
-      "loss": 0.3623,
-      "step": 7622
-    },
-    {
-      "epoch": 0.4982027318475917,
-      "grad_norm": 0.8842023611068726,
-      "learning_rate": 9.461923819579733e-06,
-      "loss": 0.3747,
-      "step": 7623
-    },
-    {
-      "epoch": 0.4982680870531338,
-      "grad_norm": 0.4726372957229614,
-      "learning_rate": 9.461766227859376e-06,
-      "loss": 0.3909,
-      "step": 7624
-    },
-    {
-      "epoch": 0.49833344225867593,
-      "grad_norm": 0.47453370690345764,
-      "learning_rate": 9.46160861437744e-06,
-      "loss": 0.4305,
-      "step": 7625
-    },
-    {
-      "epoch": 0.498398797464218,
-      "grad_norm": 0.45982107520103455,
-      "learning_rate": 9.461450979134692e-06,
-      "loss": 0.388,
-      "step": 7626
-    },
-    {
-      "epoch": 0.49846415266976013,
-      "grad_norm": 0.4556705355644226,
-      "learning_rate": 9.461293322131903e-06,
-      "loss": 0.3979,
-      "step": 7627
-    },
-    {
-      "epoch": 0.49852950787530226,
-      "grad_norm": 0.4408996105194092,
-      "learning_rate": 9.46113564336984e-06,
-      "loss": 0.3409,
-      "step": 7628
-    },
-    {
-      "epoch": 0.4985948630808444,
-      "grad_norm": 0.41892609000205994,
-      "learning_rate": 9.460977942849274e-06,
-      "loss": 0.3577,
-      "step": 7629
-    },
-    {
-      "epoch": 0.4986602182863865,
-      "grad_norm": 0.4436472952365875,
-      "learning_rate": 9.460820220570972e-06,
-      "loss": 0.35,
-      "step": 7630
-    },
-    {
-      "epoch": 0.49872557349192864,
-      "grad_norm": 0.4511732757091522,
-      "learning_rate": 9.460662476535707e-06,
-      "loss": 0.3634,
-      "step": 7631
-    },
-    {
-      "epoch": 0.49879092869747077,
-      "grad_norm": 0.44922956824302673,
-      "learning_rate": 9.460504710744243e-06,
-      "loss": 0.3675,
-      "step": 7632
-    },
-    {
-      "epoch": 0.4988562839030129,
-      "grad_norm": 0.4979376196861267,
-      "learning_rate": 9.460346923197353e-06,
-      "loss": 0.4271,
-      "step": 7633
-    },
-    {
-      "epoch": 0.49892163910855497,
-      "grad_norm": 0.45745736360549927,
-      "learning_rate": 9.460189113895805e-06,
-      "loss": 0.3868,
-      "step": 7634
-    },
-    {
-      "epoch": 0.4989869943140971,
-      "grad_norm": 0.4833010733127594,
-      "learning_rate": 9.46003128284037e-06,
-      "loss": 0.4454,
-      "step": 7635
-    },
-    {
-      "epoch": 0.4990523495196392,
-      "grad_norm": 0.4739595353603363,
-      "learning_rate": 9.459873430031819e-06,
-      "loss": 0.3982,
-      "step": 7636
-    },
-    {
-      "epoch": 0.49911770472518135,
-      "grad_norm": 0.4343498945236206,
-      "learning_rate": 9.459715555470918e-06,
-      "loss": 0.3629,
-      "step": 7637
-    },
-    {
-      "epoch": 0.4991830599307235,
-      "grad_norm": 0.41716471314430237,
-      "learning_rate": 9.45955765915844e-06,
-      "loss": 0.325,
-      "step": 7638
-    },
-    {
-      "epoch": 0.4992484151362656,
-      "grad_norm": 0.4806312620639801,
-      "learning_rate": 9.459399741095155e-06,
-      "loss": 0.4212,
-      "step": 7639
-    },
-    {
-      "epoch": 0.49931377034180774,
-      "grad_norm": 0.42395907640457153,
-      "learning_rate": 9.459241801281833e-06,
-      "loss": 0.3453,
-      "step": 7640
-    },
-    {
-      "epoch": 0.49937912554734987,
-      "grad_norm": 0.43754011392593384,
-      "learning_rate": 9.459083839719244e-06,
-      "loss": 0.4266,
-      "step": 7641
-    },
-    {
-      "epoch": 0.499444480752892,
-      "grad_norm": 0.440309077501297,
-      "learning_rate": 9.458925856408157e-06,
-      "loss": 0.3651,
-      "step": 7642
-    },
-    {
-      "epoch": 0.49950983595843407,
-      "grad_norm": 0.45112180709838867,
-      "learning_rate": 9.458767851349344e-06,
-      "loss": 0.4001,
-      "step": 7643
-    },
-    {
-      "epoch": 0.4995751911639762,
-      "grad_norm": 0.4809283912181854,
-      "learning_rate": 9.458609824543575e-06,
-      "loss": 0.4402,
-      "step": 7644
-    },
-    {
-      "epoch": 0.4996405463695183,
-      "grad_norm": 0.4792030155658722,
-      "learning_rate": 9.458451775991622e-06,
-      "loss": 0.4109,
-      "step": 7645
-    },
-    {
-      "epoch": 0.49970590157506045,
-      "grad_norm": 0.4835467040538788,
-      "learning_rate": 9.458293705694255e-06,
-      "loss": 0.4157,
-      "step": 7646
-    },
-    {
-      "epoch": 0.4997712567806026,
-      "grad_norm": 0.4652770757675171,
-      "learning_rate": 9.458135613652245e-06,
-      "loss": 0.4242,
-      "step": 7647
-    },
-    {
-      "epoch": 0.4998366119861447,
-      "grad_norm": 0.46720170974731445,
-      "learning_rate": 9.457977499866364e-06,
-      "loss": 0.4125,
-      "step": 7648
-    },
-    {
-      "epoch": 0.49990196719168684,
-      "grad_norm": 0.4799783229827881,
-      "learning_rate": 9.457819364337382e-06,
-      "loss": 0.4267,
-      "step": 7649
-    },
-    {
-      "epoch": 0.49996732239722896,
-      "grad_norm": 0.436085045337677,
-      "learning_rate": 9.45766120706607e-06,
-      "loss": 0.3551,
-      "step": 7650
-    },
-    {
-      "epoch": 0.500032677602771,
-      "grad_norm": 0.4366012215614319,
-      "learning_rate": 9.457503028053201e-06,
-      "loss": 0.3679,
-      "step": 7651
-    },
-    {
-      "epoch": 0.5000980328083132,
-      "grad_norm": 0.4659341275691986,
-      "learning_rate": 9.457344827299543e-06,
-      "loss": 0.3962,
-      "step": 7652
-    },
-    {
-      "epoch": 0.5001633880138553,
-      "grad_norm": 0.44960251450538635,
-      "learning_rate": 9.457186604805873e-06,
-      "loss": 0.3708,
-      "step": 7653
-    },
-    {
-      "epoch": 0.5002287432193975,
-      "grad_norm": 0.45768725872039795,
-      "learning_rate": 9.45702836057296e-06,
-      "loss": 0.3933,
-      "step": 7654
-    },
-    {
-      "epoch": 0.5002940984249395,
-      "grad_norm": 0.4472002685070038,
-      "learning_rate": 9.456870094601573e-06,
-      "loss": 0.4221,
-      "step": 7655
-    },
-    {
-      "epoch": 0.5003594536304816,
-      "grad_norm": 0.4778624475002289,
-      "learning_rate": 9.456711806892488e-06,
-      "loss": 0.3788,
-      "step": 7656
-    },
-    {
-      "epoch": 0.5004248088360238,
-      "grad_norm": 0.49460941553115845,
-      "learning_rate": 9.456553497446477e-06,
-      "loss": 0.402,
-      "step": 7657
-    },
-    {
-      "epoch": 0.5004901640415659,
-      "grad_norm": 0.4877833127975464,
-      "learning_rate": 9.45639516626431e-06,
-      "loss": 0.4218,
-      "step": 7658
-    },
-    {
-      "epoch": 0.5005555192471081,
-      "grad_norm": 0.6446693539619446,
-      "learning_rate": 9.456236813346758e-06,
-      "loss": 0.4076,
-      "step": 7659
-    },
-    {
-      "epoch": 0.5006208744526501,
-      "grad_norm": 0.4596126973628998,
-      "learning_rate": 9.456078438694597e-06,
-      "loss": 0.3796,
-      "step": 7660
-    },
-    {
-      "epoch": 0.5006862296581923,
-      "grad_norm": 0.483416885137558,
-      "learning_rate": 9.455920042308598e-06,
-      "loss": 0.4063,
-      "step": 7661
-    },
-    {
-      "epoch": 0.5007515848637344,
-      "grad_norm": 0.4501649737358093,
-      "learning_rate": 9.455761624189531e-06,
-      "loss": 0.3991,
-      "step": 7662
-    },
-    {
-      "epoch": 0.5008169400692766,
-      "grad_norm": 0.42372453212738037,
-      "learning_rate": 9.455603184338174e-06,
-      "loss": 0.3454,
-      "step": 7663
-    },
-    {
-      "epoch": 0.5008822952748186,
-      "grad_norm": 0.45814022421836853,
-      "learning_rate": 9.455444722755294e-06,
-      "loss": 0.3552,
-      "step": 7664
-    },
-    {
-      "epoch": 0.5009476504803607,
-      "grad_norm": 0.46859997510910034,
-      "learning_rate": 9.455286239441668e-06,
-      "loss": 0.3987,
-      "step": 7665
-    },
-    {
-      "epoch": 0.5010130056859029,
-      "grad_norm": 0.44942134618759155,
-      "learning_rate": 9.455127734398066e-06,
-      "loss": 0.3784,
-      "step": 7666
-    },
-    {
-      "epoch": 0.501078360891445,
-      "grad_norm": 0.5670586824417114,
-      "learning_rate": 9.454969207625264e-06,
-      "loss": 0.4325,
-      "step": 7667
-    },
-    {
-      "epoch": 0.5011437160969872,
-      "grad_norm": 0.4901164174079895,
-      "learning_rate": 9.454810659124035e-06,
-      "loss": 0.4114,
-      "step": 7668
-    },
-    {
-      "epoch": 0.5012090713025292,
-      "grad_norm": 0.4789154827594757,
-      "learning_rate": 9.454652088895149e-06,
-      "loss": 0.4034,
-      "step": 7669
-    },
-    {
-      "epoch": 0.5012744265080714,
-      "grad_norm": 0.5097324252128601,
-      "learning_rate": 9.454493496939383e-06,
-      "loss": 0.4225,
-      "step": 7670
-    },
-    {
-      "epoch": 0.5013397817136135,
-      "grad_norm": 0.4476543664932251,
-      "learning_rate": 9.454334883257507e-06,
-      "loss": 0.3139,
-      "step": 7671
-    },
-    {
-      "epoch": 0.5014051369191556,
-      "grad_norm": 0.5303950309753418,
-      "learning_rate": 9.4541762478503e-06,
-      "loss": 0.3982,
-      "step": 7672
-    },
-    {
-      "epoch": 0.5014704921246977,
-      "grad_norm": 0.5110664963722229,
-      "learning_rate": 9.454017590718529e-06,
-      "loss": 0.4341,
-      "step": 7673
-    },
-    {
-      "epoch": 0.5015358473302398,
-      "grad_norm": 0.4471319317817688,
-      "learning_rate": 9.453858911862972e-06,
-      "loss": 0.3898,
-      "step": 7674
-    },
-    {
-      "epoch": 0.501601202535782,
-      "grad_norm": 0.4259728789329529,
-      "learning_rate": 9.453700211284404e-06,
-      "loss": 0.3544,
-      "step": 7675
-    },
-    {
-      "epoch": 0.5016665577413241,
-      "grad_norm": 0.4339010417461395,
-      "learning_rate": 9.453541488983595e-06,
-      "loss": 0.3714,
-      "step": 7676
-    },
-    {
-      "epoch": 0.5017319129468663,
-      "grad_norm": 0.4736887216567993,
-      "learning_rate": 9.453382744961322e-06,
-      "loss": 0.3923,
-      "step": 7677
-    },
-    {
-      "epoch": 0.5017972681524083,
-      "grad_norm": 0.45846840739250183,
-      "learning_rate": 9.453223979218359e-06,
-      "loss": 0.4146,
-      "step": 7678
-    },
-    {
-      "epoch": 0.5018626233579505,
-      "grad_norm": 0.4797497093677521,
-      "learning_rate": 9.45306519175548e-06,
-      "loss": 0.4006,
-      "step": 7679
-    },
-    {
-      "epoch": 0.5019279785634926,
-      "grad_norm": 0.4832633435726166,
-      "learning_rate": 9.45290638257346e-06,
-      "loss": 0.4162,
-      "step": 7680
-    },
-    {
-      "epoch": 0.5019933337690347,
-      "grad_norm": 0.4564540982246399,
-      "learning_rate": 9.452747551673072e-06,
-      "loss": 0.3874,
-      "step": 7681
-    },
-    {
-      "epoch": 0.5020586889745768,
-      "grad_norm": 0.4771483838558197,
-      "learning_rate": 9.452588699055089e-06,
-      "loss": 0.3998,
-      "step": 7682
-    },
-    {
-      "epoch": 0.5021240441801189,
-      "grad_norm": 0.4464082419872284,
-      "learning_rate": 9.452429824720292e-06,
-      "loss": 0.3977,
-      "step": 7683
-    },
-    {
-      "epoch": 0.5021893993856611,
-      "grad_norm": 0.4652882516384125,
-      "learning_rate": 9.452270928669451e-06,
-      "loss": 0.3685,
-      "step": 7684
-    },
-    {
-      "epoch": 0.5022547545912032,
-      "grad_norm": 0.5194174647331238,
-      "learning_rate": 9.452112010903342e-06,
-      "loss": 0.4626,
-      "step": 7685
-    },
-    {
-      "epoch": 0.5023201097967454,
-      "grad_norm": 0.43637266755104065,
-      "learning_rate": 9.451953071422741e-06,
-      "loss": 0.3608,
-      "step": 7686
-    },
-    {
-      "epoch": 0.5023854650022874,
-      "grad_norm": 0.45275160670280457,
-      "learning_rate": 9.451794110228423e-06,
-      "loss": 0.3952,
-      "step": 7687
-    },
-    {
-      "epoch": 0.5024508202078296,
-      "grad_norm": 0.4362596869468689,
-      "learning_rate": 9.451635127321161e-06,
-      "loss": 0.3654,
-      "step": 7688
-    },
-    {
-      "epoch": 0.5025161754133717,
-      "grad_norm": 0.452396035194397,
-      "learning_rate": 9.451476122701735e-06,
-      "loss": 0.3529,
-      "step": 7689
-    },
-    {
-      "epoch": 0.5025815306189138,
-      "grad_norm": 0.4514504075050354,
-      "learning_rate": 9.451317096370916e-06,
-      "loss": 0.3657,
-      "step": 7690
-    },
-    {
-      "epoch": 0.5026468858244559,
-      "grad_norm": 0.48859599232673645,
-      "learning_rate": 9.451158048329483e-06,
-      "loss": 0.4624,
-      "step": 7691
-    },
-    {
-      "epoch": 0.502712241029998,
-      "grad_norm": 0.4188653528690338,
-      "learning_rate": 9.450998978578207e-06,
-      "loss": 0.3429,
-      "step": 7692
-    },
-    {
-      "epoch": 0.5027775962355402,
-      "grad_norm": 0.47802087664604187,
-      "learning_rate": 9.450839887117871e-06,
-      "loss": 0.4537,
-      "step": 7693
-    },
-    {
-      "epoch": 0.5028429514410823,
-      "grad_norm": 0.4708169996738434,
-      "learning_rate": 9.450680773949243e-06,
-      "loss": 0.4179,
-      "step": 7694
-    },
-    {
-      "epoch": 0.5029083066466244,
-      "grad_norm": 0.46108609437942505,
-      "learning_rate": 9.450521639073106e-06,
-      "loss": 0.3941,
-      "step": 7695
-    },
-    {
-      "epoch": 0.5029736618521665,
-      "grad_norm": 0.46358931064605713,
-      "learning_rate": 9.450362482490232e-06,
-      "loss": 0.4126,
-      "step": 7696
-    },
-    {
-      "epoch": 0.5030390170577086,
-      "grad_norm": 0.47584667801856995,
-      "learning_rate": 9.450203304201398e-06,
-      "loss": 0.3975,
-      "step": 7697
-    },
-    {
-      "epoch": 0.5031043722632508,
-      "grad_norm": 0.45680591464042664,
-      "learning_rate": 9.450044104207382e-06,
-      "loss": 0.3725,
-      "step": 7698
-    },
-    {
-      "epoch": 0.5031697274687928,
-      "grad_norm": 0.527522087097168,
-      "learning_rate": 9.44988488250896e-06,
-      "loss": 0.478,
-      "step": 7699
-    },
-    {
-      "epoch": 0.503235082674335,
-      "grad_norm": 0.4551538825035095,
-      "learning_rate": 9.449725639106905e-06,
-      "loss": 0.3588,
-      "step": 7700
-    },
-    {
-      "epoch": 0.5033004378798771,
-      "grad_norm": 0.47754600644111633,
-      "learning_rate": 9.449566374001998e-06,
-      "loss": 0.43,
-      "step": 7701
-    },
-    {
-      "epoch": 0.5033657930854193,
-      "grad_norm": 0.4657004475593567,
-      "learning_rate": 9.449407087195014e-06,
-      "loss": 0.43,
-      "step": 7702
-    },
-    {
-      "epoch": 0.5034311482909614,
-      "grad_norm": 0.4533071219921112,
-      "learning_rate": 9.449247778686729e-06,
-      "loss": 0.4116,
-      "step": 7703
-    },
-    {
-      "epoch": 0.5034965034965035,
-      "grad_norm": 0.4522230923175812,
-      "learning_rate": 9.449088448477924e-06,
-      "loss": 0.3815,
-      "step": 7704
-    },
-    {
-      "epoch": 0.5035618587020456,
-      "grad_norm": 0.451035737991333,
-      "learning_rate": 9.448929096569372e-06,
-      "loss": 0.3808,
-      "step": 7705
-    },
-    {
-      "epoch": 0.5036272139075877,
-      "grad_norm": 0.4907190799713135,
-      "learning_rate": 9.448769722961852e-06,
-      "loss": 0.4145,
-      "step": 7706
-    },
-    {
-      "epoch": 0.5036925691131299,
-      "grad_norm": 0.4380427300930023,
-      "learning_rate": 9.448610327656141e-06,
-      "loss": 0.3452,
-      "step": 7707
-    },
-    {
-      "epoch": 0.503757924318672,
-      "grad_norm": 0.45540523529052734,
-      "learning_rate": 9.448450910653015e-06,
-      "loss": 0.392,
-      "step": 7708
-    },
-    {
-      "epoch": 0.5038232795242141,
-      "grad_norm": 0.4351827800273895,
-      "learning_rate": 9.448291471953252e-06,
-      "loss": 0.3518,
-      "step": 7709
-    },
-    {
-      "epoch": 0.5038886347297562,
-      "grad_norm": 0.4677373468875885,
-      "learning_rate": 9.448132011557634e-06,
-      "loss": 0.4016,
-      "step": 7710
-    },
-    {
-      "epoch": 0.5039539899352984,
-      "grad_norm": 0.463039755821228,
-      "learning_rate": 9.447972529466933e-06,
-      "loss": 0.3771,
-      "step": 7711
-    },
-    {
-      "epoch": 0.5040193451408405,
-      "grad_norm": 0.4625132381916046,
-      "learning_rate": 9.447813025681929e-06,
-      "loss": 0.4114,
-      "step": 7712
-    },
-    {
-      "epoch": 0.5040847003463826,
-      "grad_norm": 0.4245222210884094,
-      "learning_rate": 9.4476535002034e-06,
-      "loss": 0.3465,
-      "step": 7713
-    },
-    {
-      "epoch": 0.5041500555519247,
-      "grad_norm": 0.4942336976528168,
-      "learning_rate": 9.447493953032124e-06,
-      "loss": 0.3807,
-      "step": 7714
-    },
-    {
-      "epoch": 0.5042154107574668,
-      "grad_norm": 0.4409201443195343,
-      "learning_rate": 9.447334384168881e-06,
-      "loss": 0.3909,
-      "step": 7715
-    },
-    {
-      "epoch": 0.504280765963009,
-      "grad_norm": 0.4462164640426636,
-      "learning_rate": 9.447174793614444e-06,
-      "loss": 0.3674,
-      "step": 7716
-    },
-    {
-      "epoch": 0.504346121168551,
-      "grad_norm": 0.4723726809024811,
-      "learning_rate": 9.447015181369597e-06,
-      "loss": 0.4121,
-      "step": 7717
-    },
-    {
-      "epoch": 0.5044114763740932,
-      "grad_norm": 0.46016383171081543,
-      "learning_rate": 9.446855547435117e-06,
-      "loss": 0.3857,
-      "step": 7718
-    },
-    {
-      "epoch": 0.5044768315796353,
-      "grad_norm": 0.4830379784107208,
-      "learning_rate": 9.446695891811781e-06,
-      "loss": 0.4014,
-      "step": 7719
-    },
-    {
-      "epoch": 0.5045421867851775,
-      "grad_norm": 0.4311649203300476,
-      "learning_rate": 9.446536214500367e-06,
-      "loss": 0.3768,
-      "step": 7720
-    },
-    {
-      "epoch": 0.5046075419907196,
-      "grad_norm": 0.4620884954929352,
-      "learning_rate": 9.446376515501659e-06,
-      "loss": 0.4095,
-      "step": 7721
-    },
-    {
-      "epoch": 0.5046728971962616,
-      "grad_norm": 0.45689043402671814,
-      "learning_rate": 9.44621679481643e-06,
-      "loss": 0.3834,
-      "step": 7722
-    },
-    {
-      "epoch": 0.5047382524018038,
-      "grad_norm": 0.4588378071784973,
-      "learning_rate": 9.446057052445463e-06,
-      "loss": 0.3915,
-      "step": 7723
-    },
-    {
-      "epoch": 0.5048036076073459,
-      "grad_norm": 0.5513094067573547,
-      "learning_rate": 9.445897288389533e-06,
-      "loss": 0.3808,
-      "step": 7724
-    },
-    {
-      "epoch": 0.5048689628128881,
-      "grad_norm": 0.4702783226966858,
-      "learning_rate": 9.445737502649425e-06,
-      "loss": 0.3584,
-      "step": 7725
-    },
-    {
-      "epoch": 0.5049343180184301,
-      "grad_norm": 0.4639442265033722,
-      "learning_rate": 9.445577695225914e-06,
-      "loss": 0.4167,
-      "step": 7726
-    },
-    {
-      "epoch": 0.5049996732239723,
-      "grad_norm": 0.4771186113357544,
-      "learning_rate": 9.445417866119779e-06,
-      "loss": 0.4129,
-      "step": 7727
-    },
-    {
-      "epoch": 0.5050650284295144,
-      "grad_norm": 0.4485853314399719,
-      "learning_rate": 9.445258015331802e-06,
-      "loss": 0.4147,
-      "step": 7728
-    },
-    {
-      "epoch": 0.5051303836350566,
-      "grad_norm": 0.47932490706443787,
-      "learning_rate": 9.44509814286276e-06,
-      "loss": 0.3547,
-      "step": 7729
-    },
-    {
-      "epoch": 0.5051957388405987,
-      "grad_norm": 0.4371171295642853,
-      "learning_rate": 9.444938248713436e-06,
-      "loss": 0.3471,
-      "step": 7730
-    },
-    {
-      "epoch": 0.5052610940461407,
-      "grad_norm": 0.45352810621261597,
-      "learning_rate": 9.44477833288461e-06,
-      "loss": 0.4164,
-      "step": 7731
-    },
-    {
-      "epoch": 0.5053264492516829,
-      "grad_norm": 0.43929749727249146,
-      "learning_rate": 9.444618395377057e-06,
-      "loss": 0.3732,
-      "step": 7732
-    },
-    {
-      "epoch": 0.505391804457225,
-      "grad_norm": 0.49312737584114075,
-      "learning_rate": 9.444458436191562e-06,
-      "loss": 0.4635,
-      "step": 7733
-    },
-    {
-      "epoch": 0.5054571596627672,
-      "grad_norm": 0.5206209421157837,
-      "learning_rate": 9.444298455328903e-06,
-      "loss": 0.3835,
-      "step": 7734
-    },
-    {
-      "epoch": 0.5055225148683092,
-      "grad_norm": 0.5052789449691772,
-      "learning_rate": 9.444138452789862e-06,
-      "loss": 0.4572,
-      "step": 7735
-    },
-    {
-      "epoch": 0.5055878700738514,
-      "grad_norm": 0.4622563123703003,
-      "learning_rate": 9.443978428575216e-06,
-      "loss": 0.3999,
-      "step": 7736
-    },
-    {
-      "epoch": 0.5056532252793935,
-      "grad_norm": 0.4477774202823639,
-      "learning_rate": 9.44381838268575e-06,
-      "loss": 0.3646,
-      "step": 7737
-    },
-    {
-      "epoch": 0.5057185804849357,
-      "grad_norm": 0.4669139087200165,
-      "learning_rate": 9.44365831512224e-06,
-      "loss": 0.3797,
-      "step": 7738
-    },
-    {
-      "epoch": 0.5057839356904777,
-      "grad_norm": 0.42422449588775635,
-      "learning_rate": 9.44349822588547e-06,
-      "loss": 0.3642,
-      "step": 7739
-    },
-    {
-      "epoch": 0.5058492908960198,
-      "grad_norm": 0.431505024433136,
-      "learning_rate": 9.443338114976222e-06,
-      "loss": 0.3287,
-      "step": 7740
-    },
-    {
-      "epoch": 0.505914646101562,
-      "grad_norm": 0.4315491318702698,
-      "learning_rate": 9.443177982395272e-06,
-      "loss": 0.3306,
-      "step": 7741
-    },
-    {
-      "epoch": 0.5059800013071041,
-      "grad_norm": 0.41238635778427124,
-      "learning_rate": 9.443017828143403e-06,
-      "loss": 0.3391,
-      "step": 7742
-    },
-    {
-      "epoch": 0.5060453565126463,
-      "grad_norm": 0.4434930682182312,
-      "learning_rate": 9.442857652221398e-06,
-      "loss": 0.3726,
-      "step": 7743
-    },
-    {
-      "epoch": 0.5061107117181883,
-      "grad_norm": 0.4321887791156769,
-      "learning_rate": 9.442697454630039e-06,
-      "loss": 0.3517,
-      "step": 7744
-    },
-    {
-      "epoch": 0.5061760669237305,
-      "grad_norm": 0.4747454524040222,
-      "learning_rate": 9.442537235370103e-06,
-      "loss": 0.4571,
-      "step": 7745
-    },
-    {
-      "epoch": 0.5062414221292726,
-      "grad_norm": 0.4539855420589447,
-      "learning_rate": 9.442376994442375e-06,
-      "loss": 0.4042,
-      "step": 7746
-    },
-    {
-      "epoch": 0.5063067773348148,
-      "grad_norm": 0.45672607421875,
-      "learning_rate": 9.442216731847635e-06,
-      "loss": 0.4225,
-      "step": 7747
-    },
-    {
-      "epoch": 0.5063721325403568,
-      "grad_norm": 0.5286438465118408,
-      "learning_rate": 9.442056447586665e-06,
-      "loss": 0.3663,
-      "step": 7748
-    },
-    {
-      "epoch": 0.5064374877458989,
-      "grad_norm": 0.42539289593696594,
-      "learning_rate": 9.441896141660246e-06,
-      "loss": 0.3283,
-      "step": 7749
-    },
-    {
-      "epoch": 0.5065028429514411,
-      "grad_norm": 0.47858306765556335,
-      "learning_rate": 9.441735814069161e-06,
-      "loss": 0.4016,
-      "step": 7750
-    },
-    {
-      "epoch": 0.5065681981569832,
-      "grad_norm": 0.4335874915122986,
-      "learning_rate": 9.441575464814193e-06,
-      "loss": 0.373,
-      "step": 7751
-    },
-    {
-      "epoch": 0.5066335533625254,
-      "grad_norm": 0.45669421553611755,
-      "learning_rate": 9.441415093896123e-06,
-      "loss": 0.4007,
-      "step": 7752
-    },
-    {
-      "epoch": 0.5066989085680674,
-      "grad_norm": 0.4713069498538971,
-      "learning_rate": 9.441254701315731e-06,
-      "loss": 0.3703,
-      "step": 7753
-    },
-    {
-      "epoch": 0.5067642637736096,
-      "grad_norm": 0.4867474138736725,
-      "learning_rate": 9.441094287073801e-06,
-      "loss": 0.4422,
-      "step": 7754
-    },
-    {
-      "epoch": 0.5068296189791517,
-      "grad_norm": 0.45350489020347595,
-      "learning_rate": 9.440933851171117e-06,
-      "loss": 0.3916,
-      "step": 7755
-    },
-    {
-      "epoch": 0.5068949741846938,
-      "grad_norm": 0.443925142288208,
-      "learning_rate": 9.44077339360846e-06,
-      "loss": 0.4058,
-      "step": 7756
-    },
-    {
-      "epoch": 0.5069603293902359,
-      "grad_norm": 0.45000678300857544,
-      "learning_rate": 9.440612914386614e-06,
-      "loss": 0.4003,
-      "step": 7757
-    },
-    {
-      "epoch": 0.507025684595778,
-      "grad_norm": 0.46410927176475525,
-      "learning_rate": 9.440452413506358e-06,
-      "loss": 0.4028,
-      "step": 7758
-    },
-    {
-      "epoch": 0.5070910398013202,
-      "grad_norm": 0.470266729593277,
-      "learning_rate": 9.440291890968479e-06,
-      "loss": 0.403,
-      "step": 7759
-    },
-    {
-      "epoch": 0.5071563950068623,
-      "grad_norm": 0.8602852821350098,
-      "learning_rate": 9.440131346773757e-06,
-      "loss": 0.4246,
-      "step": 7760
-    },
-    {
-      "epoch": 0.5072217502124045,
-      "grad_norm": 0.515372633934021,
-      "learning_rate": 9.439970780922975e-06,
-      "loss": 0.4407,
-      "step": 7761
-    },
-    {
-      "epoch": 0.5072871054179465,
-      "grad_norm": 0.43363937735557556,
-      "learning_rate": 9.43981019341692e-06,
-      "loss": 0.3447,
-      "step": 7762
-    },
-    {
-      "epoch": 0.5073524606234887,
-      "grad_norm": 0.4343889653682709,
-      "learning_rate": 9.439649584256372e-06,
-      "loss": 0.3647,
-      "step": 7763
-    },
-    {
-      "epoch": 0.5074178158290308,
-      "grad_norm": 0.43482160568237305,
-      "learning_rate": 9.439488953442114e-06,
-      "loss": 0.3586,
-      "step": 7764
-    },
-    {
-      "epoch": 0.5074831710345729,
-      "grad_norm": 0.43836653232574463,
-      "learning_rate": 9.43932830097493e-06,
-      "loss": 0.365,
-      "step": 7765
-    },
-    {
-      "epoch": 0.507548526240115,
-      "grad_norm": 0.49678322672843933,
-      "learning_rate": 9.439167626855605e-06,
-      "loss": 0.429,
-      "step": 7766
-    },
-    {
-      "epoch": 0.5076138814456571,
-      "grad_norm": 0.47993841767311096,
-      "learning_rate": 9.439006931084921e-06,
-      "loss": 0.4042,
-      "step": 7767
-    },
-    {
-      "epoch": 0.5076792366511993,
-      "grad_norm": 0.4363028407096863,
-      "learning_rate": 9.438846213663664e-06,
-      "loss": 0.3648,
-      "step": 7768
-    },
-    {
-      "epoch": 0.5077445918567414,
-      "grad_norm": 0.48934856057167053,
-      "learning_rate": 9.438685474592615e-06,
-      "loss": 0.4221,
-      "step": 7769
-    },
-    {
-      "epoch": 0.5078099470622836,
-      "grad_norm": 0.5310774445533752,
-      "learning_rate": 9.43852471387256e-06,
-      "loss": 0.4706,
-      "step": 7770
-    },
-    {
-      "epoch": 0.5078753022678256,
-      "grad_norm": 0.41704484820365906,
-      "learning_rate": 9.438363931504282e-06,
-      "loss": 0.3225,
-      "step": 7771
-    },
-    {
-      "epoch": 0.5079406574733678,
-      "grad_norm": 0.4534618854522705,
-      "learning_rate": 9.438203127488564e-06,
-      "loss": 0.372,
-      "step": 7772
-    },
-    {
-      "epoch": 0.5080060126789099,
-      "grad_norm": 0.43370527029037476,
-      "learning_rate": 9.438042301826193e-06,
-      "loss": 0.3825,
-      "step": 7773
-    },
-    {
-      "epoch": 0.508071367884452,
-      "grad_norm": 0.4758480489253998,
-      "learning_rate": 9.437881454517952e-06,
-      "loss": 0.4192,
-      "step": 7774
-    },
-    {
-      "epoch": 0.5081367230899941,
-      "grad_norm": 0.46820536255836487,
-      "learning_rate": 9.437720585564628e-06,
-      "loss": 0.4079,
-      "step": 7775
-    },
-    {
-      "epoch": 0.5082020782955362,
-      "grad_norm": 0.43157297372817993,
-      "learning_rate": 9.437559694967e-06,
-      "loss": 0.3535,
-      "step": 7776
-    },
-    {
-      "epoch": 0.5082674335010784,
-      "grad_norm": 0.4799809157848358,
-      "learning_rate": 9.437398782725857e-06,
-      "loss": 0.3967,
-      "step": 7777
-    },
-    {
-      "epoch": 0.5083327887066205,
-      "grad_norm": 0.5004230737686157,
-      "learning_rate": 9.437237848841982e-06,
-      "loss": 0.3799,
-      "step": 7778
-    },
-    {
-      "epoch": 0.5083981439121626,
-      "grad_norm": 0.47232672572135925,
-      "learning_rate": 9.437076893316163e-06,
-      "loss": 0.3857,
-      "step": 7779
-    },
-    {
-      "epoch": 0.5084634991177047,
-      "grad_norm": 0.4624641239643097,
-      "learning_rate": 9.436915916149181e-06,
-      "loss": 0.3518,
-      "step": 7780
-    },
-    {
-      "epoch": 0.5085288543232468,
-      "grad_norm": 0.44828665256500244,
-      "learning_rate": 9.436754917341823e-06,
-      "loss": 0.3246,
-      "step": 7781
-    },
-    {
-      "epoch": 0.508594209528789,
-      "grad_norm": 0.46594393253326416,
-      "learning_rate": 9.436593896894876e-06,
-      "loss": 0.3773,
-      "step": 7782
-    },
-    {
-      "epoch": 0.508659564734331,
-      "grad_norm": 0.5000684261322021,
-      "learning_rate": 9.436432854809124e-06,
-      "loss": 0.4605,
-      "step": 7783
-    },
-    {
-      "epoch": 0.5087249199398732,
-      "grad_norm": 0.4356588125228882,
-      "learning_rate": 9.43627179108535e-06,
-      "loss": 0.3663,
-      "step": 7784
-    },
-    {
-      "epoch": 0.5087902751454153,
-      "grad_norm": 0.4482481777667999,
-      "learning_rate": 9.436110705724341e-06,
-      "loss": 0.3675,
-      "step": 7785
-    },
-    {
-      "epoch": 0.5088556303509575,
-      "grad_norm": 0.4329874813556671,
-      "learning_rate": 9.435949598726885e-06,
-      "loss": 0.356,
-      "step": 7786
-    },
-    {
-      "epoch": 0.5089209855564996,
-      "grad_norm": 0.45425376296043396,
-      "learning_rate": 9.435788470093766e-06,
-      "loss": 0.3541,
-      "step": 7787
-    },
-    {
-      "epoch": 0.5089863407620417,
-      "grad_norm": 0.47848591208457947,
-      "learning_rate": 9.435627319825769e-06,
-      "loss": 0.4598,
-      "step": 7788
-    },
-    {
-      "epoch": 0.5090516959675838,
-      "grad_norm": 0.4244730472564697,
-      "learning_rate": 9.435466147923682e-06,
-      "loss": 0.342,
-      "step": 7789
-    },
-    {
-      "epoch": 0.5091170511731259,
-      "grad_norm": 0.48533114790916443,
-      "learning_rate": 9.435304954388288e-06,
-      "loss": 0.4343,
-      "step": 7790
-    },
-    {
-      "epoch": 0.5091824063786681,
-      "grad_norm": 0.46364644169807434,
-      "learning_rate": 9.435143739220378e-06,
-      "loss": 0.371,
-      "step": 7791
-    },
-    {
-      "epoch": 0.5092477615842101,
-      "grad_norm": 0.4621589481830597,
-      "learning_rate": 9.434982502420733e-06,
-      "loss": 0.3861,
-      "step": 7792
-    },
-    {
-      "epoch": 0.5093131167897523,
-      "grad_norm": 0.4631290137767792,
-      "learning_rate": 9.434821243990145e-06,
-      "loss": 0.3889,
-      "step": 7793
-    },
-    {
-      "epoch": 0.5093784719952944,
-      "grad_norm": 0.47467970848083496,
-      "learning_rate": 9.434659963929394e-06,
-      "loss": 0.4032,
-      "step": 7794
-    },
-    {
-      "epoch": 0.5094438272008366,
-      "grad_norm": 0.4567209482192993,
-      "learning_rate": 9.434498662239271e-06,
-      "loss": 0.4168,
-      "step": 7795
-    },
-    {
-      "epoch": 0.5095091824063787,
-      "grad_norm": 0.45567426085472107,
-      "learning_rate": 9.434337338920562e-06,
-      "loss": 0.3438,
-      "step": 7796
-    },
-    {
-      "epoch": 0.5095745376119208,
-      "grad_norm": 0.4654828608036041,
-      "learning_rate": 9.434175993974055e-06,
-      "loss": 0.3891,
-      "step": 7797
-    },
-    {
-      "epoch": 0.5096398928174629,
-      "grad_norm": 0.449202299118042,
-      "learning_rate": 9.434014627400534e-06,
-      "loss": 0.3865,
-      "step": 7798
-    },
-    {
-      "epoch": 0.509705248023005,
-      "grad_norm": 0.4640216827392578,
-      "learning_rate": 9.433853239200787e-06,
-      "loss": 0.4083,
-      "step": 7799
-    },
-    {
-      "epoch": 0.5097706032285472,
-      "grad_norm": 0.4358929395675659,
-      "learning_rate": 9.433691829375605e-06,
-      "loss": 0.3395,
-      "step": 7800
-    },
-    {
-      "epoch": 0.5098359584340892,
-      "grad_norm": 0.48115190863609314,
-      "learning_rate": 9.433530397925768e-06,
-      "loss": 0.3927,
-      "step": 7801
-    },
-    {
-      "epoch": 0.5099013136396314,
-      "grad_norm": 0.4378347396850586,
-      "learning_rate": 9.433368944852069e-06,
-      "loss": 0.3703,
-      "step": 7802
-    },
-    {
-      "epoch": 0.5099666688451735,
-      "grad_norm": 0.5631244778633118,
-      "learning_rate": 9.433207470155294e-06,
-      "loss": 0.3516,
-      "step": 7803
-    },
-    {
-      "epoch": 0.5100320240507157,
-      "grad_norm": 0.4904472827911377,
-      "learning_rate": 9.43304597383623e-06,
-      "loss": 0.4362,
-      "step": 7804
-    },
-    {
-      "epoch": 0.5100973792562578,
-      "grad_norm": 0.45113232731819153,
-      "learning_rate": 9.432884455895665e-06,
-      "loss": 0.3536,
-      "step": 7805
-    },
-    {
-      "epoch": 0.5101627344617998,
-      "grad_norm": 0.4851573705673218,
-      "learning_rate": 9.432722916334387e-06,
-      "loss": 0.4351,
-      "step": 7806
-    },
-    {
-      "epoch": 0.510228089667342,
-      "grad_norm": 0.4441041350364685,
-      "learning_rate": 9.432561355153183e-06,
-      "loss": 0.3571,
-      "step": 7807
-    },
-    {
-      "epoch": 0.5102934448728841,
-      "grad_norm": 0.4777660369873047,
-      "learning_rate": 9.432399772352843e-06,
-      "loss": 0.4167,
-      "step": 7808
-    },
-    {
-      "epoch": 0.5103588000784263,
-      "grad_norm": 0.5215006470680237,
-      "learning_rate": 9.432238167934153e-06,
-      "loss": 0.4379,
-      "step": 7809
-    },
-    {
-      "epoch": 0.5104241552839683,
-      "grad_norm": 0.45751526951789856,
-      "learning_rate": 9.432076541897902e-06,
-      "loss": 0.406,
-      "step": 7810
-    },
-    {
-      "epoch": 0.5104895104895105,
-      "grad_norm": 0.49232354760169983,
-      "learning_rate": 9.431914894244878e-06,
-      "loss": 0.445,
-      "step": 7811
-    },
-    {
-      "epoch": 0.5105548656950526,
-      "grad_norm": 0.4416586458683014,
-      "learning_rate": 9.43175322497587e-06,
-      "loss": 0.3829,
-      "step": 7812
-    },
-    {
-      "epoch": 0.5106202209005948,
-      "grad_norm": 0.46103307604789734,
-      "learning_rate": 9.431591534091666e-06,
-      "loss": 0.3629,
-      "step": 7813
-    },
-    {
-      "epoch": 0.5106855761061369,
-      "grad_norm": 0.4570455551147461,
-      "learning_rate": 9.431429821593055e-06,
-      "loss": 0.4046,
-      "step": 7814
-    },
-    {
-      "epoch": 0.5107509313116789,
-      "grad_norm": 0.4625372886657715,
-      "learning_rate": 9.431268087480826e-06,
-      "loss": 0.4266,
-      "step": 7815
-    },
-    {
-      "epoch": 0.5108162865172211,
-      "grad_norm": 0.44160646200180054,
-      "learning_rate": 9.431106331755766e-06,
-      "loss": 0.3812,
-      "step": 7816
-    },
-    {
-      "epoch": 0.5108816417227632,
-      "grad_norm": 0.4473101496696472,
-      "learning_rate": 9.430944554418668e-06,
-      "loss": 0.3827,
-      "step": 7817
-    },
-    {
-      "epoch": 0.5109469969283054,
-      "grad_norm": 0.458060085773468,
-      "learning_rate": 9.430782755470316e-06,
-      "loss": 0.3877,
-      "step": 7818
-    },
-    {
-      "epoch": 0.5110123521338474,
-      "grad_norm": 0.4500366151332855,
-      "learning_rate": 9.430620934911503e-06,
-      "loss": 0.3948,
-      "step": 7819
-    },
-    {
-      "epoch": 0.5110777073393896,
-      "grad_norm": 0.45969077944755554,
-      "learning_rate": 9.430459092743016e-06,
-      "loss": 0.3902,
-      "step": 7820
-    },
-    {
-      "epoch": 0.5111430625449317,
-      "grad_norm": 0.43781375885009766,
-      "learning_rate": 9.430297228965645e-06,
-      "loss": 0.3538,
-      "step": 7821
-    },
-    {
-      "epoch": 0.5112084177504739,
-      "grad_norm": 0.4211559593677521,
-      "learning_rate": 9.430135343580181e-06,
-      "loss": 0.354,
-      "step": 7822
-    },
-    {
-      "epoch": 0.511273772956016,
-      "grad_norm": 0.4532836079597473,
-      "learning_rate": 9.429973436587409e-06,
-      "loss": 0.3727,
-      "step": 7823
-    },
-    {
-      "epoch": 0.511339128161558,
-      "grad_norm": 0.484272837638855,
-      "learning_rate": 9.429811507988124e-06,
-      "loss": 0.4122,
-      "step": 7824
-    },
-    {
-      "epoch": 0.5114044833671002,
-      "grad_norm": 0.5284202098846436,
-      "learning_rate": 9.429649557783115e-06,
-      "loss": 0.4599,
-      "step": 7825
-    },
-    {
-      "epoch": 0.5114698385726423,
-      "grad_norm": 0.43804940581321716,
-      "learning_rate": 9.429487585973167e-06,
-      "loss": 0.3355,
-      "step": 7826
-    },
-    {
-      "epoch": 0.5115351937781845,
-      "grad_norm": 0.4836186468601227,
-      "learning_rate": 9.429325592559077e-06,
-      "loss": 0.444,
-      "step": 7827
-    },
-    {
-      "epoch": 0.5116005489837265,
-      "grad_norm": 0.43310973048210144,
-      "learning_rate": 9.42916357754163e-06,
-      "loss": 0.3764,
-      "step": 7828
-    },
-    {
-      "epoch": 0.5116659041892687,
-      "grad_norm": 0.43801456689834595,
-      "learning_rate": 9.429001540921618e-06,
-      "loss": 0.4043,
-      "step": 7829
-    },
-    {
-      "epoch": 0.5117312593948108,
-      "grad_norm": 0.43409159779548645,
-      "learning_rate": 9.428839482699831e-06,
-      "loss": 0.3789,
-      "step": 7830
-    },
-    {
-      "epoch": 0.511796614600353,
-      "grad_norm": 0.45470452308654785,
-      "learning_rate": 9.42867740287706e-06,
-      "loss": 0.4111,
-      "step": 7831
-    },
-    {
-      "epoch": 0.511861969805895,
-      "grad_norm": 0.44637343287467957,
-      "learning_rate": 9.428515301454095e-06,
-      "loss": 0.3717,
-      "step": 7832
-    },
-    {
-      "epoch": 0.5119273250114371,
-      "grad_norm": 0.45838499069213867,
-      "learning_rate": 9.428353178431726e-06,
-      "loss": 0.4231,
-      "step": 7833
-    },
-    {
-      "epoch": 0.5119926802169793,
-      "grad_norm": 0.485531747341156,
-      "learning_rate": 9.428191033810746e-06,
-      "loss": 0.4097,
-      "step": 7834
-    },
-    {
-      "epoch": 0.5120580354225214,
-      "grad_norm": 0.4642377197742462,
-      "learning_rate": 9.428028867591943e-06,
-      "loss": 0.3873,
-      "step": 7835
-    },
-    {
-      "epoch": 0.5121233906280636,
-      "grad_norm": 0.46806132793426514,
-      "learning_rate": 9.427866679776109e-06,
-      "loss": 0.3881,
-      "step": 7836
-    },
-    {
-      "epoch": 0.5121887458336056,
-      "grad_norm": 0.4533544182777405,
-      "learning_rate": 9.427704470364035e-06,
-      "loss": 0.3792,
-      "step": 7837
-    },
-    {
-      "epoch": 0.5122541010391478,
-      "grad_norm": 0.4343569278717041,
-      "learning_rate": 9.427542239356512e-06,
-      "loss": 0.3568,
-      "step": 7838
-    },
-    {
-      "epoch": 0.5123194562446899,
-      "grad_norm": 0.45173031091690063,
-      "learning_rate": 9.427379986754333e-06,
-      "loss": 0.3966,
-      "step": 7839
-    },
-    {
-      "epoch": 0.512384811450232,
-      "grad_norm": 0.45075392723083496,
-      "learning_rate": 9.427217712558288e-06,
-      "loss": 0.3993,
-      "step": 7840
-    },
-    {
-      "epoch": 0.5124501666557741,
-      "grad_norm": 0.45193302631378174,
-      "learning_rate": 9.427055416769169e-06,
-      "loss": 0.4206,
-      "step": 7841
-    },
-    {
-      "epoch": 0.5125155218613162,
-      "grad_norm": 0.4493101239204407,
-      "learning_rate": 9.426893099387767e-06,
-      "loss": 0.3855,
-      "step": 7842
-    },
-    {
-      "epoch": 0.5125808770668584,
-      "grad_norm": 0.44628816843032837,
-      "learning_rate": 9.426730760414871e-06,
-      "loss": 0.387,
-      "step": 7843
-    },
-    {
-      "epoch": 0.5126462322724005,
-      "grad_norm": 0.45157426595687866,
-      "learning_rate": 9.426568399851277e-06,
-      "loss": 0.4002,
-      "step": 7844
-    },
-    {
-      "epoch": 0.5127115874779427,
-      "grad_norm": 0.4979647994041443,
-      "learning_rate": 9.426406017697777e-06,
-      "loss": 0.4369,
-      "step": 7845
-    },
-    {
-      "epoch": 0.5127769426834847,
-      "grad_norm": 0.4294925034046173,
-      "learning_rate": 9.42624361395516e-06,
-      "loss": 0.3506,
-      "step": 7846
-    },
-    {
-      "epoch": 0.5128422978890269,
-      "grad_norm": 0.4421248137950897,
-      "learning_rate": 9.42608118862422e-06,
-      "loss": 0.3887,
-      "step": 7847
-    },
-    {
-      "epoch": 0.512907653094569,
-      "grad_norm": 0.4316932260990143,
-      "learning_rate": 9.425918741705749e-06,
-      "loss": 0.3528,
-      "step": 7848
-    },
-    {
-      "epoch": 0.512973008300111,
-      "grad_norm": 0.45609912276268005,
-      "learning_rate": 9.425756273200536e-06,
-      "loss": 0.4086,
-      "step": 7849
-    },
-    {
-      "epoch": 0.5130383635056532,
-      "grad_norm": 0.44079044461250305,
-      "learning_rate": 9.42559378310938e-06,
-      "loss": 0.3338,
-      "step": 7850
-    },
-    {
-      "epoch": 0.5131037187111953,
-      "grad_norm": 0.4636673033237457,
-      "learning_rate": 9.42543127143307e-06,
-      "loss": 0.3685,
-      "step": 7851
-    },
-    {
-      "epoch": 0.5131690739167375,
-      "grad_norm": 0.46439874172210693,
-      "learning_rate": 9.425268738172397e-06,
-      "loss": 0.4018,
-      "step": 7852
-    },
-    {
-      "epoch": 0.5132344291222796,
-      "grad_norm": 0.48195880651474,
-      "learning_rate": 9.425106183328156e-06,
-      "loss": 0.4504,
-      "step": 7853
-    },
-    {
-      "epoch": 0.5132997843278218,
-      "grad_norm": 0.4582245349884033,
-      "learning_rate": 9.424943606901137e-06,
-      "loss": 0.3982,
-      "step": 7854
-    },
-    {
-      "epoch": 0.5133651395333638,
-      "grad_norm": 0.4763210415840149,
-      "learning_rate": 9.424781008892138e-06,
-      "loss": 0.4384,
-      "step": 7855
-    },
-    {
-      "epoch": 0.513430494738906,
-      "grad_norm": 0.959374189376831,
-      "learning_rate": 9.424618389301947e-06,
-      "loss": 0.4676,
-      "step": 7856
-    },
-    {
-      "epoch": 0.5134958499444481,
-      "grad_norm": 0.41970404982566833,
-      "learning_rate": 9.424455748131362e-06,
-      "loss": 0.3329,
-      "step": 7857
-    },
-    {
-      "epoch": 0.5135612051499902,
-      "grad_norm": 0.49007922410964966,
-      "learning_rate": 9.424293085381172e-06,
-      "loss": 0.4307,
-      "step": 7858
-    },
-    {
-      "epoch": 0.5136265603555323,
-      "grad_norm": 0.4442076086997986,
-      "learning_rate": 9.424130401052171e-06,
-      "loss": 0.352,
-      "step": 7859
-    },
-    {
-      "epoch": 0.5136919155610744,
-      "grad_norm": 0.4454544484615326,
-      "learning_rate": 9.423967695145154e-06,
-      "loss": 0.3823,
-      "step": 7860
-    },
-    {
-      "epoch": 0.5137572707666166,
-      "grad_norm": 0.490543931722641,
-      "learning_rate": 9.423804967660914e-06,
-      "loss": 0.4189,
-      "step": 7861
-    },
-    {
-      "epoch": 0.5138226259721587,
-      "grad_norm": 0.45555487275123596,
-      "learning_rate": 9.423642218600244e-06,
-      "loss": 0.3853,
-      "step": 7862
-    },
-    {
-      "epoch": 0.5138879811777008,
-      "grad_norm": 0.4604410231113434,
-      "learning_rate": 9.42347944796394e-06,
-      "loss": 0.381,
-      "step": 7863
-    },
-    {
-      "epoch": 0.5139533363832429,
-      "grad_norm": 0.4191436171531677,
-      "learning_rate": 9.423316655752793e-06,
-      "loss": 0.3562,
-      "step": 7864
-    },
-    {
-      "epoch": 0.514018691588785,
-      "grad_norm": 0.482517272233963,
-      "learning_rate": 9.423153841967598e-06,
-      "loss": 0.4471,
-      "step": 7865
-    },
-    {
-      "epoch": 0.5140840467943272,
-      "grad_norm": 0.42155736684799194,
-      "learning_rate": 9.42299100660915e-06,
-      "loss": 0.3321,
-      "step": 7866
-    },
-    {
-      "epoch": 0.5141494019998692,
-      "grad_norm": 0.4597262442111969,
-      "learning_rate": 9.422828149678244e-06,
-      "loss": 0.3689,
-      "step": 7867
-    },
-    {
-      "epoch": 0.5142147572054114,
-      "grad_norm": 0.4536152184009552,
-      "learning_rate": 9.42266527117567e-06,
-      "loss": 0.3856,
-      "step": 7868
-    },
-    {
-      "epoch": 0.5142801124109535,
-      "grad_norm": 0.46638286113739014,
-      "learning_rate": 9.422502371102228e-06,
-      "loss": 0.4212,
-      "step": 7869
-    },
-    {
-      "epoch": 0.5143454676164957,
-      "grad_norm": 0.48029616475105286,
-      "learning_rate": 9.422339449458709e-06,
-      "loss": 0.4453,
-      "step": 7870
-    },
-    {
-      "epoch": 0.5144108228220378,
-      "grad_norm": 0.4586377739906311,
-      "learning_rate": 9.422176506245908e-06,
-      "loss": 0.3542,
-      "step": 7871
-    },
-    {
-      "epoch": 0.51447617802758,
-      "grad_norm": 0.4356926381587982,
-      "learning_rate": 9.42201354146462e-06,
-      "loss": 0.3269,
-      "step": 7872
-    },
-    {
-      "epoch": 0.514541533233122,
-      "grad_norm": 0.46058180928230286,
-      "learning_rate": 9.421850555115641e-06,
-      "loss": 0.419,
-      "step": 7873
-    },
-    {
-      "epoch": 0.5146068884386641,
-      "grad_norm": 0.46969467401504517,
-      "learning_rate": 9.421687547199763e-06,
-      "loss": 0.414,
-      "step": 7874
-    },
-    {
-      "epoch": 0.5146722436442063,
-      "grad_norm": 0.5013342499732971,
-      "learning_rate": 9.421524517717784e-06,
-      "loss": 0.463,
-      "step": 7875
-    },
-    {
-      "epoch": 0.5147375988497483,
-      "grad_norm": 0.4446825087070465,
-      "learning_rate": 9.4213614666705e-06,
-      "loss": 0.3951,
-      "step": 7876
-    },
-    {
-      "epoch": 0.5148029540552905,
-      "grad_norm": 0.4600411355495453,
-      "learning_rate": 9.421198394058702e-06,
-      "loss": 0.4068,
-      "step": 7877
-    },
-    {
-      "epoch": 0.5148683092608326,
-      "grad_norm": 0.4615230858325958,
-      "learning_rate": 9.421035299883188e-06,
-      "loss": 0.3538,
-      "step": 7878
-    },
-    {
-      "epoch": 0.5149336644663748,
-      "grad_norm": 0.4540770351886749,
-      "learning_rate": 9.420872184144754e-06,
-      "loss": 0.3812,
-      "step": 7879
-    },
-    {
-      "epoch": 0.5149990196719169,
-      "grad_norm": 0.46806600689888,
-      "learning_rate": 9.420709046844196e-06,
-      "loss": 0.4016,
-      "step": 7880
-    },
-    {
-      "epoch": 0.515064374877459,
-      "grad_norm": 0.44293394684791565,
-      "learning_rate": 9.420545887982307e-06,
-      "loss": 0.3761,
-      "step": 7881
-    },
-    {
-      "epoch": 0.5151297300830011,
-      "grad_norm": 0.44064861536026,
-      "learning_rate": 9.420382707559885e-06,
-      "loss": 0.3823,
-      "step": 7882
-    },
-    {
-      "epoch": 0.5151950852885432,
-      "grad_norm": 0.46114489436149597,
-      "learning_rate": 9.420219505577724e-06,
-      "loss": 0.3906,
-      "step": 7883
-    },
-    {
-      "epoch": 0.5152604404940854,
-      "grad_norm": 0.45865270495414734,
-      "learning_rate": 9.420056282036623e-06,
-      "loss": 0.416,
-      "step": 7884
-    },
-    {
-      "epoch": 0.5153257956996274,
-      "grad_norm": 0.43503662943840027,
-      "learning_rate": 9.419893036937374e-06,
-      "loss": 0.3688,
-      "step": 7885
-    },
-    {
-      "epoch": 0.5153911509051696,
-      "grad_norm": 0.46031102538108826,
-      "learning_rate": 9.419729770280776e-06,
-      "loss": 0.4123,
-      "step": 7886
-    },
-    {
-      "epoch": 0.5154565061107117,
-      "grad_norm": 0.4093955159187317,
-      "learning_rate": 9.419566482067625e-06,
-      "loss": 0.3158,
-      "step": 7887
-    },
-    {
-      "epoch": 0.5155218613162539,
-      "grad_norm": 0.4443757236003876,
-      "learning_rate": 9.419403172298719e-06,
-      "loss": 0.3815,
-      "step": 7888
-    },
-    {
-      "epoch": 0.515587216521796,
-      "grad_norm": 0.45519816875457764,
-      "learning_rate": 9.41923984097485e-06,
-      "loss": 0.4094,
-      "step": 7889
-    },
-    {
-      "epoch": 0.515652571727338,
-      "grad_norm": 0.4669530689716339,
-      "learning_rate": 9.419076488096819e-06,
-      "loss": 0.3936,
-      "step": 7890
-    },
-    {
-      "epoch": 0.5157179269328802,
-      "grad_norm": 0.4384254217147827,
-      "learning_rate": 9.41891311366542e-06,
-      "loss": 0.359,
-      "step": 7891
-    },
-    {
-      "epoch": 0.5157832821384223,
-      "grad_norm": 0.45849543809890747,
-      "learning_rate": 9.418749717681452e-06,
-      "loss": 0.3967,
-      "step": 7892
-    },
-    {
-      "epoch": 0.5158486373439645,
-      "grad_norm": 0.4708442986011505,
-      "learning_rate": 9.418586300145711e-06,
-      "loss": 0.4218,
-      "step": 7893
-    },
-    {
-      "epoch": 0.5159139925495065,
-      "grad_norm": 0.45116758346557617,
-      "learning_rate": 9.418422861058993e-06,
-      "loss": 0.3908,
-      "step": 7894
-    },
-    {
-      "epoch": 0.5159793477550487,
-      "grad_norm": 0.45099860429763794,
-      "learning_rate": 9.418259400422095e-06,
-      "loss": 0.3619,
-      "step": 7895
-    },
-    {
-      "epoch": 0.5160447029605908,
-      "grad_norm": 0.836210310459137,
-      "learning_rate": 9.418095918235818e-06,
-      "loss": 0.3915,
-      "step": 7896
-    },
-    {
-      "epoch": 0.516110058166133,
-      "grad_norm": 0.4576779305934906,
-      "learning_rate": 9.417932414500954e-06,
-      "loss": 0.4526,
-      "step": 7897
-    },
-    {
-      "epoch": 0.516175413371675,
-      "grad_norm": 0.43634334206581116,
-      "learning_rate": 9.417768889218306e-06,
-      "loss": 0.3891,
-      "step": 7898
-    },
-    {
-      "epoch": 0.5162407685772171,
-      "grad_norm": 0.5078834891319275,
-      "learning_rate": 9.417605342388666e-06,
-      "loss": 0.4677,
-      "step": 7899
-    },
-    {
-      "epoch": 0.5163061237827593,
-      "grad_norm": 0.4795784056186676,
-      "learning_rate": 9.417441774012835e-06,
-      "loss": 0.409,
-      "step": 7900
-    },
-    {
-      "epoch": 0.5163714789883014,
-      "grad_norm": 0.46211862564086914,
-      "learning_rate": 9.41727818409161e-06,
-      "loss": 0.3788,
-      "step": 7901
-    },
-    {
-      "epoch": 0.5164368341938436,
-      "grad_norm": 0.45512014627456665,
-      "learning_rate": 9.417114572625789e-06,
-      "loss": 0.4262,
-      "step": 7902
-    },
-    {
-      "epoch": 0.5165021893993856,
-      "grad_norm": 0.46899402141571045,
-      "learning_rate": 9.416950939616172e-06,
-      "loss": 0.4267,
-      "step": 7903
-    },
-    {
-      "epoch": 0.5165675446049278,
-      "grad_norm": 0.4571188688278198,
-      "learning_rate": 9.416787285063553e-06,
-      "loss": 0.3428,
-      "step": 7904
-    },
-    {
-      "epoch": 0.5166328998104699,
-      "grad_norm": 0.45445504784584045,
-      "learning_rate": 9.416623608968732e-06,
-      "loss": 0.3989,
-      "step": 7905
-    },
-    {
-      "epoch": 0.5166982550160121,
-      "grad_norm": 0.4811440706253052,
-      "learning_rate": 9.416459911332509e-06,
-      "loss": 0.3949,
-      "step": 7906
-    },
-    {
-      "epoch": 0.5167636102215541,
-      "grad_norm": 0.45945218205451965,
-      "learning_rate": 9.416296192155681e-06,
-      "loss": 0.3999,
-      "step": 7907
-    },
-    {
-      "epoch": 0.5168289654270962,
-      "grad_norm": 0.4790816307067871,
-      "learning_rate": 9.416132451439046e-06,
-      "loss": 0.4115,
-      "step": 7908
-    },
-    {
-      "epoch": 0.5168943206326384,
-      "grad_norm": 0.5333146452903748,
-      "learning_rate": 9.4159686891834e-06,
-      "loss": 0.4436,
-      "step": 7909
-    },
-    {
-      "epoch": 0.5169596758381805,
-      "grad_norm": 0.4660441279411316,
-      "learning_rate": 9.41580490538955e-06,
-      "loss": 0.3939,
-      "step": 7910
-    },
-    {
-      "epoch": 0.5170250310437227,
-      "grad_norm": 0.4453851878643036,
-      "learning_rate": 9.415641100058287e-06,
-      "loss": 0.4342,
-      "step": 7911
-    },
-    {
-      "epoch": 0.5170903862492647,
-      "grad_norm": 0.47244173288345337,
-      "learning_rate": 9.415477273190415e-06,
-      "loss": 0.3913,
-      "step": 7912
-    },
-    {
-      "epoch": 0.5171557414548069,
-      "grad_norm": 0.4236837327480316,
-      "learning_rate": 9.415313424786727e-06,
-      "loss": 0.3579,
-      "step": 7913
-    },
-    {
-      "epoch": 0.517221096660349,
-      "grad_norm": 0.4299204349517822,
-      "learning_rate": 9.415149554848029e-06,
-      "loss": 0.3561,
-      "step": 7914
-    },
-    {
-      "epoch": 0.5172864518658912,
-      "grad_norm": 0.4726879894733429,
-      "learning_rate": 9.414985663375114e-06,
-      "loss": 0.4011,
-      "step": 7915
-    },
-    {
-      "epoch": 0.5173518070714332,
-      "grad_norm": 0.4675091803073883,
-      "learning_rate": 9.414821750368786e-06,
-      "loss": 0.3765,
-      "step": 7916
-    },
-    {
-      "epoch": 0.5174171622769753,
-      "grad_norm": 0.4623364210128784,
-      "learning_rate": 9.414657815829845e-06,
-      "loss": 0.3961,
-      "step": 7917
-    },
-    {
-      "epoch": 0.5174825174825175,
-      "grad_norm": 0.4778570532798767,
-      "learning_rate": 9.414493859759086e-06,
-      "loss": 0.419,
-      "step": 7918
-    },
-    {
-      "epoch": 0.5175478726880596,
-      "grad_norm": 0.4654449224472046,
-      "learning_rate": 9.414329882157311e-06,
-      "loss": 0.3996,
-      "step": 7919
-    },
-    {
-      "epoch": 0.5176132278936018,
-      "grad_norm": 0.48707547783851624,
-      "learning_rate": 9.41416588302532e-06,
-      "loss": 0.4329,
-      "step": 7920
-    },
-    {
-      "epoch": 0.5176785830991438,
-      "grad_norm": 0.4768628478050232,
-      "learning_rate": 9.414001862363913e-06,
-      "loss": 0.428,
-      "step": 7921
-    },
-    {
-      "epoch": 0.517743938304686,
-      "grad_norm": 0.4502464830875397,
-      "learning_rate": 9.41383782017389e-06,
-      "loss": 0.397,
-      "step": 7922
-    },
-    {
-      "epoch": 0.5178092935102281,
-      "grad_norm": 0.4830716550350189,
-      "learning_rate": 9.413673756456052e-06,
-      "loss": 0.4565,
-      "step": 7923
-    },
-    {
-      "epoch": 0.5178746487157702,
-      "grad_norm": 0.4762672483921051,
-      "learning_rate": 9.413509671211196e-06,
-      "loss": 0.4157,
-      "step": 7924
-    },
-    {
-      "epoch": 0.5179400039213123,
-      "grad_norm": 0.4664156138896942,
-      "learning_rate": 9.413345564440127e-06,
-      "loss": 0.3971,
-      "step": 7925
-    },
-    {
-      "epoch": 0.5180053591268544,
-      "grad_norm": 0.44348761439323425,
-      "learning_rate": 9.413181436143639e-06,
-      "loss": 0.3612,
-      "step": 7926
-    },
-    {
-      "epoch": 0.5180707143323966,
-      "grad_norm": 0.5411385297775269,
-      "learning_rate": 9.413017286322538e-06,
-      "loss": 0.4534,
-      "step": 7927
-    },
-    {
-      "epoch": 0.5181360695379387,
-      "grad_norm": 0.4380732476711273,
-      "learning_rate": 9.412853114977625e-06,
-      "loss": 0.3546,
-      "step": 7928
-    },
-    {
-      "epoch": 0.5182014247434809,
-      "grad_norm": 0.48528042435646057,
-      "learning_rate": 9.412688922109697e-06,
-      "loss": 0.4118,
-      "step": 7929
-    },
-    {
-      "epoch": 0.5182667799490229,
-      "grad_norm": 0.49747562408447266,
-      "learning_rate": 9.412524707719555e-06,
-      "loss": 0.4028,
-      "step": 7930
-    },
-    {
-      "epoch": 0.5183321351545651,
-      "grad_norm": 0.42012205719947815,
-      "learning_rate": 9.412360471808003e-06,
-      "loss": 0.3336,
-      "step": 7931
-    },
-    {
-      "epoch": 0.5183974903601072,
-      "grad_norm": 0.44400691986083984,
-      "learning_rate": 9.41219621437584e-06,
-      "loss": 0.3805,
-      "step": 7932
-    },
-    {
-      "epoch": 0.5184628455656493,
-      "grad_norm": 0.46124425530433655,
-      "learning_rate": 9.412031935423866e-06,
-      "loss": 0.3947,
-      "step": 7933
-    },
-    {
-      "epoch": 0.5185282007711914,
-      "grad_norm": 0.4640866816043854,
-      "learning_rate": 9.411867634952886e-06,
-      "loss": 0.3785,
-      "step": 7934
-    },
-    {
-      "epoch": 0.5185935559767335,
-      "grad_norm": 0.4519132077693939,
-      "learning_rate": 9.411703312963698e-06,
-      "loss": 0.3854,
-      "step": 7935
-    },
-    {
-      "epoch": 0.5186589111822757,
-      "grad_norm": 0.4467296600341797,
-      "learning_rate": 9.411538969457106e-06,
-      "loss": 0.3821,
-      "step": 7936
-    },
-    {
-      "epoch": 0.5187242663878178,
-      "grad_norm": 0.5134698748588562,
-      "learning_rate": 9.411374604433909e-06,
-      "loss": 0.4588,
-      "step": 7937
-    },
-    {
-      "epoch": 0.51878962159336,
-      "grad_norm": 0.460938960313797,
-      "learning_rate": 9.411210217894909e-06,
-      "loss": 0.408,
-      "step": 7938
-    },
-    {
-      "epoch": 0.518854976798902,
-      "grad_norm": 0.4559634327888489,
-      "learning_rate": 9.41104580984091e-06,
-      "loss": 0.3571,
-      "step": 7939
-    },
-    {
-      "epoch": 0.5189203320044442,
-      "grad_norm": 0.46090608835220337,
-      "learning_rate": 9.410881380272712e-06,
-      "loss": 0.4042,
-      "step": 7940
-    },
-    {
-      "epoch": 0.5189856872099863,
-      "grad_norm": 0.43631237745285034,
-      "learning_rate": 9.410716929191116e-06,
-      "loss": 0.4032,
-      "step": 7941
-    },
-    {
-      "epoch": 0.5190510424155284,
-      "grad_norm": 0.45229873061180115,
-      "learning_rate": 9.410552456596928e-06,
-      "loss": 0.3882,
-      "step": 7942
-    },
-    {
-      "epoch": 0.5191163976210705,
-      "grad_norm": 0.42981332540512085,
-      "learning_rate": 9.410387962490946e-06,
-      "loss": 0.3639,
-      "step": 7943
-    },
-    {
-      "epoch": 0.5191817528266126,
-      "grad_norm": 0.4539669156074524,
-      "learning_rate": 9.410223446873974e-06,
-      "loss": 0.3791,
-      "step": 7944
-    },
-    {
-      "epoch": 0.5192471080321548,
-      "grad_norm": 0.46513253450393677,
-      "learning_rate": 9.410058909746816e-06,
-      "loss": 0.4007,
-      "step": 7945
-    },
-    {
-      "epoch": 0.5193124632376969,
-      "grad_norm": 0.4235931932926178,
-      "learning_rate": 9.40989435111027e-06,
-      "loss": 0.3416,
-      "step": 7946
-    },
-    {
-      "epoch": 0.519377818443239,
-      "grad_norm": 0.5087857246398926,
-      "learning_rate": 9.409729770965145e-06,
-      "loss": 0.4563,
-      "step": 7947
-    },
-    {
-      "epoch": 0.5194431736487811,
-      "grad_norm": 0.4316421151161194,
-      "learning_rate": 9.40956516931224e-06,
-      "loss": 0.3859,
-      "step": 7948
-    },
-    {
-      "epoch": 0.5195085288543232,
-      "grad_norm": 0.5023858547210693,
-      "learning_rate": 9.409400546152357e-06,
-      "loss": 0.4635,
-      "step": 7949
-    },
-    {
-      "epoch": 0.5195738840598654,
-      "grad_norm": 0.45761409401893616,
-      "learning_rate": 9.409235901486297e-06,
-      "loss": 0.4269,
-      "step": 7950
-    },
-    {
-      "epoch": 0.5196392392654074,
-      "grad_norm": 0.4780981242656708,
-      "learning_rate": 9.40907123531487e-06,
-      "loss": 0.39,
-      "step": 7951
-    },
-    {
-      "epoch": 0.5197045944709496,
-      "grad_norm": 0.47008877992630005,
-      "learning_rate": 9.408906547638875e-06,
-      "loss": 0.4001,
-      "step": 7952
-    },
-    {
-      "epoch": 0.5197699496764917,
-      "grad_norm": 0.4629535973072052,
-      "learning_rate": 9.408741838459113e-06,
-      "loss": 0.4072,
-      "step": 7953
-    },
-    {
-      "epoch": 0.5198353048820339,
-      "grad_norm": 0.5037320852279663,
-      "learning_rate": 9.408577107776391e-06,
-      "loss": 0.465,
-      "step": 7954
-    },
-    {
-      "epoch": 0.519900660087576,
-      "grad_norm": 0.43817412853240967,
-      "learning_rate": 9.408412355591512e-06,
-      "loss": 0.3561,
-      "step": 7955
-    },
-    {
-      "epoch": 0.5199660152931181,
-      "grad_norm": 0.506892204284668,
-      "learning_rate": 9.408247581905276e-06,
-      "loss": 0.4385,
-      "step": 7956
-    },
-    {
-      "epoch": 0.5200313704986602,
-      "grad_norm": 0.4520317614078522,
-      "learning_rate": 9.40808278671849e-06,
-      "loss": 0.3834,
-      "step": 7957
-    },
-    {
-      "epoch": 0.5200967257042023,
-      "grad_norm": 0.5231791734695435,
-      "learning_rate": 9.40791797003196e-06,
-      "loss": 0.4623,
-      "step": 7958
-    },
-    {
-      "epoch": 0.5201620809097445,
-      "grad_norm": 0.45876771211624146,
-      "learning_rate": 9.407753131846485e-06,
-      "loss": 0.3952,
-      "step": 7959
-    },
-    {
-      "epoch": 0.5202274361152865,
-      "grad_norm": 0.4685318171977997,
-      "learning_rate": 9.40758827216287e-06,
-      "loss": 0.4043,
-      "step": 7960
-    },
-    {
-      "epoch": 0.5202927913208287,
-      "grad_norm": 0.4444715082645416,
-      "learning_rate": 9.407423390981922e-06,
-      "loss": 0.3723,
-      "step": 7961
-    },
-    {
-      "epoch": 0.5203581465263708,
-      "grad_norm": 0.5097283124923706,
-      "learning_rate": 9.407258488304444e-06,
-      "loss": 0.4508,
-      "step": 7962
-    },
-    {
-      "epoch": 0.520423501731913,
-      "grad_norm": 0.41760626435279846,
-      "learning_rate": 9.407093564131238e-06,
-      "loss": 0.3569,
-      "step": 7963
-    },
-    {
-      "epoch": 0.5204888569374551,
-      "grad_norm": 0.4615936279296875,
-      "learning_rate": 9.406928618463108e-06,
-      "loss": 0.3916,
-      "step": 7964
-    },
-    {
-      "epoch": 0.5205542121429972,
-      "grad_norm": 0.43561121821403503,
-      "learning_rate": 9.406763651300863e-06,
-      "loss": 0.3186,
-      "step": 7965
-    },
-    {
-      "epoch": 0.5206195673485393,
-      "grad_norm": 0.492966890335083,
-      "learning_rate": 9.406598662645305e-06,
-      "loss": 0.3873,
-      "step": 7966
-    },
-    {
-      "epoch": 0.5206849225540814,
-      "grad_norm": 0.4438392221927643,
-      "learning_rate": 9.406433652497239e-06,
-      "loss": 0.3748,
-      "step": 7967
-    },
-    {
-      "epoch": 0.5207502777596236,
-      "grad_norm": 0.44614288210868835,
-      "learning_rate": 9.406268620857469e-06,
-      "loss": 0.3828,
-      "step": 7968
-    },
-    {
-      "epoch": 0.5208156329651656,
-      "grad_norm": 0.5281800627708435,
-      "learning_rate": 9.4061035677268e-06,
-      "loss": 0.4166,
-      "step": 7969
-    },
-    {
-      "epoch": 0.5208809881707078,
-      "grad_norm": 0.4884756803512573,
-      "learning_rate": 9.405938493106037e-06,
-      "loss": 0.464,
-      "step": 7970
-    },
-    {
-      "epoch": 0.5209463433762499,
-      "grad_norm": 0.48060768842697144,
-      "learning_rate": 9.405773396995987e-06,
-      "loss": 0.3952,
-      "step": 7971
-    },
-    {
-      "epoch": 0.5210116985817921,
-      "grad_norm": 0.4539071023464203,
-      "learning_rate": 9.405608279397454e-06,
-      "loss": 0.3619,
-      "step": 7972
-    },
-    {
-      "epoch": 0.5210770537873342,
-      "grad_norm": 0.46185436844825745,
-      "learning_rate": 9.40544314031124e-06,
-      "loss": 0.4034,
-      "step": 7973
-    },
-    {
-      "epoch": 0.5211424089928762,
-      "grad_norm": 0.4426986277103424,
-      "learning_rate": 9.405277979738156e-06,
-      "loss": 0.3442,
-      "step": 7974
-    },
-    {
-      "epoch": 0.5212077641984184,
-      "grad_norm": 0.49941545724868774,
-      "learning_rate": 9.405112797679006e-06,
-      "loss": 0.4346,
-      "step": 7975
-    },
-    {
-      "epoch": 0.5212731194039605,
-      "grad_norm": 0.49342066049575806,
-      "learning_rate": 9.404947594134595e-06,
-      "loss": 0.4008,
-      "step": 7976
-    },
-    {
-      "epoch": 0.5213384746095027,
-      "grad_norm": 0.4608862102031708,
-      "learning_rate": 9.404782369105728e-06,
-      "loss": 0.4065,
-      "step": 7977
-    },
-    {
-      "epoch": 0.5214038298150447,
-      "grad_norm": 0.45691150426864624,
-      "learning_rate": 9.404617122593209e-06,
-      "loss": 0.3841,
-      "step": 7978
-    },
-    {
-      "epoch": 0.5214691850205869,
-      "grad_norm": 0.4427545964717865,
-      "learning_rate": 9.40445185459785e-06,
-      "loss": 0.3374,
-      "step": 7979
-    },
-    {
-      "epoch": 0.521534540226129,
-      "grad_norm": 0.47346892952919006,
-      "learning_rate": 9.40428656512045e-06,
-      "loss": 0.3938,
-      "step": 7980
-    },
-    {
-      "epoch": 0.5215998954316712,
-      "grad_norm": 0.4658379852771759,
-      "learning_rate": 9.40412125416182e-06,
-      "loss": 0.3964,
-      "step": 7981
-    },
-    {
-      "epoch": 0.5216652506372133,
-      "grad_norm": 0.4694722294807434,
-      "learning_rate": 9.403955921722766e-06,
-      "loss": 0.4014,
-      "step": 7982
-    },
-    {
-      "epoch": 0.5217306058427553,
-      "grad_norm": 0.4675805866718292,
-      "learning_rate": 9.403790567804092e-06,
-      "loss": 0.3909,
-      "step": 7983
-    },
-    {
-      "epoch": 0.5217959610482975,
-      "grad_norm": 0.4329233467578888,
-      "learning_rate": 9.403625192406606e-06,
-      "loss": 0.3259,
-      "step": 7984
-    },
-    {
-      "epoch": 0.5218613162538396,
-      "grad_norm": 0.4623168706893921,
-      "learning_rate": 9.403459795531117e-06,
-      "loss": 0.3807,
-      "step": 7985
-    },
-    {
-      "epoch": 0.5219266714593818,
-      "grad_norm": 0.4252798855304718,
-      "learning_rate": 9.403294377178425e-06,
-      "loss": 0.3704,
-      "step": 7986
-    },
-    {
-      "epoch": 0.5219920266649238,
-      "grad_norm": 0.44873565435409546,
-      "learning_rate": 9.403128937349344e-06,
-      "loss": 0.3797,
-      "step": 7987
-    },
-    {
-      "epoch": 0.522057381870466,
-      "grad_norm": 0.46728047728538513,
-      "learning_rate": 9.402963476044675e-06,
-      "loss": 0.3972,
-      "step": 7988
-    },
-    {
-      "epoch": 0.5221227370760081,
-      "grad_norm": 0.4808270037174225,
-      "learning_rate": 9.40279799326523e-06,
-      "loss": 0.3908,
-      "step": 7989
-    },
-    {
-      "epoch": 0.5221880922815503,
-      "grad_norm": 0.44836002588272095,
-      "learning_rate": 9.402632489011814e-06,
-      "loss": 0.3486,
-      "step": 7990
-    },
-    {
-      "epoch": 0.5222534474870923,
-      "grad_norm": 0.45154696702957153,
-      "learning_rate": 9.402466963285233e-06,
-      "loss": 0.3857,
-      "step": 7991
-    },
-    {
-      "epoch": 0.5223188026926344,
-      "grad_norm": 0.43766096234321594,
-      "learning_rate": 9.402301416086295e-06,
-      "loss": 0.4051,
-      "step": 7992
-    },
-    {
-      "epoch": 0.5223841578981766,
-      "grad_norm": 0.47397372126579285,
-      "learning_rate": 9.402135847415808e-06,
-      "loss": 0.4441,
-      "step": 7993
-    },
-    {
-      "epoch": 0.5224495131037187,
-      "grad_norm": 0.47660985589027405,
-      "learning_rate": 9.401970257274581e-06,
-      "loss": 0.4099,
-      "step": 7994
-    },
-    {
-      "epoch": 0.5225148683092609,
-      "grad_norm": 0.4678085446357727,
-      "learning_rate": 9.40180464566342e-06,
-      "loss": 0.4004,
-      "step": 7995
-    },
-    {
-      "epoch": 0.5225802235148029,
-      "grad_norm": 0.41487210988998413,
-      "learning_rate": 9.40163901258313e-06,
-      "loss": 0.3407,
-      "step": 7996
-    },
-    {
-      "epoch": 0.5226455787203451,
-      "grad_norm": 0.44676473736763,
-      "learning_rate": 9.401473358034526e-06,
-      "loss": 0.3804,
-      "step": 7997
-    },
-    {
-      "epoch": 0.5227109339258872,
-      "grad_norm": 0.4447825253009796,
-      "learning_rate": 9.401307682018408e-06,
-      "loss": 0.3553,
-      "step": 7998
-    },
-    {
-      "epoch": 0.5227762891314294,
-      "grad_norm": 0.46299442648887634,
-      "learning_rate": 9.40114198453559e-06,
-      "loss": 0.3994,
-      "step": 7999
-    },
-    {
-      "epoch": 0.5228416443369714,
-      "grad_norm": 0.488105446100235,
-      "learning_rate": 9.400976265586875e-06,
-      "loss": 0.3883,
-      "step": 8000
-    },
-    {
-      "epoch": 0.5229069995425135,
-      "grad_norm": 0.4538838267326355,
-      "learning_rate": 9.400810525173076e-06,
-      "loss": 0.3945,
-      "step": 8001
-    },
-    {
-      "epoch": 0.5229723547480557,
-      "grad_norm": 0.43405792117118835,
-      "learning_rate": 9.400644763294999e-06,
-      "loss": 0.3575,
-      "step": 8002
-    },
-    {
-      "epoch": 0.5230377099535978,
-      "grad_norm": 0.4692228138446808,
-      "learning_rate": 9.400478979953454e-06,
-      "loss": 0.4311,
-      "step": 8003
-    },
-    {
-      "epoch": 0.52310306515914,
-      "grad_norm": 0.4506242871284485,
-      "learning_rate": 9.400313175149247e-06,
-      "loss": 0.3687,
-      "step": 8004
-    },
-    {
-      "epoch": 0.523168420364682,
-      "grad_norm": 0.45200034976005554,
-      "learning_rate": 9.40014734888319e-06,
-      "loss": 0.3768,
-      "step": 8005
-    },
-    {
-      "epoch": 0.5232337755702242,
-      "grad_norm": 0.41679516434669495,
-      "learning_rate": 9.399981501156087e-06,
-      "loss": 0.3496,
-      "step": 8006
-    },
-    {
-      "epoch": 0.5232991307757663,
-      "grad_norm": 0.5029006004333496,
-      "learning_rate": 9.399815631968753e-06,
-      "loss": 0.43,
-      "step": 8007
-    },
-    {
-      "epoch": 0.5233644859813084,
-      "grad_norm": 0.548403799533844,
-      "learning_rate": 9.399649741321993e-06,
-      "loss": 0.4606,
-      "step": 8008
-    },
-    {
-      "epoch": 0.5234298411868505,
-      "grad_norm": 0.4558612108230591,
-      "learning_rate": 9.399483829216618e-06,
-      "loss": 0.4198,
-      "step": 8009
-    },
-    {
-      "epoch": 0.5234951963923926,
-      "grad_norm": 0.4317144751548767,
-      "learning_rate": 9.399317895653434e-06,
-      "loss": 0.3945,
-      "step": 8010
-    },
-    {
-      "epoch": 0.5235605515979348,
-      "grad_norm": 0.4775659143924713,
-      "learning_rate": 9.399151940633255e-06,
-      "loss": 0.4085,
-      "step": 8011
-    },
-    {
-      "epoch": 0.5236259068034769,
-      "grad_norm": 0.45099619030952454,
-      "learning_rate": 9.398985964156885e-06,
-      "loss": 0.3799,
-      "step": 8012
-    },
-    {
-      "epoch": 0.523691262009019,
-      "grad_norm": 0.4405990242958069,
-      "learning_rate": 9.39881996622514e-06,
-      "loss": 0.3953,
-      "step": 8013
-    },
-    {
-      "epoch": 0.5237566172145611,
-      "grad_norm": 0.43898147344589233,
-      "learning_rate": 9.398653946838824e-06,
-      "loss": 0.3527,
-      "step": 8014
-    },
-    {
-      "epoch": 0.5238219724201033,
-      "grad_norm": 0.43351277709007263,
-      "learning_rate": 9.398487905998749e-06,
-      "loss": 0.3583,
-      "step": 8015
-    },
-    {
-      "epoch": 0.5238873276256454,
-      "grad_norm": 0.43379583954811096,
-      "learning_rate": 9.398321843705724e-06,
-      "loss": 0.3586,
-      "step": 8016
-    },
-    {
-      "epoch": 0.5239526828311875,
-      "grad_norm": 0.5770723223686218,
-      "learning_rate": 9.398155759960562e-06,
-      "loss": 0.4713,
-      "step": 8017
-    },
-    {
-      "epoch": 0.5240180380367296,
-      "grad_norm": 0.4691585898399353,
-      "learning_rate": 9.397989654764068e-06,
-      "loss": 0.396,
-      "step": 8018
-    },
-    {
-      "epoch": 0.5240833932422717,
-      "grad_norm": 0.47056636214256287,
-      "learning_rate": 9.397823528117056e-06,
-      "loss": 0.3272,
-      "step": 8019
-    },
-    {
-      "epoch": 0.5241487484478139,
-      "grad_norm": 0.46248659491539,
-      "learning_rate": 9.397657380020335e-06,
-      "loss": 0.3404,
-      "step": 8020
-    },
-    {
-      "epoch": 0.524214103653356,
-      "grad_norm": 0.4465576708316803,
-      "learning_rate": 9.397491210474716e-06,
-      "loss": 0.378,
-      "step": 8021
-    },
-    {
-      "epoch": 0.5242794588588982,
-      "grad_norm": 0.43117213249206543,
-      "learning_rate": 9.39732501948101e-06,
-      "loss": 0.3901,
-      "step": 8022
-    },
-    {
-      "epoch": 0.5243448140644402,
-      "grad_norm": 0.4837399125099182,
-      "learning_rate": 9.397158807040023e-06,
-      "loss": 0.4328,
-      "step": 8023
-    },
-    {
-      "epoch": 0.5244101692699824,
-      "grad_norm": 0.4933948218822479,
-      "learning_rate": 9.396992573152572e-06,
-      "loss": 0.4332,
-      "step": 8024
-    },
-    {
-      "epoch": 0.5244755244755245,
-      "grad_norm": 0.4807804822921753,
-      "learning_rate": 9.396826317819465e-06,
-      "loss": 0.3978,
-      "step": 8025
-    },
-    {
-      "epoch": 0.5245408796810666,
-      "grad_norm": 0.44339725375175476,
-      "learning_rate": 9.396660041041511e-06,
-      "loss": 0.3732,
-      "step": 8026
-    },
-    {
-      "epoch": 0.5246062348866087,
-      "grad_norm": 0.46488943696022034,
-      "learning_rate": 9.396493742819524e-06,
-      "loss": 0.4205,
-      "step": 8027
-    },
-    {
-      "epoch": 0.5246715900921508,
-      "grad_norm": 0.461679607629776,
-      "learning_rate": 9.396327423154315e-06,
-      "loss": 0.3635,
-      "step": 8028
-    },
-    {
-      "epoch": 0.524736945297693,
-      "grad_norm": 0.4390772879123688,
-      "learning_rate": 9.396161082046693e-06,
-      "loss": 0.3654,
-      "step": 8029
-    },
-    {
-      "epoch": 0.5248023005032351,
-      "grad_norm": 0.5087174773216248,
-      "learning_rate": 9.39599471949747e-06,
-      "loss": 0.4039,
-      "step": 8030
-    },
-    {
-      "epoch": 0.5248676557087772,
-      "grad_norm": 0.49739664793014526,
-      "learning_rate": 9.395828335507457e-06,
-      "loss": 0.49,
-      "step": 8031
-    },
-    {
-      "epoch": 0.5249330109143193,
-      "grad_norm": 0.475120484828949,
-      "learning_rate": 9.395661930077469e-06,
-      "loss": 0.4052,
-      "step": 8032
-    },
-    {
-      "epoch": 0.5249983661198614,
-      "grad_norm": 0.4692741930484772,
-      "learning_rate": 9.395495503208311e-06,
-      "loss": 0.3939,
-      "step": 8033
-    },
-    {
-      "epoch": 0.5250637213254036,
-      "grad_norm": 0.5037876963615417,
-      "learning_rate": 9.395329054900802e-06,
-      "loss": 0.389,
-      "step": 8034
-    },
-    {
-      "epoch": 0.5251290765309456,
-      "grad_norm": 0.45203882455825806,
-      "learning_rate": 9.395162585155751e-06,
-      "loss": 0.3557,
-      "step": 8035
-    },
-    {
-      "epoch": 0.5251944317364878,
-      "grad_norm": 0.4454691708087921,
-      "learning_rate": 9.394996093973967e-06,
-      "loss": 0.3501,
-      "step": 8036
-    },
-    {
-      "epoch": 0.5252597869420299,
-      "grad_norm": 0.4618183672428131,
-      "learning_rate": 9.394829581356267e-06,
-      "loss": 0.4097,
-      "step": 8037
-    },
-    {
-      "epoch": 0.5253251421475721,
-      "grad_norm": 0.4642794728279114,
-      "learning_rate": 9.394663047303458e-06,
-      "loss": 0.4295,
-      "step": 8038
-    },
-    {
-      "epoch": 0.5253904973531142,
-      "grad_norm": 0.4464781582355499,
-      "learning_rate": 9.394496491816356e-06,
-      "loss": 0.3716,
-      "step": 8039
-    },
-    {
-      "epoch": 0.5254558525586563,
-      "grad_norm": 0.47627368569374084,
-      "learning_rate": 9.394329914895772e-06,
-      "loss": 0.4116,
-      "step": 8040
-    },
-    {
-      "epoch": 0.5255212077641984,
-      "grad_norm": 0.4596094787120819,
-      "learning_rate": 9.394163316542519e-06,
-      "loss": 0.4002,
-      "step": 8041
-    },
-    {
-      "epoch": 0.5255865629697405,
-      "grad_norm": 0.4332751929759979,
-      "learning_rate": 9.393996696757411e-06,
-      "loss": 0.3552,
-      "step": 8042
-    },
-    {
-      "epoch": 0.5256519181752827,
-      "grad_norm": 0.4469105303287506,
-      "learning_rate": 9.393830055541255e-06,
-      "loss": 0.35,
-      "step": 8043
-    },
-    {
-      "epoch": 0.5257172733808247,
-      "grad_norm": 0.43737125396728516,
-      "learning_rate": 9.39366339289487e-06,
-      "loss": 0.3871,
-      "step": 8044
-    },
-    {
-      "epoch": 0.5257826285863669,
-      "grad_norm": 0.46573638916015625,
-      "learning_rate": 9.393496708819066e-06,
-      "loss": 0.3775,
-      "step": 8045
-    },
-    {
-      "epoch": 0.525847983791909,
-      "grad_norm": 0.4459984600543976,
-      "learning_rate": 9.393330003314658e-06,
-      "loss": 0.3727,
-      "step": 8046
-    },
-    {
-      "epoch": 0.5259133389974512,
-      "grad_norm": 0.42989417910575867,
-      "learning_rate": 9.393163276382455e-06,
-      "loss": 0.332,
-      "step": 8047
-    },
-    {
-      "epoch": 0.5259786942029933,
-      "grad_norm": 0.444750040769577,
-      "learning_rate": 9.392996528023275e-06,
-      "loss": 0.3936,
-      "step": 8048
-    },
-    {
-      "epoch": 0.5260440494085354,
-      "grad_norm": 0.4220130443572998,
-      "learning_rate": 9.392829758237928e-06,
-      "loss": 0.3357,
-      "step": 8049
-    },
-    {
-      "epoch": 0.5261094046140775,
-      "grad_norm": 0.5116159319877625,
-      "learning_rate": 9.392662967027228e-06,
-      "loss": 0.3774,
-      "step": 8050
-    },
-    {
-      "epoch": 0.5261747598196196,
-      "grad_norm": 0.4552266299724579,
-      "learning_rate": 9.39249615439199e-06,
-      "loss": 0.3766,
-      "step": 8051
-    },
-    {
-      "epoch": 0.5262401150251618,
-      "grad_norm": 0.4574126899242401,
-      "learning_rate": 9.392329320333027e-06,
-      "loss": 0.3709,
-      "step": 8052
-    },
-    {
-      "epoch": 0.5263054702307038,
-      "grad_norm": 0.47054770588874817,
-      "learning_rate": 9.392162464851152e-06,
-      "loss": 0.4175,
-      "step": 8053
-    },
-    {
-      "epoch": 0.526370825436246,
-      "grad_norm": 0.48097535967826843,
-      "learning_rate": 9.391995587947175e-06,
-      "loss": 0.4163,
-      "step": 8054
-    },
-    {
-      "epoch": 0.5264361806417881,
-      "grad_norm": 0.49322423338890076,
-      "learning_rate": 9.391828689621918e-06,
-      "loss": 0.417,
-      "step": 8055
-    },
-    {
-      "epoch": 0.5265015358473303,
-      "grad_norm": 0.47926315665245056,
-      "learning_rate": 9.391661769876191e-06,
-      "loss": 0.4539,
-      "step": 8056
-    },
-    {
-      "epoch": 0.5265668910528724,
-      "grad_norm": 0.44520673155784607,
-      "learning_rate": 9.391494828710808e-06,
-      "loss": 0.3855,
-      "step": 8057
-    },
-    {
-      "epoch": 0.5266322462584145,
-      "grad_norm": 0.4670080840587616,
-      "learning_rate": 9.391327866126583e-06,
-      "loss": 0.3967,
-      "step": 8058
-    },
-    {
-      "epoch": 0.5266976014639566,
-      "grad_norm": 0.453642338514328,
-      "learning_rate": 9.391160882124331e-06,
-      "loss": 0.3721,
-      "step": 8059
-    },
-    {
-      "epoch": 0.5267629566694987,
-      "grad_norm": 0.44446754455566406,
-      "learning_rate": 9.390993876704865e-06,
-      "loss": 0.4194,
-      "step": 8060
-    },
-    {
-      "epoch": 0.5268283118750409,
-      "grad_norm": 0.4614487886428833,
-      "learning_rate": 9.390826849869001e-06,
-      "loss": 0.4017,
-      "step": 8061
-    },
-    {
-      "epoch": 0.5268936670805829,
-      "grad_norm": 0.48017358779907227,
-      "learning_rate": 9.390659801617554e-06,
-      "loss": 0.4488,
-      "step": 8062
-    },
-    {
-      "epoch": 0.5269590222861251,
-      "grad_norm": 0.4257350564002991,
-      "learning_rate": 9.390492731951337e-06,
-      "loss": 0.3588,
-      "step": 8063
-    },
-    {
-      "epoch": 0.5270243774916672,
-      "grad_norm": 0.5147051811218262,
-      "learning_rate": 9.390325640871168e-06,
-      "loss": 0.395,
-      "step": 8064
-    },
-    {
-      "epoch": 0.5270897326972094,
-      "grad_norm": 0.4848407804965973,
-      "learning_rate": 9.390158528377857e-06,
-      "loss": 0.4287,
-      "step": 8065
-    },
-    {
-      "epoch": 0.5271550879027515,
-      "grad_norm": 0.48413023352622986,
-      "learning_rate": 9.389991394472224e-06,
-      "loss": 0.4434,
-      "step": 8066
-    },
-    {
-      "epoch": 0.5272204431082935,
-      "grad_norm": 0.4852887988090515,
-      "learning_rate": 9.389824239155084e-06,
-      "loss": 0.4065,
-      "step": 8067
-    },
-    {
-      "epoch": 0.5272857983138357,
-      "grad_norm": 0.5066052079200745,
-      "learning_rate": 9.389657062427246e-06,
-      "loss": 0.3988,
-      "step": 8068
-    },
-    {
-      "epoch": 0.5273511535193778,
-      "grad_norm": 0.42358142137527466,
-      "learning_rate": 9.389489864289533e-06,
-      "loss": 0.3144,
-      "step": 8069
-    },
-    {
-      "epoch": 0.52741650872492,
-      "grad_norm": 0.44607630372047424,
-      "learning_rate": 9.389322644742755e-06,
-      "loss": 0.3655,
-      "step": 8070
-    },
-    {
-      "epoch": 0.527481863930462,
-      "grad_norm": 0.4566013514995575,
-      "learning_rate": 9.389155403787731e-06,
-      "loss": 0.369,
-      "step": 8071
-    },
-    {
-      "epoch": 0.5275472191360042,
-      "grad_norm": 0.4796139597892761,
-      "learning_rate": 9.388988141425276e-06,
-      "loss": 0.3988,
-      "step": 8072
-    },
-    {
-      "epoch": 0.5276125743415463,
-      "grad_norm": 0.48625293374061584,
-      "learning_rate": 9.388820857656204e-06,
-      "loss": 0.4024,
-      "step": 8073
-    },
-    {
-      "epoch": 0.5276779295470885,
-      "grad_norm": 0.47770261764526367,
-      "learning_rate": 9.388653552481335e-06,
-      "loss": 0.4033,
-      "step": 8074
-    },
-    {
-      "epoch": 0.5277432847526305,
-      "grad_norm": 0.4452119469642639,
-      "learning_rate": 9.388486225901478e-06,
-      "loss": 0.3679,
-      "step": 8075
-    },
-    {
-      "epoch": 0.5278086399581726,
-      "grad_norm": 0.4728561043739319,
-      "learning_rate": 9.388318877917458e-06,
-      "loss": 0.3997,
-      "step": 8076
-    },
-    {
-      "epoch": 0.5278739951637148,
-      "grad_norm": 0.4540014863014221,
-      "learning_rate": 9.388151508530083e-06,
-      "loss": 0.387,
-      "step": 8077
-    },
-    {
-      "epoch": 0.5279393503692569,
-      "grad_norm": 0.45731332898139954,
-      "learning_rate": 9.387984117740173e-06,
-      "loss": 0.36,
-      "step": 8078
-    },
-    {
-      "epoch": 0.5280047055747991,
-      "grad_norm": 0.45137470960617065,
-      "learning_rate": 9.387816705548547e-06,
-      "loss": 0.3594,
-      "step": 8079
-    },
-    {
-      "epoch": 0.5280700607803411,
-      "grad_norm": 0.4522440433502197,
-      "learning_rate": 9.387649271956017e-06,
-      "loss": 0.3452,
-      "step": 8080
-    },
-    {
-      "epoch": 0.5281354159858833,
-      "grad_norm": 0.7184723019599915,
-      "learning_rate": 9.387481816963402e-06,
-      "loss": 0.4073,
-      "step": 8081
-    },
-    {
-      "epoch": 0.5282007711914254,
-      "grad_norm": 0.42924800515174866,
-      "learning_rate": 9.387314340571518e-06,
-      "loss": 0.3835,
-      "step": 8082
-    },
-    {
-      "epoch": 0.5282661263969676,
-      "grad_norm": 0.4880116581916809,
-      "learning_rate": 9.387146842781184e-06,
-      "loss": 0.4272,
-      "step": 8083
-    },
-    {
-      "epoch": 0.5283314816025096,
-      "grad_norm": 0.4179949462413788,
-      "learning_rate": 9.386979323593212e-06,
-      "loss": 0.3205,
-      "step": 8084
-    },
-    {
-      "epoch": 0.5283968368080517,
-      "grad_norm": 0.4548265039920807,
-      "learning_rate": 9.386811783008423e-06,
-      "loss": 0.3805,
-      "step": 8085
-    },
-    {
-      "epoch": 0.5284621920135939,
-      "grad_norm": 0.4920929968357086,
-      "learning_rate": 9.386644221027633e-06,
-      "loss": 0.3991,
-      "step": 8086
-    },
-    {
-      "epoch": 0.528527547219136,
-      "grad_norm": 0.4421020746231079,
-      "learning_rate": 9.386476637651661e-06,
-      "loss": 0.3675,
-      "step": 8087
-    },
-    {
-      "epoch": 0.5285929024246782,
-      "grad_norm": 0.45332884788513184,
-      "learning_rate": 9.38630903288132e-06,
-      "loss": 0.3875,
-      "step": 8088
-    },
-    {
-      "epoch": 0.5286582576302202,
-      "grad_norm": 0.42376795411109924,
-      "learning_rate": 9.386141406717432e-06,
-      "loss": 0.3548,
-      "step": 8089
-    },
-    {
-      "epoch": 0.5287236128357624,
-      "grad_norm": 0.40692609548568726,
-      "learning_rate": 9.385973759160813e-06,
-      "loss": 0.3424,
-      "step": 8090
-    },
-    {
-      "epoch": 0.5287889680413045,
-      "grad_norm": 0.498879611492157,
-      "learning_rate": 9.38580609021228e-06,
-      "loss": 0.3947,
-      "step": 8091
-    },
-    {
-      "epoch": 0.5288543232468466,
-      "grad_norm": 0.4672560691833496,
-      "learning_rate": 9.385638399872651e-06,
-      "loss": 0.3551,
-      "step": 8092
-    },
-    {
-      "epoch": 0.5289196784523887,
-      "grad_norm": 0.4360695481300354,
-      "learning_rate": 9.385470688142746e-06,
-      "loss": 0.3975,
-      "step": 8093
-    },
-    {
-      "epoch": 0.5289850336579308,
-      "grad_norm": 0.4511704742908478,
-      "learning_rate": 9.385302955023379e-06,
-      "loss": 0.376,
-      "step": 8094
-    },
-    {
-      "epoch": 0.529050388863473,
-      "grad_norm": 0.48188722133636475,
-      "learning_rate": 9.385135200515372e-06,
-      "loss": 0.3924,
-      "step": 8095
-    },
-    {
-      "epoch": 0.5291157440690151,
-      "grad_norm": 0.4877548813819885,
-      "learning_rate": 9.38496742461954e-06,
-      "loss": 0.4149,
-      "step": 8096
-    },
-    {
-      "epoch": 0.5291810992745573,
-      "grad_norm": 0.4314827024936676,
-      "learning_rate": 9.384799627336703e-06,
-      "loss": 0.3474,
-      "step": 8097
-    },
-    {
-      "epoch": 0.5292464544800993,
-      "grad_norm": 0.44380849599838257,
-      "learning_rate": 9.38463180866768e-06,
-      "loss": 0.365,
-      "step": 8098
-    },
-    {
-      "epoch": 0.5293118096856415,
-      "grad_norm": 0.4292486011981964,
-      "learning_rate": 9.384463968613289e-06,
-      "loss": 0.3236,
-      "step": 8099
-    },
-    {
-      "epoch": 0.5293771648911836,
-      "grad_norm": 0.4576188027858734,
-      "learning_rate": 9.384296107174347e-06,
-      "loss": 0.3888,
-      "step": 8100
-    },
-    {
-      "epoch": 0.5294425200967257,
-      "grad_norm": 0.4467967748641968,
-      "learning_rate": 9.384128224351676e-06,
-      "loss": 0.4106,
-      "step": 8101
-    },
-    {
-      "epoch": 0.5295078753022678,
-      "grad_norm": 0.4577098786830902,
-      "learning_rate": 9.38396032014609e-06,
-      "loss": 0.3942,
-      "step": 8102
-    },
-    {
-      "epoch": 0.5295732305078099,
-      "grad_norm": 0.48310619592666626,
-      "learning_rate": 9.383792394558412e-06,
-      "loss": 0.4446,
-      "step": 8103
-    },
-    {
-      "epoch": 0.5296385857133521,
-      "grad_norm": 0.4269142746925354,
-      "learning_rate": 9.383624447589462e-06,
-      "loss": 0.3499,
-      "step": 8104
-    },
-    {
-      "epoch": 0.5297039409188942,
-      "grad_norm": 0.4764975905418396,
-      "learning_rate": 9.383456479240054e-06,
-      "loss": 0.4144,
-      "step": 8105
-    },
-    {
-      "epoch": 0.5297692961244364,
-      "grad_norm": 0.45994073152542114,
-      "learning_rate": 9.38328848951101e-06,
-      "loss": 0.3947,
-      "step": 8106
-    },
-    {
-      "epoch": 0.5298346513299784,
-      "grad_norm": 0.42793282866477966,
-      "learning_rate": 9.383120478403151e-06,
-      "loss": 0.3367,
-      "step": 8107
-    },
-    {
-      "epoch": 0.5299000065355206,
-      "grad_norm": 0.45754751563072205,
-      "learning_rate": 9.382952445917295e-06,
-      "loss": 0.4171,
-      "step": 8108
-    },
-    {
-      "epoch": 0.5299653617410627,
-      "grad_norm": 0.47818121314048767,
-      "learning_rate": 9.382784392054262e-06,
-      "loss": 0.4069,
-      "step": 8109
-    },
-    {
-      "epoch": 0.5300307169466048,
-      "grad_norm": 0.43848422169685364,
-      "learning_rate": 9.38261631681487e-06,
-      "loss": 0.3821,
-      "step": 8110
-    },
-    {
-      "epoch": 0.5300960721521469,
-      "grad_norm": 0.4554431736469269,
-      "learning_rate": 9.38244822019994e-06,
-      "loss": 0.37,
-      "step": 8111
-    },
-    {
-      "epoch": 0.530161427357689,
-      "grad_norm": 0.47809508442878723,
-      "learning_rate": 9.382280102210292e-06,
-      "loss": 0.3948,
-      "step": 8112
-    },
-    {
-      "epoch": 0.5302267825632312,
-      "grad_norm": 0.4500177502632141,
-      "learning_rate": 9.382111962846745e-06,
-      "loss": 0.3757,
-      "step": 8113
-    },
-    {
-      "epoch": 0.5302921377687733,
-      "grad_norm": 0.45078301429748535,
-      "learning_rate": 9.381943802110121e-06,
-      "loss": 0.3622,
-      "step": 8114
-    },
-    {
-      "epoch": 0.5303574929743154,
-      "grad_norm": 0.4388760030269623,
-      "learning_rate": 9.381775620001238e-06,
-      "loss": 0.3573,
-      "step": 8115
-    },
-    {
-      "epoch": 0.5304228481798575,
-      "grad_norm": 0.444981187582016,
-      "learning_rate": 9.38160741652092e-06,
-      "loss": 0.3788,
-      "step": 8116
-    },
-    {
-      "epoch": 0.5304882033853996,
-      "grad_norm": 0.39920949935913086,
-      "learning_rate": 9.381439191669983e-06,
-      "loss": 0.3128,
-      "step": 8117
-    },
-    {
-      "epoch": 0.5305535585909418,
-      "grad_norm": 0.4420296251773834,
-      "learning_rate": 9.381270945449247e-06,
-      "loss": 0.3753,
-      "step": 8118
-    },
-    {
-      "epoch": 0.5306189137964838,
-      "grad_norm": 0.5094230771064758,
-      "learning_rate": 9.381102677859537e-06,
-      "loss": 0.4332,
-      "step": 8119
-    },
-    {
-      "epoch": 0.530684269002026,
-      "grad_norm": 0.42383527755737305,
-      "learning_rate": 9.380934388901673e-06,
-      "loss": 0.3833,
-      "step": 8120
-    },
-    {
-      "epoch": 0.5307496242075681,
-      "grad_norm": 0.4477185606956482,
-      "learning_rate": 9.38076607857647e-06,
-      "loss": 0.4222,
-      "step": 8121
-    },
-    {
-      "epoch": 0.5308149794131103,
-      "grad_norm": 0.48360368609428406,
-      "learning_rate": 9.380597746884757e-06,
-      "loss": 0.4133,
-      "step": 8122
-    },
-    {
-      "epoch": 0.5308803346186524,
-      "grad_norm": 0.4472695589065552,
-      "learning_rate": 9.38042939382735e-06,
-      "loss": 0.368,
-      "step": 8123
-    },
-    {
-      "epoch": 0.5309456898241945,
-      "grad_norm": 0.44982022047042847,
-      "learning_rate": 9.380261019405071e-06,
-      "loss": 0.3726,
-      "step": 8124
-    },
-    {
-      "epoch": 0.5310110450297366,
-      "grad_norm": 0.4698355793952942,
-      "learning_rate": 9.380092623618743e-06,
-      "loss": 0.4508,
-      "step": 8125
-    },
-    {
-      "epoch": 0.5310764002352787,
-      "grad_norm": 0.4377356469631195,
-      "learning_rate": 9.379924206469184e-06,
-      "loss": 0.3549,
-      "step": 8126
-    },
-    {
-      "epoch": 0.5311417554408209,
-      "grad_norm": 0.4417353868484497,
-      "learning_rate": 9.379755767957217e-06,
-      "loss": 0.3952,
-      "step": 8127
-    },
-    {
-      "epoch": 0.531207110646363,
-      "grad_norm": 0.4596727192401886,
-      "learning_rate": 9.379587308083666e-06,
-      "loss": 0.3701,
-      "step": 8128
-    },
-    {
-      "epoch": 0.5312724658519051,
-      "grad_norm": 0.4739995300769806,
-      "learning_rate": 9.37941882684935e-06,
-      "loss": 0.42,
-      "step": 8129
-    },
-    {
-      "epoch": 0.5313378210574472,
-      "grad_norm": 0.437948077917099,
-      "learning_rate": 9.37925032425509e-06,
-      "loss": 0.38,
-      "step": 8130
-    },
-    {
-      "epoch": 0.5314031762629894,
-      "grad_norm": 0.45688772201538086,
-      "learning_rate": 9.379081800301709e-06,
-      "loss": 0.3973,
-      "step": 8131
-    },
-    {
-      "epoch": 0.5314685314685315,
-      "grad_norm": 0.45127663016319275,
-      "learning_rate": 9.37891325499003e-06,
-      "loss": 0.3685,
-      "step": 8132
-    },
-    {
-      "epoch": 0.5315338866740736,
-      "grad_norm": 0.4748782515525818,
-      "learning_rate": 9.378744688320871e-06,
-      "loss": 0.424,
-      "step": 8133
-    },
-    {
-      "epoch": 0.5315992418796157,
-      "grad_norm": 0.43217262625694275,
-      "learning_rate": 9.37857610029506e-06,
-      "loss": 0.3562,
-      "step": 8134
-    },
-    {
-      "epoch": 0.5316645970851578,
-      "grad_norm": 0.4683869481086731,
-      "learning_rate": 9.378407490913417e-06,
-      "loss": 0.4006,
-      "step": 8135
-    },
-    {
-      "epoch": 0.5317299522907,
-      "grad_norm": 0.4753198027610779,
-      "learning_rate": 9.378238860176762e-06,
-      "loss": 0.3981,
-      "step": 8136
-    },
-    {
-      "epoch": 0.531795307496242,
-      "grad_norm": 0.44580692052841187,
-      "learning_rate": 9.378070208085921e-06,
-      "loss": 0.368,
-      "step": 8137
-    },
-    {
-      "epoch": 0.5318606627017842,
-      "grad_norm": 0.42868709564208984,
-      "learning_rate": 9.377901534641714e-06,
-      "loss": 0.3439,
-      "step": 8138
-    },
-    {
-      "epoch": 0.5319260179073263,
-      "grad_norm": 0.5087711811065674,
-      "learning_rate": 9.377732839844966e-06,
-      "loss": 0.4215,
-      "step": 8139
-    },
-    {
-      "epoch": 0.5319913731128685,
-      "grad_norm": 0.46497300267219543,
-      "learning_rate": 9.377564123696497e-06,
-      "loss": 0.3777,
-      "step": 8140
-    },
-    {
-      "epoch": 0.5320567283184106,
-      "grad_norm": 0.4794805347919464,
-      "learning_rate": 9.37739538619713e-06,
-      "loss": 0.3762,
-      "step": 8141
-    },
-    {
-      "epoch": 0.5321220835239527,
-      "grad_norm": 0.5074891448020935,
-      "learning_rate": 9.377226627347692e-06,
-      "loss": 0.4516,
-      "step": 8142
-    },
-    {
-      "epoch": 0.5321874387294948,
-      "grad_norm": 0.41453054547309875,
-      "learning_rate": 9.377057847149002e-06,
-      "loss": 0.3397,
-      "step": 8143
-    },
-    {
-      "epoch": 0.5322527939350369,
-      "grad_norm": 0.4522910714149475,
-      "learning_rate": 9.376889045601885e-06,
-      "loss": 0.3591,
-      "step": 8144
-    },
-    {
-      "epoch": 0.5323181491405791,
-      "grad_norm": 0.46257084608078003,
-      "learning_rate": 9.376720222707163e-06,
-      "loss": 0.338,
-      "step": 8145
-    },
-    {
-      "epoch": 0.5323835043461211,
-      "grad_norm": 0.46426162123680115,
-      "learning_rate": 9.37655137846566e-06,
-      "loss": 0.4139,
-      "step": 8146
-    },
-    {
-      "epoch": 0.5324488595516633,
-      "grad_norm": 0.449640691280365,
-      "learning_rate": 9.3763825128782e-06,
-      "loss": 0.3962,
-      "step": 8147
-    },
-    {
-      "epoch": 0.5325142147572054,
-      "grad_norm": 0.44668322801589966,
-      "learning_rate": 9.376213625945607e-06,
-      "loss": 0.4036,
-      "step": 8148
-    },
-    {
-      "epoch": 0.5325795699627476,
-      "grad_norm": 0.48974403738975525,
-      "learning_rate": 9.376044717668704e-06,
-      "loss": 0.3797,
-      "step": 8149
-    },
-    {
-      "epoch": 0.5326449251682897,
-      "grad_norm": 0.4639590084552765,
-      "learning_rate": 9.375875788048315e-06,
-      "loss": 0.351,
-      "step": 8150
-    },
-    {
-      "epoch": 0.5327102803738317,
-      "grad_norm": 0.4961114823818207,
-      "learning_rate": 9.375706837085262e-06,
-      "loss": 0.4348,
-      "step": 8151
-    },
-    {
-      "epoch": 0.5327756355793739,
-      "grad_norm": 0.4123740494251251,
-      "learning_rate": 9.375537864780373e-06,
-      "loss": 0.3225,
-      "step": 8152
-    },
-    {
-      "epoch": 0.532840990784916,
-      "grad_norm": 0.4464639723300934,
-      "learning_rate": 9.37536887113447e-06,
-      "loss": 0.3851,
-      "step": 8153
-    },
-    {
-      "epoch": 0.5329063459904582,
-      "grad_norm": 0.4487189054489136,
-      "learning_rate": 9.375199856148376e-06,
-      "loss": 0.3333,
-      "step": 8154
-    },
-    {
-      "epoch": 0.5329717011960002,
-      "grad_norm": 0.4613710641860962,
-      "learning_rate": 9.375030819822916e-06,
-      "loss": 0.3953,
-      "step": 8155
-    },
-    {
-      "epoch": 0.5330370564015424,
-      "grad_norm": 0.4397062659263611,
-      "learning_rate": 9.374861762158917e-06,
-      "loss": 0.3552,
-      "step": 8156
-    },
-    {
-      "epoch": 0.5331024116070845,
-      "grad_norm": 0.47692734003067017,
-      "learning_rate": 9.3746926831572e-06,
-      "loss": 0.3797,
-      "step": 8157
-    },
-    {
-      "epoch": 0.5331677668126267,
-      "grad_norm": 0.535108745098114,
-      "learning_rate": 9.374523582818591e-06,
-      "loss": 0.4581,
-      "step": 8158
-    },
-    {
-      "epoch": 0.5332331220181687,
-      "grad_norm": 0.4900972247123718,
-      "learning_rate": 9.374354461143917e-06,
-      "loss": 0.4235,
-      "step": 8159
-    },
-    {
-      "epoch": 0.5332984772237108,
-      "grad_norm": 0.4511905908584595,
-      "learning_rate": 9.374185318134e-06,
-      "loss": 0.3701,
-      "step": 8160
-    },
-    {
-      "epoch": 0.533363832429253,
-      "grad_norm": 0.4689152240753174,
-      "learning_rate": 9.374016153789666e-06,
-      "loss": 0.339,
-      "step": 8161
-    },
-    {
-      "epoch": 0.5334291876347951,
-      "grad_norm": 0.46662044525146484,
-      "learning_rate": 9.373846968111739e-06,
-      "loss": 0.3868,
-      "step": 8162
-    },
-    {
-      "epoch": 0.5334945428403373,
-      "grad_norm": 0.4650476574897766,
-      "learning_rate": 9.373677761101045e-06,
-      "loss": 0.4015,
-      "step": 8163
-    },
-    {
-      "epoch": 0.5335598980458793,
-      "grad_norm": 0.45298993587493896,
-      "learning_rate": 9.37350853275841e-06,
-      "loss": 0.3488,
-      "step": 8164
-    },
-    {
-      "epoch": 0.5336252532514215,
-      "grad_norm": 0.4997517764568329,
-      "learning_rate": 9.37333928308466e-06,
-      "loss": 0.4216,
-      "step": 8165
-    },
-    {
-      "epoch": 0.5336906084569636,
-      "grad_norm": 0.4577353894710541,
-      "learning_rate": 9.373170012080618e-06,
-      "loss": 0.3896,
-      "step": 8166
-    },
-    {
-      "epoch": 0.5337559636625058,
-      "grad_norm": 0.45616233348846436,
-      "learning_rate": 9.37300071974711e-06,
-      "loss": 0.3477,
-      "step": 8167
-    },
-    {
-      "epoch": 0.5338213188680478,
-      "grad_norm": 0.5024468898773193,
-      "learning_rate": 9.372831406084965e-06,
-      "loss": 0.4235,
-      "step": 8168
-    },
-    {
-      "epoch": 0.5338866740735899,
-      "grad_norm": 0.4371213912963867,
-      "learning_rate": 9.372662071095004e-06,
-      "loss": 0.3692,
-      "step": 8169
-    },
-    {
-      "epoch": 0.5339520292791321,
-      "grad_norm": 0.4440755248069763,
-      "learning_rate": 9.372492714778057e-06,
-      "loss": 0.3826,
-      "step": 8170
-    },
-    {
-      "epoch": 0.5340173844846742,
-      "grad_norm": 0.47860926389694214,
-      "learning_rate": 9.372323337134947e-06,
-      "loss": 0.3847,
-      "step": 8171
-    },
-    {
-      "epoch": 0.5340827396902164,
-      "grad_norm": 0.4336561858654022,
-      "learning_rate": 9.372153938166502e-06,
-      "loss": 0.3769,
-      "step": 8172
-    },
-    {
-      "epoch": 0.5341480948957584,
-      "grad_norm": 0.4629087448120117,
-      "learning_rate": 9.371984517873547e-06,
-      "loss": 0.3691,
-      "step": 8173
-    },
-    {
-      "epoch": 0.5342134501013006,
-      "grad_norm": 0.4819185435771942,
-      "learning_rate": 9.37181507625691e-06,
-      "loss": 0.3814,
-      "step": 8174
-    },
-    {
-      "epoch": 0.5342788053068427,
-      "grad_norm": 0.47867029905319214,
-      "learning_rate": 9.371645613317414e-06,
-      "loss": 0.4223,
-      "step": 8175
-    },
-    {
-      "epoch": 0.5343441605123848,
-      "grad_norm": 0.47211819887161255,
-      "learning_rate": 9.371476129055889e-06,
-      "loss": 0.3672,
-      "step": 8176
-    },
-    {
-      "epoch": 0.5344095157179269,
-      "grad_norm": 0.5237996578216553,
-      "learning_rate": 9.37130662347316e-06,
-      "loss": 0.4479,
-      "step": 8177
-    },
-    {
-      "epoch": 0.534474870923469,
-      "grad_norm": 0.453158438205719,
-      "learning_rate": 9.371137096570056e-06,
-      "loss": 0.3836,
-      "step": 8178
-    },
-    {
-      "epoch": 0.5345402261290112,
-      "grad_norm": 0.4489046335220337,
-      "learning_rate": 9.3709675483474e-06,
-      "loss": 0.3997,
-      "step": 8179
-    },
-    {
-      "epoch": 0.5346055813345533,
-      "grad_norm": 0.451736718416214,
-      "learning_rate": 9.370797978806024e-06,
-      "loss": 0.3501,
-      "step": 8180
-    },
-    {
-      "epoch": 0.5346709365400955,
-      "grad_norm": 0.4901093542575836,
-      "learning_rate": 9.37062838794675e-06,
-      "loss": 0.3646,
-      "step": 8181
-    },
-    {
-      "epoch": 0.5347362917456375,
-      "grad_norm": 0.4496528208255768,
-      "learning_rate": 9.370458775770406e-06,
-      "loss": 0.3418,
-      "step": 8182
-    },
-    {
-      "epoch": 0.5348016469511797,
-      "grad_norm": 0.47359248995780945,
-      "learning_rate": 9.370289142277822e-06,
-      "loss": 0.3506,
-      "step": 8183
-    },
-    {
-      "epoch": 0.5348670021567218,
-      "grad_norm": 0.4595816135406494,
-      "learning_rate": 9.370119487469825e-06,
-      "loss": 0.3715,
-      "step": 8184
-    },
-    {
-      "epoch": 0.5349323573622639,
-      "grad_norm": 0.4757009744644165,
-      "learning_rate": 9.36994981134724e-06,
-      "loss": 0.4078,
-      "step": 8185
-    },
-    {
-      "epoch": 0.534997712567806,
-      "grad_norm": 0.49687087535858154,
-      "learning_rate": 9.369780113910897e-06,
-      "loss": 0.4572,
-      "step": 8186
-    },
-    {
-      "epoch": 0.5350630677733481,
-      "grad_norm": 0.4593808054924011,
-      "learning_rate": 9.36961039516162e-06,
-      "loss": 0.3805,
-      "step": 8187
-    },
-    {
-      "epoch": 0.5351284229788903,
-      "grad_norm": 0.5112813711166382,
-      "learning_rate": 9.369440655100241e-06,
-      "loss": 0.4195,
-      "step": 8188
-    },
-    {
-      "epoch": 0.5351937781844324,
-      "grad_norm": 0.4520493745803833,
-      "learning_rate": 9.369270893727586e-06,
-      "loss": 0.386,
-      "step": 8189
-    },
-    {
-      "epoch": 0.5352591333899746,
-      "grad_norm": 0.48702430725097656,
-      "learning_rate": 9.369101111044484e-06,
-      "loss": 0.4211,
-      "step": 8190
-    },
-    {
-      "epoch": 0.5353244885955166,
-      "grad_norm": 0.45532673597335815,
-      "learning_rate": 9.36893130705176e-06,
-      "loss": 0.3717,
-      "step": 8191
-    },
-    {
-      "epoch": 0.5353898438010588,
-      "grad_norm": 0.44316256046295166,
-      "learning_rate": 9.368761481750245e-06,
-      "loss": 0.3524,
-      "step": 8192
-    },
-    {
-      "epoch": 0.5354551990066009,
-      "grad_norm": 0.4726574420928955,
-      "learning_rate": 9.36859163514077e-06,
-      "loss": 0.4144,
-      "step": 8193
-    },
-    {
-      "epoch": 0.535520554212143,
-      "grad_norm": 0.44861721992492676,
-      "learning_rate": 9.368421767224157e-06,
-      "loss": 0.3529,
-      "step": 8194
-    },
-    {
-      "epoch": 0.5355859094176851,
-      "grad_norm": 0.44170206785202026,
-      "learning_rate": 9.368251878001238e-06,
-      "loss": 0.362,
-      "step": 8195
-    },
-    {
-      "epoch": 0.5356512646232272,
-      "grad_norm": 0.42073050141334534,
-      "learning_rate": 9.368081967472842e-06,
-      "loss": 0.3255,
-      "step": 8196
-    },
-    {
-      "epoch": 0.5357166198287694,
-      "grad_norm": 0.6058770418167114,
-      "learning_rate": 9.367912035639797e-06,
-      "loss": 0.541,
-      "step": 8197
-    },
-    {
-      "epoch": 0.5357819750343115,
-      "grad_norm": 0.451310932636261,
-      "learning_rate": 9.367742082502932e-06,
-      "loss": 0.3696,
-      "step": 8198
-    },
-    {
-      "epoch": 0.5358473302398536,
-      "grad_norm": 0.5183236598968506,
-      "learning_rate": 9.367572108063076e-06,
-      "loss": 0.4422,
-      "step": 8199
-    },
-    {
-      "epoch": 0.5359126854453957,
-      "grad_norm": 0.45446789264678955,
-      "learning_rate": 9.367402112321056e-06,
-      "loss": 0.3738,
-      "step": 8200
-    },
-    {
-      "epoch": 0.5359780406509378,
-      "grad_norm": 0.45923709869384766,
-      "learning_rate": 9.367232095277705e-06,
-      "loss": 0.4003,
-      "step": 8201
-    },
-    {
-      "epoch": 0.53604339585648,
-      "grad_norm": 0.4609838128089905,
-      "learning_rate": 9.36706205693385e-06,
-      "loss": 0.3745,
-      "step": 8202
-    },
-    {
-      "epoch": 0.536108751062022,
-      "grad_norm": 0.506086528301239,
-      "learning_rate": 9.366891997290318e-06,
-      "loss": 0.4049,
-      "step": 8203
-    },
-    {
-      "epoch": 0.5361741062675642,
-      "grad_norm": 0.4835215210914612,
-      "learning_rate": 9.366721916347942e-06,
-      "loss": 0.4267,
-      "step": 8204
-    },
-    {
-      "epoch": 0.5362394614731063,
-      "grad_norm": 0.526479184627533,
-      "learning_rate": 9.366551814107552e-06,
-      "loss": 0.419,
-      "step": 8205
-    },
-    {
-      "epoch": 0.5363048166786485,
-      "grad_norm": 0.5380675792694092,
-      "learning_rate": 9.366381690569974e-06,
-      "loss": 0.4852,
-      "step": 8206
-    },
-    {
-      "epoch": 0.5363701718841906,
-      "grad_norm": 0.48616766929626465,
-      "learning_rate": 9.366211545736042e-06,
-      "loss": 0.4343,
-      "step": 8207
-    },
-    {
-      "epoch": 0.5364355270897327,
-      "grad_norm": 0.41571882367134094,
-      "learning_rate": 9.366041379606582e-06,
-      "loss": 0.3588,
-      "step": 8208
-    },
-    {
-      "epoch": 0.5365008822952748,
-      "grad_norm": 0.5063007473945618,
-      "learning_rate": 9.365871192182428e-06,
-      "loss": 0.4522,
-      "step": 8209
-    },
-    {
-      "epoch": 0.5365662375008169,
-      "grad_norm": 0.4544987380504608,
-      "learning_rate": 9.365700983464406e-06,
-      "loss": 0.3496,
-      "step": 8210
-    },
-    {
-      "epoch": 0.5366315927063591,
-      "grad_norm": 0.4544180929660797,
-      "learning_rate": 9.365530753453346e-06,
-      "loss": 0.351,
-      "step": 8211
-    },
-    {
-      "epoch": 0.5366969479119011,
-      "grad_norm": 0.4394286572933197,
-      "learning_rate": 9.365360502150083e-06,
-      "loss": 0.3874,
-      "step": 8212
-    },
-    {
-      "epoch": 0.5367623031174433,
-      "grad_norm": 0.47277411818504333,
-      "learning_rate": 9.365190229555443e-06,
-      "loss": 0.3853,
-      "step": 8213
-    },
-    {
-      "epoch": 0.5368276583229854,
-      "grad_norm": 0.4341241419315338,
-      "learning_rate": 9.365019935670259e-06,
-      "loss": 0.3141,
-      "step": 8214
-    },
-    {
-      "epoch": 0.5368930135285276,
-      "grad_norm": 0.4787592589855194,
-      "learning_rate": 9.36484962049536e-06,
-      "loss": 0.3888,
-      "step": 8215
-    },
-    {
-      "epoch": 0.5369583687340697,
-      "grad_norm": 0.48868072032928467,
-      "learning_rate": 9.364679284031577e-06,
-      "loss": 0.4195,
-      "step": 8216
-    },
-    {
-      "epoch": 0.5370237239396118,
-      "grad_norm": 0.4858606457710266,
-      "learning_rate": 9.36450892627974e-06,
-      "loss": 0.4051,
-      "step": 8217
-    },
-    {
-      "epoch": 0.5370890791451539,
-      "grad_norm": 0.46490201354026794,
-      "learning_rate": 9.364338547240684e-06,
-      "loss": 0.4149,
-      "step": 8218
-    },
-    {
-      "epoch": 0.537154434350696,
-      "grad_norm": 0.49066856503486633,
-      "learning_rate": 9.364168146915234e-06,
-      "loss": 0.4001,
-      "step": 8219
-    },
-    {
-      "epoch": 0.5372197895562382,
-      "grad_norm": 0.48547351360321045,
-      "learning_rate": 9.363997725304225e-06,
-      "loss": 0.422,
-      "step": 8220
-    },
-    {
-      "epoch": 0.5372851447617802,
-      "grad_norm": 0.4581473767757416,
-      "learning_rate": 9.363827282408488e-06,
-      "loss": 0.361,
-      "step": 8221
-    },
-    {
-      "epoch": 0.5373504999673224,
-      "grad_norm": 0.44778013229370117,
-      "learning_rate": 9.363656818228853e-06,
-      "loss": 0.3722,
-      "step": 8222
-    },
-    {
-      "epoch": 0.5374158551728645,
-      "grad_norm": 0.4640384316444397,
-      "learning_rate": 9.363486332766153e-06,
-      "loss": 0.392,
-      "step": 8223
-    },
-    {
-      "epoch": 0.5374812103784067,
-      "grad_norm": 0.4378077983856201,
-      "learning_rate": 9.363315826021218e-06,
-      "loss": 0.3851,
-      "step": 8224
-    },
-    {
-      "epoch": 0.5375465655839488,
-      "grad_norm": 0.47820448875427246,
-      "learning_rate": 9.36314529799488e-06,
-      "loss": 0.4294,
-      "step": 8225
-    },
-    {
-      "epoch": 0.5376119207894909,
-      "grad_norm": 0.5370205044746399,
-      "learning_rate": 9.362974748687969e-06,
-      "loss": 0.4164,
-      "step": 8226
-    },
-    {
-      "epoch": 0.537677275995033,
-      "grad_norm": 0.47790366411209106,
-      "learning_rate": 9.36280417810132e-06,
-      "loss": 0.4198,
-      "step": 8227
-    },
-    {
-      "epoch": 0.5377426312005751,
-      "grad_norm": 0.49692991375923157,
-      "learning_rate": 9.362633586235766e-06,
-      "loss": 0.4289,
-      "step": 8228
-    },
-    {
-      "epoch": 0.5378079864061173,
-      "grad_norm": 0.4387872815132141,
-      "learning_rate": 9.362462973092134e-06,
-      "loss": 0.3348,
-      "step": 8229
-    },
-    {
-      "epoch": 0.5378733416116593,
-      "grad_norm": 0.47035783529281616,
-      "learning_rate": 9.36229233867126e-06,
-      "loss": 0.3845,
-      "step": 8230
-    },
-    {
-      "epoch": 0.5379386968172015,
-      "grad_norm": 0.46790122985839844,
-      "learning_rate": 9.362121682973972e-06,
-      "loss": 0.4011,
-      "step": 8231
-    },
-    {
-      "epoch": 0.5380040520227436,
-      "grad_norm": 0.4822106659412384,
-      "learning_rate": 9.361951006001109e-06,
-      "loss": 0.4041,
-      "step": 8232
-    },
-    {
-      "epoch": 0.5380694072282858,
-      "grad_norm": 0.5083541870117188,
-      "learning_rate": 9.361780307753498e-06,
-      "loss": 0.4704,
-      "step": 8233
-    },
-    {
-      "epoch": 0.5381347624338279,
-      "grad_norm": 0.48938223719596863,
-      "learning_rate": 9.361609588231974e-06,
-      "loss": 0.416,
-      "step": 8234
-    },
-    {
-      "epoch": 0.5382001176393699,
-      "grad_norm": 0.45437952876091003,
-      "learning_rate": 9.36143884743737e-06,
-      "loss": 0.3507,
-      "step": 8235
-    },
-    {
-      "epoch": 0.5382654728449121,
-      "grad_norm": 0.4548284411430359,
-      "learning_rate": 9.361268085370517e-06,
-      "loss": 0.3796,
-      "step": 8236
-    },
-    {
-      "epoch": 0.5383308280504542,
-      "grad_norm": 0.4518842399120331,
-      "learning_rate": 9.361097302032246e-06,
-      "loss": 0.3668,
-      "step": 8237
-    },
-    {
-      "epoch": 0.5383961832559964,
-      "grad_norm": 0.44761666655540466,
-      "learning_rate": 9.360926497423397e-06,
-      "loss": 0.3683,
-      "step": 8238
-    },
-    {
-      "epoch": 0.5384615384615384,
-      "grad_norm": 0.4721292555332184,
-      "learning_rate": 9.360755671544797e-06,
-      "loss": 0.3791,
-      "step": 8239
-    },
-    {
-      "epoch": 0.5385268936670806,
-      "grad_norm": 0.46845686435699463,
-      "learning_rate": 9.36058482439728e-06,
-      "loss": 0.3935,
-      "step": 8240
-    },
-    {
-      "epoch": 0.5385922488726227,
-      "grad_norm": 0.44099998474121094,
-      "learning_rate": 9.360413955981679e-06,
-      "loss": 0.3935,
-      "step": 8241
-    },
-    {
-      "epoch": 0.5386576040781649,
-      "grad_norm": 0.4769476056098938,
-      "learning_rate": 9.36024306629883e-06,
-      "loss": 0.4344,
-      "step": 8242
-    },
-    {
-      "epoch": 0.538722959283707,
-      "grad_norm": 0.45490676164627075,
-      "learning_rate": 9.360072155349567e-06,
-      "loss": 0.4212,
-      "step": 8243
-    },
-    {
-      "epoch": 0.538788314489249,
-      "grad_norm": 0.4358070492744446,
-      "learning_rate": 9.35990122313472e-06,
-      "loss": 0.3922,
-      "step": 8244
-    },
-    {
-      "epoch": 0.5388536696947912,
-      "grad_norm": 0.4774872064590454,
-      "learning_rate": 9.359730269655124e-06,
-      "loss": 0.416,
-      "step": 8245
-    },
-    {
-      "epoch": 0.5389190249003333,
-      "grad_norm": 0.4705999791622162,
-      "learning_rate": 9.359559294911613e-06,
-      "loss": 0.4095,
-      "step": 8246
-    },
-    {
-      "epoch": 0.5389843801058755,
-      "grad_norm": 0.4116984009742737,
-      "learning_rate": 9.359388298905023e-06,
-      "loss": 0.3215,
-      "step": 8247
-    },
-    {
-      "epoch": 0.5390497353114175,
-      "grad_norm": 0.4325437843799591,
-      "learning_rate": 9.359217281636183e-06,
-      "loss": 0.3766,
-      "step": 8248
-    },
-    {
-      "epoch": 0.5391150905169597,
-      "grad_norm": 0.4551779329776764,
-      "learning_rate": 9.359046243105932e-06,
-      "loss": 0.4135,
-      "step": 8249
-    },
-    {
-      "epoch": 0.5391804457225018,
-      "grad_norm": 0.4630483090877533,
-      "learning_rate": 9.358875183315102e-06,
-      "loss": 0.4103,
-      "step": 8250
-    },
-    {
-      "epoch": 0.539245800928044,
-      "grad_norm": 0.46130189299583435,
-      "learning_rate": 9.358704102264527e-06,
-      "loss": 0.3982,
-      "step": 8251
-    },
-    {
-      "epoch": 0.539311156133586,
-      "grad_norm": 0.4228488802909851,
-      "learning_rate": 9.358532999955043e-06,
-      "loss": 0.3311,
-      "step": 8252
-    },
-    {
-      "epoch": 0.5393765113391281,
-      "grad_norm": 0.4424111247062683,
-      "learning_rate": 9.358361876387482e-06,
-      "loss": 0.3562,
-      "step": 8253
-    },
-    {
-      "epoch": 0.5394418665446703,
-      "grad_norm": 0.48746955394744873,
-      "learning_rate": 9.358190731562682e-06,
-      "loss": 0.4175,
-      "step": 8254
-    },
-    {
-      "epoch": 0.5395072217502124,
-      "grad_norm": 0.43694406747817993,
-      "learning_rate": 9.358019565481477e-06,
-      "loss": 0.3566,
-      "step": 8255
-    },
-    {
-      "epoch": 0.5395725769557546,
-      "grad_norm": 0.44341006875038147,
-      "learning_rate": 9.357848378144698e-06,
-      "loss": 0.3726,
-      "step": 8256
-    },
-    {
-      "epoch": 0.5396379321612966,
-      "grad_norm": 0.47627320885658264,
-      "learning_rate": 9.357677169553186e-06,
-      "loss": 0.4182,
-      "step": 8257
-    },
-    {
-      "epoch": 0.5397032873668388,
-      "grad_norm": 0.4702316224575043,
-      "learning_rate": 9.357505939707769e-06,
-      "loss": 0.4434,
-      "step": 8258
-    },
-    {
-      "epoch": 0.5397686425723809,
-      "grad_norm": 0.45931556820869446,
-      "learning_rate": 9.35733468860929e-06,
-      "loss": 0.3788,
-      "step": 8259
-    },
-    {
-      "epoch": 0.539833997777923,
-      "grad_norm": 0.44867175817489624,
-      "learning_rate": 9.357163416258577e-06,
-      "loss": 0.382,
-      "step": 8260
-    },
-    {
-      "epoch": 0.5398993529834651,
-      "grad_norm": 0.4833263158798218,
-      "learning_rate": 9.35699212265647e-06,
-      "loss": 0.4326,
-      "step": 8261
-    },
-    {
-      "epoch": 0.5399647081890072,
-      "grad_norm": 0.45278486609458923,
-      "learning_rate": 9.356820807803802e-06,
-      "loss": 0.3846,
-      "step": 8262
-    },
-    {
-      "epoch": 0.5400300633945494,
-      "grad_norm": 0.5050371289253235,
-      "learning_rate": 9.35664947170141e-06,
-      "loss": 0.4079,
-      "step": 8263
-    },
-    {
-      "epoch": 0.5400954186000915,
-      "grad_norm": 0.4680522382259369,
-      "learning_rate": 9.35647811435013e-06,
-      "loss": 0.4307,
-      "step": 8264
-    },
-    {
-      "epoch": 0.5401607738056337,
-      "grad_norm": 0.4409938156604767,
-      "learning_rate": 9.356306735750796e-06,
-      "loss": 0.3881,
-      "step": 8265
-    },
-    {
-      "epoch": 0.5402261290111757,
-      "grad_norm": 0.450053870677948,
-      "learning_rate": 9.356135335904247e-06,
-      "loss": 0.4038,
-      "step": 8266
-    },
-    {
-      "epoch": 0.5402914842167179,
-      "grad_norm": 0.46919873356819153,
-      "learning_rate": 9.355963914811313e-06,
-      "loss": 0.419,
-      "step": 8267
-    },
-    {
-      "epoch": 0.54035683942226,
-      "grad_norm": 0.4970439672470093,
-      "learning_rate": 9.355792472472836e-06,
-      "loss": 0.436,
-      "step": 8268
-    },
-    {
-      "epoch": 0.540422194627802,
-      "grad_norm": 0.468803346157074,
-      "learning_rate": 9.355621008889651e-06,
-      "loss": 0.3586,
-      "step": 8269
-    },
-    {
-      "epoch": 0.5404875498333442,
-      "grad_norm": 0.5252040028572083,
-      "learning_rate": 9.355449524062592e-06,
-      "loss": 0.4647,
-      "step": 8270
-    },
-    {
-      "epoch": 0.5405529050388863,
-      "grad_norm": 0.46350598335266113,
-      "learning_rate": 9.355278017992498e-06,
-      "loss": 0.4013,
-      "step": 8271
-    },
-    {
-      "epoch": 0.5406182602444285,
-      "grad_norm": 0.44724878668785095,
-      "learning_rate": 9.355106490680204e-06,
-      "loss": 0.3655,
-      "step": 8272
-    },
-    {
-      "epoch": 0.5406836154499706,
-      "grad_norm": 0.43497711420059204,
-      "learning_rate": 9.354934942126545e-06,
-      "loss": 0.3421,
-      "step": 8273
-    },
-    {
-      "epoch": 0.5407489706555128,
-      "grad_norm": 0.46079549193382263,
-      "learning_rate": 9.354763372332362e-06,
-      "loss": 0.3899,
-      "step": 8274
-    },
-    {
-      "epoch": 0.5408143258610548,
-      "grad_norm": 0.4535283148288727,
-      "learning_rate": 9.35459178129849e-06,
-      "loss": 0.4033,
-      "step": 8275
-    },
-    {
-      "epoch": 0.540879681066597,
-      "grad_norm": 0.450179785490036,
-      "learning_rate": 9.354420169025763e-06,
-      "loss": 0.4097,
-      "step": 8276
-    },
-    {
-      "epoch": 0.5409450362721391,
-      "grad_norm": 0.4596855640411377,
-      "learning_rate": 9.354248535515021e-06,
-      "loss": 0.4309,
-      "step": 8277
-    },
-    {
-      "epoch": 0.5410103914776812,
-      "grad_norm": 0.41704466938972473,
-      "learning_rate": 9.354076880767102e-06,
-      "loss": 0.3716,
-      "step": 8278
-    },
-    {
-      "epoch": 0.5410757466832233,
-      "grad_norm": 0.44406241178512573,
-      "learning_rate": 9.35390520478284e-06,
-      "loss": 0.4148,
-      "step": 8279
-    },
-    {
-      "epoch": 0.5411411018887654,
-      "grad_norm": 0.461429238319397,
-      "learning_rate": 9.353733507563074e-06,
-      "loss": 0.4248,
-      "step": 8280
-    },
-    {
-      "epoch": 0.5412064570943076,
-      "grad_norm": 0.4740643799304962,
-      "learning_rate": 9.353561789108641e-06,
-      "loss": 0.3924,
-      "step": 8281
-    },
-    {
-      "epoch": 0.5412718122998497,
-      "grad_norm": 0.4778648018836975,
-      "learning_rate": 9.35339004942038e-06,
-      "loss": 0.4251,
-      "step": 8282
-    },
-    {
-      "epoch": 0.5413371675053918,
-      "grad_norm": 0.45684945583343506,
-      "learning_rate": 9.353218288499127e-06,
-      "loss": 0.3904,
-      "step": 8283
-    },
-    {
-      "epoch": 0.5414025227109339,
-      "grad_norm": 0.4378284811973572,
-      "learning_rate": 9.35304650634572e-06,
-      "loss": 0.3769,
-      "step": 8284
-    },
-    {
-      "epoch": 0.541467877916476,
-      "grad_norm": 0.4128952920436859,
-      "learning_rate": 9.352874702960998e-06,
-      "loss": 0.3363,
-      "step": 8285
-    },
-    {
-      "epoch": 0.5415332331220182,
-      "grad_norm": 0.4351802170276642,
-      "learning_rate": 9.352702878345799e-06,
-      "loss": 0.384,
-      "step": 8286
-    },
-    {
-      "epoch": 0.5415985883275602,
-      "grad_norm": 0.4509058892726898,
-      "learning_rate": 9.352531032500958e-06,
-      "loss": 0.3828,
-      "step": 8287
-    },
-    {
-      "epoch": 0.5416639435331024,
-      "grad_norm": 0.44013240933418274,
-      "learning_rate": 9.352359165427317e-06,
-      "loss": 0.3556,
-      "step": 8288
-    },
-    {
-      "epoch": 0.5417292987386445,
-      "grad_norm": 0.4528917074203491,
-      "learning_rate": 9.35218727712571e-06,
-      "loss": 0.3944,
-      "step": 8289
-    },
-    {
-      "epoch": 0.5417946539441867,
-      "grad_norm": 0.43234896659851074,
-      "learning_rate": 9.35201536759698e-06,
-      "loss": 0.3683,
-      "step": 8290
-    },
-    {
-      "epoch": 0.5418600091497288,
-      "grad_norm": 0.5005369186401367,
-      "learning_rate": 9.351843436841964e-06,
-      "loss": 0.4255,
-      "step": 8291
-    },
-    {
-      "epoch": 0.5419253643552709,
-      "grad_norm": 0.45198288559913635,
-      "learning_rate": 9.3516714848615e-06,
-      "loss": 0.3764,
-      "step": 8292
-    },
-    {
-      "epoch": 0.541990719560813,
-      "grad_norm": 0.45872727036476135,
-      "learning_rate": 9.351499511656424e-06,
-      "loss": 0.3912,
-      "step": 8293
-    },
-    {
-      "epoch": 0.5420560747663551,
-      "grad_norm": 0.48523542284965515,
-      "learning_rate": 9.35132751722758e-06,
-      "loss": 0.4227,
-      "step": 8294
-    },
-    {
-      "epoch": 0.5421214299718973,
-      "grad_norm": 0.4680887460708618,
-      "learning_rate": 9.351155501575803e-06,
-      "loss": 0.4333,
-      "step": 8295
-    },
-    {
-      "epoch": 0.5421867851774393,
-      "grad_norm": 0.45950746536254883,
-      "learning_rate": 9.350983464701932e-06,
-      "loss": 0.3704,
-      "step": 8296
-    },
-    {
-      "epoch": 0.5422521403829815,
-      "grad_norm": 0.4689929187297821,
-      "learning_rate": 9.35081140660681e-06,
-      "loss": 0.4157,
-      "step": 8297
-    },
-    {
-      "epoch": 0.5423174955885236,
-      "grad_norm": 0.4797610640525818,
-      "learning_rate": 9.35063932729127e-06,
-      "loss": 0.4177,
-      "step": 8298
-    },
-    {
-      "epoch": 0.5423828507940658,
-      "grad_norm": 0.4746837317943573,
-      "learning_rate": 9.350467226756159e-06,
-      "loss": 0.4244,
-      "step": 8299
-    },
-    {
-      "epoch": 0.5424482059996079,
-      "grad_norm": 0.4472707211971283,
-      "learning_rate": 9.350295105002311e-06,
-      "loss": 0.3941,
-      "step": 8300
-    },
-    {
-      "epoch": 0.54251356120515,
-      "grad_norm": 0.41988304257392883,
-      "learning_rate": 9.350122962030566e-06,
-      "loss": 0.3321,
-      "step": 8301
-    },
-    {
-      "epoch": 0.5425789164106921,
-      "grad_norm": 0.4819179177284241,
-      "learning_rate": 9.349950797841763e-06,
-      "loss": 0.4132,
-      "step": 8302
-    },
-    {
-      "epoch": 0.5426442716162342,
-      "grad_norm": 0.41595616936683655,
-      "learning_rate": 9.349778612436743e-06,
-      "loss": 0.3283,
-      "step": 8303
-    },
-    {
-      "epoch": 0.5427096268217764,
-      "grad_norm": 0.4350241720676422,
-      "learning_rate": 9.349606405816349e-06,
-      "loss": 0.3859,
-      "step": 8304
-    },
-    {
-      "epoch": 0.5427749820273184,
-      "grad_norm": 0.448354035615921,
-      "learning_rate": 9.349434177981416e-06,
-      "loss": 0.3946,
-      "step": 8305
-    },
-    {
-      "epoch": 0.5428403372328606,
-      "grad_norm": 0.4604875147342682,
-      "learning_rate": 9.349261928932785e-06,
-      "loss": 0.3571,
-      "step": 8306
-    },
-    {
-      "epoch": 0.5429056924384027,
-      "grad_norm": 0.47523099184036255,
-      "learning_rate": 9.349089658671297e-06,
-      "loss": 0.3528,
-      "step": 8307
-    },
-    {
-      "epoch": 0.5429710476439449,
-      "grad_norm": 0.43087831139564514,
-      "learning_rate": 9.348917367197791e-06,
-      "loss": 0.3273,
-      "step": 8308
-    },
-    {
-      "epoch": 0.543036402849487,
-      "grad_norm": 0.48309123516082764,
-      "learning_rate": 9.348745054513112e-06,
-      "loss": 0.407,
-      "step": 8309
-    },
-    {
-      "epoch": 0.5431017580550291,
-      "grad_norm": 0.432910293340683,
-      "learning_rate": 9.348572720618095e-06,
-      "loss": 0.354,
-      "step": 8310
-    },
-    {
-      "epoch": 0.5431671132605712,
-      "grad_norm": 0.4773752689361572,
-      "learning_rate": 9.348400365513582e-06,
-      "loss": 0.3865,
-      "step": 8311
-    },
-    {
-      "epoch": 0.5432324684661133,
-      "grad_norm": 0.48800089955329895,
-      "learning_rate": 9.348227989200413e-06,
-      "loss": 0.4368,
-      "step": 8312
-    },
-    {
-      "epoch": 0.5432978236716555,
-      "grad_norm": 0.45648813247680664,
-      "learning_rate": 9.34805559167943e-06,
-      "loss": 0.3918,
-      "step": 8313
-    },
-    {
-      "epoch": 0.5433631788771975,
-      "grad_norm": 0.4994303286075592,
-      "learning_rate": 9.347883172951474e-06,
-      "loss": 0.4546,
-      "step": 8314
-    },
-    {
-      "epoch": 0.5434285340827397,
-      "grad_norm": 0.46923038363456726,
-      "learning_rate": 9.347710733017386e-06,
-      "loss": 0.374,
-      "step": 8315
-    },
-    {
-      "epoch": 0.5434938892882818,
-      "grad_norm": 0.4745754301548004,
-      "learning_rate": 9.347538271878007e-06,
-      "loss": 0.3916,
-      "step": 8316
-    },
-    {
-      "epoch": 0.543559244493824,
-      "grad_norm": 0.43778276443481445,
-      "learning_rate": 9.347365789534176e-06,
-      "loss": 0.3438,
-      "step": 8317
-    },
-    {
-      "epoch": 0.543624599699366,
-      "grad_norm": 0.473345011472702,
-      "learning_rate": 9.347193285986738e-06,
-      "loss": 0.434,
-      "step": 8318
-    },
-    {
-      "epoch": 0.5436899549049081,
-      "grad_norm": 0.41520264744758606,
-      "learning_rate": 9.347020761236531e-06,
-      "loss": 0.3155,
-      "step": 8319
-    },
-    {
-      "epoch": 0.5437553101104503,
-      "grad_norm": 0.4335134029388428,
-      "learning_rate": 9.346848215284397e-06,
-      "loss": 0.3402,
-      "step": 8320
-    },
-    {
-      "epoch": 0.5438206653159924,
-      "grad_norm": 0.46292465925216675,
-      "learning_rate": 9.346675648131181e-06,
-      "loss": 0.3918,
-      "step": 8321
-    },
-    {
-      "epoch": 0.5438860205215346,
-      "grad_norm": 0.46981081366539,
-      "learning_rate": 9.34650305977772e-06,
-      "loss": 0.3766,
-      "step": 8322
-    },
-    {
-      "epoch": 0.5439513757270766,
-      "grad_norm": 0.47351184487342834,
-      "learning_rate": 9.346330450224858e-06,
-      "loss": 0.4171,
-      "step": 8323
-    },
-    {
-      "epoch": 0.5440167309326188,
-      "grad_norm": 0.44288018345832825,
-      "learning_rate": 9.346157819473437e-06,
-      "loss": 0.4062,
-      "step": 8324
-    },
-    {
-      "epoch": 0.5440820861381609,
-      "grad_norm": 0.5060564875602722,
-      "learning_rate": 9.345985167524298e-06,
-      "loss": 0.4093,
-      "step": 8325
-    },
-    {
-      "epoch": 0.5441474413437031,
-      "grad_norm": 0.492012619972229,
-      "learning_rate": 9.345812494378285e-06,
-      "loss": 0.3613,
-      "step": 8326
-    },
-    {
-      "epoch": 0.5442127965492451,
-      "grad_norm": 0.45168179273605347,
-      "learning_rate": 9.345639800036238e-06,
-      "loss": 0.3714,
-      "step": 8327
-    },
-    {
-      "epoch": 0.5442781517547872,
-      "grad_norm": 0.48553574085235596,
-      "learning_rate": 9.345467084499e-06,
-      "loss": 0.4131,
-      "step": 8328
-    },
-    {
-      "epoch": 0.5443435069603294,
-      "grad_norm": 0.4387553334236145,
-      "learning_rate": 9.345294347767415e-06,
-      "loss": 0.3659,
-      "step": 8329
-    },
-    {
-      "epoch": 0.5444088621658715,
-      "grad_norm": 0.419382244348526,
-      "learning_rate": 9.345121589842323e-06,
-      "loss": 0.3591,
-      "step": 8330
-    },
-    {
-      "epoch": 0.5444742173714137,
-      "grad_norm": 0.47618135809898376,
-      "learning_rate": 9.344948810724567e-06,
-      "loss": 0.3983,
-      "step": 8331
-    },
-    {
-      "epoch": 0.5445395725769557,
-      "grad_norm": 0.48393312096595764,
-      "learning_rate": 9.344776010414994e-06,
-      "loss": 0.4282,
-      "step": 8332
-    },
-    {
-      "epoch": 0.5446049277824979,
-      "grad_norm": 0.4687272906303406,
-      "learning_rate": 9.344603188914438e-06,
-      "loss": 0.4076,
-      "step": 8333
-    },
-    {
-      "epoch": 0.54467028298804,
-      "grad_norm": 0.4399765729904175,
-      "learning_rate": 9.34443034622375e-06,
-      "loss": 0.3748,
-      "step": 8334
-    },
-    {
-      "epoch": 0.5447356381935822,
-      "grad_norm": 0.4569462239742279,
-      "learning_rate": 9.344257482343771e-06,
-      "loss": 0.4102,
-      "step": 8335
-    },
-    {
-      "epoch": 0.5448009933991242,
-      "grad_norm": 0.4702807664871216,
-      "learning_rate": 9.34408459727534e-06,
-      "loss": 0.4229,
-      "step": 8336
-    },
-    {
-      "epoch": 0.5448663486046663,
-      "grad_norm": 0.4702181816101074,
-      "learning_rate": 9.343911691019308e-06,
-      "loss": 0.406,
-      "step": 8337
-    },
-    {
-      "epoch": 0.5449317038102085,
-      "grad_norm": 0.44575101137161255,
-      "learning_rate": 9.343738763576511e-06,
-      "loss": 0.4027,
-      "step": 8338
-    },
-    {
-      "epoch": 0.5449970590157506,
-      "grad_norm": 0.469691663980484,
-      "learning_rate": 9.343565814947796e-06,
-      "loss": 0.388,
-      "step": 8339
-    },
-    {
-      "epoch": 0.5450624142212928,
-      "grad_norm": 0.425841748714447,
-      "learning_rate": 9.343392845134005e-06,
-      "loss": 0.3468,
-      "step": 8340
-    },
-    {
-      "epoch": 0.5451277694268348,
-      "grad_norm": 0.43938112258911133,
-      "learning_rate": 9.343219854135984e-06,
-      "loss": 0.3291,
-      "step": 8341
-    },
-    {
-      "epoch": 0.545193124632377,
-      "grad_norm": 0.44285666942596436,
-      "learning_rate": 9.343046841954572e-06,
-      "loss": 0.3704,
-      "step": 8342
-    },
-    {
-      "epoch": 0.5452584798379191,
-      "grad_norm": 0.4263397455215454,
-      "learning_rate": 9.342873808590617e-06,
-      "loss": 0.3635,
-      "step": 8343
-    },
-    {
-      "epoch": 0.5453238350434612,
-      "grad_norm": 0.4233444929122925,
-      "learning_rate": 9.342700754044965e-06,
-      "loss": 0.3191,
-      "step": 8344
-    },
-    {
-      "epoch": 0.5453891902490033,
-      "grad_norm": 0.4461345374584198,
-      "learning_rate": 9.342527678318454e-06,
-      "loss": 0.3777,
-      "step": 8345
-    },
-    {
-      "epoch": 0.5454545454545454,
-      "grad_norm": 0.4585961699485779,
-      "learning_rate": 9.342354581411932e-06,
-      "loss": 0.4125,
-      "step": 8346
-    },
-    {
-      "epoch": 0.5455199006600876,
-      "grad_norm": 0.5012494921684265,
-      "learning_rate": 9.342181463326243e-06,
-      "loss": 0.4523,
-      "step": 8347
-    },
-    {
-      "epoch": 0.5455852558656297,
-      "grad_norm": 0.4398760199546814,
-      "learning_rate": 9.342008324062229e-06,
-      "loss": 0.3686,
-      "step": 8348
-    },
-    {
-      "epoch": 0.5456506110711719,
-      "grad_norm": 0.4209686815738678,
-      "learning_rate": 9.341835163620738e-06,
-      "loss": 0.3687,
-      "step": 8349
-    },
-    {
-      "epoch": 0.5457159662767139,
-      "grad_norm": 0.5082706212997437,
-      "learning_rate": 9.341661982002612e-06,
-      "loss": 0.4654,
-      "step": 8350
-    },
-    {
-      "epoch": 0.5457813214822561,
-      "grad_norm": 0.43471962213516235,
-      "learning_rate": 9.341488779208696e-06,
-      "loss": 0.3444,
-      "step": 8351
-    },
-    {
-      "epoch": 0.5458466766877982,
-      "grad_norm": 0.45494237542152405,
-      "learning_rate": 9.341315555239835e-06,
-      "loss": 0.4041,
-      "step": 8352
-    },
-    {
-      "epoch": 0.5459120318933403,
-      "grad_norm": 0.46279793977737427,
-      "learning_rate": 9.341142310096876e-06,
-      "loss": 0.3932,
-      "step": 8353
-    },
-    {
-      "epoch": 0.5459773870988824,
-      "grad_norm": 0.45075204968452454,
-      "learning_rate": 9.34096904378066e-06,
-      "loss": 0.3802,
-      "step": 8354
-    },
-    {
-      "epoch": 0.5460427423044245,
-      "grad_norm": 0.4650687277317047,
-      "learning_rate": 9.340795756292036e-06,
-      "loss": 0.3837,
-      "step": 8355
-    },
-    {
-      "epoch": 0.5461080975099667,
-      "grad_norm": 0.46025142073631287,
-      "learning_rate": 9.340622447631844e-06,
-      "loss": 0.3972,
-      "step": 8356
-    },
-    {
-      "epoch": 0.5461734527155088,
-      "grad_norm": 0.46367308497428894,
-      "learning_rate": 9.340449117800936e-06,
-      "loss": 0.3789,
-      "step": 8357
-    },
-    {
-      "epoch": 0.546238807921051,
-      "grad_norm": 0.4615289270877838,
-      "learning_rate": 9.340275766800153e-06,
-      "loss": 0.3768,
-      "step": 8358
-    },
-    {
-      "epoch": 0.546304163126593,
-      "grad_norm": 0.4335293471813202,
-      "learning_rate": 9.34010239463034e-06,
-      "loss": 0.365,
-      "step": 8359
-    },
-    {
-      "epoch": 0.5463695183321352,
-      "grad_norm": 0.4475209414958954,
-      "learning_rate": 9.339929001292345e-06,
-      "loss": 0.3904,
-      "step": 8360
-    },
-    {
-      "epoch": 0.5464348735376773,
-      "grad_norm": 0.45525529980659485,
-      "learning_rate": 9.339755586787014e-06,
-      "loss": 0.3787,
-      "step": 8361
-    },
-    {
-      "epoch": 0.5465002287432194,
-      "grad_norm": 0.47592693567276,
-      "learning_rate": 9.33958215111519e-06,
-      "loss": 0.4347,
-      "step": 8362
-    },
-    {
-      "epoch": 0.5465655839487615,
-      "grad_norm": 0.4509865939617157,
-      "learning_rate": 9.33940869427772e-06,
-      "loss": 0.4164,
-      "step": 8363
-    },
-    {
-      "epoch": 0.5466309391543036,
-      "grad_norm": 0.49032536149024963,
-      "learning_rate": 9.339235216275453e-06,
-      "loss": 0.4464,
-      "step": 8364
-    },
-    {
-      "epoch": 0.5466962943598458,
-      "grad_norm": 0.46070197224617004,
-      "learning_rate": 9.33906171710923e-06,
-      "loss": 0.3816,
-      "step": 8365
-    },
-    {
-      "epoch": 0.5467616495653879,
-      "grad_norm": 0.4691459536552429,
-      "learning_rate": 9.338888196779901e-06,
-      "loss": 0.4051,
-      "step": 8366
-    },
-    {
-      "epoch": 0.54682700477093,
-      "grad_norm": 0.4831182062625885,
-      "learning_rate": 9.338714655288311e-06,
-      "loss": 0.4048,
-      "step": 8367
-    },
-    {
-      "epoch": 0.5468923599764721,
-      "grad_norm": 0.4534238576889038,
-      "learning_rate": 9.338541092635307e-06,
-      "loss": 0.3946,
-      "step": 8368
-    },
-    {
-      "epoch": 0.5469577151820142,
-      "grad_norm": 0.448739230632782,
-      "learning_rate": 9.338367508821734e-06,
-      "loss": 0.3623,
-      "step": 8369
-    },
-    {
-      "epoch": 0.5470230703875564,
-      "grad_norm": 0.4333237409591675,
-      "learning_rate": 9.33819390384844e-06,
-      "loss": 0.3204,
-      "step": 8370
-    },
-    {
-      "epoch": 0.5470884255930984,
-      "grad_norm": 0.40983936190605164,
-      "learning_rate": 9.338020277716273e-06,
-      "loss": 0.3431,
-      "step": 8371
-    },
-    {
-      "epoch": 0.5471537807986406,
-      "grad_norm": 0.4308077394962311,
-      "learning_rate": 9.337846630426077e-06,
-      "loss": 0.3358,
-      "step": 8372
-    },
-    {
-      "epoch": 0.5472191360041827,
-      "grad_norm": 0.4654064178466797,
-      "learning_rate": 9.3376729619787e-06,
-      "loss": 0.3808,
-      "step": 8373
-    },
-    {
-      "epoch": 0.5472844912097249,
-      "grad_norm": 0.46433669328689575,
-      "learning_rate": 9.33749927237499e-06,
-      "loss": 0.4165,
-      "step": 8374
-    },
-    {
-      "epoch": 0.547349846415267,
-      "grad_norm": 0.4505382180213928,
-      "learning_rate": 9.337325561615793e-06,
-      "loss": 0.3827,
-      "step": 8375
-    },
-    {
-      "epoch": 0.5474152016208091,
-      "grad_norm": 0.4646647870540619,
-      "learning_rate": 9.337151829701955e-06,
-      "loss": 0.385,
-      "step": 8376
-    },
-    {
-      "epoch": 0.5474805568263512,
-      "grad_norm": 0.47459107637405396,
-      "learning_rate": 9.336978076634327e-06,
-      "loss": 0.4016,
-      "step": 8377
-    },
-    {
-      "epoch": 0.5475459120318933,
-      "grad_norm": 0.4663126468658447,
-      "learning_rate": 9.336804302413755e-06,
-      "loss": 0.4476,
-      "step": 8378
-    },
-    {
-      "epoch": 0.5476112672374355,
-      "grad_norm": 0.44067710638046265,
-      "learning_rate": 9.336630507041085e-06,
-      "loss": 0.3719,
-      "step": 8379
-    },
-    {
-      "epoch": 0.5476766224429775,
-      "grad_norm": 0.4350115954875946,
-      "learning_rate": 9.336456690517165e-06,
-      "loss": 0.3466,
-      "step": 8380
-    },
-    {
-      "epoch": 0.5477419776485197,
-      "grad_norm": 0.42475584149360657,
-      "learning_rate": 9.336282852842844e-06,
-      "loss": 0.3543,
-      "step": 8381
-    },
-    {
-      "epoch": 0.5478073328540618,
-      "grad_norm": 0.459639310836792,
-      "learning_rate": 9.33610899401897e-06,
-      "loss": 0.3779,
-      "step": 8382
-    },
-    {
-      "epoch": 0.547872688059604,
-      "grad_norm": 0.5365619659423828,
-      "learning_rate": 9.335935114046389e-06,
-      "loss": 0.405,
-      "step": 8383
-    },
-    {
-      "epoch": 0.5479380432651461,
-      "grad_norm": 0.4446386992931366,
-      "learning_rate": 9.335761212925951e-06,
-      "loss": 0.4018,
-      "step": 8384
-    },
-    {
-      "epoch": 0.5480033984706882,
-      "grad_norm": 0.4449726641178131,
-      "learning_rate": 9.335587290658504e-06,
-      "loss": 0.3521,
-      "step": 8385
-    },
-    {
-      "epoch": 0.5480687536762303,
-      "grad_norm": 0.4633252024650574,
-      "learning_rate": 9.335413347244895e-06,
-      "loss": 0.3893,
-      "step": 8386
-    },
-    {
-      "epoch": 0.5481341088817724,
-      "grad_norm": 0.4171968400478363,
-      "learning_rate": 9.335239382685974e-06,
-      "loss": 0.3183,
-      "step": 8387
-    },
-    {
-      "epoch": 0.5481994640873146,
-      "grad_norm": 0.43910449743270874,
-      "learning_rate": 9.335065396982588e-06,
-      "loss": 0.3865,
-      "step": 8388
-    },
-    {
-      "epoch": 0.5482648192928566,
-      "grad_norm": 0.4020332098007202,
-      "learning_rate": 9.334891390135586e-06,
-      "loss": 0.3498,
-      "step": 8389
-    },
-    {
-      "epoch": 0.5483301744983988,
-      "grad_norm": 0.4654817283153534,
-      "learning_rate": 9.334717362145818e-06,
-      "loss": 0.4165,
-      "step": 8390
-    },
-    {
-      "epoch": 0.5483955297039409,
-      "grad_norm": 0.4696250557899475,
-      "learning_rate": 9.33454331301413e-06,
-      "loss": 0.3939,
-      "step": 8391
-    },
-    {
-      "epoch": 0.5484608849094831,
-      "grad_norm": 0.5158215761184692,
-      "learning_rate": 9.334369242741374e-06,
-      "loss": 0.4496,
-      "step": 8392
-    },
-    {
-      "epoch": 0.5485262401150252,
-      "grad_norm": 0.4201121926307678,
-      "learning_rate": 9.334195151328398e-06,
-      "loss": 0.3564,
-      "step": 8393
-    },
-    {
-      "epoch": 0.5485915953205673,
-      "grad_norm": 0.4531320035457611,
-      "learning_rate": 9.334021038776048e-06,
-      "loss": 0.4175,
-      "step": 8394
-    },
-    {
-      "epoch": 0.5486569505261094,
-      "grad_norm": 0.42785894870758057,
-      "learning_rate": 9.33384690508518e-06,
-      "loss": 0.3401,
-      "step": 8395
-    },
-    {
-      "epoch": 0.5487223057316515,
-      "grad_norm": 0.4672718942165375,
-      "learning_rate": 9.333672750256636e-06,
-      "loss": 0.4006,
-      "step": 8396
-    },
-    {
-      "epoch": 0.5487876609371937,
-      "grad_norm": 0.455115407705307,
-      "learning_rate": 9.333498574291272e-06,
-      "loss": 0.3861,
-      "step": 8397
-    },
-    {
-      "epoch": 0.5488530161427357,
-      "grad_norm": 0.4673006534576416,
-      "learning_rate": 9.333324377189931e-06,
-      "loss": 0.4229,
-      "step": 8398
-    },
-    {
-      "epoch": 0.5489183713482779,
-      "grad_norm": 0.4351649880409241,
-      "learning_rate": 9.333150158953467e-06,
-      "loss": 0.3819,
-      "step": 8399
-    },
-    {
-      "epoch": 0.54898372655382,
-      "grad_norm": 0.44421637058258057,
-      "learning_rate": 9.332975919582727e-06,
-      "loss": 0.3912,
-      "step": 8400
-    },
-    {
-      "epoch": 0.5490490817593622,
-      "grad_norm": 0.4454924166202545,
-      "learning_rate": 9.332801659078565e-06,
-      "loss": 0.3946,
-      "step": 8401
-    },
-    {
-      "epoch": 0.5491144369649043,
-      "grad_norm": 0.47957855463027954,
-      "learning_rate": 9.332627377441827e-06,
-      "loss": 0.4334,
-      "step": 8402
-    },
-    {
-      "epoch": 0.5491797921704463,
-      "grad_norm": 0.48958173394203186,
-      "learning_rate": 9.332453074673365e-06,
-      "loss": 0.4617,
-      "step": 8403
-    },
-    {
-      "epoch": 0.5492451473759885,
-      "grad_norm": 0.46007731556892395,
-      "learning_rate": 9.332278750774026e-06,
-      "loss": 0.3761,
-      "step": 8404
-    },
-    {
-      "epoch": 0.5493105025815306,
-      "grad_norm": 0.48850181698799133,
-      "learning_rate": 9.332104405744666e-06,
-      "loss": 0.4474,
-      "step": 8405
-    },
-    {
-      "epoch": 0.5493758577870728,
-      "grad_norm": 0.4466870129108429,
-      "learning_rate": 9.33193003958613e-06,
-      "loss": 0.3646,
-      "step": 8406
-    },
-    {
-      "epoch": 0.5494412129926148,
-      "grad_norm": 0.49112334847450256,
-      "learning_rate": 9.331755652299271e-06,
-      "loss": 0.4152,
-      "step": 8407
-    },
-    {
-      "epoch": 0.549506568198157,
-      "grad_norm": 0.4598344564437866,
-      "learning_rate": 9.33158124388494e-06,
-      "loss": 0.3989,
-      "step": 8408
-    },
-    {
-      "epoch": 0.5495719234036991,
-      "grad_norm": 0.42260119318962097,
-      "learning_rate": 9.331406814343986e-06,
-      "loss": 0.3463,
-      "step": 8409
-    },
-    {
-      "epoch": 0.5496372786092413,
-      "grad_norm": 0.48907238245010376,
-      "learning_rate": 9.331232363677259e-06,
-      "loss": 0.427,
-      "step": 8410
-    },
-    {
-      "epoch": 0.5497026338147833,
-      "grad_norm": 0.4614992141723633,
-      "learning_rate": 9.331057891885614e-06,
-      "loss": 0.373,
-      "step": 8411
-    },
-    {
-      "epoch": 0.5497679890203254,
-      "grad_norm": 0.44425952434539795,
-      "learning_rate": 9.330883398969897e-06,
-      "loss": 0.3683,
-      "step": 8412
-    },
-    {
-      "epoch": 0.5498333442258676,
-      "grad_norm": 0.48977503180503845,
-      "learning_rate": 9.330708884930962e-06,
-      "loss": 0.4347,
-      "step": 8413
-    },
-    {
-      "epoch": 0.5498986994314097,
-      "grad_norm": 0.46309882402420044,
-      "learning_rate": 9.330534349769658e-06,
-      "loss": 0.3942,
-      "step": 8414
-    },
-    {
-      "epoch": 0.5499640546369519,
-      "grad_norm": 0.4486205577850342,
-      "learning_rate": 9.330359793486839e-06,
-      "loss": 0.3779,
-      "step": 8415
-    },
-    {
-      "epoch": 0.5500294098424939,
-      "grad_norm": 0.4534038007259369,
-      "learning_rate": 9.330185216083356e-06,
-      "loss": 0.3812,
-      "step": 8416
-    },
-    {
-      "epoch": 0.5500947650480361,
-      "grad_norm": 0.4830993413925171,
-      "learning_rate": 9.33001061756006e-06,
-      "loss": 0.4427,
-      "step": 8417
-    },
-    {
-      "epoch": 0.5501601202535782,
-      "grad_norm": 0.49837592244148254,
-      "learning_rate": 9.3298359979178e-06,
-      "loss": 0.4591,
-      "step": 8418
-    },
-    {
-      "epoch": 0.5502254754591204,
-      "grad_norm": 0.43504467606544495,
-      "learning_rate": 9.32966135715743e-06,
-      "loss": 0.3704,
-      "step": 8419
-    },
-    {
-      "epoch": 0.5502908306646624,
-      "grad_norm": 0.45898109674453735,
-      "learning_rate": 9.329486695279803e-06,
-      "loss": 0.378,
-      "step": 8420
-    },
-    {
-      "epoch": 0.5503561858702045,
-      "grad_norm": 0.42473548650741577,
-      "learning_rate": 9.32931201228577e-06,
-      "loss": 0.3357,
-      "step": 8421
-    },
-    {
-      "epoch": 0.5504215410757467,
-      "grad_norm": 0.43120864033699036,
-      "learning_rate": 9.32913730817618e-06,
-      "loss": 0.3832,
-      "step": 8422
-    },
-    {
-      "epoch": 0.5504868962812888,
-      "grad_norm": 0.4985445439815521,
-      "learning_rate": 9.328962582951889e-06,
-      "loss": 0.4308,
-      "step": 8423
-    },
-    {
-      "epoch": 0.550552251486831,
-      "grad_norm": 0.45426177978515625,
-      "learning_rate": 9.328787836613748e-06,
-      "loss": 0.4016,
-      "step": 8424
-    },
-    {
-      "epoch": 0.550617606692373,
-      "grad_norm": 0.440967321395874,
-      "learning_rate": 9.328613069162608e-06,
-      "loss": 0.3838,
-      "step": 8425
-    },
-    {
-      "epoch": 0.5506829618979152,
-      "grad_norm": 0.45935264229774475,
-      "learning_rate": 9.328438280599326e-06,
-      "loss": 0.4127,
-      "step": 8426
-    },
-    {
-      "epoch": 0.5507483171034573,
-      "grad_norm": 0.43026798963546753,
-      "learning_rate": 9.328263470924747e-06,
-      "loss": 0.3232,
-      "step": 8427
-    },
-    {
-      "epoch": 0.5508136723089994,
-      "grad_norm": 0.463697224855423,
-      "learning_rate": 9.328088640139729e-06,
-      "loss": 0.3898,
-      "step": 8428
-    },
-    {
-      "epoch": 0.5508790275145415,
-      "grad_norm": 0.46323880553245544,
-      "learning_rate": 9.327913788245125e-06,
-      "loss": 0.4116,
-      "step": 8429
-    },
-    {
-      "epoch": 0.5509443827200836,
-      "grad_norm": 0.4392174780368805,
-      "learning_rate": 9.327738915241782e-06,
-      "loss": 0.3827,
-      "step": 8430
-    },
-    {
-      "epoch": 0.5510097379256258,
-      "grad_norm": 0.44167813658714294,
-      "learning_rate": 9.32756402113056e-06,
-      "loss": 0.3553,
-      "step": 8431
-    },
-    {
-      "epoch": 0.5510750931311679,
-      "grad_norm": 0.4294022023677826,
-      "learning_rate": 9.327389105912308e-06,
-      "loss": 0.3603,
-      "step": 8432
-    },
-    {
-      "epoch": 0.55114044833671,
-      "grad_norm": 0.4425641596317291,
-      "learning_rate": 9.327214169587881e-06,
-      "loss": 0.3888,
-      "step": 8433
-    },
-    {
-      "epoch": 0.5512058035422521,
-      "grad_norm": 0.43455198407173157,
-      "learning_rate": 9.32703921215813e-06,
-      "loss": 0.3916,
-      "step": 8434
-    },
-    {
-      "epoch": 0.5512711587477943,
-      "grad_norm": 0.4327811598777771,
-      "learning_rate": 9.326864233623912e-06,
-      "loss": 0.3907,
-      "step": 8435
-    },
-    {
-      "epoch": 0.5513365139533364,
-      "grad_norm": 0.427016019821167,
-      "learning_rate": 9.326689233986077e-06,
-      "loss": 0.3403,
-      "step": 8436
-    },
-    {
-      "epoch": 0.5514018691588785,
-      "grad_norm": 0.48279014229774475,
-      "learning_rate": 9.32651421324548e-06,
-      "loss": 0.3987,
-      "step": 8437
-    },
-    {
-      "epoch": 0.5514672243644206,
-      "grad_norm": 0.46045464277267456,
-      "learning_rate": 9.326339171402974e-06,
-      "loss": 0.4103,
-      "step": 8438
-    },
-    {
-      "epoch": 0.5515325795699627,
-      "grad_norm": 0.5111485123634338,
-      "learning_rate": 9.326164108459413e-06,
-      "loss": 0.4348,
-      "step": 8439
-    },
-    {
-      "epoch": 0.5515979347755049,
-      "grad_norm": 0.4589383602142334,
-      "learning_rate": 9.325989024415652e-06,
-      "loss": 0.402,
-      "step": 8440
-    },
-    {
-      "epoch": 0.551663289981047,
-      "grad_norm": 0.4229415953159332,
-      "learning_rate": 9.325813919272542e-06,
-      "loss": 0.3224,
-      "step": 8441
-    },
-    {
-      "epoch": 0.5517286451865892,
-      "grad_norm": 0.4934961497783661,
-      "learning_rate": 9.32563879303094e-06,
-      "loss": 0.4481,
-      "step": 8442
-    },
-    {
-      "epoch": 0.5517940003921312,
-      "grad_norm": 0.45548489689826965,
-      "learning_rate": 9.325463645691699e-06,
-      "loss": 0.3794,
-      "step": 8443
-    },
-    {
-      "epoch": 0.5518593555976734,
-      "grad_norm": 0.42007023096084595,
-      "learning_rate": 9.325288477255673e-06,
-      "loss": 0.34,
-      "step": 8444
-    },
-    {
-      "epoch": 0.5519247108032155,
-      "grad_norm": 0.4694366753101349,
-      "learning_rate": 9.325113287723718e-06,
-      "loss": 0.3862,
-      "step": 8445
-    },
-    {
-      "epoch": 0.5519900660087576,
-      "grad_norm": 0.4611228108406067,
-      "learning_rate": 9.324938077096685e-06,
-      "loss": 0.414,
-      "step": 8446
-    },
-    {
-      "epoch": 0.5520554212142997,
-      "grad_norm": 0.48163169622421265,
-      "learning_rate": 9.324762845375433e-06,
-      "loss": 0.3617,
-      "step": 8447
-    },
-    {
-      "epoch": 0.5521207764198418,
-      "grad_norm": 0.42597824335098267,
-      "learning_rate": 9.324587592560812e-06,
-      "loss": 0.3365,
-      "step": 8448
-    },
-    {
-      "epoch": 0.552186131625384,
-      "grad_norm": 0.4741002023220062,
-      "learning_rate": 9.324412318653679e-06,
-      "loss": 0.4429,
-      "step": 8449
-    },
-    {
-      "epoch": 0.5522514868309261,
-      "grad_norm": 0.48227596282958984,
-      "learning_rate": 9.324237023654892e-06,
-      "loss": 0.4309,
-      "step": 8450
-    },
-    {
-      "epoch": 0.5523168420364682,
-      "grad_norm": 0.49901261925697327,
-      "learning_rate": 9.324061707565299e-06,
-      "loss": 0.3886,
-      "step": 8451
-    },
-    {
-      "epoch": 0.5523821972420103,
-      "grad_norm": 0.4953928589820862,
-      "learning_rate": 9.323886370385762e-06,
-      "loss": 0.4208,
-      "step": 8452
-    },
-    {
-      "epoch": 0.5524475524475524,
-      "grad_norm": 0.45888665318489075,
-      "learning_rate": 9.323711012117134e-06,
-      "loss": 0.3819,
-      "step": 8453
-    },
-    {
-      "epoch": 0.5525129076530946,
-      "grad_norm": 0.4151547849178314,
-      "learning_rate": 9.323535632760267e-06,
-      "loss": 0.3553,
-      "step": 8454
-    },
-    {
-      "epoch": 0.5525782628586366,
-      "grad_norm": 0.4288499057292938,
-      "learning_rate": 9.32336023231602e-06,
-      "loss": 0.3603,
-      "step": 8455
-    },
-    {
-      "epoch": 0.5526436180641788,
-      "grad_norm": 0.4717220962047577,
-      "learning_rate": 9.323184810785247e-06,
-      "loss": 0.4287,
-      "step": 8456
-    },
-    {
-      "epoch": 0.5527089732697209,
-      "grad_norm": 0.47539812326431274,
-      "learning_rate": 9.323009368168805e-06,
-      "loss": 0.4089,
-      "step": 8457
-    },
-    {
-      "epoch": 0.5527743284752631,
-      "grad_norm": 0.4591185748577118,
-      "learning_rate": 9.322833904467548e-06,
-      "loss": 0.3922,
-      "step": 8458
-    },
-    {
-      "epoch": 0.5528396836808052,
-      "grad_norm": 0.45324063301086426,
-      "learning_rate": 9.322658419682331e-06,
-      "loss": 0.4169,
-      "step": 8459
-    },
-    {
-      "epoch": 0.5529050388863473,
-      "grad_norm": 0.4990595281124115,
-      "learning_rate": 9.322482913814012e-06,
-      "loss": 0.4469,
-      "step": 8460
-    },
-    {
-      "epoch": 0.5529703940918894,
-      "grad_norm": 0.4342723786830902,
-      "learning_rate": 9.322307386863448e-06,
-      "loss": 0.3744,
-      "step": 8461
-    },
-    {
-      "epoch": 0.5530357492974315,
-      "grad_norm": 0.44127944111824036,
-      "learning_rate": 9.322131838831493e-06,
-      "loss": 0.3821,
-      "step": 8462
-    },
-    {
-      "epoch": 0.5531011045029737,
-      "grad_norm": 0.479885995388031,
-      "learning_rate": 9.321956269719003e-06,
-      "loss": 0.3973,
-      "step": 8463
-    },
-    {
-      "epoch": 0.5531664597085157,
-      "grad_norm": 0.4902941584587097,
-      "learning_rate": 9.321780679526835e-06,
-      "loss": 0.4142,
-      "step": 8464
-    },
-    {
-      "epoch": 0.5532318149140579,
-      "grad_norm": 0.5010184645652771,
-      "learning_rate": 9.321605068255845e-06,
-      "loss": 0.4483,
-      "step": 8465
-    },
-    {
-      "epoch": 0.5532971701196,
-      "grad_norm": 0.4486129879951477,
-      "learning_rate": 9.321429435906893e-06,
-      "loss": 0.3705,
-      "step": 8466
-    },
-    {
-      "epoch": 0.5533625253251422,
-      "grad_norm": 0.4655238389968872,
-      "learning_rate": 9.321253782480829e-06,
-      "loss": 0.4214,
-      "step": 8467
-    },
-    {
-      "epoch": 0.5534278805306843,
-      "grad_norm": 0.5725885033607483,
-      "learning_rate": 9.321078107978514e-06,
-      "loss": 0.3714,
-      "step": 8468
-    },
-    {
-      "epoch": 0.5534932357362264,
-      "grad_norm": 0.4592706263065338,
-      "learning_rate": 9.320902412400806e-06,
-      "loss": 0.417,
-      "step": 8469
-    },
-    {
-      "epoch": 0.5535585909417685,
-      "grad_norm": 0.4308174252510071,
-      "learning_rate": 9.320726695748558e-06,
-      "loss": 0.3735,
-      "step": 8470
-    },
-    {
-      "epoch": 0.5536239461473106,
-      "grad_norm": 0.45350342988967896,
-      "learning_rate": 9.320550958022629e-06,
-      "loss": 0.4151,
-      "step": 8471
-    },
-    {
-      "epoch": 0.5536893013528528,
-      "grad_norm": 0.43352949619293213,
-      "learning_rate": 9.320375199223877e-06,
-      "loss": 0.3383,
-      "step": 8472
-    },
-    {
-      "epoch": 0.5537546565583948,
-      "grad_norm": 0.48943135142326355,
-      "learning_rate": 9.32019941935316e-06,
-      "loss": 0.4346,
-      "step": 8473
-    },
-    {
-      "epoch": 0.553820011763937,
-      "grad_norm": 0.42777833342552185,
-      "learning_rate": 9.320023618411332e-06,
-      "loss": 0.3468,
-      "step": 8474
-    },
-    {
-      "epoch": 0.5538853669694791,
-      "grad_norm": 0.44771796464920044,
-      "learning_rate": 9.319847796399251e-06,
-      "loss": 0.3561,
-      "step": 8475
-    },
-    {
-      "epoch": 0.5539507221750213,
-      "grad_norm": 0.5173906087875366,
-      "learning_rate": 9.319671953317776e-06,
-      "loss": 0.4371,
-      "step": 8476
-    },
-    {
-      "epoch": 0.5540160773805634,
-      "grad_norm": 0.5046651363372803,
-      "learning_rate": 9.319496089167767e-06,
-      "loss": 0.3692,
-      "step": 8477
-    },
-    {
-      "epoch": 0.5540814325861055,
-      "grad_norm": 0.4478219151496887,
-      "learning_rate": 9.319320203950077e-06,
-      "loss": 0.3511,
-      "step": 8478
-    },
-    {
-      "epoch": 0.5541467877916476,
-      "grad_norm": 0.4522320032119751,
-      "learning_rate": 9.319144297665566e-06,
-      "loss": 0.4043,
-      "step": 8479
-    },
-    {
-      "epoch": 0.5542121429971897,
-      "grad_norm": 0.4755884110927582,
-      "learning_rate": 9.318968370315094e-06,
-      "loss": 0.397,
-      "step": 8480
-    },
-    {
-      "epoch": 0.5542774982027319,
-      "grad_norm": 0.4711938500404358,
-      "learning_rate": 9.318792421899516e-06,
-      "loss": 0.4588,
-      "step": 8481
-    },
-    {
-      "epoch": 0.5543428534082739,
-      "grad_norm": 0.4489021301269531,
-      "learning_rate": 9.31861645241969e-06,
-      "loss": 0.3607,
-      "step": 8482
-    },
-    {
-      "epoch": 0.5544082086138161,
-      "grad_norm": 0.49408623576164246,
-      "learning_rate": 9.318440461876476e-06,
-      "loss": 0.4324,
-      "step": 8483
-    },
-    {
-      "epoch": 0.5544735638193582,
-      "grad_norm": 0.5169716477394104,
-      "learning_rate": 9.318264450270733e-06,
-      "loss": 0.4674,
-      "step": 8484
-    },
-    {
-      "epoch": 0.5545389190249004,
-      "grad_norm": 0.4246198534965515,
-      "learning_rate": 9.318088417603317e-06,
-      "loss": 0.3611,
-      "step": 8485
-    },
-    {
-      "epoch": 0.5546042742304425,
-      "grad_norm": 0.4757101535797119,
-      "learning_rate": 9.317912363875089e-06,
-      "loss": 0.4551,
-      "step": 8486
-    },
-    {
-      "epoch": 0.5546696294359845,
-      "grad_norm": 0.4473075866699219,
-      "learning_rate": 9.317736289086904e-06,
-      "loss": 0.4081,
-      "step": 8487
-    },
-    {
-      "epoch": 0.5547349846415267,
-      "grad_norm": 0.45982101559638977,
-      "learning_rate": 9.317560193239626e-06,
-      "loss": 0.379,
-      "step": 8488
-    },
-    {
-      "epoch": 0.5548003398470688,
-      "grad_norm": 0.45575153827667236,
-      "learning_rate": 9.317384076334109e-06,
-      "loss": 0.3517,
-      "step": 8489
-    },
-    {
-      "epoch": 0.554865695052611,
-      "grad_norm": 0.46721091866493225,
-      "learning_rate": 9.317207938371216e-06,
-      "loss": 0.4049,
-      "step": 8490
-    },
-    {
-      "epoch": 0.554931050258153,
-      "grad_norm": 0.46285346150398254,
-      "learning_rate": 9.317031779351803e-06,
-      "loss": 0.3927,
-      "step": 8491
-    },
-    {
-      "epoch": 0.5549964054636952,
-      "grad_norm": 0.4649997055530548,
-      "learning_rate": 9.316855599276733e-06,
-      "loss": 0.4246,
-      "step": 8492
-    },
-    {
-      "epoch": 0.5550617606692373,
-      "grad_norm": 0.44084203243255615,
-      "learning_rate": 9.316679398146859e-06,
-      "loss": 0.351,
-      "step": 8493
-    },
-    {
-      "epoch": 0.5551271158747795,
-      "grad_norm": 0.477680504322052,
-      "learning_rate": 9.316503175963048e-06,
-      "loss": 0.3938,
-      "step": 8494
-    },
-    {
-      "epoch": 0.5551924710803215,
-      "grad_norm": 0.4778280258178711,
-      "learning_rate": 9.316326932726151e-06,
-      "loss": 0.422,
-      "step": 8495
-    },
-    {
-      "epoch": 0.5552578262858636,
-      "grad_norm": 0.4426637291908264,
-      "learning_rate": 9.316150668437037e-06,
-      "loss": 0.3715,
-      "step": 8496
-    },
-    {
-      "epoch": 0.5553231814914058,
-      "grad_norm": 0.42743223905563354,
-      "learning_rate": 9.315974383096557e-06,
-      "loss": 0.3376,
-      "step": 8497
-    },
-    {
-      "epoch": 0.5553885366969479,
-      "grad_norm": 0.45268023014068604,
-      "learning_rate": 9.315798076705578e-06,
-      "loss": 0.3829,
-      "step": 8498
-    },
-    {
-      "epoch": 0.5554538919024901,
-      "grad_norm": 0.4755485951900482,
-      "learning_rate": 9.315621749264956e-06,
-      "loss": 0.4254,
-      "step": 8499
-    },
-    {
-      "epoch": 0.5555192471080321,
-      "grad_norm": 0.4836467206478119,
-      "learning_rate": 9.315445400775549e-06,
-      "loss": 0.3897,
-      "step": 8500
-    },
-    {
-      "epoch": 0.5555846023135743,
-      "grad_norm": 0.4236466586589813,
-      "learning_rate": 9.315269031238222e-06,
-      "loss": 0.3356,
-      "step": 8501
-    },
-    {
-      "epoch": 0.5556499575191164,
-      "grad_norm": 0.4499092996120453,
-      "learning_rate": 9.315092640653834e-06,
-      "loss": 0.401,
-      "step": 8502
-    },
-    {
-      "epoch": 0.5557153127246586,
-      "grad_norm": 0.4628172516822815,
-      "learning_rate": 9.314916229023242e-06,
-      "loss": 0.382,
-      "step": 8503
-    },
-    {
-      "epoch": 0.5557806679302006,
-      "grad_norm": 0.47890061140060425,
-      "learning_rate": 9.31473979634731e-06,
-      "loss": 0.4077,
-      "step": 8504
-    },
-    {
-      "epoch": 0.5558460231357427,
-      "grad_norm": 0.479810506105423,
-      "learning_rate": 9.314563342626897e-06,
-      "loss": 0.4386,
-      "step": 8505
-    },
-    {
-      "epoch": 0.5559113783412849,
-      "grad_norm": 0.4570308327674866,
-      "learning_rate": 9.314386867862863e-06,
-      "loss": 0.3808,
-      "step": 8506
-    },
-    {
-      "epoch": 0.555976733546827,
-      "grad_norm": 0.4701350927352905,
-      "learning_rate": 9.31421037205607e-06,
-      "loss": 0.3972,
-      "step": 8507
-    },
-    {
-      "epoch": 0.5560420887523692,
-      "grad_norm": 0.44427141547203064,
-      "learning_rate": 9.314033855207379e-06,
-      "loss": 0.3435,
-      "step": 8508
-    },
-    {
-      "epoch": 0.5561074439579112,
-      "grad_norm": 0.4655913710594177,
-      "learning_rate": 9.31385731731765e-06,
-      "loss": 0.3739,
-      "step": 8509
-    },
-    {
-      "epoch": 0.5561727991634534,
-      "grad_norm": 0.45355671644210815,
-      "learning_rate": 9.313680758387745e-06,
-      "loss": 0.3977,
-      "step": 8510
-    },
-    {
-      "epoch": 0.5562381543689955,
-      "grad_norm": 0.43664395809173584,
-      "learning_rate": 9.313504178418524e-06,
-      "loss": 0.3566,
-      "step": 8511
-    },
-    {
-      "epoch": 0.5563035095745376,
-      "grad_norm": 0.42067351937294006,
-      "learning_rate": 9.313327577410849e-06,
-      "loss": 0.3508,
-      "step": 8512
-    },
-    {
-      "epoch": 0.5563688647800797,
-      "grad_norm": 0.48400741815567017,
-      "learning_rate": 9.31315095536558e-06,
-      "loss": 0.4489,
-      "step": 8513
-    },
-    {
-      "epoch": 0.5564342199856218,
-      "grad_norm": 0.4556126296520233,
-      "learning_rate": 9.31297431228358e-06,
-      "loss": 0.4071,
-      "step": 8514
-    },
-    {
-      "epoch": 0.556499575191164,
-      "grad_norm": 0.4724879264831543,
-      "learning_rate": 9.312797648165712e-06,
-      "loss": 0.3754,
-      "step": 8515
-    },
-    {
-      "epoch": 0.5565649303967061,
-      "grad_norm": 0.43408671021461487,
-      "learning_rate": 9.312620963012833e-06,
-      "loss": 0.3568,
-      "step": 8516
-    },
-    {
-      "epoch": 0.5566302856022483,
-      "grad_norm": 0.4010029435157776,
-      "learning_rate": 9.31244425682581e-06,
-      "loss": 0.3112,
-      "step": 8517
-    },
-    {
-      "epoch": 0.5566956408077903,
-      "grad_norm": 0.4599505662918091,
-      "learning_rate": 9.3122675296055e-06,
-      "loss": 0.356,
-      "step": 8518
-    },
-    {
-      "epoch": 0.5567609960133325,
-      "grad_norm": 0.41994336247444153,
-      "learning_rate": 9.31209078135277e-06,
-      "loss": 0.3447,
-      "step": 8519
-    },
-    {
-      "epoch": 0.5568263512188746,
-      "grad_norm": 0.46921485662460327,
-      "learning_rate": 9.311914012068478e-06,
-      "loss": 0.3964,
-      "step": 8520
-    },
-    {
-      "epoch": 0.5568917064244167,
-      "grad_norm": 0.4661003351211548,
-      "learning_rate": 9.311737221753487e-06,
-      "loss": 0.3939,
-      "step": 8521
-    },
-    {
-      "epoch": 0.5569570616299588,
-      "grad_norm": 0.4298548698425293,
-      "learning_rate": 9.31156041040866e-06,
-      "loss": 0.3406,
-      "step": 8522
-    },
-    {
-      "epoch": 0.5570224168355009,
-      "grad_norm": 0.47248151898384094,
-      "learning_rate": 9.31138357803486e-06,
-      "loss": 0.3952,
-      "step": 8523
-    },
-    {
-      "epoch": 0.5570877720410431,
-      "grad_norm": 0.4456464350223541,
-      "learning_rate": 9.311206724632949e-06,
-      "loss": 0.3926,
-      "step": 8524
-    },
-    {
-      "epoch": 0.5571531272465852,
-      "grad_norm": 0.4399384558200836,
-      "learning_rate": 9.311029850203788e-06,
-      "loss": 0.3757,
-      "step": 8525
-    },
-    {
-      "epoch": 0.5572184824521274,
-      "grad_norm": 0.4460543394088745,
-      "learning_rate": 9.310852954748241e-06,
-      "loss": 0.3574,
-      "step": 8526
-    },
-    {
-      "epoch": 0.5572838376576694,
-      "grad_norm": 0.4757439196109772,
-      "learning_rate": 9.310676038267173e-06,
-      "loss": 0.4155,
-      "step": 8527
-    },
-    {
-      "epoch": 0.5573491928632116,
-      "grad_norm": 0.45683062076568604,
-      "learning_rate": 9.310499100761443e-06,
-      "loss": 0.3955,
-      "step": 8528
-    },
-    {
-      "epoch": 0.5574145480687537,
-      "grad_norm": 0.43982434272766113,
-      "learning_rate": 9.310322142231916e-06,
-      "loss": 0.3825,
-      "step": 8529
-    },
-    {
-      "epoch": 0.5574799032742958,
-      "grad_norm": 0.4997251629829407,
-      "learning_rate": 9.310145162679454e-06,
-      "loss": 0.455,
-      "step": 8530
-    },
-    {
-      "epoch": 0.5575452584798379,
-      "grad_norm": 0.4151521325111389,
-      "learning_rate": 9.309968162104921e-06,
-      "loss": 0.3374,
-      "step": 8531
-    },
-    {
-      "epoch": 0.55761061368538,
-      "grad_norm": 0.4819811284542084,
-      "learning_rate": 9.309791140509178e-06,
-      "loss": 0.4262,
-      "step": 8532
-    },
-    {
-      "epoch": 0.5576759688909222,
-      "grad_norm": 0.4586372375488281,
-      "learning_rate": 9.309614097893093e-06,
-      "loss": 0.4032,
-      "step": 8533
-    },
-    {
-      "epoch": 0.5577413240964643,
-      "grad_norm": 0.43472665548324585,
-      "learning_rate": 9.309437034257526e-06,
-      "loss": 0.3699,
-      "step": 8534
-    },
-    {
-      "epoch": 0.5578066793020064,
-      "grad_norm": 0.45355939865112305,
-      "learning_rate": 9.309259949603344e-06,
-      "loss": 0.4229,
-      "step": 8535
-    },
-    {
-      "epoch": 0.5578720345075485,
-      "grad_norm": 0.48780471086502075,
-      "learning_rate": 9.309082843931407e-06,
-      "loss": 0.4228,
-      "step": 8536
-    },
-    {
-      "epoch": 0.5579373897130906,
-      "grad_norm": 0.49163946509361267,
-      "learning_rate": 9.30890571724258e-06,
-      "loss": 0.4377,
-      "step": 8537
-    },
-    {
-      "epoch": 0.5580027449186328,
-      "grad_norm": 0.4605172276496887,
-      "learning_rate": 9.308728569537728e-06,
-      "loss": 0.4274,
-      "step": 8538
-    },
-    {
-      "epoch": 0.5580681001241748,
-      "grad_norm": 0.45451197028160095,
-      "learning_rate": 9.308551400817712e-06,
-      "loss": 0.4312,
-      "step": 8539
-    },
-    {
-      "epoch": 0.558133455329717,
-      "grad_norm": 0.44905582070350647,
-      "learning_rate": 9.3083742110834e-06,
-      "loss": 0.376,
-      "step": 8540
-    },
-    {
-      "epoch": 0.5581988105352591,
-      "grad_norm": 0.43554186820983887,
-      "learning_rate": 9.308197000335652e-06,
-      "loss": 0.3728,
-      "step": 8541
-    },
-    {
-      "epoch": 0.5582641657408013,
-      "grad_norm": 0.4635400176048279,
-      "learning_rate": 9.308019768575338e-06,
-      "loss": 0.4027,
-      "step": 8542
-    },
-    {
-      "epoch": 0.5583295209463434,
-      "grad_norm": 0.4505763649940491,
-      "learning_rate": 9.307842515803318e-06,
-      "loss": 0.3809,
-      "step": 8543
-    },
-    {
-      "epoch": 0.5583948761518855,
-      "grad_norm": 0.46907296776771545,
-      "learning_rate": 9.307665242020456e-06,
-      "loss": 0.4243,
-      "step": 8544
-    },
-    {
-      "epoch": 0.5584602313574276,
-      "grad_norm": 0.4555659592151642,
-      "learning_rate": 9.30748794722762e-06,
-      "loss": 0.3883,
-      "step": 8545
-    },
-    {
-      "epoch": 0.5585255865629697,
-      "grad_norm": 0.4730190932750702,
-      "learning_rate": 9.307310631425673e-06,
-      "loss": 0.4317,
-      "step": 8546
-    },
-    {
-      "epoch": 0.5585909417685119,
-      "grad_norm": 0.42633357644081116,
-      "learning_rate": 9.307133294615479e-06,
-      "loss": 0.3149,
-      "step": 8547
-    },
-    {
-      "epoch": 0.558656296974054,
-      "grad_norm": 0.4453902542591095,
-      "learning_rate": 9.306955936797904e-06,
-      "loss": 0.387,
-      "step": 8548
-    },
-    {
-      "epoch": 0.5587216521795961,
-      "grad_norm": 0.4490743577480316,
-      "learning_rate": 9.306778557973813e-06,
-      "loss": 0.3792,
-      "step": 8549
-    },
-    {
-      "epoch": 0.5587870073851382,
-      "grad_norm": 0.40580257773399353,
-      "learning_rate": 9.306601158144071e-06,
-      "loss": 0.2995,
-      "step": 8550
-    },
-    {
-      "epoch": 0.5588523625906804,
-      "grad_norm": 0.4768867492675781,
-      "learning_rate": 9.306423737309544e-06,
-      "loss": 0.4078,
-      "step": 8551
-    },
-    {
-      "epoch": 0.5589177177962225,
-      "grad_norm": 0.5181326866149902,
-      "learning_rate": 9.306246295471096e-06,
-      "loss": 0.4095,
-      "step": 8552
-    },
-    {
-      "epoch": 0.5589830730017646,
-      "grad_norm": 0.4448120594024658,
-      "learning_rate": 9.30606883262959e-06,
-      "loss": 0.3416,
-      "step": 8553
-    },
-    {
-      "epoch": 0.5590484282073067,
-      "grad_norm": 0.4278298020362854,
-      "learning_rate": 9.305891348785898e-06,
-      "loss": 0.3481,
-      "step": 8554
-    },
-    {
-      "epoch": 0.5591137834128488,
-      "grad_norm": 0.48393508791923523,
-      "learning_rate": 9.305713843940883e-06,
-      "loss": 0.4162,
-      "step": 8555
-    },
-    {
-      "epoch": 0.559179138618391,
-      "grad_norm": 0.48160406947135925,
-      "learning_rate": 9.305536318095408e-06,
-      "loss": 0.3878,
-      "step": 8556
-    },
-    {
-      "epoch": 0.559244493823933,
-      "grad_norm": 0.510643482208252,
-      "learning_rate": 9.305358771250342e-06,
-      "loss": 0.4474,
-      "step": 8557
-    },
-    {
-      "epoch": 0.5593098490294752,
-      "grad_norm": 0.4668619930744171,
-      "learning_rate": 9.305181203406548e-06,
-      "loss": 0.3687,
-      "step": 8558
-    },
-    {
-      "epoch": 0.5593752042350173,
-      "grad_norm": 0.44012463092803955,
-      "learning_rate": 9.305003614564895e-06,
-      "loss": 0.3548,
-      "step": 8559
-    },
-    {
-      "epoch": 0.5594405594405595,
-      "grad_norm": 0.4808661937713623,
-      "learning_rate": 9.304826004726248e-06,
-      "loss": 0.4233,
-      "step": 8560
-    },
-    {
-      "epoch": 0.5595059146461016,
-      "grad_norm": 0.4720933139324188,
-      "learning_rate": 9.304648373891472e-06,
-      "loss": 0.3555,
-      "step": 8561
-    },
-    {
-      "epoch": 0.5595712698516437,
-      "grad_norm": 0.44952312111854553,
-      "learning_rate": 9.304470722061437e-06,
-      "loss": 0.3903,
-      "step": 8562
-    },
-    {
-      "epoch": 0.5596366250571858,
-      "grad_norm": 0.4470025300979614,
-      "learning_rate": 9.304293049237005e-06,
-      "loss": 0.368,
-      "step": 8563
-    },
-    {
-      "epoch": 0.5597019802627279,
-      "grad_norm": 0.46034061908721924,
-      "learning_rate": 9.304115355419045e-06,
-      "loss": 0.3909,
-      "step": 8564
-    },
-    {
-      "epoch": 0.5597673354682701,
-      "grad_norm": 0.48923662304878235,
-      "learning_rate": 9.303937640608423e-06,
-      "loss": 0.4617,
-      "step": 8565
-    },
-    {
-      "epoch": 0.5598326906738121,
-      "grad_norm": 0.4368457794189453,
-      "learning_rate": 9.303759904806007e-06,
-      "loss": 0.3402,
-      "step": 8566
-    },
-    {
-      "epoch": 0.5598980458793543,
-      "grad_norm": 0.5113939046859741,
-      "learning_rate": 9.303582148012663e-06,
-      "loss": 0.4111,
-      "step": 8567
-    },
-    {
-      "epoch": 0.5599634010848964,
-      "grad_norm": 0.43663060665130615,
-      "learning_rate": 9.303404370229257e-06,
-      "loss": 0.3611,
-      "step": 8568
-    },
-    {
-      "epoch": 0.5600287562904386,
-      "grad_norm": 0.4688543379306793,
-      "learning_rate": 9.303226571456658e-06,
-      "loss": 0.4179,
-      "step": 8569
-    },
-    {
-      "epoch": 0.5600941114959807,
-      "grad_norm": 0.47538647055625916,
-      "learning_rate": 9.303048751695732e-06,
-      "loss": 0.3674,
-      "step": 8570
-    },
-    {
-      "epoch": 0.5601594667015227,
-      "grad_norm": 0.48249363899230957,
-      "learning_rate": 9.302870910947346e-06,
-      "loss": 0.4418,
-      "step": 8571
-    },
-    {
-      "epoch": 0.5602248219070649,
-      "grad_norm": 0.4834454655647278,
-      "learning_rate": 9.30269304921237e-06,
-      "loss": 0.4054,
-      "step": 8572
-    },
-    {
-      "epoch": 0.560290177112607,
-      "grad_norm": 0.4889773726463318,
-      "learning_rate": 9.302515166491667e-06,
-      "loss": 0.4349,
-      "step": 8573
-    },
-    {
-      "epoch": 0.5603555323181492,
-      "grad_norm": 0.6230728030204773,
-      "learning_rate": 9.302337262786107e-06,
-      "loss": 0.4034,
-      "step": 8574
-    },
-    {
-      "epoch": 0.5604208875236912,
-      "grad_norm": 0.43193289637565613,
-      "learning_rate": 9.302159338096559e-06,
-      "loss": 0.3391,
-      "step": 8575
-    },
-    {
-      "epoch": 0.5604862427292334,
-      "grad_norm": 0.45328015089035034,
-      "learning_rate": 9.30198139242389e-06,
-      "loss": 0.3668,
-      "step": 8576
-    },
-    {
-      "epoch": 0.5605515979347755,
-      "grad_norm": 0.46770498156547546,
-      "learning_rate": 9.301803425768964e-06,
-      "loss": 0.3915,
-      "step": 8577
-    },
-    {
-      "epoch": 0.5606169531403177,
-      "grad_norm": 0.4672142267227173,
-      "learning_rate": 9.301625438132655e-06,
-      "loss": 0.3989,
-      "step": 8578
-    },
-    {
-      "epoch": 0.5606823083458597,
-      "grad_norm": 0.427051842212677,
-      "learning_rate": 9.301447429515829e-06,
-      "loss": 0.3246,
-      "step": 8579
-    },
-    {
-      "epoch": 0.5607476635514018,
-      "grad_norm": 0.41258934140205383,
-      "learning_rate": 9.301269399919352e-06,
-      "loss": 0.3339,
-      "step": 8580
-    },
-    {
-      "epoch": 0.560813018756944,
-      "grad_norm": 0.45607173442840576,
-      "learning_rate": 9.301091349344096e-06,
-      "loss": 0.372,
-      "step": 8581
-    },
-    {
-      "epoch": 0.5608783739624861,
-      "grad_norm": 0.4394467771053314,
-      "learning_rate": 9.300913277790926e-06,
-      "loss": 0.3935,
-      "step": 8582
-    },
-    {
-      "epoch": 0.5609437291680283,
-      "grad_norm": 0.42783913016319275,
-      "learning_rate": 9.300735185260713e-06,
-      "loss": 0.327,
-      "step": 8583
-    },
-    {
-      "epoch": 0.5610090843735703,
-      "grad_norm": 0.46981081366539,
-      "learning_rate": 9.300557071754324e-06,
-      "loss": 0.4184,
-      "step": 8584
-    },
-    {
-      "epoch": 0.5610744395791125,
-      "grad_norm": 0.4887408912181854,
-      "learning_rate": 9.300378937272629e-06,
-      "loss": 0.4318,
-      "step": 8585
-    },
-    {
-      "epoch": 0.5611397947846546,
-      "grad_norm": 0.4351556897163391,
-      "learning_rate": 9.300200781816495e-06,
-      "loss": 0.3603,
-      "step": 8586
-    },
-    {
-      "epoch": 0.5612051499901968,
-      "grad_norm": 0.42420145869255066,
-      "learning_rate": 9.300022605386793e-06,
-      "loss": 0.3405,
-      "step": 8587
-    },
-    {
-      "epoch": 0.5612705051957388,
-      "grad_norm": 0.4372050166130066,
-      "learning_rate": 9.29984440798439e-06,
-      "loss": 0.39,
-      "step": 8588
-    },
-    {
-      "epoch": 0.5613358604012809,
-      "grad_norm": 0.46907809376716614,
-      "learning_rate": 9.299666189610157e-06,
-      "loss": 0.3912,
-      "step": 8589
-    },
-    {
-      "epoch": 0.5614012156068231,
-      "grad_norm": 0.4684206545352936,
-      "learning_rate": 9.299487950264963e-06,
-      "loss": 0.3737,
-      "step": 8590
-    },
-    {
-      "epoch": 0.5614665708123652,
-      "grad_norm": 0.4382263422012329,
-      "learning_rate": 9.299309689949676e-06,
-      "loss": 0.3957,
-      "step": 8591
-    },
-    {
-      "epoch": 0.5615319260179074,
-      "grad_norm": 0.43257221579551697,
-      "learning_rate": 9.299131408665166e-06,
-      "loss": 0.3683,
-      "step": 8592
-    },
-    {
-      "epoch": 0.5615972812234494,
-      "grad_norm": 0.4597841203212738,
-      "learning_rate": 9.298953106412304e-06,
-      "loss": 0.4233,
-      "step": 8593
-    },
-    {
-      "epoch": 0.5616626364289916,
-      "grad_norm": 0.4499904215335846,
-      "learning_rate": 9.298774783191956e-06,
-      "loss": 0.3917,
-      "step": 8594
-    },
-    {
-      "epoch": 0.5617279916345337,
-      "grad_norm": 0.44387346506118774,
-      "learning_rate": 9.298596439004996e-06,
-      "loss": 0.354,
-      "step": 8595
-    },
-    {
-      "epoch": 0.5617933468400758,
-      "grad_norm": 0.4785110354423523,
-      "learning_rate": 9.298418073852291e-06,
-      "loss": 0.3891,
-      "step": 8596
-    },
-    {
-      "epoch": 0.5618587020456179,
-      "grad_norm": 0.4568953216075897,
-      "learning_rate": 9.298239687734712e-06,
-      "loss": 0.3957,
-      "step": 8597
-    },
-    {
-      "epoch": 0.56192405725116,
-      "grad_norm": 0.4183341860771179,
-      "learning_rate": 9.29806128065313e-06,
-      "loss": 0.3388,
-      "step": 8598
-    },
-    {
-      "epoch": 0.5619894124567022,
-      "grad_norm": 0.413400799036026,
-      "learning_rate": 9.297882852608413e-06,
-      "loss": 0.3108,
-      "step": 8599
-    },
-    {
-      "epoch": 0.5620547676622443,
-      "grad_norm": 0.5093491077423096,
-      "learning_rate": 9.297704403601433e-06,
-      "loss": 0.4676,
-      "step": 8600
-    },
-    {
-      "epoch": 0.5621201228677865,
-      "grad_norm": 0.47162389755249023,
-      "learning_rate": 9.297525933633061e-06,
-      "loss": 0.4524,
-      "step": 8601
-    },
-    {
-      "epoch": 0.5621854780733285,
-      "grad_norm": 0.4399264454841614,
-      "learning_rate": 9.297347442704164e-06,
-      "loss": 0.358,
-      "step": 8602
-    },
-    {
-      "epoch": 0.5622508332788707,
-      "grad_norm": 0.49042192101478577,
-      "learning_rate": 9.297168930815616e-06,
-      "loss": 0.4067,
-      "step": 8603
-    },
-    {
-      "epoch": 0.5623161884844128,
-      "grad_norm": 0.44997796416282654,
-      "learning_rate": 9.296990397968286e-06,
-      "loss": 0.3824,
-      "step": 8604
-    },
-    {
-      "epoch": 0.5623815436899549,
-      "grad_norm": 0.45137375593185425,
-      "learning_rate": 9.296811844163046e-06,
-      "loss": 0.3753,
-      "step": 8605
-    },
-    {
-      "epoch": 0.562446898895497,
-      "grad_norm": 0.46417152881622314,
-      "learning_rate": 9.296633269400766e-06,
-      "loss": 0.3899,
-      "step": 8606
-    },
-    {
-      "epoch": 0.5625122541010391,
-      "grad_norm": 0.44375890493392944,
-      "learning_rate": 9.296454673682316e-06,
-      "loss": 0.3849,
-      "step": 8607
-    },
-    {
-      "epoch": 0.5625776093065813,
-      "grad_norm": 0.46177980303764343,
-      "learning_rate": 9.296276057008569e-06,
-      "loss": 0.3794,
-      "step": 8608
-    },
-    {
-      "epoch": 0.5626429645121234,
-      "grad_norm": 0.4350912570953369,
-      "learning_rate": 9.296097419380394e-06,
-      "loss": 0.3384,
-      "step": 8609
-    },
-    {
-      "epoch": 0.5627083197176656,
-      "grad_norm": 0.4928731918334961,
-      "learning_rate": 9.295918760798665e-06,
-      "loss": 0.401,
-      "step": 8610
-    },
-    {
-      "epoch": 0.5627736749232076,
-      "grad_norm": 0.46854060888290405,
-      "learning_rate": 9.295740081264252e-06,
-      "loss": 0.3889,
-      "step": 8611
-    },
-    {
-      "epoch": 0.5628390301287498,
-      "grad_norm": 0.4220615327358246,
-      "learning_rate": 9.295561380778025e-06,
-      "loss": 0.3453,
-      "step": 8612
-    },
-    {
-      "epoch": 0.5629043853342919,
-      "grad_norm": 0.4408263564109802,
-      "learning_rate": 9.29538265934086e-06,
-      "loss": 0.3374,
-      "step": 8613
-    },
-    {
-      "epoch": 0.562969740539834,
-      "grad_norm": 0.4586383104324341,
-      "learning_rate": 9.295203916953622e-06,
-      "loss": 0.3845,
-      "step": 8614
-    },
-    {
-      "epoch": 0.5630350957453761,
-      "grad_norm": 0.47495511174201965,
-      "learning_rate": 9.295025153617189e-06,
-      "loss": 0.3718,
-      "step": 8615
-    },
-    {
-      "epoch": 0.5631004509509182,
-      "grad_norm": 0.46429243683815,
-      "learning_rate": 9.294846369332429e-06,
-      "loss": 0.3959,
-      "step": 8616
-    },
-    {
-      "epoch": 0.5631658061564604,
-      "grad_norm": 0.43938446044921875,
-      "learning_rate": 9.294667564100216e-06,
-      "loss": 0.3659,
-      "step": 8617
-    },
-    {
-      "epoch": 0.5632311613620025,
-      "grad_norm": 0.4337911307811737,
-      "learning_rate": 9.294488737921422e-06,
-      "loss": 0.3944,
-      "step": 8618
-    },
-    {
-      "epoch": 0.5632965165675446,
-      "grad_norm": 0.5173372030258179,
-      "learning_rate": 9.294309890796918e-06,
-      "loss": 0.4011,
-      "step": 8619
-    },
-    {
-      "epoch": 0.5633618717730867,
-      "grad_norm": 0.4294881820678711,
-      "learning_rate": 9.294131022727576e-06,
-      "loss": 0.3556,
-      "step": 8620
-    },
-    {
-      "epoch": 0.5634272269786288,
-      "grad_norm": 0.44056859612464905,
-      "learning_rate": 9.293952133714271e-06,
-      "loss": 0.3869,
-      "step": 8621
-    },
-    {
-      "epoch": 0.563492582184171,
-      "grad_norm": 0.43963128328323364,
-      "learning_rate": 9.293773223757873e-06,
-      "loss": 0.3359,
-      "step": 8622
-    },
-    {
-      "epoch": 0.563557937389713,
-      "grad_norm": 0.45700833201408386,
-      "learning_rate": 9.293594292859257e-06,
-      "loss": 0.4142,
-      "step": 8623
-    },
-    {
-      "epoch": 0.5636232925952552,
-      "grad_norm": 0.4210696518421173,
-      "learning_rate": 9.293415341019292e-06,
-      "loss": 0.3292,
-      "step": 8624
-    },
-    {
-      "epoch": 0.5636886478007973,
-      "grad_norm": 0.463018000125885,
-      "learning_rate": 9.293236368238854e-06,
-      "loss": 0.424,
-      "step": 8625
-    },
-    {
-      "epoch": 0.5637540030063395,
-      "grad_norm": 0.44253072142601013,
-      "learning_rate": 9.293057374518814e-06,
-      "loss": 0.4123,
-      "step": 8626
-    },
-    {
-      "epoch": 0.5638193582118816,
-      "grad_norm": 0.5666806697845459,
-      "learning_rate": 9.292878359860047e-06,
-      "loss": 0.4112,
-      "step": 8627
-    },
-    {
-      "epoch": 0.5638847134174237,
-      "grad_norm": 0.47362110018730164,
-      "learning_rate": 9.292699324263425e-06,
-      "loss": 0.4513,
-      "step": 8628
-    },
-    {
-      "epoch": 0.5639500686229658,
-      "grad_norm": 0.43033018708229065,
-      "learning_rate": 9.292520267729821e-06,
-      "loss": 0.3745,
-      "step": 8629
-    },
-    {
-      "epoch": 0.5640154238285079,
-      "grad_norm": 0.47930434346199036,
-      "learning_rate": 9.292341190260108e-06,
-      "loss": 0.4228,
-      "step": 8630
-    },
-    {
-      "epoch": 0.5640807790340501,
-      "grad_norm": 0.48313361406326294,
-      "learning_rate": 9.292162091855161e-06,
-      "loss": 0.3644,
-      "step": 8631
-    },
-    {
-      "epoch": 0.5641461342395921,
-      "grad_norm": 0.4595695734024048,
-      "learning_rate": 9.291982972515853e-06,
-      "loss": 0.4022,
-      "step": 8632
-    },
-    {
-      "epoch": 0.5642114894451343,
-      "grad_norm": 0.41800472140312195,
-      "learning_rate": 9.291803832243055e-06,
-      "loss": 0.3411,
-      "step": 8633
-    },
-    {
-      "epoch": 0.5642768446506764,
-      "grad_norm": 0.4657498896121979,
-      "learning_rate": 9.291624671037646e-06,
-      "loss": 0.3902,
-      "step": 8634
-    },
-    {
-      "epoch": 0.5643421998562186,
-      "grad_norm": 0.4660916328430176,
-      "learning_rate": 9.291445488900494e-06,
-      "loss": 0.3701,
-      "step": 8635
-    },
-    {
-      "epoch": 0.5644075550617607,
-      "grad_norm": 0.41811466217041016,
-      "learning_rate": 9.291266285832477e-06,
-      "loss": 0.3809,
-      "step": 8636
-    },
-    {
-      "epoch": 0.5644729102673028,
-      "grad_norm": 0.4972060024738312,
-      "learning_rate": 9.291087061834467e-06,
-      "loss": 0.4304,
-      "step": 8637
-    },
-    {
-      "epoch": 0.5645382654728449,
-      "grad_norm": 0.46764495968818665,
-      "learning_rate": 9.29090781690734e-06,
-      "loss": 0.4293,
-      "step": 8638
-    },
-    {
-      "epoch": 0.564603620678387,
-      "grad_norm": 0.4698627293109894,
-      "learning_rate": 9.290728551051969e-06,
-      "loss": 0.3943,
-      "step": 8639
-    },
-    {
-      "epoch": 0.5646689758839292,
-      "grad_norm": 0.4544629752635956,
-      "learning_rate": 9.290549264269227e-06,
-      "loss": 0.3561,
-      "step": 8640
-    },
-    {
-      "epoch": 0.5647343310894712,
-      "grad_norm": 0.468403697013855,
-      "learning_rate": 9.29036995655999e-06,
-      "loss": 0.3841,
-      "step": 8641
-    },
-    {
-      "epoch": 0.5647996862950134,
-      "grad_norm": 0.4759049713611603,
-      "learning_rate": 9.290190627925133e-06,
-      "loss": 0.4015,
-      "step": 8642
-    },
-    {
-      "epoch": 0.5648650415005555,
-      "grad_norm": 0.48529309034347534,
-      "learning_rate": 9.290011278365529e-06,
-      "loss": 0.409,
-      "step": 8643
-    },
-    {
-      "epoch": 0.5649303967060977,
-      "grad_norm": 0.5249614715576172,
-      "learning_rate": 9.289831907882055e-06,
-      "loss": 0.5118,
-      "step": 8644
-    },
-    {
-      "epoch": 0.5649957519116398,
-      "grad_norm": 0.4921436905860901,
-      "learning_rate": 9.289652516475584e-06,
-      "loss": 0.4659,
-      "step": 8645
-    },
-    {
-      "epoch": 0.5650611071171819,
-      "grad_norm": 0.45720160007476807,
-      "learning_rate": 9.289473104146993e-06,
-      "loss": 0.3783,
-      "step": 8646
-    },
-    {
-      "epoch": 0.565126462322724,
-      "grad_norm": 0.4505588114261627,
-      "learning_rate": 9.289293670897156e-06,
-      "loss": 0.3748,
-      "step": 8647
-    },
-    {
-      "epoch": 0.5651918175282661,
-      "grad_norm": 0.495913028717041,
-      "learning_rate": 9.289114216726944e-06,
-      "loss": 0.4725,
-      "step": 8648
-    },
-    {
-      "epoch": 0.5652571727338083,
-      "grad_norm": 0.4198303520679474,
-      "learning_rate": 9.288934741637239e-06,
-      "loss": 0.3522,
-      "step": 8649
-    },
-    {
-      "epoch": 0.5653225279393503,
-      "grad_norm": 0.43457698822021484,
-      "learning_rate": 9.288755245628913e-06,
-      "loss": 0.3483,
-      "step": 8650
-    },
-    {
-      "epoch": 0.5653878831448925,
-      "grad_norm": 0.47232887148857117,
-      "learning_rate": 9.288575728702842e-06,
-      "loss": 0.4,
-      "step": 8651
-    },
-    {
-      "epoch": 0.5654532383504346,
-      "grad_norm": 0.4795187711715698,
-      "learning_rate": 9.288396190859901e-06,
-      "loss": 0.4176,
-      "step": 8652
-    },
-    {
-      "epoch": 0.5655185935559768,
-      "grad_norm": 0.43719935417175293,
-      "learning_rate": 9.288216632100965e-06,
-      "loss": 0.3512,
-      "step": 8653
-    },
-    {
-      "epoch": 0.5655839487615189,
-      "grad_norm": 0.49101826548576355,
-      "learning_rate": 9.288037052426914e-06,
-      "loss": 0.4094,
-      "step": 8654
-    },
-    {
-      "epoch": 0.5656493039670609,
-      "grad_norm": 0.4694311022758484,
-      "learning_rate": 9.287857451838618e-06,
-      "loss": 0.4255,
-      "step": 8655
-    },
-    {
-      "epoch": 0.5657146591726031,
-      "grad_norm": 0.4479016363620758,
-      "learning_rate": 9.287677830336958e-06,
-      "loss": 0.3751,
-      "step": 8656
-    },
-    {
-      "epoch": 0.5657800143781452,
-      "grad_norm": 0.4267028272151947,
-      "learning_rate": 9.287498187922806e-06,
-      "loss": 0.3211,
-      "step": 8657
-    },
-    {
-      "epoch": 0.5658453695836874,
-      "grad_norm": 0.4701300263404846,
-      "learning_rate": 9.287318524597041e-06,
-      "loss": 0.3711,
-      "step": 8658
-    },
-    {
-      "epoch": 0.5659107247892294,
-      "grad_norm": 0.49085545539855957,
-      "learning_rate": 9.287138840360539e-06,
-      "loss": 0.444,
-      "step": 8659
-    },
-    {
-      "epoch": 0.5659760799947716,
-      "grad_norm": 0.4548317492008209,
-      "learning_rate": 9.286959135214174e-06,
-      "loss": 0.3932,
-      "step": 8660
-    },
-    {
-      "epoch": 0.5660414352003137,
-      "grad_norm": 0.5431153774261475,
-      "learning_rate": 9.286779409158825e-06,
-      "loss": 0.4172,
-      "step": 8661
-    },
-    {
-      "epoch": 0.5661067904058559,
-      "grad_norm": 0.7028505206108093,
-      "learning_rate": 9.286599662195366e-06,
-      "loss": 0.3475,
-      "step": 8662
-    },
-    {
-      "epoch": 0.566172145611398,
-      "grad_norm": 0.4354560375213623,
-      "learning_rate": 9.286419894324679e-06,
-      "loss": 0.3459,
-      "step": 8663
-    },
-    {
-      "epoch": 0.56623750081694,
-      "grad_norm": 0.47522449493408203,
-      "learning_rate": 9.286240105547634e-06,
-      "loss": 0.4194,
-      "step": 8664
-    },
-    {
-      "epoch": 0.5663028560224822,
-      "grad_norm": 0.4407506585121155,
-      "learning_rate": 9.286060295865113e-06,
-      "loss": 0.3603,
-      "step": 8665
-    },
-    {
-      "epoch": 0.5663682112280243,
-      "grad_norm": 0.4717441201210022,
-      "learning_rate": 9.28588046527799e-06,
-      "loss": 0.3719,
-      "step": 8666
-    },
-    {
-      "epoch": 0.5664335664335665,
-      "grad_norm": 0.4331178069114685,
-      "learning_rate": 9.285700613787143e-06,
-      "loss": 0.366,
-      "step": 8667
-    },
-    {
-      "epoch": 0.5664989216391085,
-      "grad_norm": 0.4383242428302765,
-      "learning_rate": 9.28552074139345e-06,
-      "loss": 0.3901,
-      "step": 8668
-    },
-    {
-      "epoch": 0.5665642768446507,
-      "grad_norm": 0.4694553315639496,
-      "learning_rate": 9.285340848097788e-06,
-      "loss": 0.4356,
-      "step": 8669
-    },
-    {
-      "epoch": 0.5666296320501928,
-      "grad_norm": 0.49563634395599365,
-      "learning_rate": 9.285160933901034e-06,
-      "loss": 0.4208,
-      "step": 8670
-    },
-    {
-      "epoch": 0.566694987255735,
-      "grad_norm": 0.4513710141181946,
-      "learning_rate": 9.284980998804065e-06,
-      "loss": 0.3626,
-      "step": 8671
-    },
-    {
-      "epoch": 0.566760342461277,
-      "grad_norm": 0.47239089012145996,
-      "learning_rate": 9.28480104280776e-06,
-      "loss": 0.4108,
-      "step": 8672
-    },
-    {
-      "epoch": 0.5668256976668191,
-      "grad_norm": 0.42150744795799255,
-      "learning_rate": 9.284621065912995e-06,
-      "loss": 0.3659,
-      "step": 8673
-    },
-    {
-      "epoch": 0.5668910528723613,
-      "grad_norm": 0.46837830543518066,
-      "learning_rate": 9.284441068120649e-06,
-      "loss": 0.43,
-      "step": 8674
-    },
-    {
-      "epoch": 0.5669564080779034,
-      "grad_norm": 0.45562320947647095,
-      "learning_rate": 9.2842610494316e-06,
-      "loss": 0.3655,
-      "step": 8675
-    },
-    {
-      "epoch": 0.5670217632834456,
-      "grad_norm": 0.5412616729736328,
-      "learning_rate": 9.284081009846725e-06,
-      "loss": 0.3798,
-      "step": 8676
-    },
-    {
-      "epoch": 0.5670871184889876,
-      "grad_norm": 0.45065397024154663,
-      "learning_rate": 9.283900949366903e-06,
-      "loss": 0.3324,
-      "step": 8677
-    },
-    {
-      "epoch": 0.5671524736945298,
-      "grad_norm": 0.4478610157966614,
-      "learning_rate": 9.283720867993013e-06,
-      "loss": 0.3807,
-      "step": 8678
-    },
-    {
-      "epoch": 0.5672178289000719,
-      "grad_norm": 0.48112455010414124,
-      "learning_rate": 9.28354076572593e-06,
-      "loss": 0.4082,
-      "step": 8679
-    },
-    {
-      "epoch": 0.567283184105614,
-      "grad_norm": 0.4559597074985504,
-      "learning_rate": 9.28336064256654e-06,
-      "loss": 0.3602,
-      "step": 8680
-    },
-    {
-      "epoch": 0.5673485393111561,
-      "grad_norm": 0.4585297703742981,
-      "learning_rate": 9.283180498515711e-06,
-      "loss": 0.3839,
-      "step": 8681
-    },
-    {
-      "epoch": 0.5674138945166982,
-      "grad_norm": 0.4685385823249817,
-      "learning_rate": 9.283000333574328e-06,
-      "loss": 0.3856,
-      "step": 8682
-    },
-    {
-      "epoch": 0.5674792497222404,
-      "grad_norm": 0.45225974917411804,
-      "learning_rate": 9.28282014774327e-06,
-      "loss": 0.394,
-      "step": 8683
-    },
-    {
-      "epoch": 0.5675446049277825,
-      "grad_norm": 0.4868801236152649,
-      "learning_rate": 9.282639941023413e-06,
-      "loss": 0.4484,
-      "step": 8684
-    },
-    {
-      "epoch": 0.5676099601333247,
-      "grad_norm": 0.43020132184028625,
-      "learning_rate": 9.282459713415637e-06,
-      "loss": 0.3681,
-      "step": 8685
-    },
-    {
-      "epoch": 0.5676753153388667,
-      "grad_norm": 0.4511564075946808,
-      "learning_rate": 9.282279464920823e-06,
-      "loss": 0.4038,
-      "step": 8686
-    },
-    {
-      "epoch": 0.5677406705444089,
-      "grad_norm": 0.40919962525367737,
-      "learning_rate": 9.282099195539847e-06,
-      "loss": 0.3372,
-      "step": 8687
-    },
-    {
-      "epoch": 0.567806025749951,
-      "grad_norm": 0.4809260666370392,
-      "learning_rate": 9.28191890527359e-06,
-      "loss": 0.4269,
-      "step": 8688
-    },
-    {
-      "epoch": 0.567871380955493,
-      "grad_norm": 0.4630745053291321,
-      "learning_rate": 9.281738594122933e-06,
-      "loss": 0.3846,
-      "step": 8689
-    },
-    {
-      "epoch": 0.5679367361610352,
-      "grad_norm": 0.4574970304965973,
-      "learning_rate": 9.281558262088753e-06,
-      "loss": 0.381,
-      "step": 8690
-    },
-    {
-      "epoch": 0.5680020913665773,
-      "grad_norm": 0.457764208316803,
-      "learning_rate": 9.281377909171927e-06,
-      "loss": 0.3929,
-      "step": 8691
-    },
-    {
-      "epoch": 0.5680674465721195,
-      "grad_norm": 0.39306744933128357,
-      "learning_rate": 9.281197535373341e-06,
-      "loss": 0.2826,
-      "step": 8692
-    },
-    {
-      "epoch": 0.5681328017776616,
-      "grad_norm": 0.4456539750099182,
-      "learning_rate": 9.28101714069387e-06,
-      "loss": 0.3812,
-      "step": 8693
-    },
-    {
-      "epoch": 0.5681981569832038,
-      "grad_norm": 0.4824658930301666,
-      "learning_rate": 9.280836725134396e-06,
-      "loss": 0.4175,
-      "step": 8694
-    },
-    {
-      "epoch": 0.5682635121887458,
-      "grad_norm": 0.43890756368637085,
-      "learning_rate": 9.280656288695797e-06,
-      "loss": 0.357,
-      "step": 8695
-    },
-    {
-      "epoch": 0.568328867394288,
-      "grad_norm": 0.4513535797595978,
-      "learning_rate": 9.280475831378957e-06,
-      "loss": 0.3599,
-      "step": 8696
-    },
-    {
-      "epoch": 0.5683942225998301,
-      "grad_norm": 0.4885607063770294,
-      "learning_rate": 9.280295353184751e-06,
-      "loss": 0.414,
-      "step": 8697
-    },
-    {
-      "epoch": 0.5684595778053722,
-      "grad_norm": 0.4631774127483368,
-      "learning_rate": 9.280114854114062e-06,
-      "loss": 0.3978,
-      "step": 8698
-    },
-    {
-      "epoch": 0.5685249330109143,
-      "grad_norm": 0.4900452494621277,
-      "learning_rate": 9.279934334167771e-06,
-      "loss": 0.4221,
-      "step": 8699
-    },
-    {
-      "epoch": 0.5685902882164564,
-      "grad_norm": 0.47625797986984253,
-      "learning_rate": 9.279753793346758e-06,
-      "loss": 0.4086,
-      "step": 8700
-    },
-    {
-      "epoch": 0.5686556434219986,
-      "grad_norm": 0.46046555042266846,
-      "learning_rate": 9.279573231651901e-06,
-      "loss": 0.3931,
-      "step": 8701
-    },
-    {
-      "epoch": 0.5687209986275407,
-      "grad_norm": 0.4615381956100464,
-      "learning_rate": 9.279392649084084e-06,
-      "loss": 0.4088,
-      "step": 8702
-    },
-    {
-      "epoch": 0.5687863538330828,
-      "grad_norm": 0.4618852138519287,
-      "learning_rate": 9.279212045644185e-06,
-      "loss": 0.4472,
-      "step": 8703
-    },
-    {
-      "epoch": 0.5688517090386249,
-      "grad_norm": 0.45024463534355164,
-      "learning_rate": 9.279031421333089e-06,
-      "loss": 0.3928,
-      "step": 8704
-    },
-    {
-      "epoch": 0.5689170642441671,
-      "grad_norm": 0.45709168910980225,
-      "learning_rate": 9.278850776151673e-06,
-      "loss": 0.4237,
-      "step": 8705
-    },
-    {
-      "epoch": 0.5689824194497092,
-      "grad_norm": 0.44051453471183777,
-      "learning_rate": 9.27867011010082e-06,
-      "loss": 0.3821,
-      "step": 8706
-    },
-    {
-      "epoch": 0.5690477746552512,
-      "grad_norm": 0.4611460566520691,
-      "learning_rate": 9.278489423181412e-06,
-      "loss": 0.3872,
-      "step": 8707
-    },
-    {
-      "epoch": 0.5691131298607934,
-      "grad_norm": 0.44791755080223083,
-      "learning_rate": 9.278308715394326e-06,
-      "loss": 0.3552,
-      "step": 8708
-    },
-    {
-      "epoch": 0.5691784850663355,
-      "grad_norm": 0.4372178018093109,
-      "learning_rate": 9.278127986740448e-06,
-      "loss": 0.3553,
-      "step": 8709
-    },
-    {
-      "epoch": 0.5692438402718777,
-      "grad_norm": 0.4763945937156677,
-      "learning_rate": 9.277947237220657e-06,
-      "loss": 0.4066,
-      "step": 8710
-    },
-    {
-      "epoch": 0.5693091954774198,
-      "grad_norm": 0.4812767207622528,
-      "learning_rate": 9.277766466835836e-06,
-      "loss": 0.3918,
-      "step": 8711
-    },
-    {
-      "epoch": 0.5693745506829619,
-      "grad_norm": 0.44954633712768555,
-      "learning_rate": 9.277585675586865e-06,
-      "loss": 0.3622,
-      "step": 8712
-    },
-    {
-      "epoch": 0.569439905888504,
-      "grad_norm": 0.465044766664505,
-      "learning_rate": 9.277404863474625e-06,
-      "loss": 0.3895,
-      "step": 8713
-    },
-    {
-      "epoch": 0.5695052610940461,
-      "grad_norm": 0.469230979681015,
-      "learning_rate": 9.277224030500004e-06,
-      "loss": 0.4119,
-      "step": 8714
-    },
-    {
-      "epoch": 0.5695706162995883,
-      "grad_norm": 0.4770100712776184,
-      "learning_rate": 9.277043176663877e-06,
-      "loss": 0.4144,
-      "step": 8715
-    },
-    {
-      "epoch": 0.5696359715051303,
-      "grad_norm": 0.42255279421806335,
-      "learning_rate": 9.276862301967129e-06,
-      "loss": 0.3563,
-      "step": 8716
-    },
-    {
-      "epoch": 0.5697013267106725,
-      "grad_norm": 0.45624223351478577,
-      "learning_rate": 9.276681406410642e-06,
-      "loss": 0.4024,
-      "step": 8717
-    },
-    {
-      "epoch": 0.5697666819162146,
-      "grad_norm": 0.4329855442047119,
-      "learning_rate": 9.276500489995299e-06,
-      "loss": 0.3916,
-      "step": 8718
-    },
-    {
-      "epoch": 0.5698320371217568,
-      "grad_norm": 0.4552854299545288,
-      "learning_rate": 9.27631955272198e-06,
-      "loss": 0.3883,
-      "step": 8719
-    },
-    {
-      "epoch": 0.5698973923272989,
-      "grad_norm": 0.4230673313140869,
-      "learning_rate": 9.27613859459157e-06,
-      "loss": 0.3505,
-      "step": 8720
-    },
-    {
-      "epoch": 0.569962747532841,
-      "grad_norm": 0.4523487091064453,
-      "learning_rate": 9.275957615604952e-06,
-      "loss": 0.4033,
-      "step": 8721
-    },
-    {
-      "epoch": 0.5700281027383831,
-      "grad_norm": 0.4278339743614197,
-      "learning_rate": 9.275776615763004e-06,
-      "loss": 0.3723,
-      "step": 8722
-    },
-    {
-      "epoch": 0.5700934579439252,
-      "grad_norm": 0.44875356554985046,
-      "learning_rate": 9.275595595066615e-06,
-      "loss": 0.3686,
-      "step": 8723
-    },
-    {
-      "epoch": 0.5701588131494674,
-      "grad_norm": 0.44797366857528687,
-      "learning_rate": 9.275414553516664e-06,
-      "loss": 0.3539,
-      "step": 8724
-    },
-    {
-      "epoch": 0.5702241683550094,
-      "grad_norm": 0.456897109746933,
-      "learning_rate": 9.275233491114035e-06,
-      "loss": 0.3608,
-      "step": 8725
-    },
-    {
-      "epoch": 0.5702895235605516,
-      "grad_norm": 0.4676462411880493,
-      "learning_rate": 9.275052407859612e-06,
-      "loss": 0.4194,
-      "step": 8726
-    },
-    {
-      "epoch": 0.5703548787660937,
-      "grad_norm": 0.4540182948112488,
-      "learning_rate": 9.274871303754277e-06,
-      "loss": 0.3996,
-      "step": 8727
-    },
-    {
-      "epoch": 0.5704202339716359,
-      "grad_norm": 0.45552679896354675,
-      "learning_rate": 9.274690178798913e-06,
-      "loss": 0.4044,
-      "step": 8728
-    },
-    {
-      "epoch": 0.570485589177178,
-      "grad_norm": 0.42875367403030396,
-      "learning_rate": 9.274509032994405e-06,
-      "loss": 0.355,
-      "step": 8729
-    },
-    {
-      "epoch": 0.5705509443827201,
-      "grad_norm": 0.47671782970428467,
-      "learning_rate": 9.274327866341635e-06,
-      "loss": 0.423,
-      "step": 8730
-    },
-    {
-      "epoch": 0.5706162995882622,
-      "grad_norm": 0.4635762870311737,
-      "learning_rate": 9.274146678841488e-06,
-      "loss": 0.4069,
-      "step": 8731
-    },
-    {
-      "epoch": 0.5706816547938043,
-      "grad_norm": 0.4347746670246124,
-      "learning_rate": 9.273965470494846e-06,
-      "loss": 0.3863,
-      "step": 8732
-    },
-    {
-      "epoch": 0.5707470099993465,
-      "grad_norm": 0.44302892684936523,
-      "learning_rate": 9.273784241302594e-06,
-      "loss": 0.3877,
-      "step": 8733
-    },
-    {
-      "epoch": 0.5708123652048885,
-      "grad_norm": 0.4646381139755249,
-      "learning_rate": 9.273602991265617e-06,
-      "loss": 0.386,
-      "step": 8734
-    },
-    {
-      "epoch": 0.5708777204104307,
-      "grad_norm": 0.490407258272171,
-      "learning_rate": 9.273421720384797e-06,
-      "loss": 0.4026,
-      "step": 8735
-    },
-    {
-      "epoch": 0.5709430756159728,
-      "grad_norm": 0.4498681128025055,
-      "learning_rate": 9.273240428661017e-06,
-      "loss": 0.3773,
-      "step": 8736
-    },
-    {
-      "epoch": 0.571008430821515,
-      "grad_norm": 0.48290586471557617,
-      "learning_rate": 9.273059116095165e-06,
-      "loss": 0.409,
-      "step": 8737
-    },
-    {
-      "epoch": 0.571073786027057,
-      "grad_norm": 0.47024261951446533,
-      "learning_rate": 9.272877782688123e-06,
-      "loss": 0.4237,
-      "step": 8738
-    },
-    {
-      "epoch": 0.5711391412325991,
-      "grad_norm": 0.4955271780490875,
-      "learning_rate": 9.272696428440775e-06,
-      "loss": 0.4309,
-      "step": 8739
-    },
-    {
-      "epoch": 0.5712044964381413,
-      "grad_norm": 0.4528626501560211,
-      "learning_rate": 9.272515053354008e-06,
-      "loss": 0.3896,
-      "step": 8740
-    },
-    {
-      "epoch": 0.5712698516436834,
-      "grad_norm": 0.4436284005641937,
-      "learning_rate": 9.272333657428703e-06,
-      "loss": 0.377,
-      "step": 8741
-    },
-    {
-      "epoch": 0.5713352068492256,
-      "grad_norm": 0.45342323184013367,
-      "learning_rate": 9.272152240665748e-06,
-      "loss": 0.3724,
-      "step": 8742
-    },
-    {
-      "epoch": 0.5714005620547676,
-      "grad_norm": 0.47489410638809204,
-      "learning_rate": 9.271970803066027e-06,
-      "loss": 0.3683,
-      "step": 8743
-    },
-    {
-      "epoch": 0.5714659172603098,
-      "grad_norm": 0.4638662040233612,
-      "learning_rate": 9.271789344630421e-06,
-      "loss": 0.4109,
-      "step": 8744
-    },
-    {
-      "epoch": 0.5715312724658519,
-      "grad_norm": 0.4163479208946228,
-      "learning_rate": 9.271607865359822e-06,
-      "loss": 0.3541,
-      "step": 8745
-    },
-    {
-      "epoch": 0.5715966276713941,
-      "grad_norm": 0.4930134117603302,
-      "learning_rate": 9.27142636525511e-06,
-      "loss": 0.4605,
-      "step": 8746
-    },
-    {
-      "epoch": 0.5716619828769361,
-      "grad_norm": 0.4300340712070465,
-      "learning_rate": 9.271244844317174e-06,
-      "loss": 0.3688,
-      "step": 8747
-    },
-    {
-      "epoch": 0.5717273380824782,
-      "grad_norm": 0.43568021059036255,
-      "learning_rate": 9.271063302546895e-06,
-      "loss": 0.3679,
-      "step": 8748
-    },
-    {
-      "epoch": 0.5717926932880204,
-      "grad_norm": 0.46499136090278625,
-      "learning_rate": 9.270881739945161e-06,
-      "loss": 0.4136,
-      "step": 8749
-    },
-    {
-      "epoch": 0.5718580484935625,
-      "grad_norm": 0.4746057987213135,
-      "learning_rate": 9.270700156512857e-06,
-      "loss": 0.4093,
-      "step": 8750
-    },
-    {
-      "epoch": 0.5719234036991047,
-      "grad_norm": 0.418390691280365,
-      "learning_rate": 9.270518552250868e-06,
-      "loss": 0.348,
-      "step": 8751
-    },
-    {
-      "epoch": 0.5719887589046467,
-      "grad_norm": 0.46858885884284973,
-      "learning_rate": 9.270336927160084e-06,
-      "loss": 0.3771,
-      "step": 8752
-    },
-    {
-      "epoch": 0.5720541141101889,
-      "grad_norm": 0.44084256887435913,
-      "learning_rate": 9.270155281241385e-06,
-      "loss": 0.3901,
-      "step": 8753
-    },
-    {
-      "epoch": 0.572119469315731,
-      "grad_norm": 0.41746649146080017,
-      "learning_rate": 9.26997361449566e-06,
-      "loss": 0.3739,
-      "step": 8754
-    },
-    {
-      "epoch": 0.5721848245212732,
-      "grad_norm": 0.4664878249168396,
-      "learning_rate": 9.269791926923793e-06,
-      "loss": 0.4085,
-      "step": 8755
-    },
-    {
-      "epoch": 0.5722501797268152,
-      "grad_norm": 0.4303506016731262,
-      "learning_rate": 9.269610218526673e-06,
-      "loss": 0.3734,
-      "step": 8756
-    },
-    {
-      "epoch": 0.5723155349323573,
-      "grad_norm": 0.4895287752151489,
-      "learning_rate": 9.269428489305185e-06,
-      "loss": 0.4059,
-      "step": 8757
-    },
-    {
-      "epoch": 0.5723808901378995,
-      "grad_norm": 0.45096322894096375,
-      "learning_rate": 9.269246739260214e-06,
-      "loss": 0.3922,
-      "step": 8758
-    },
-    {
-      "epoch": 0.5724462453434416,
-      "grad_norm": 0.44489356875419617,
-      "learning_rate": 9.269064968392649e-06,
-      "loss": 0.3794,
-      "step": 8759
-    },
-    {
-      "epoch": 0.5725116005489838,
-      "grad_norm": 0.42755281925201416,
-      "learning_rate": 9.268883176703374e-06,
-      "loss": 0.3288,
-      "step": 8760
-    },
-    {
-      "epoch": 0.5725769557545258,
-      "grad_norm": 0.47941091656684875,
-      "learning_rate": 9.268701364193277e-06,
-      "loss": 0.3859,
-      "step": 8761
-    },
-    {
-      "epoch": 0.572642310960068,
-      "grad_norm": 0.464324414730072,
-      "learning_rate": 9.268519530863244e-06,
-      "loss": 0.3824,
-      "step": 8762
-    },
-    {
-      "epoch": 0.5727076661656101,
-      "grad_norm": 0.4192202389240265,
-      "learning_rate": 9.268337676714165e-06,
-      "loss": 0.3234,
-      "step": 8763
-    },
-    {
-      "epoch": 0.5727730213711522,
-      "grad_norm": 0.4357430934906006,
-      "learning_rate": 9.268155801746923e-06,
-      "loss": 0.3375,
-      "step": 8764
-    },
-    {
-      "epoch": 0.5728383765766943,
-      "grad_norm": 0.5227752923965454,
-      "learning_rate": 9.267973905962406e-06,
-      "loss": 0.4523,
-      "step": 8765
-    },
-    {
-      "epoch": 0.5729037317822364,
-      "grad_norm": 0.44509416818618774,
-      "learning_rate": 9.267791989361501e-06,
-      "loss": 0.3794,
-      "step": 8766
-    },
-    {
-      "epoch": 0.5729690869877786,
-      "grad_norm": 0.4422967731952667,
-      "learning_rate": 9.267610051945097e-06,
-      "loss": 0.3316,
-      "step": 8767
-    },
-    {
-      "epoch": 0.5730344421933207,
-      "grad_norm": 0.452650785446167,
-      "learning_rate": 9.26742809371408e-06,
-      "loss": 0.3588,
-      "step": 8768
-    },
-    {
-      "epoch": 0.5730997973988629,
-      "grad_norm": 0.4376973807811737,
-      "learning_rate": 9.26724611466934e-06,
-      "loss": 0.3946,
-      "step": 8769
-    },
-    {
-      "epoch": 0.5731651526044049,
-      "grad_norm": 0.4528738558292389,
-      "learning_rate": 9.26706411481176e-06,
-      "loss": 0.4099,
-      "step": 8770
-    },
-    {
-      "epoch": 0.5732305078099471,
-      "grad_norm": 0.4566257894039154,
-      "learning_rate": 9.266882094142232e-06,
-      "loss": 0.3689,
-      "step": 8771
-    },
-    {
-      "epoch": 0.5732958630154892,
-      "grad_norm": 0.49401339888572693,
-      "learning_rate": 9.26670005266164e-06,
-      "loss": 0.4464,
-      "step": 8772
-    },
-    {
-      "epoch": 0.5733612182210313,
-      "grad_norm": 0.4841618537902832,
-      "learning_rate": 9.266517990370873e-06,
-      "loss": 0.4405,
-      "step": 8773
-    },
-    {
-      "epoch": 0.5734265734265734,
-      "grad_norm": 0.42835187911987305,
-      "learning_rate": 9.26633590727082e-06,
-      "loss": 0.341,
-      "step": 8774
-    },
-    {
-      "epoch": 0.5734919286321155,
-      "grad_norm": 0.42060476541519165,
-      "learning_rate": 9.26615380336237e-06,
-      "loss": 0.3694,
-      "step": 8775
-    },
-    {
-      "epoch": 0.5735572838376577,
-      "grad_norm": 0.44720593094825745,
-      "learning_rate": 9.26597167864641e-06,
-      "loss": 0.3982,
-      "step": 8776
-    },
-    {
-      "epoch": 0.5736226390431998,
-      "grad_norm": 0.4220096468925476,
-      "learning_rate": 9.265789533123828e-06,
-      "loss": 0.3518,
-      "step": 8777
-    },
-    {
-      "epoch": 0.573687994248742,
-      "grad_norm": 0.5278857350349426,
-      "learning_rate": 9.265607366795513e-06,
-      "loss": 0.4637,
-      "step": 8778
-    },
-    {
-      "epoch": 0.573753349454284,
-      "grad_norm": 0.4478168487548828,
-      "learning_rate": 9.265425179662354e-06,
-      "loss": 0.3902,
-      "step": 8779
-    },
-    {
-      "epoch": 0.5738187046598262,
-      "grad_norm": 0.44495293498039246,
-      "learning_rate": 9.265242971725235e-06,
-      "loss": 0.3646,
-      "step": 8780
-    },
-    {
-      "epoch": 0.5738840598653683,
-      "grad_norm": 0.45817679166793823,
-      "learning_rate": 9.265060742985052e-06,
-      "loss": 0.4093,
-      "step": 8781
-    },
-    {
-      "epoch": 0.5739494150709104,
-      "grad_norm": 0.45591941475868225,
-      "learning_rate": 9.264878493442689e-06,
-      "loss": 0.3696,
-      "step": 8782
-    },
-    {
-      "epoch": 0.5740147702764525,
-      "grad_norm": 0.4401930570602417,
-      "learning_rate": 9.264696223099036e-06,
-      "loss": 0.3476,
-      "step": 8783
-    },
-    {
-      "epoch": 0.5740801254819946,
-      "grad_norm": 0.44114094972610474,
-      "learning_rate": 9.264513931954981e-06,
-      "loss": 0.3577,
-      "step": 8784
-    },
-    {
-      "epoch": 0.5741454806875368,
-      "grad_norm": 0.4722091555595398,
-      "learning_rate": 9.264331620011416e-06,
-      "loss": 0.3775,
-      "step": 8785
-    },
-    {
-      "epoch": 0.5742108358930789,
-      "grad_norm": 0.4401685297489166,
-      "learning_rate": 9.264149287269229e-06,
-      "loss": 0.3753,
-      "step": 8786
-    },
-    {
-      "epoch": 0.574276191098621,
-      "grad_norm": 0.47649145126342773,
-      "learning_rate": 9.263966933729307e-06,
-      "loss": 0.4598,
-      "step": 8787
-    },
-    {
-      "epoch": 0.5743415463041631,
-      "grad_norm": 0.5930619239807129,
-      "learning_rate": 9.263784559392541e-06,
-      "loss": 0.4158,
-      "step": 8788
-    },
-    {
-      "epoch": 0.5744069015097053,
-      "grad_norm": 0.5872689485549927,
-      "learning_rate": 9.263602164259822e-06,
-      "loss": 0.3999,
-      "step": 8789
-    },
-    {
-      "epoch": 0.5744722567152474,
-      "grad_norm": 0.43598517775535583,
-      "learning_rate": 9.263419748332037e-06,
-      "loss": 0.3516,
-      "step": 8790
-    },
-    {
-      "epoch": 0.5745376119207894,
-      "grad_norm": 0.4877925217151642,
-      "learning_rate": 9.263237311610076e-06,
-      "loss": 0.4097,
-      "step": 8791
-    },
-    {
-      "epoch": 0.5746029671263316,
-      "grad_norm": 0.4351055920124054,
-      "learning_rate": 9.26305485409483e-06,
-      "loss": 0.3431,
-      "step": 8792
-    },
-    {
-      "epoch": 0.5746683223318737,
-      "grad_norm": 0.48612111806869507,
-      "learning_rate": 9.262872375787191e-06,
-      "loss": 0.4377,
-      "step": 8793
-    },
-    {
-      "epoch": 0.5747336775374159,
-      "grad_norm": 0.4468783140182495,
-      "learning_rate": 9.262689876688045e-06,
-      "loss": 0.407,
-      "step": 8794
-    },
-    {
-      "epoch": 0.574799032742958,
-      "grad_norm": 0.465574711561203,
-      "learning_rate": 9.262507356798284e-06,
-      "loss": 0.4172,
-      "step": 8795
-    },
-    {
-      "epoch": 0.5748643879485001,
-      "grad_norm": 0.44357535243034363,
-      "learning_rate": 9.262324816118798e-06,
-      "loss": 0.3849,
-      "step": 8796
-    },
-    {
-      "epoch": 0.5749297431540422,
-      "grad_norm": 0.46302565932273865,
-      "learning_rate": 9.262142254650476e-06,
-      "loss": 0.3913,
-      "step": 8797
-    },
-    {
-      "epoch": 0.5749950983595843,
-      "grad_norm": 0.44863948225975037,
-      "learning_rate": 9.261959672394212e-06,
-      "loss": 0.3917,
-      "step": 8798
-    },
-    {
-      "epoch": 0.5750604535651265,
-      "grad_norm": 0.4717997908592224,
-      "learning_rate": 9.261777069350892e-06,
-      "loss": 0.3964,
-      "step": 8799
-    },
-    {
-      "epoch": 0.5751258087706685,
-      "grad_norm": 0.4403414726257324,
-      "learning_rate": 9.26159444552141e-06,
-      "loss": 0.347,
-      "step": 8800
-    },
-    {
-      "epoch": 0.5751911639762107,
-      "grad_norm": 0.4735489785671234,
-      "learning_rate": 9.261411800906656e-06,
-      "loss": 0.3702,
-      "step": 8801
-    },
-    {
-      "epoch": 0.5752565191817528,
-      "grad_norm": 0.468436062335968,
-      "learning_rate": 9.261229135507519e-06,
-      "loss": 0.3949,
-      "step": 8802
-    },
-    {
-      "epoch": 0.575321874387295,
-      "grad_norm": 0.4591672122478485,
-      "learning_rate": 9.261046449324892e-06,
-      "loss": 0.3921,
-      "step": 8803
-    },
-    {
-      "epoch": 0.5753872295928371,
-      "grad_norm": 0.5436349511146545,
-      "learning_rate": 9.260863742359665e-06,
-      "loss": 0.3809,
-      "step": 8804
-    },
-    {
-      "epoch": 0.5754525847983792,
-      "grad_norm": 0.45340773463249207,
-      "learning_rate": 9.260681014612728e-06,
-      "loss": 0.4302,
-      "step": 8805
-    },
-    {
-      "epoch": 0.5755179400039213,
-      "grad_norm": 0.47672516107559204,
-      "learning_rate": 9.260498266084977e-06,
-      "loss": 0.4347,
-      "step": 8806
-    },
-    {
-      "epoch": 0.5755832952094634,
-      "grad_norm": 0.4370521605014801,
-      "learning_rate": 9.260315496777298e-06,
-      "loss": 0.3618,
-      "step": 8807
-    },
-    {
-      "epoch": 0.5756486504150056,
-      "grad_norm": 0.4761723279953003,
-      "learning_rate": 9.260132706690586e-06,
-      "loss": 0.3582,
-      "step": 8808
-    },
-    {
-      "epoch": 0.5757140056205476,
-      "grad_norm": 0.44726479053497314,
-      "learning_rate": 9.259949895825728e-06,
-      "loss": 0.3629,
-      "step": 8809
-    },
-    {
-      "epoch": 0.5757793608260898,
-      "grad_norm": 0.4513629972934723,
-      "learning_rate": 9.25976706418362e-06,
-      "loss": 0.3572,
-      "step": 8810
-    },
-    {
-      "epoch": 0.5758447160316319,
-      "grad_norm": 0.4882752299308777,
-      "learning_rate": 9.259584211765153e-06,
-      "loss": 0.4025,
-      "step": 8811
-    },
-    {
-      "epoch": 0.5759100712371741,
-      "grad_norm": 0.5161880254745483,
-      "learning_rate": 9.259401338571216e-06,
-      "loss": 0.4447,
-      "step": 8812
-    },
-    {
-      "epoch": 0.5759754264427162,
-      "grad_norm": 0.46421363949775696,
-      "learning_rate": 9.259218444602707e-06,
-      "loss": 0.3547,
-      "step": 8813
-    },
-    {
-      "epoch": 0.5760407816482583,
-      "grad_norm": 0.44217050075531006,
-      "learning_rate": 9.259035529860511e-06,
-      "loss": 0.3898,
-      "step": 8814
-    },
-    {
-      "epoch": 0.5761061368538004,
-      "grad_norm": 0.44459763169288635,
-      "learning_rate": 9.258852594345524e-06,
-      "loss": 0.3863,
-      "step": 8815
-    },
-    {
-      "epoch": 0.5761714920593425,
-      "grad_norm": 0.4126439392566681,
-      "learning_rate": 9.258669638058637e-06,
-      "loss": 0.3076,
-      "step": 8816
-    },
-    {
-      "epoch": 0.5762368472648847,
-      "grad_norm": 0.45681625604629517,
-      "learning_rate": 9.258486661000744e-06,
-      "loss": 0.3973,
-      "step": 8817
-    },
-    {
-      "epoch": 0.5763022024704267,
-      "grad_norm": 0.4271981418132782,
-      "learning_rate": 9.258303663172736e-06,
-      "loss": 0.3545,
-      "step": 8818
-    },
-    {
-      "epoch": 0.5763675576759689,
-      "grad_norm": 0.451896071434021,
-      "learning_rate": 9.258120644575505e-06,
-      "loss": 0.3967,
-      "step": 8819
-    },
-    {
-      "epoch": 0.576432912881511,
-      "grad_norm": 0.4758460223674774,
-      "learning_rate": 9.257937605209946e-06,
-      "loss": 0.4104,
-      "step": 8820
-    },
-    {
-      "epoch": 0.5764982680870532,
-      "grad_norm": 0.47071176767349243,
-      "learning_rate": 9.257754545076947e-06,
-      "loss": 0.3605,
-      "step": 8821
-    },
-    {
-      "epoch": 0.5765636232925953,
-      "grad_norm": 0.525622546672821,
-      "learning_rate": 9.257571464177408e-06,
-      "loss": 0.3436,
-      "step": 8822
-    },
-    {
-      "epoch": 0.5766289784981373,
-      "grad_norm": 0.5055347681045532,
-      "learning_rate": 9.257388362512215e-06,
-      "loss": 0.4054,
-      "step": 8823
-    },
-    {
-      "epoch": 0.5766943337036795,
-      "grad_norm": 0.46902206540107727,
-      "learning_rate": 9.257205240082265e-06,
-      "loss": 0.3875,
-      "step": 8824
-    },
-    {
-      "epoch": 0.5767596889092216,
-      "grad_norm": 0.44883987307548523,
-      "learning_rate": 9.257022096888451e-06,
-      "loss": 0.3924,
-      "step": 8825
-    },
-    {
-      "epoch": 0.5768250441147638,
-      "grad_norm": 0.46128007769584656,
-      "learning_rate": 9.256838932931664e-06,
-      "loss": 0.4266,
-      "step": 8826
-    },
-    {
-      "epoch": 0.5768903993203058,
-      "grad_norm": 0.4396490156650543,
-      "learning_rate": 9.2566557482128e-06,
-      "loss": 0.3625,
-      "step": 8827
-    },
-    {
-      "epoch": 0.576955754525848,
-      "grad_norm": 0.44678547978401184,
-      "learning_rate": 9.256472542732751e-06,
-      "loss": 0.3454,
-      "step": 8828
-    },
-    {
-      "epoch": 0.5770211097313901,
-      "grad_norm": 0.4691782295703888,
-      "learning_rate": 9.25628931649241e-06,
-      "loss": 0.3752,
-      "step": 8829
-    },
-    {
-      "epoch": 0.5770864649369323,
-      "grad_norm": 0.45704370737075806,
-      "learning_rate": 9.256106069492674e-06,
-      "loss": 0.4151,
-      "step": 8830
-    },
-    {
-      "epoch": 0.5771518201424743,
-      "grad_norm": 0.44608741998672485,
-      "learning_rate": 9.255922801734432e-06,
-      "loss": 0.3936,
-      "step": 8831
-    },
-    {
-      "epoch": 0.5772171753480164,
-      "grad_norm": 0.450510710477829,
-      "learning_rate": 9.255739513218581e-06,
-      "loss": 0.426,
-      "step": 8832
-    },
-    {
-      "epoch": 0.5772825305535586,
-      "grad_norm": 0.4668146073818207,
-      "learning_rate": 9.255556203946013e-06,
-      "loss": 0.4163,
-      "step": 8833
-    },
-    {
-      "epoch": 0.5773478857591007,
-      "grad_norm": 0.4742022752761841,
-      "learning_rate": 9.255372873917624e-06,
-      "loss": 0.405,
-      "step": 8834
-    },
-    {
-      "epoch": 0.5774132409646429,
-      "grad_norm": 0.41478219628334045,
-      "learning_rate": 9.255189523134307e-06,
-      "loss": 0.3562,
-      "step": 8835
-    },
-    {
-      "epoch": 0.5774785961701849,
-      "grad_norm": 0.45639121532440186,
-      "learning_rate": 9.255006151596956e-06,
-      "loss": 0.4122,
-      "step": 8836
-    },
-    {
-      "epoch": 0.5775439513757271,
-      "grad_norm": 0.48286187648773193,
-      "learning_rate": 9.254822759306468e-06,
-      "loss": 0.4057,
-      "step": 8837
-    },
-    {
-      "epoch": 0.5776093065812692,
-      "grad_norm": 0.4757295548915863,
-      "learning_rate": 9.254639346263734e-06,
-      "loss": 0.3853,
-      "step": 8838
-    },
-    {
-      "epoch": 0.5776746617868114,
-      "grad_norm": 0.45264384150505066,
-      "learning_rate": 9.25445591246965e-06,
-      "loss": 0.3896,
-      "step": 8839
-    },
-    {
-      "epoch": 0.5777400169923534,
-      "grad_norm": 0.44126197695732117,
-      "learning_rate": 9.25427245792511e-06,
-      "loss": 0.3699,
-      "step": 8840
-    },
-    {
-      "epoch": 0.5778053721978955,
-      "grad_norm": 0.4655766785144806,
-      "learning_rate": 9.25408898263101e-06,
-      "loss": 0.383,
-      "step": 8841
-    },
-    {
-      "epoch": 0.5778707274034377,
-      "grad_norm": 0.4766494631767273,
-      "learning_rate": 9.253905486588244e-06,
-      "loss": 0.4421,
-      "step": 8842
-    },
-    {
-      "epoch": 0.5779360826089798,
-      "grad_norm": 0.4508941173553467,
-      "learning_rate": 9.253721969797708e-06,
-      "loss": 0.3814,
-      "step": 8843
-    },
-    {
-      "epoch": 0.578001437814522,
-      "grad_norm": 0.4765235483646393,
-      "learning_rate": 9.253538432260296e-06,
-      "loss": 0.3952,
-      "step": 8844
-    },
-    {
-      "epoch": 0.578066793020064,
-      "grad_norm": 0.4527175724506378,
-      "learning_rate": 9.253354873976904e-06,
-      "loss": 0.3758,
-      "step": 8845
-    },
-    {
-      "epoch": 0.5781321482256062,
-      "grad_norm": 0.4632437527179718,
-      "learning_rate": 9.253171294948425e-06,
-      "loss": 0.3519,
-      "step": 8846
-    },
-    {
-      "epoch": 0.5781975034311483,
-      "grad_norm": 0.43092381954193115,
-      "learning_rate": 9.252987695175756e-06,
-      "loss": 0.3487,
-      "step": 8847
-    },
-    {
-      "epoch": 0.5782628586366904,
-      "grad_norm": 0.4423348903656006,
-      "learning_rate": 9.252804074659796e-06,
-      "loss": 0.4189,
-      "step": 8848
-    },
-    {
-      "epoch": 0.5783282138422325,
-      "grad_norm": 0.4612898826599121,
-      "learning_rate": 9.252620433401434e-06,
-      "loss": 0.4038,
-      "step": 8849
-    },
-    {
-      "epoch": 0.5783935690477746,
-      "grad_norm": 0.45391908288002014,
-      "learning_rate": 9.25243677140157e-06,
-      "loss": 0.3802,
-      "step": 8850
-    },
-    {
-      "epoch": 0.5784589242533168,
-      "grad_norm": 0.49826616048812866,
-      "learning_rate": 9.252253088661098e-06,
-      "loss": 0.4735,
-      "step": 8851
-    },
-    {
-      "epoch": 0.5785242794588589,
-      "grad_norm": 0.44494590163230896,
-      "learning_rate": 9.252069385180917e-06,
-      "loss": 0.3881,
-      "step": 8852
-    },
-    {
-      "epoch": 0.578589634664401,
-      "grad_norm": 0.40570759773254395,
-      "learning_rate": 9.251885660961917e-06,
-      "loss": 0.3431,
-      "step": 8853
-    },
-    {
-      "epoch": 0.5786549898699431,
-      "grad_norm": 0.4479600787162781,
-      "learning_rate": 9.251701916005e-06,
-      "loss": 0.3886,
-      "step": 8854
-    },
-    {
-      "epoch": 0.5787203450754853,
-      "grad_norm": 0.44045430421829224,
-      "learning_rate": 9.251518150311059e-06,
-      "loss": 0.3906,
-      "step": 8855
-    },
-    {
-      "epoch": 0.5787857002810274,
-      "grad_norm": 0.46660858392715454,
-      "learning_rate": 9.251334363880992e-06,
-      "loss": 0.3719,
-      "step": 8856
-    },
-    {
-      "epoch": 0.5788510554865695,
-      "grad_norm": 0.49139517545700073,
-      "learning_rate": 9.251150556715693e-06,
-      "loss": 0.3944,
-      "step": 8857
-    },
-    {
-      "epoch": 0.5789164106921116,
-      "grad_norm": 0.44801798462867737,
-      "learning_rate": 9.250966728816062e-06,
-      "loss": 0.3677,
-      "step": 8858
-    },
-    {
-      "epoch": 0.5789817658976537,
-      "grad_norm": 0.48235008120536804,
-      "learning_rate": 9.250782880182991e-06,
-      "loss": 0.4145,
-      "step": 8859
-    },
-    {
-      "epoch": 0.5790471211031959,
-      "grad_norm": 0.4536688029766083,
-      "learning_rate": 9.250599010817381e-06,
-      "loss": 0.4246,
-      "step": 8860
-    },
-    {
-      "epoch": 0.579112476308738,
-      "grad_norm": 0.4723179042339325,
-      "learning_rate": 9.250415120720128e-06,
-      "loss": 0.4068,
-      "step": 8861
-    },
-    {
-      "epoch": 0.5791778315142802,
-      "grad_norm": 0.44327279925346375,
-      "learning_rate": 9.250231209892126e-06,
-      "loss": 0.3807,
-      "step": 8862
-    },
-    {
-      "epoch": 0.5792431867198222,
-      "grad_norm": 0.45332780480384827,
-      "learning_rate": 9.250047278334276e-06,
-      "loss": 0.3733,
-      "step": 8863
-    },
-    {
-      "epoch": 0.5793085419253644,
-      "grad_norm": 0.44817081093788147,
-      "learning_rate": 9.24986332604747e-06,
-      "loss": 0.3986,
-      "step": 8864
-    },
-    {
-      "epoch": 0.5793738971309065,
-      "grad_norm": 0.4547162652015686,
-      "learning_rate": 9.24967935303261e-06,
-      "loss": 0.3977,
-      "step": 8865
-    },
-    {
-      "epoch": 0.5794392523364486,
-      "grad_norm": 0.48845168948173523,
-      "learning_rate": 9.249495359290592e-06,
-      "loss": 0.5043,
-      "step": 8866
-    },
-    {
-      "epoch": 0.5795046075419907,
-      "grad_norm": 0.4500035345554352,
-      "learning_rate": 9.249311344822313e-06,
-      "loss": 0.3944,
-      "step": 8867
-    },
-    {
-      "epoch": 0.5795699627475328,
-      "grad_norm": 0.47348421812057495,
-      "learning_rate": 9.24912730962867e-06,
-      "loss": 0.4127,
-      "step": 8868
-    },
-    {
-      "epoch": 0.579635317953075,
-      "grad_norm": 0.45496344566345215,
-      "learning_rate": 9.248943253710562e-06,
-      "loss": 0.4082,
-      "step": 8869
-    },
-    {
-      "epoch": 0.5797006731586171,
-      "grad_norm": 0.4660952091217041,
-      "learning_rate": 9.248759177068885e-06,
-      "loss": 0.3981,
-      "step": 8870
-    },
-    {
-      "epoch": 0.5797660283641592,
-      "grad_norm": 0.4442198872566223,
-      "learning_rate": 9.248575079704536e-06,
-      "loss": 0.38,
-      "step": 8871
-    },
-    {
-      "epoch": 0.5798313835697013,
-      "grad_norm": 0.46022218465805054,
-      "learning_rate": 9.248390961618416e-06,
-      "loss": 0.3583,
-      "step": 8872
-    },
-    {
-      "epoch": 0.5798967387752435,
-      "grad_norm": 0.5202801823616028,
-      "learning_rate": 9.248206822811422e-06,
-      "loss": 0.478,
-      "step": 8873
-    },
-    {
-      "epoch": 0.5799620939807856,
-      "grad_norm": 0.48663321137428284,
-      "learning_rate": 9.24802266328445e-06,
-      "loss": 0.4302,
-      "step": 8874
-    },
-    {
-      "epoch": 0.5800274491863276,
-      "grad_norm": 0.46237891912460327,
-      "learning_rate": 9.247838483038403e-06,
-      "loss": 0.4005,
-      "step": 8875
-    },
-    {
-      "epoch": 0.5800928043918698,
-      "grad_norm": 0.4653884470462799,
-      "learning_rate": 9.247654282074173e-06,
-      "loss": 0.3756,
-      "step": 8876
-    },
-    {
-      "epoch": 0.5801581595974119,
-      "grad_norm": 0.4650568962097168,
-      "learning_rate": 9.247470060392662e-06,
-      "loss": 0.4372,
-      "step": 8877
-    },
-    {
-      "epoch": 0.5802235148029541,
-      "grad_norm": 0.43710896372795105,
-      "learning_rate": 9.247285817994768e-06,
-      "loss": 0.3749,
-      "step": 8878
-    },
-    {
-      "epoch": 0.5802888700084962,
-      "grad_norm": 0.4592326581478119,
-      "learning_rate": 9.247101554881393e-06,
-      "loss": 0.4226,
-      "step": 8879
-    },
-    {
-      "epoch": 0.5803542252140383,
-      "grad_norm": 0.4464440643787384,
-      "learning_rate": 9.246917271053427e-06,
-      "loss": 0.3805,
-      "step": 8880
-    },
-    {
-      "epoch": 0.5804195804195804,
-      "grad_norm": 0.4540195167064667,
-      "learning_rate": 9.246732966511778e-06,
-      "loss": 0.3794,
-      "step": 8881
-    },
-    {
-      "epoch": 0.5804849356251225,
-      "grad_norm": 0.4573756754398346,
-      "learning_rate": 9.24654864125734e-06,
-      "loss": 0.3881,
-      "step": 8882
-    },
-    {
-      "epoch": 0.5805502908306647,
-      "grad_norm": 0.4530811309814453,
-      "learning_rate": 9.246364295291013e-06,
-      "loss": 0.3868,
-      "step": 8883
-    },
-    {
-      "epoch": 0.5806156460362067,
-      "grad_norm": 0.4787803888320923,
-      "learning_rate": 9.246179928613696e-06,
-      "loss": 0.3761,
-      "step": 8884
-    },
-    {
-      "epoch": 0.5806810012417489,
-      "grad_norm": 0.4541168510913849,
-      "learning_rate": 9.24599554122629e-06,
-      "loss": 0.3444,
-      "step": 8885
-    },
-    {
-      "epoch": 0.580746356447291,
-      "grad_norm": 0.4370727241039276,
-      "learning_rate": 9.245811133129692e-06,
-      "loss": 0.3431,
-      "step": 8886
-    },
-    {
-      "epoch": 0.5808117116528332,
-      "grad_norm": 0.4194653034210205,
-      "learning_rate": 9.245626704324802e-06,
-      "loss": 0.3466,
-      "step": 8887
-    },
-    {
-      "epoch": 0.5808770668583753,
-      "grad_norm": 0.4648381769657135,
-      "learning_rate": 9.24544225481252e-06,
-      "loss": 0.3866,
-      "step": 8888
-    },
-    {
-      "epoch": 0.5809424220639174,
-      "grad_norm": 0.4517500400543213,
-      "learning_rate": 9.245257784593744e-06,
-      "loss": 0.3616,
-      "step": 8889
-    },
-    {
-      "epoch": 0.5810077772694595,
-      "grad_norm": 0.45734766125679016,
-      "learning_rate": 9.245073293669376e-06,
-      "loss": 0.3746,
-      "step": 8890
-    },
-    {
-      "epoch": 0.5810731324750016,
-      "grad_norm": 0.4588703215122223,
-      "learning_rate": 9.244888782040315e-06,
-      "loss": 0.4092,
-      "step": 8891
-    },
-    {
-      "epoch": 0.5811384876805438,
-      "grad_norm": 0.46255338191986084,
-      "learning_rate": 9.24470424970746e-06,
-      "loss": 0.4092,
-      "step": 8892
-    },
-    {
-      "epoch": 0.5812038428860858,
-      "grad_norm": 0.47326725721359253,
-      "learning_rate": 9.244519696671712e-06,
-      "loss": 0.4156,
-      "step": 8893
-    },
-    {
-      "epoch": 0.581269198091628,
-      "grad_norm": 0.4809836745262146,
-      "learning_rate": 9.244335122933972e-06,
-      "loss": 0.4302,
-      "step": 8894
-    },
-    {
-      "epoch": 0.5813345532971701,
-      "grad_norm": 0.4386519491672516,
-      "learning_rate": 9.244150528495139e-06,
-      "loss": 0.3686,
-      "step": 8895
-    },
-    {
-      "epoch": 0.5813999085027123,
-      "grad_norm": 0.48370644450187683,
-      "learning_rate": 9.243965913356114e-06,
-      "loss": 0.4589,
-      "step": 8896
-    },
-    {
-      "epoch": 0.5814652637082544,
-      "grad_norm": 0.48877575993537903,
-      "learning_rate": 9.243781277517796e-06,
-      "loss": 0.4587,
-      "step": 8897
-    },
-    {
-      "epoch": 0.5815306189137965,
-      "grad_norm": 0.4940846860408783,
-      "learning_rate": 9.243596620981089e-06,
-      "loss": 0.4141,
-      "step": 8898
-    },
-    {
-      "epoch": 0.5815959741193386,
-      "grad_norm": 0.47367990016937256,
-      "learning_rate": 9.243411943746887e-06,
-      "loss": 0.4045,
-      "step": 8899
-    },
-    {
-      "epoch": 0.5816613293248807,
-      "grad_norm": 0.45567116141319275,
-      "learning_rate": 9.243227245816097e-06,
-      "loss": 0.3817,
-      "step": 8900
-    },
-    {
-      "epoch": 0.5817266845304229,
-      "grad_norm": 0.4673498272895813,
-      "learning_rate": 9.243042527189618e-06,
-      "loss": 0.4207,
-      "step": 8901
-    },
-    {
-      "epoch": 0.5817920397359649,
-      "grad_norm": 0.46828028559684753,
-      "learning_rate": 9.24285778786835e-06,
-      "loss": 0.3906,
-      "step": 8902
-    },
-    {
-      "epoch": 0.5818573949415071,
-      "grad_norm": 0.4705347716808319,
-      "learning_rate": 9.242673027853194e-06,
-      "loss": 0.3665,
-      "step": 8903
-    },
-    {
-      "epoch": 0.5819227501470492,
-      "grad_norm": 0.4391564428806305,
-      "learning_rate": 9.242488247145055e-06,
-      "loss": 0.3892,
-      "step": 8904
-    },
-    {
-      "epoch": 0.5819881053525914,
-      "grad_norm": 0.4890103340148926,
-      "learning_rate": 9.242303445744828e-06,
-      "loss": 0.4675,
-      "step": 8905
-    },
-    {
-      "epoch": 0.5820534605581335,
-      "grad_norm": 0.4659532606601715,
-      "learning_rate": 9.242118623653418e-06,
-      "loss": 0.4274,
-      "step": 8906
-    },
-    {
-      "epoch": 0.5821188157636755,
-      "grad_norm": 0.4810665547847748,
-      "learning_rate": 9.241933780871726e-06,
-      "loss": 0.3743,
-      "step": 8907
-    },
-    {
-      "epoch": 0.5821841709692177,
-      "grad_norm": 0.4475446343421936,
-      "learning_rate": 9.241748917400654e-06,
-      "loss": 0.3464,
-      "step": 8908
-    },
-    {
-      "epoch": 0.5822495261747598,
-      "grad_norm": 0.4875513017177582,
-      "learning_rate": 9.241564033241105e-06,
-      "loss": 0.3873,
-      "step": 8909
-    },
-    {
-      "epoch": 0.582314881380302,
-      "grad_norm": 0.4478124976158142,
-      "learning_rate": 9.241379128393974e-06,
-      "loss": 0.3583,
-      "step": 8910
-    },
-    {
-      "epoch": 0.582380236585844,
-      "grad_norm": 0.4584704041481018,
-      "learning_rate": 9.24119420286017e-06,
-      "loss": 0.3975,
-      "step": 8911
-    },
-    {
-      "epoch": 0.5824455917913862,
-      "grad_norm": 0.42169299721717834,
-      "learning_rate": 9.241009256640595e-06,
-      "loss": 0.345,
-      "step": 8912
-    },
-    {
-      "epoch": 0.5825109469969283,
-      "grad_norm": 0.4419157803058624,
-      "learning_rate": 9.240824289736146e-06,
-      "loss": 0.3586,
-      "step": 8913
-    },
-    {
-      "epoch": 0.5825763022024705,
-      "grad_norm": 0.45704102516174316,
-      "learning_rate": 9.240639302147727e-06,
-      "loss": 0.3507,
-      "step": 8914
-    },
-    {
-      "epoch": 0.5826416574080125,
-      "grad_norm": 0.41148611903190613,
-      "learning_rate": 9.240454293876243e-06,
-      "loss": 0.3338,
-      "step": 8915
-    },
-    {
-      "epoch": 0.5827070126135546,
-      "grad_norm": 0.4673536717891693,
-      "learning_rate": 9.240269264922592e-06,
-      "loss": 0.3775,
-      "step": 8916
-    },
-    {
-      "epoch": 0.5827723678190968,
-      "grad_norm": 0.4489016532897949,
-      "learning_rate": 9.240084215287682e-06,
-      "loss": 0.3975,
-      "step": 8917
-    },
-    {
-      "epoch": 0.5828377230246389,
-      "grad_norm": 0.453816682100296,
-      "learning_rate": 9.23989914497241e-06,
-      "loss": 0.4052,
-      "step": 8918
-    },
-    {
-      "epoch": 0.5829030782301811,
-      "grad_norm": 0.43146708607673645,
-      "learning_rate": 9.239714053977682e-06,
-      "loss": 0.331,
-      "step": 8919
-    },
-    {
-      "epoch": 0.5829684334357231,
-      "grad_norm": 0.45014113187789917,
-      "learning_rate": 9.2395289423044e-06,
-      "loss": 0.3412,
-      "step": 8920
-    },
-    {
-      "epoch": 0.5830337886412653,
-      "grad_norm": 0.4720291495323181,
-      "learning_rate": 9.239343809953464e-06,
-      "loss": 0.364,
-      "step": 8921
-    },
-    {
-      "epoch": 0.5830991438468074,
-      "grad_norm": 0.45932865142822266,
-      "learning_rate": 9.239158656925784e-06,
-      "loss": 0.3985,
-      "step": 8922
-    },
-    {
-      "epoch": 0.5831644990523496,
-      "grad_norm": 0.4266372323036194,
-      "learning_rate": 9.238973483222256e-06,
-      "loss": 0.3164,
-      "step": 8923
-    },
-    {
-      "epoch": 0.5832298542578916,
-      "grad_norm": 0.43437832593917847,
-      "learning_rate": 9.238788288843786e-06,
-      "loss": 0.3351,
-      "step": 8924
-    },
-    {
-      "epoch": 0.5832952094634337,
-      "grad_norm": 0.4799064099788666,
-      "learning_rate": 9.238603073791277e-06,
-      "loss": 0.3891,
-      "step": 8925
-    },
-    {
-      "epoch": 0.5833605646689759,
-      "grad_norm": 0.47381865978240967,
-      "learning_rate": 9.238417838065632e-06,
-      "loss": 0.3977,
-      "step": 8926
-    },
-    {
-      "epoch": 0.583425919874518,
-      "grad_norm": 0.45300525426864624,
-      "learning_rate": 9.238232581667757e-06,
-      "loss": 0.3913,
-      "step": 8927
-    },
-    {
-      "epoch": 0.5834912750800602,
-      "grad_norm": 0.4447530210018158,
-      "learning_rate": 9.23804730459855e-06,
-      "loss": 0.3714,
-      "step": 8928
-    },
-    {
-      "epoch": 0.5835566302856022,
-      "grad_norm": 0.452631413936615,
-      "learning_rate": 9.237862006858922e-06,
-      "loss": 0.4009,
-      "step": 8929
-    },
-    {
-      "epoch": 0.5836219854911444,
-      "grad_norm": 0.49760764837265015,
-      "learning_rate": 9.237676688449771e-06,
-      "loss": 0.3932,
-      "step": 8930
-    },
-    {
-      "epoch": 0.5836873406966865,
-      "grad_norm": 0.4606693387031555,
-      "learning_rate": 9.237491349372002e-06,
-      "loss": 0.3827,
-      "step": 8931
-    },
-    {
-      "epoch": 0.5837526959022286,
-      "grad_norm": 0.4380769729614258,
-      "learning_rate": 9.23730598962652e-06,
-      "loss": 0.3647,
-      "step": 8932
-    },
-    {
-      "epoch": 0.5838180511077707,
-      "grad_norm": 0.5160754919052124,
-      "learning_rate": 9.237120609214229e-06,
-      "loss": 0.4451,
-      "step": 8933
-    },
-    {
-      "epoch": 0.5838834063133128,
-      "grad_norm": 0.4520348310470581,
-      "learning_rate": 9.236935208136033e-06,
-      "loss": 0.4205,
-      "step": 8934
-    },
-    {
-      "epoch": 0.583948761518855,
-      "grad_norm": 0.452730268239975,
-      "learning_rate": 9.236749786392835e-06,
-      "loss": 0.4088,
-      "step": 8935
-    },
-    {
-      "epoch": 0.5840141167243971,
-      "grad_norm": 0.43783777952194214,
-      "learning_rate": 9.236564343985544e-06,
-      "loss": 0.3854,
-      "step": 8936
-    },
-    {
-      "epoch": 0.5840794719299393,
-      "grad_norm": 0.5229910016059875,
-      "learning_rate": 9.236378880915058e-06,
-      "loss": 0.4549,
-      "step": 8937
-    },
-    {
-      "epoch": 0.5841448271354813,
-      "grad_norm": 0.47226646542549133,
-      "learning_rate": 9.236193397182284e-06,
-      "loss": 0.3156,
-      "step": 8938
-    },
-    {
-      "epoch": 0.5842101823410235,
-      "grad_norm": 0.46665552258491516,
-      "learning_rate": 9.236007892788129e-06,
-      "loss": 0.3517,
-      "step": 8939
-    },
-    {
-      "epoch": 0.5842755375465656,
-      "grad_norm": 0.44889697432518005,
-      "learning_rate": 9.235822367733495e-06,
-      "loss": 0.3474,
-      "step": 8940
-    },
-    {
-      "epoch": 0.5843408927521077,
-      "grad_norm": 0.46805232763290405,
-      "learning_rate": 9.235636822019289e-06,
-      "loss": 0.4556,
-      "step": 8941
-    },
-    {
-      "epoch": 0.5844062479576498,
-      "grad_norm": 0.457375168800354,
-      "learning_rate": 9.235451255646415e-06,
-      "loss": 0.4091,
-      "step": 8942
-    },
-    {
-      "epoch": 0.5844716031631919,
-      "grad_norm": 0.4257332980632782,
-      "learning_rate": 9.235265668615778e-06,
-      "loss": 0.3381,
-      "step": 8943
-    },
-    {
-      "epoch": 0.5845369583687341,
-      "grad_norm": 0.4659012258052826,
-      "learning_rate": 9.23508006092828e-06,
-      "loss": 0.3906,
-      "step": 8944
-    },
-    {
-      "epoch": 0.5846023135742762,
-      "grad_norm": 0.432725727558136,
-      "learning_rate": 9.234894432584833e-06,
-      "loss": 0.3374,
-      "step": 8945
-    },
-    {
-      "epoch": 0.5846676687798184,
-      "grad_norm": 0.4783468246459961,
-      "learning_rate": 9.234708783586337e-06,
-      "loss": 0.4302,
-      "step": 8946
-    },
-    {
-      "epoch": 0.5847330239853604,
-      "grad_norm": 0.45998015999794006,
-      "learning_rate": 9.2345231139337e-06,
-      "loss": 0.4149,
-      "step": 8947
-    },
-    {
-      "epoch": 0.5847983791909026,
-      "grad_norm": 0.4069378674030304,
-      "learning_rate": 9.234337423627826e-06,
-      "loss": 0.302,
-      "step": 8948
-    },
-    {
-      "epoch": 0.5848637343964447,
-      "grad_norm": 0.438897043466568,
-      "learning_rate": 9.234151712669622e-06,
-      "loss": 0.3593,
-      "step": 8949
-    },
-    {
-      "epoch": 0.5849290896019868,
-      "grad_norm": 0.4755827784538269,
-      "learning_rate": 9.233965981059993e-06,
-      "loss": 0.4013,
-      "step": 8950
-    },
-    {
-      "epoch": 0.5849944448075289,
-      "grad_norm": 0.42014962434768677,
-      "learning_rate": 9.233780228799847e-06,
-      "loss": 0.3524,
-      "step": 8951
-    },
-    {
-      "epoch": 0.585059800013071,
-      "grad_norm": 0.4726681113243103,
-      "learning_rate": 9.233594455890084e-06,
-      "loss": 0.4245,
-      "step": 8952
-    },
-    {
-      "epoch": 0.5851251552186132,
-      "grad_norm": 0.442443311214447,
-      "learning_rate": 9.233408662331618e-06,
-      "loss": 0.3687,
-      "step": 8953
-    },
-    {
-      "epoch": 0.5851905104241553,
-      "grad_norm": 0.4709050953388214,
-      "learning_rate": 9.233222848125349e-06,
-      "loss": 0.3857,
-      "step": 8954
-    },
-    {
-      "epoch": 0.5852558656296974,
-      "grad_norm": 0.5213170051574707,
-      "learning_rate": 9.233037013272189e-06,
-      "loss": 0.4629,
-      "step": 8955
-    },
-    {
-      "epoch": 0.5853212208352395,
-      "grad_norm": 0.44747892022132874,
-      "learning_rate": 9.232851157773038e-06,
-      "loss": 0.3957,
-      "step": 8956
-    },
-    {
-      "epoch": 0.5853865760407817,
-      "grad_norm": 0.44545143842697144,
-      "learning_rate": 9.232665281628805e-06,
-      "loss": 0.4221,
-      "step": 8957
-    },
-    {
-      "epoch": 0.5854519312463238,
-      "grad_norm": 0.49116387963294983,
-      "learning_rate": 9.232479384840398e-06,
-      "loss": 0.4635,
-      "step": 8958
-    },
-    {
-      "epoch": 0.5855172864518658,
-      "grad_norm": 0.4828069806098938,
-      "learning_rate": 9.232293467408724e-06,
-      "loss": 0.3756,
-      "step": 8959
-    },
-    {
-      "epoch": 0.585582641657408,
-      "grad_norm": 0.48043420910835266,
-      "learning_rate": 9.232107529334687e-06,
-      "loss": 0.4416,
-      "step": 8960
-    },
-    {
-      "epoch": 0.5856479968629501,
-      "grad_norm": 0.43332362174987793,
-      "learning_rate": 9.231921570619196e-06,
-      "loss": 0.3349,
-      "step": 8961
-    },
-    {
-      "epoch": 0.5857133520684923,
-      "grad_norm": 0.48748546838760376,
-      "learning_rate": 9.231735591263157e-06,
-      "loss": 0.4007,
-      "step": 8962
-    },
-    {
-      "epoch": 0.5857787072740344,
-      "grad_norm": 0.54531329870224,
-      "learning_rate": 9.231549591267478e-06,
-      "loss": 0.4761,
-      "step": 8963
-    },
-    {
-      "epoch": 0.5858440624795765,
-      "grad_norm": 0.47210320830345154,
-      "learning_rate": 9.231363570633066e-06,
-      "loss": 0.4098,
-      "step": 8964
-    },
-    {
-      "epoch": 0.5859094176851186,
-      "grad_norm": 0.44633516669273376,
-      "learning_rate": 9.231177529360826e-06,
-      "loss": 0.3824,
-      "step": 8965
-    },
-    {
-      "epoch": 0.5859747728906607,
-      "grad_norm": 0.4741937816143036,
-      "learning_rate": 9.23099146745167e-06,
-      "loss": 0.3827,
-      "step": 8966
-    },
-    {
-      "epoch": 0.5860401280962029,
-      "grad_norm": 0.4759604036808014,
-      "learning_rate": 9.230805384906501e-06,
-      "loss": 0.3943,
-      "step": 8967
-    },
-    {
-      "epoch": 0.586105483301745,
-      "grad_norm": 0.4685611128807068,
-      "learning_rate": 9.23061928172623e-06,
-      "loss": 0.4034,
-      "step": 8968
-    },
-    {
-      "epoch": 0.5861708385072871,
-      "grad_norm": 0.46542221307754517,
-      "learning_rate": 9.230433157911762e-06,
-      "loss": 0.3717,
-      "step": 8969
-    },
-    {
-      "epoch": 0.5862361937128292,
-      "grad_norm": 0.4728912115097046,
-      "learning_rate": 9.230247013464006e-06,
-      "loss": 0.4212,
-      "step": 8970
-    },
-    {
-      "epoch": 0.5863015489183714,
-      "grad_norm": 0.4357195198535919,
-      "learning_rate": 9.23006084838387e-06,
-      "loss": 0.3659,
-      "step": 8971
-    },
-    {
-      "epoch": 0.5863669041239135,
-      "grad_norm": 0.62079918384552,
-      "learning_rate": 9.229874662672262e-06,
-      "loss": 0.3539,
-      "step": 8972
-    },
-    {
-      "epoch": 0.5864322593294556,
-      "grad_norm": 0.4491276144981384,
-      "learning_rate": 9.22968845633009e-06,
-      "loss": 0.392,
-      "step": 8973
-    },
-    {
-      "epoch": 0.5864976145349977,
-      "grad_norm": 0.4587726294994354,
-      "learning_rate": 9.229502229358263e-06,
-      "loss": 0.3987,
-      "step": 8974
-    },
-    {
-      "epoch": 0.5865629697405398,
-      "grad_norm": 0.4528464674949646,
-      "learning_rate": 9.229315981757686e-06,
-      "loss": 0.3658,
-      "step": 8975
-    },
-    {
-      "epoch": 0.586628324946082,
-      "grad_norm": 0.4658888876438141,
-      "learning_rate": 9.229129713529271e-06,
-      "loss": 0.3783,
-      "step": 8976
-    },
-    {
-      "epoch": 0.586693680151624,
-      "grad_norm": 0.467284619808197,
-      "learning_rate": 9.228943424673925e-06,
-      "loss": 0.3779,
-      "step": 8977
-    },
-    {
-      "epoch": 0.5867590353571662,
-      "grad_norm": 0.4674723148345947,
-      "learning_rate": 9.228757115192559e-06,
-      "loss": 0.397,
-      "step": 8978
-    },
-    {
-      "epoch": 0.5868243905627083,
-      "grad_norm": 0.48817887902259827,
-      "learning_rate": 9.228570785086077e-06,
-      "loss": 0.4502,
-      "step": 8979
-    },
-    {
-      "epoch": 0.5868897457682505,
-      "grad_norm": 0.4686819612979889,
-      "learning_rate": 9.228384434355391e-06,
-      "loss": 0.3895,
-      "step": 8980
-    },
-    {
-      "epoch": 0.5869551009737926,
-      "grad_norm": 0.45721468329429626,
-      "learning_rate": 9.22819806300141e-06,
-      "loss": 0.4018,
-      "step": 8981
-    },
-    {
-      "epoch": 0.5870204561793347,
-      "grad_norm": 0.46017321944236755,
-      "learning_rate": 9.228011671025041e-06,
-      "loss": 0.3517,
-      "step": 8982
-    },
-    {
-      "epoch": 0.5870858113848768,
-      "grad_norm": 0.49878695607185364,
-      "learning_rate": 9.227825258427194e-06,
-      "loss": 0.4122,
-      "step": 8983
-    },
-    {
-      "epoch": 0.5871511665904189,
-      "grad_norm": 0.4482685327529907,
-      "learning_rate": 9.22763882520878e-06,
-      "loss": 0.3812,
-      "step": 8984
-    },
-    {
-      "epoch": 0.5872165217959611,
-      "grad_norm": 0.4409409761428833,
-      "learning_rate": 9.227452371370706e-06,
-      "loss": 0.3757,
-      "step": 8985
-    },
-    {
-      "epoch": 0.5872818770015031,
-      "grad_norm": 0.46814683079719543,
-      "learning_rate": 9.227265896913884e-06,
-      "loss": 0.3781,
-      "step": 8986
-    },
-    {
-      "epoch": 0.5873472322070453,
-      "grad_norm": 0.44779160618782043,
-      "learning_rate": 9.22707940183922e-06,
-      "loss": 0.375,
-      "step": 8987
-    },
-    {
-      "epoch": 0.5874125874125874,
-      "grad_norm": 0.4393285810947418,
-      "learning_rate": 9.226892886147625e-06,
-      "loss": 0.3697,
-      "step": 8988
-    },
-    {
-      "epoch": 0.5874779426181296,
-      "grad_norm": 0.4500323235988617,
-      "learning_rate": 9.22670634984001e-06,
-      "loss": 0.3948,
-      "step": 8989
-    },
-    {
-      "epoch": 0.5875432978236717,
-      "grad_norm": 0.4728008806705475,
-      "learning_rate": 9.226519792917284e-06,
-      "loss": 0.3592,
-      "step": 8990
-    },
-    {
-      "epoch": 0.5876086530292137,
-      "grad_norm": 0.45040658116340637,
-      "learning_rate": 9.226333215380357e-06,
-      "loss": 0.3843,
-      "step": 8991
-    },
-    {
-      "epoch": 0.5876740082347559,
-      "grad_norm": 0.47039368748664856,
-      "learning_rate": 9.226146617230138e-06,
-      "loss": 0.3772,
-      "step": 8992
-    },
-    {
-      "epoch": 0.587739363440298,
-      "grad_norm": 0.4496113359928131,
-      "learning_rate": 9.225959998467538e-06,
-      "loss": 0.3652,
-      "step": 8993
-    },
-    {
-      "epoch": 0.5878047186458402,
-      "grad_norm": 0.45141497254371643,
-      "learning_rate": 9.225773359093467e-06,
-      "loss": 0.3887,
-      "step": 8994
-    },
-    {
-      "epoch": 0.5878700738513822,
-      "grad_norm": 0.4537845253944397,
-      "learning_rate": 9.225586699108835e-06,
-      "loss": 0.3885,
-      "step": 8995
-    },
-    {
-      "epoch": 0.5879354290569244,
-      "grad_norm": 0.46830037236213684,
-      "learning_rate": 9.225400018514554e-06,
-      "loss": 0.3788,
-      "step": 8996
-    },
-    {
-      "epoch": 0.5880007842624665,
-      "grad_norm": 0.4767850339412689,
-      "learning_rate": 9.225213317311532e-06,
-      "loss": 0.4265,
-      "step": 8997
-    },
-    {
-      "epoch": 0.5880661394680087,
-      "grad_norm": 0.4945718050003052,
-      "learning_rate": 9.225026595500683e-06,
-      "loss": 0.4166,
-      "step": 8998
-    },
-    {
-      "epoch": 0.5881314946735507,
-      "grad_norm": 0.4451162815093994,
-      "learning_rate": 9.224839853082912e-06,
-      "loss": 0.379,
-      "step": 8999
-    },
-    {
-      "epoch": 0.5881968498790928,
-      "grad_norm": 0.37969881296157837,
-      "learning_rate": 9.224653090059136e-06,
-      "loss": 0.2887,
-      "step": 9000
-    },
-    {
-      "epoch": 0.588262205084635,
-      "grad_norm": 0.42612314224243164,
-      "learning_rate": 9.224466306430264e-06,
-      "loss": 0.4071,
-      "step": 9001
-    },
-    {
-      "epoch": 0.5883275602901771,
-      "grad_norm": 0.47960203886032104,
-      "learning_rate": 9.224279502197205e-06,
-      "loss": 0.3894,
-      "step": 9002
-    },
-    {
-      "epoch": 0.5883929154957193,
-      "grad_norm": 0.4567774832248688,
-      "learning_rate": 9.224092677360872e-06,
-      "loss": 0.3386,
-      "step": 9003
-    },
-    {
-      "epoch": 0.5884582707012613,
-      "grad_norm": 0.4322126805782318,
-      "learning_rate": 9.223905831922174e-06,
-      "loss": 0.3557,
-      "step": 9004
-    },
-    {
-      "epoch": 0.5885236259068035,
-      "grad_norm": 0.4723544418811798,
-      "learning_rate": 9.223718965882026e-06,
-      "loss": 0.4089,
-      "step": 9005
-    },
-    {
-      "epoch": 0.5885889811123456,
-      "grad_norm": 0.4543313980102539,
-      "learning_rate": 9.223532079241336e-06,
-      "loss": 0.4024,
-      "step": 9006
-    },
-    {
-      "epoch": 0.5886543363178878,
-      "grad_norm": 0.41359829902648926,
-      "learning_rate": 9.223345172001018e-06,
-      "loss": 0.3174,
-      "step": 9007
-    },
-    {
-      "epoch": 0.5887196915234298,
-      "grad_norm": 0.4719434678554535,
-      "learning_rate": 9.223158244161982e-06,
-      "loss": 0.4227,
-      "step": 9008
-    },
-    {
-      "epoch": 0.5887850467289719,
-      "grad_norm": 0.47162926197052,
-      "learning_rate": 9.22297129572514e-06,
-      "loss": 0.4045,
-      "step": 9009
-    },
-    {
-      "epoch": 0.5888504019345141,
-      "grad_norm": 0.4442574977874756,
-      "learning_rate": 9.222784326691404e-06,
-      "loss": 0.3603,
-      "step": 9010
-    },
-    {
-      "epoch": 0.5889157571400562,
-      "grad_norm": 0.45001548528671265,
-      "learning_rate": 9.222597337061686e-06,
-      "loss": 0.3363,
-      "step": 9011
-    },
-    {
-      "epoch": 0.5889811123455984,
-      "grad_norm": 0.48516085743904114,
-      "learning_rate": 9.2224103268369e-06,
-      "loss": 0.4837,
-      "step": 9012
-    },
-    {
-      "epoch": 0.5890464675511404,
-      "grad_norm": 0.4140465557575226,
-      "learning_rate": 9.222223296017953e-06,
-      "loss": 0.3525,
-      "step": 9013
-    },
-    {
-      "epoch": 0.5891118227566826,
-      "grad_norm": 0.4737408459186554,
-      "learning_rate": 9.22203624460576e-06,
-      "loss": 0.4215,
-      "step": 9014
-    },
-    {
-      "epoch": 0.5891771779622247,
-      "grad_norm": 0.46026068925857544,
-      "learning_rate": 9.221849172601236e-06,
-      "loss": 0.4183,
-      "step": 9015
-    },
-    {
-      "epoch": 0.5892425331677668,
-      "grad_norm": 0.4173218309879303,
-      "learning_rate": 9.22166208000529e-06,
-      "loss": 0.3561,
-      "step": 9016
-    },
-    {
-      "epoch": 0.5893078883733089,
-      "grad_norm": 0.4410687983036041,
-      "learning_rate": 9.221474966818836e-06,
-      "loss": 0.3888,
-      "step": 9017
-    },
-    {
-      "epoch": 0.589373243578851,
-      "grad_norm": 0.45688557624816895,
-      "learning_rate": 9.221287833042784e-06,
-      "loss": 0.4263,
-      "step": 9018
-    },
-    {
-      "epoch": 0.5894385987843932,
-      "grad_norm": 0.4256860017776489,
-      "learning_rate": 9.221100678678051e-06,
-      "loss": 0.3654,
-      "step": 9019
-    },
-    {
-      "epoch": 0.5895039539899353,
-      "grad_norm": 0.4474024772644043,
-      "learning_rate": 9.220913503725548e-06,
-      "loss": 0.367,
-      "step": 9020
-    },
-    {
-      "epoch": 0.5895693091954775,
-      "grad_norm": 0.46031689643859863,
-      "learning_rate": 9.220726308186186e-06,
-      "loss": 0.3868,
-      "step": 9021
-    },
-    {
-      "epoch": 0.5896346644010195,
-      "grad_norm": 0.4893234074115753,
-      "learning_rate": 9.220539092060881e-06,
-      "loss": 0.3974,
-      "step": 9022
-    },
-    {
-      "epoch": 0.5897000196065617,
-      "grad_norm": 0.44660866260528564,
-      "learning_rate": 9.220351855350543e-06,
-      "loss": 0.3825,
-      "step": 9023
-    },
-    {
-      "epoch": 0.5897653748121038,
-      "grad_norm": 0.5068235397338867,
-      "learning_rate": 9.220164598056088e-06,
-      "loss": 0.4405,
-      "step": 9024
-    },
-    {
-      "epoch": 0.5898307300176459,
-      "grad_norm": 0.45615455508232117,
-      "learning_rate": 9.219977320178429e-06,
-      "loss": 0.3599,
-      "step": 9025
-    },
-    {
-      "epoch": 0.589896085223188,
-      "grad_norm": 0.4334595501422882,
-      "learning_rate": 9.219790021718477e-06,
-      "loss": 0.3544,
-      "step": 9026
-    },
-    {
-      "epoch": 0.5899614404287301,
-      "grad_norm": 0.5093353986740112,
-      "learning_rate": 9.219602702677148e-06,
-      "loss": 0.4572,
-      "step": 9027
-    },
-    {
-      "epoch": 0.5900267956342723,
-      "grad_norm": 0.4460398256778717,
-      "learning_rate": 9.219415363055355e-06,
-      "loss": 0.3717,
-      "step": 9028
-    },
-    {
-      "epoch": 0.5900921508398144,
-      "grad_norm": 0.44923314452171326,
-      "learning_rate": 9.219228002854011e-06,
-      "loss": 0.3922,
-      "step": 9029
-    },
-    {
-      "epoch": 0.5901575060453566,
-      "grad_norm": 0.45911285281181335,
-      "learning_rate": 9.219040622074031e-06,
-      "loss": 0.3819,
-      "step": 9030
-    },
-    {
-      "epoch": 0.5902228612508986,
-      "grad_norm": 0.44375184178352356,
-      "learning_rate": 9.218853220716329e-06,
-      "loss": 0.3863,
-      "step": 9031
-    },
-    {
-      "epoch": 0.5902882164564408,
-      "grad_norm": 0.4464426338672638,
-      "learning_rate": 9.218665798781817e-06,
-      "loss": 0.3765,
-      "step": 9032
-    },
-    {
-      "epoch": 0.5903535716619829,
-      "grad_norm": 0.4511531889438629,
-      "learning_rate": 9.21847835627141e-06,
-      "loss": 0.3684,
-      "step": 9033
-    },
-    {
-      "epoch": 0.590418926867525,
-      "grad_norm": 0.44682687520980835,
-      "learning_rate": 9.218290893186023e-06,
-      "loss": 0.4265,
-      "step": 9034
-    },
-    {
-      "epoch": 0.5904842820730671,
-      "grad_norm": 0.46610260009765625,
-      "learning_rate": 9.218103409526571e-06,
-      "loss": 0.4444,
-      "step": 9035
-    },
-    {
-      "epoch": 0.5905496372786092,
-      "grad_norm": 0.46726521849632263,
-      "learning_rate": 9.217915905293965e-06,
-      "loss": 0.3995,
-      "step": 9036
-    },
-    {
-      "epoch": 0.5906149924841514,
-      "grad_norm": 0.46023795008659363,
-      "learning_rate": 9.217728380489124e-06,
-      "loss": 0.3712,
-      "step": 9037
-    },
-    {
-      "epoch": 0.5906803476896935,
-      "grad_norm": 0.4558008909225464,
-      "learning_rate": 9.217540835112961e-06,
-      "loss": 0.3887,
-      "step": 9038
-    },
-    {
-      "epoch": 0.5907457028952356,
-      "grad_norm": 0.4319905936717987,
-      "learning_rate": 9.217353269166388e-06,
-      "loss": 0.342,
-      "step": 9039
-    },
-    {
-      "epoch": 0.5908110581007777,
-      "grad_norm": 0.488342821598053,
-      "learning_rate": 9.217165682650323e-06,
-      "loss": 0.3755,
-      "step": 9040
-    },
-    {
-      "epoch": 0.5908764133063199,
-      "grad_norm": 0.4749290943145752,
-      "learning_rate": 9.216978075565681e-06,
-      "loss": 0.4095,
-      "step": 9041
-    },
-    {
-      "epoch": 0.590941768511862,
-      "grad_norm": 0.44274985790252686,
-      "learning_rate": 9.216790447913376e-06,
-      "loss": 0.3559,
-      "step": 9042
-    },
-    {
-      "epoch": 0.591007123717404,
-      "grad_norm": 0.4593304693698883,
-      "learning_rate": 9.216602799694324e-06,
-      "loss": 0.3815,
-      "step": 9043
-    },
-    {
-      "epoch": 0.5910724789229462,
-      "grad_norm": 0.4413754642009735,
-      "learning_rate": 9.216415130909438e-06,
-      "loss": 0.3791,
-      "step": 9044
-    },
-    {
-      "epoch": 0.5911378341284883,
-      "grad_norm": 0.5065827369689941,
-      "learning_rate": 9.216227441559633e-06,
-      "loss": 0.4888,
-      "step": 9045
-    },
-    {
-      "epoch": 0.5912031893340305,
-      "grad_norm": 0.4619041085243225,
-      "learning_rate": 9.216039731645828e-06,
-      "loss": 0.3981,
-      "step": 9046
-    },
-    {
-      "epoch": 0.5912685445395726,
-      "grad_norm": 0.48635900020599365,
-      "learning_rate": 9.215852001168937e-06,
-      "loss": 0.4214,
-      "step": 9047
-    },
-    {
-      "epoch": 0.5913338997451147,
-      "grad_norm": 0.40824180841445923,
-      "learning_rate": 9.215664250129875e-06,
-      "loss": 0.3628,
-      "step": 9048
-    },
-    {
-      "epoch": 0.5913992549506568,
-      "grad_norm": 0.44060084223747253,
-      "learning_rate": 9.215476478529557e-06,
-      "loss": 0.3846,
-      "step": 9049
-    },
-    {
-      "epoch": 0.5914646101561989,
-      "grad_norm": 0.4784317910671234,
-      "learning_rate": 9.2152886863689e-06,
-      "loss": 0.4068,
-      "step": 9050
-    },
-    {
-      "epoch": 0.5915299653617411,
-      "grad_norm": 0.43177342414855957,
-      "learning_rate": 9.21510087364882e-06,
-      "loss": 0.3494,
-      "step": 9051
-    },
-    {
-      "epoch": 0.5915953205672831,
-      "grad_norm": 0.44881895184516907,
-      "learning_rate": 9.214913040370233e-06,
-      "loss": 0.3568,
-      "step": 9052
-    },
-    {
-      "epoch": 0.5916606757728253,
-      "grad_norm": 0.45579829812049866,
-      "learning_rate": 9.214725186534057e-06,
-      "loss": 0.4396,
-      "step": 9053
-    },
-    {
-      "epoch": 0.5917260309783674,
-      "grad_norm": 0.4880259335041046,
-      "learning_rate": 9.214537312141203e-06,
-      "loss": 0.4365,
-      "step": 9054
-    },
-    {
-      "epoch": 0.5917913861839096,
-      "grad_norm": 0.4429958760738373,
-      "learning_rate": 9.214349417192592e-06,
-      "loss": 0.3688,
-      "step": 9055
-    },
-    {
-      "epoch": 0.5918567413894517,
-      "grad_norm": 0.4387759566307068,
-      "learning_rate": 9.214161501689138e-06,
-      "loss": 0.3733,
-      "step": 9056
-    },
-    {
-      "epoch": 0.5919220965949938,
-      "grad_norm": 0.4535773694515228,
-      "learning_rate": 9.21397356563176e-06,
-      "loss": 0.4017,
-      "step": 9057
-    },
-    {
-      "epoch": 0.5919874518005359,
-      "grad_norm": 0.49486440420150757,
-      "learning_rate": 9.21378560902137e-06,
-      "loss": 0.4568,
-      "step": 9058
-    },
-    {
-      "epoch": 0.592052807006078,
-      "grad_norm": 0.46412143111228943,
-      "learning_rate": 9.21359763185889e-06,
-      "loss": 0.399,
-      "step": 9059
-    },
-    {
-      "epoch": 0.5921181622116202,
-      "grad_norm": 0.46160778403282166,
-      "learning_rate": 9.213409634145236e-06,
-      "loss": 0.4126,
-      "step": 9060
-    },
-    {
-      "epoch": 0.5921835174171622,
-      "grad_norm": 0.4399988055229187,
-      "learning_rate": 9.213221615881321e-06,
-      "loss": 0.3995,
-      "step": 9061
-    },
-    {
-      "epoch": 0.5922488726227044,
-      "grad_norm": 0.44348669052124023,
-      "learning_rate": 9.213033577068065e-06,
-      "loss": 0.3928,
-      "step": 9062
-    },
-    {
-      "epoch": 0.5923142278282465,
-      "grad_norm": 0.46584224700927734,
-      "learning_rate": 9.212845517706386e-06,
-      "loss": 0.4136,
-      "step": 9063
-    },
-    {
-      "epoch": 0.5923795830337887,
-      "grad_norm": 0.439632385969162,
-      "learning_rate": 9.212657437797198e-06,
-      "loss": 0.3576,
-      "step": 9064
-    },
-    {
-      "epoch": 0.5924449382393308,
-      "grad_norm": 0.4497891366481781,
-      "learning_rate": 9.212469337341422e-06,
-      "loss": 0.3729,
-      "step": 9065
-    },
-    {
-      "epoch": 0.5925102934448729,
-      "grad_norm": 0.43726828694343567,
-      "learning_rate": 9.212281216339975e-06,
-      "loss": 0.3618,
-      "step": 9066
-    },
-    {
-      "epoch": 0.592575648650415,
-      "grad_norm": 0.5010649561882019,
-      "learning_rate": 9.21209307479377e-06,
-      "loss": 0.4298,
-      "step": 9067
-    },
-    {
-      "epoch": 0.5926410038559571,
-      "grad_norm": 0.4569692015647888,
-      "learning_rate": 9.21190491270373e-06,
-      "loss": 0.3834,
-      "step": 9068
-    },
-    {
-      "epoch": 0.5927063590614993,
-      "grad_norm": 0.42680060863494873,
-      "learning_rate": 9.21171673007077e-06,
-      "loss": 0.3463,
-      "step": 9069
-    },
-    {
-      "epoch": 0.5927717142670413,
-      "grad_norm": 0.4854176938533783,
-      "learning_rate": 9.211528526895808e-06,
-      "loss": 0.4097,
-      "step": 9070
-    },
-    {
-      "epoch": 0.5928370694725835,
-      "grad_norm": 0.4481373727321625,
-      "learning_rate": 9.211340303179764e-06,
-      "loss": 0.3551,
-      "step": 9071
-    },
-    {
-      "epoch": 0.5929024246781256,
-      "grad_norm": 0.44821399450302124,
-      "learning_rate": 9.211152058923552e-06,
-      "loss": 0.3415,
-      "step": 9072
-    },
-    {
-      "epoch": 0.5929677798836678,
-      "grad_norm": 0.4925774037837982,
-      "learning_rate": 9.210963794128094e-06,
-      "loss": 0.4952,
-      "step": 9073
-    },
-    {
-      "epoch": 0.5930331350892099,
-      "grad_norm": 0.43282485008239746,
-      "learning_rate": 9.210775508794306e-06,
-      "loss": 0.3667,
-      "step": 9074
-    },
-    {
-      "epoch": 0.5930984902947519,
-      "grad_norm": 0.43615296483039856,
-      "learning_rate": 9.210587202923106e-06,
-      "loss": 0.3415,
-      "step": 9075
-    },
-    {
-      "epoch": 0.5931638455002941,
-      "grad_norm": 0.4541088342666626,
-      "learning_rate": 9.210398876515417e-06,
-      "loss": 0.4235,
-      "step": 9076
-    },
-    {
-      "epoch": 0.5932292007058362,
-      "grad_norm": 0.47272148728370667,
-      "learning_rate": 9.21021052957215e-06,
-      "loss": 0.4112,
-      "step": 9077
-    },
-    {
-      "epoch": 0.5932945559113784,
-      "grad_norm": 0.4851337671279907,
-      "learning_rate": 9.21002216209423e-06,
-      "loss": 0.4457,
-      "step": 9078
-    },
-    {
-      "epoch": 0.5933599111169204,
-      "grad_norm": 0.4812110960483551,
-      "learning_rate": 9.209833774082573e-06,
-      "loss": 0.4372,
-      "step": 9079
-    },
-    {
-      "epoch": 0.5934252663224626,
-      "grad_norm": 0.43391725420951843,
-      "learning_rate": 9.209645365538099e-06,
-      "loss": 0.3508,
-      "step": 9080
-    },
-    {
-      "epoch": 0.5934906215280047,
-      "grad_norm": 0.4428863823413849,
-      "learning_rate": 9.209456936461725e-06,
-      "loss": 0.3487,
-      "step": 9081
-    },
-    {
-      "epoch": 0.5935559767335469,
-      "grad_norm": 0.4442507326602936,
-      "learning_rate": 9.209268486854373e-06,
-      "loss": 0.3763,
-      "step": 9082
-    },
-    {
-      "epoch": 0.593621331939089,
-      "grad_norm": 0.41742801666259766,
-      "learning_rate": 9.209080016716957e-06,
-      "loss": 0.3612,
-      "step": 9083
-    },
-    {
-      "epoch": 0.593686687144631,
-      "grad_norm": 0.4596056640148163,
-      "learning_rate": 9.208891526050403e-06,
-      "loss": 0.4116,
-      "step": 9084
-    },
-    {
-      "epoch": 0.5937520423501732,
-      "grad_norm": 0.4446125030517578,
-      "learning_rate": 9.208703014855627e-06,
-      "loss": 0.3938,
-      "step": 9085
-    },
-    {
-      "epoch": 0.5938173975557153,
-      "grad_norm": 0.43721136450767517,
-      "learning_rate": 9.208514483133546e-06,
-      "loss": 0.3712,
-      "step": 9086
-    },
-    {
-      "epoch": 0.5938827527612575,
-      "grad_norm": 0.4259456396102905,
-      "learning_rate": 9.208325930885082e-06,
-      "loss": 0.3404,
-      "step": 9087
-    },
-    {
-      "epoch": 0.5939481079667995,
-      "grad_norm": 0.4259801506996155,
-      "learning_rate": 9.208137358111156e-06,
-      "loss": 0.3758,
-      "step": 9088
-    },
-    {
-      "epoch": 0.5940134631723417,
-      "grad_norm": 0.46871837973594666,
-      "learning_rate": 9.207948764812686e-06,
-      "loss": 0.3545,
-      "step": 9089
-    },
-    {
-      "epoch": 0.5940788183778838,
-      "grad_norm": 0.4888988733291626,
-      "learning_rate": 9.207760150990593e-06,
-      "loss": 0.4493,
-      "step": 9090
-    },
-    {
-      "epoch": 0.594144173583426,
-      "grad_norm": 0.4476315975189209,
-      "learning_rate": 9.207571516645795e-06,
-      "loss": 0.3781,
-      "step": 9091
-    },
-    {
-      "epoch": 0.594209528788968,
-      "grad_norm": 0.4323274493217468,
-      "learning_rate": 9.207382861779213e-06,
-      "loss": 0.3454,
-      "step": 9092
-    },
-    {
-      "epoch": 0.5942748839945101,
-      "grad_norm": 0.45750895142555237,
-      "learning_rate": 9.207194186391766e-06,
-      "loss": 0.3713,
-      "step": 9093
-    },
-    {
-      "epoch": 0.5943402392000523,
-      "grad_norm": 0.5198811292648315,
-      "learning_rate": 9.207005490484376e-06,
-      "loss": 0.3899,
-      "step": 9094
-    },
-    {
-      "epoch": 0.5944055944055944,
-      "grad_norm": 0.5023952126502991,
-      "learning_rate": 9.206816774057964e-06,
-      "loss": 0.4233,
-      "step": 9095
-    },
-    {
-      "epoch": 0.5944709496111366,
-      "grad_norm": 0.4601942002773285,
-      "learning_rate": 9.206628037113447e-06,
-      "loss": 0.3839,
-      "step": 9096
-    },
-    {
-      "epoch": 0.5945363048166786,
-      "grad_norm": 0.5169584155082703,
-      "learning_rate": 9.206439279651752e-06,
-      "loss": 0.4461,
-      "step": 9097
-    },
-    {
-      "epoch": 0.5946016600222208,
-      "grad_norm": 0.466145783662796,
-      "learning_rate": 9.206250501673791e-06,
-      "loss": 0.3887,
-      "step": 9098
-    },
-    {
-      "epoch": 0.5946670152277629,
-      "grad_norm": 0.44113680720329285,
-      "learning_rate": 9.206061703180491e-06,
-      "loss": 0.3796,
-      "step": 9099
-    },
-    {
-      "epoch": 0.594732370433305,
-      "grad_norm": 0.5136611461639404,
-      "learning_rate": 9.20587288417277e-06,
-      "loss": 0.4524,
-      "step": 9100
-    },
-    {
-      "epoch": 0.5947977256388471,
-      "grad_norm": 0.4596930742263794,
-      "learning_rate": 9.205684044651552e-06,
-      "loss": 0.3715,
-      "step": 9101
-    },
-    {
-      "epoch": 0.5948630808443892,
-      "grad_norm": 0.4330142140388489,
-      "learning_rate": 9.205495184617754e-06,
-      "loss": 0.3459,
-      "step": 9102
-    },
-    {
-      "epoch": 0.5949284360499314,
-      "grad_norm": 0.4458671808242798,
-      "learning_rate": 9.2053063040723e-06,
-      "loss": 0.3339,
-      "step": 9103
-    },
-    {
-      "epoch": 0.5949937912554735,
-      "grad_norm": 0.44250407814979553,
-      "learning_rate": 9.20511740301611e-06,
-      "loss": 0.3793,
-      "step": 9104
-    },
-    {
-      "epoch": 0.5950591464610157,
-      "grad_norm": 0.42624783515930176,
-      "learning_rate": 9.204928481450106e-06,
-      "loss": 0.3605,
-      "step": 9105
-    },
-    {
-      "epoch": 0.5951245016665577,
-      "grad_norm": 0.4709337055683136,
-      "learning_rate": 9.204739539375207e-06,
-      "loss": 0.4033,
-      "step": 9106
-    },
-    {
-      "epoch": 0.5951898568720999,
-      "grad_norm": 0.4625800549983978,
-      "learning_rate": 9.204550576792339e-06,
-      "loss": 0.3996,
-      "step": 9107
-    },
-    {
-      "epoch": 0.595255212077642,
-      "grad_norm": 0.46261706948280334,
-      "learning_rate": 9.204361593702421e-06,
-      "loss": 0.4234,
-      "step": 9108
-    },
-    {
-      "epoch": 0.595320567283184,
-      "grad_norm": 0.4749184250831604,
-      "learning_rate": 9.204172590106374e-06,
-      "loss": 0.4274,
-      "step": 9109
-    },
-    {
-      "epoch": 0.5953859224887262,
-      "grad_norm": 0.46007803082466125,
-      "learning_rate": 9.20398356600512e-06,
-      "loss": 0.4046,
-      "step": 9110
-    },
-    {
-      "epoch": 0.5954512776942683,
-      "grad_norm": 0.4391649663448334,
-      "learning_rate": 9.203794521399584e-06,
-      "loss": 0.3892,
-      "step": 9111
-    },
-    {
-      "epoch": 0.5955166328998105,
-      "grad_norm": 0.46174532175064087,
-      "learning_rate": 9.203605456290685e-06,
-      "loss": 0.421,
-      "step": 9112
-    },
-    {
-      "epoch": 0.5955819881053526,
-      "grad_norm": 0.43722209334373474,
-      "learning_rate": 9.203416370679346e-06,
-      "loss": 0.4242,
-      "step": 9113
-    },
-    {
-      "epoch": 0.5956473433108948,
-      "grad_norm": 0.4329133629798889,
-      "learning_rate": 9.20322726456649e-06,
-      "loss": 0.3767,
-      "step": 9114
-    },
-    {
-      "epoch": 0.5957126985164368,
-      "grad_norm": 0.6488149166107178,
-      "learning_rate": 9.203038137953036e-06,
-      "loss": 0.3931,
-      "step": 9115
-    },
-    {
-      "epoch": 0.595778053721979,
-      "grad_norm": 0.4795227646827698,
-      "learning_rate": 9.20284899083991e-06,
-      "loss": 0.3944,
-      "step": 9116
-    },
-    {
-      "epoch": 0.5958434089275211,
-      "grad_norm": 0.43959277868270874,
-      "learning_rate": 9.202659823228035e-06,
-      "loss": 0.3887,
-      "step": 9117
-    },
-    {
-      "epoch": 0.5959087641330632,
-      "grad_norm": 0.4309938848018646,
-      "learning_rate": 9.20247063511833e-06,
-      "loss": 0.357,
-      "step": 9118
-    },
-    {
-      "epoch": 0.5959741193386053,
-      "grad_norm": 0.46862471103668213,
-      "learning_rate": 9.20228142651172e-06,
-      "loss": 0.4304,
-      "step": 9119
-    },
-    {
-      "epoch": 0.5960394745441474,
-      "grad_norm": 0.42521026730537415,
-      "learning_rate": 9.202092197409129e-06,
-      "loss": 0.3472,
-      "step": 9120
-    },
-    {
-      "epoch": 0.5961048297496896,
-      "grad_norm": 0.46872183680534363,
-      "learning_rate": 9.201902947811478e-06,
-      "loss": 0.4417,
-      "step": 9121
-    },
-    {
-      "epoch": 0.5961701849552317,
-      "grad_norm": 0.4616164565086365,
-      "learning_rate": 9.201713677719692e-06,
-      "loss": 0.373,
-      "step": 9122
-    },
-    {
-      "epoch": 0.5962355401607738,
-      "grad_norm": 0.4247811734676361,
-      "learning_rate": 9.20152438713469e-06,
-      "loss": 0.3607,
-      "step": 9123
-    },
-    {
-      "epoch": 0.5963008953663159,
-      "grad_norm": 0.42037004232406616,
-      "learning_rate": 9.201335076057401e-06,
-      "loss": 0.3131,
-      "step": 9124
-    },
-    {
-      "epoch": 0.5963662505718581,
-      "grad_norm": 0.47550836205482483,
-      "learning_rate": 9.201145744488744e-06,
-      "loss": 0.4208,
-      "step": 9125
-    },
-    {
-      "epoch": 0.5964316057774002,
-      "grad_norm": 0.4449000358581543,
-      "learning_rate": 9.200956392429643e-06,
-      "loss": 0.3642,
-      "step": 9126
-    },
-    {
-      "epoch": 0.5964969609829422,
-      "grad_norm": 0.4218422472476959,
-      "learning_rate": 9.200767019881023e-06,
-      "loss": 0.3638,
-      "step": 9127
-    },
-    {
-      "epoch": 0.5965623161884844,
-      "grad_norm": 0.47300466895103455,
-      "learning_rate": 9.200577626843807e-06,
-      "loss": 0.4264,
-      "step": 9128
-    },
-    {
-      "epoch": 0.5966276713940265,
-      "grad_norm": 0.4390457570552826,
-      "learning_rate": 9.200388213318918e-06,
-      "loss": 0.3791,
-      "step": 9129
-    },
-    {
-      "epoch": 0.5966930265995687,
-      "grad_norm": 0.45548853278160095,
-      "learning_rate": 9.200198779307281e-06,
-      "loss": 0.4235,
-      "step": 9130
-    },
-    {
-      "epoch": 0.5967583818051108,
-      "grad_norm": 0.41298967599868774,
-      "learning_rate": 9.20000932480982e-06,
-      "loss": 0.355,
-      "step": 9131
-    },
-    {
-      "epoch": 0.5968237370106529,
-      "grad_norm": 0.46050941944122314,
-      "learning_rate": 9.199819849827458e-06,
-      "loss": 0.4065,
-      "step": 9132
-    },
-    {
-      "epoch": 0.596889092216195,
-      "grad_norm": 0.4615412950515747,
-      "learning_rate": 9.19963035436112e-06,
-      "loss": 0.4242,
-      "step": 9133
-    },
-    {
-      "epoch": 0.5969544474217371,
-      "grad_norm": 0.423781156539917,
-      "learning_rate": 9.199440838411729e-06,
-      "loss": 0.3362,
-      "step": 9134
-    },
-    {
-      "epoch": 0.5970198026272793,
-      "grad_norm": 0.48159605264663696,
-      "learning_rate": 9.19925130198021e-06,
-      "loss": 0.3747,
-      "step": 9135
-    },
-    {
-      "epoch": 0.5970851578328213,
-      "grad_norm": 0.43601229786872864,
-      "learning_rate": 9.199061745067488e-06,
-      "loss": 0.3831,
-      "step": 9136
-    },
-    {
-      "epoch": 0.5971505130383635,
-      "grad_norm": 0.44950616359710693,
-      "learning_rate": 9.198872167674488e-06,
-      "loss": 0.3983,
-      "step": 9137
-    },
-    {
-      "epoch": 0.5972158682439056,
-      "grad_norm": 0.4608835279941559,
-      "learning_rate": 9.198682569802135e-06,
-      "loss": 0.4323,
-      "step": 9138
-    },
-    {
-      "epoch": 0.5972812234494478,
-      "grad_norm": 0.4185160994529724,
-      "learning_rate": 9.19849295145135e-06,
-      "loss": 0.3578,
-      "step": 9139
-    },
-    {
-      "epoch": 0.5973465786549899,
-      "grad_norm": 0.4930746555328369,
-      "learning_rate": 9.198303312623062e-06,
-      "loss": 0.4701,
-      "step": 9140
-    },
-    {
-      "epoch": 0.597411933860532,
-      "grad_norm": 0.46320006251335144,
-      "learning_rate": 9.198113653318193e-06,
-      "loss": 0.3934,
-      "step": 9141
-    },
-    {
-      "epoch": 0.5974772890660741,
-      "grad_norm": 0.43301257491111755,
-      "learning_rate": 9.19792397353767e-06,
-      "loss": 0.3805,
-      "step": 9142
-    },
-    {
-      "epoch": 0.5975426442716162,
-      "grad_norm": 0.4329968988895416,
-      "learning_rate": 9.197734273282417e-06,
-      "loss": 0.3671,
-      "step": 9143
-    },
-    {
-      "epoch": 0.5976079994771584,
-      "grad_norm": 0.44462713599205017,
-      "learning_rate": 9.197544552553361e-06,
-      "loss": 0.3635,
-      "step": 9144
-    },
-    {
-      "epoch": 0.5976733546827004,
-      "grad_norm": 0.48472943902015686,
-      "learning_rate": 9.197354811351424e-06,
-      "loss": 0.4656,
-      "step": 9145
-    },
-    {
-      "epoch": 0.5977387098882426,
-      "grad_norm": 0.4545370936393738,
-      "learning_rate": 9.197165049677535e-06,
-      "loss": 0.4047,
-      "step": 9146
-    },
-    {
-      "epoch": 0.5978040650937847,
-      "grad_norm": 0.4666481018066406,
-      "learning_rate": 9.196975267532617e-06,
-      "loss": 0.4337,
-      "step": 9147
-    },
-    {
-      "epoch": 0.5978694202993269,
-      "grad_norm": 0.46113815903663635,
-      "learning_rate": 9.1967854649176e-06,
-      "loss": 0.3764,
-      "step": 9148
-    },
-    {
-      "epoch": 0.597934775504869,
-      "grad_norm": 0.4551337659358978,
-      "learning_rate": 9.196595641833402e-06,
-      "loss": 0.3893,
-      "step": 9149
-    },
-    {
-      "epoch": 0.5980001307104111,
-      "grad_norm": 0.44840535521507263,
-      "learning_rate": 9.196405798280956e-06,
-      "loss": 0.3531,
-      "step": 9150
-    },
-    {
-      "epoch": 0.5980654859159532,
-      "grad_norm": 0.42207643389701843,
-      "learning_rate": 9.196215934261184e-06,
-      "loss": 0.3542,
-      "step": 9151
-    },
-    {
-      "epoch": 0.5981308411214953,
-      "grad_norm": 0.45021456480026245,
-      "learning_rate": 9.196026049775013e-06,
-      "loss": 0.3805,
-      "step": 9152
-    },
-    {
-      "epoch": 0.5981961963270375,
-      "grad_norm": 0.43388471007347107,
-      "learning_rate": 9.195836144823368e-06,
-      "loss": 0.3322,
-      "step": 9153
-    },
-    {
-      "epoch": 0.5982615515325795,
-      "grad_norm": 0.49034833908081055,
-      "learning_rate": 9.19564621940718e-06,
-      "loss": 0.4043,
-      "step": 9154
-    },
-    {
-      "epoch": 0.5983269067381217,
-      "grad_norm": 0.44993117451667786,
-      "learning_rate": 9.195456273527369e-06,
-      "loss": 0.3794,
-      "step": 9155
-    },
-    {
-      "epoch": 0.5983922619436638,
-      "grad_norm": 0.4515838325023651,
-      "learning_rate": 9.195266307184866e-06,
-      "loss": 0.3834,
-      "step": 9156
-    },
-    {
-      "epoch": 0.598457617149206,
-      "grad_norm": 0.4347997307777405,
-      "learning_rate": 9.195076320380596e-06,
-      "loss": 0.3463,
-      "step": 9157
-    },
-    {
-      "epoch": 0.598522972354748,
-      "grad_norm": 0.4673284888267517,
-      "learning_rate": 9.194886313115482e-06,
-      "loss": 0.3996,
-      "step": 9158
-    },
-    {
-      "epoch": 0.5985883275602901,
-      "grad_norm": 0.4702996611595154,
-      "learning_rate": 9.194696285390458e-06,
-      "loss": 0.3775,
-      "step": 9159
-    },
-    {
-      "epoch": 0.5986536827658323,
-      "grad_norm": 0.4738173186779022,
-      "learning_rate": 9.194506237206447e-06,
-      "loss": 0.3839,
-      "step": 9160
-    },
-    {
-      "epoch": 0.5987190379713744,
-      "grad_norm": 0.4467931389808655,
-      "learning_rate": 9.194316168564374e-06,
-      "loss": 0.377,
-      "step": 9161
-    },
-    {
-      "epoch": 0.5987843931769166,
-      "grad_norm": 0.43088576197624207,
-      "learning_rate": 9.194126079465169e-06,
-      "loss": 0.354,
-      "step": 9162
-    },
-    {
-      "epoch": 0.5988497483824586,
-      "grad_norm": 0.4529799222946167,
-      "learning_rate": 9.193935969909758e-06,
-      "loss": 0.4335,
-      "step": 9163
-    },
-    {
-      "epoch": 0.5989151035880008,
-      "grad_norm": 0.44353675842285156,
-      "learning_rate": 9.19374583989907e-06,
-      "loss": 0.3328,
-      "step": 9164
-    },
-    {
-      "epoch": 0.5989804587935429,
-      "grad_norm": 0.4639052450656891,
-      "learning_rate": 9.193555689434026e-06,
-      "loss": 0.4115,
-      "step": 9165
-    },
-    {
-      "epoch": 0.5990458139990851,
-      "grad_norm": 0.4736430048942566,
-      "learning_rate": 9.193365518515562e-06,
-      "loss": 0.4268,
-      "step": 9166
-    },
-    {
-      "epoch": 0.5991111692046271,
-      "grad_norm": 0.44214147329330444,
-      "learning_rate": 9.1931753271446e-06,
-      "loss": 0.3593,
-      "step": 9167
-    },
-    {
-      "epoch": 0.5991765244101692,
-      "grad_norm": 0.4463392496109009,
-      "learning_rate": 9.192985115322071e-06,
-      "loss": 0.3657,
-      "step": 9168
-    },
-    {
-      "epoch": 0.5992418796157114,
-      "grad_norm": 0.45829901099205017,
-      "learning_rate": 9.1927948830489e-06,
-      "loss": 0.3923,
-      "step": 9169
-    },
-    {
-      "epoch": 0.5993072348212535,
-      "grad_norm": 0.4527924954891205,
-      "learning_rate": 9.192604630326017e-06,
-      "loss": 0.3811,
-      "step": 9170
-    },
-    {
-      "epoch": 0.5993725900267957,
-      "grad_norm": 0.43915653228759766,
-      "learning_rate": 9.192414357154346e-06,
-      "loss": 0.3959,
-      "step": 9171
-    },
-    {
-      "epoch": 0.5994379452323377,
-      "grad_norm": 0.4366587698459625,
-      "learning_rate": 9.19222406353482e-06,
-      "loss": 0.3524,
-      "step": 9172
-    },
-    {
-      "epoch": 0.5995033004378799,
-      "grad_norm": 0.4628661870956421,
-      "learning_rate": 9.192033749468365e-06,
-      "loss": 0.3899,
-      "step": 9173
-    },
-    {
-      "epoch": 0.599568655643422,
-      "grad_norm": 0.46450647711753845,
-      "learning_rate": 9.191843414955908e-06,
-      "loss": 0.3521,
-      "step": 9174
-    },
-    {
-      "epoch": 0.5996340108489642,
-      "grad_norm": 0.40289822220802307,
-      "learning_rate": 9.191653059998378e-06,
-      "loss": 0.3255,
-      "step": 9175
-    },
-    {
-      "epoch": 0.5996993660545062,
-      "grad_norm": 0.48823636770248413,
-      "learning_rate": 9.191462684596707e-06,
-      "loss": 0.4141,
-      "step": 9176
-    },
-    {
-      "epoch": 0.5997647212600483,
-      "grad_norm": 0.47811248898506165,
-      "learning_rate": 9.191272288751817e-06,
-      "loss": 0.4024,
-      "step": 9177
-    },
-    {
-      "epoch": 0.5998300764655905,
-      "grad_norm": 0.4462015926837921,
-      "learning_rate": 9.191081872464641e-06,
-      "loss": 0.4219,
-      "step": 9178
-    },
-    {
-      "epoch": 0.5998954316711326,
-      "grad_norm": 0.47095420956611633,
-      "learning_rate": 9.190891435736107e-06,
-      "loss": 0.4276,
-      "step": 9179
-    },
-    {
-      "epoch": 0.5999607868766748,
-      "grad_norm": 0.46351829171180725,
-      "learning_rate": 9.190700978567144e-06,
-      "loss": 0.3848,
-      "step": 9180
-    },
-    {
-      "epoch": 0.6000261420822168,
-      "grad_norm": 0.4518841803073883,
-      "learning_rate": 9.19051050095868e-06,
-      "loss": 0.4166,
-      "step": 9181
-    },
-    {
-      "epoch": 0.600091497287759,
-      "grad_norm": 0.4417521059513092,
-      "learning_rate": 9.190320002911644e-06,
-      "loss": 0.3937,
-      "step": 9182
-    },
-    {
-      "epoch": 0.6001568524933011,
-      "grad_norm": 0.43405234813690186,
-      "learning_rate": 9.190129484426967e-06,
-      "loss": 0.3838,
-      "step": 9183
-    },
-    {
-      "epoch": 0.6002222076988432,
-      "grad_norm": 0.4737446904182434,
-      "learning_rate": 9.189938945505576e-06,
-      "loss": 0.4147,
-      "step": 9184
-    },
-    {
-      "epoch": 0.6002875629043853,
-      "grad_norm": 0.428546279668808,
-      "learning_rate": 9.189748386148403e-06,
-      "loss": 0.3824,
-      "step": 9185
-    },
-    {
-      "epoch": 0.6003529181099274,
-      "grad_norm": 0.449856698513031,
-      "learning_rate": 9.189557806356374e-06,
-      "loss": 0.3733,
-      "step": 9186
-    },
-    {
-      "epoch": 0.6004182733154696,
-      "grad_norm": 0.41608351469039917,
-      "learning_rate": 9.18936720613042e-06,
-      "loss": 0.3484,
-      "step": 9187
-    },
-    {
-      "epoch": 0.6004836285210117,
-      "grad_norm": 0.4258476793766022,
-      "learning_rate": 9.189176585471471e-06,
-      "loss": 0.3671,
-      "step": 9188
-    },
-    {
-      "epoch": 0.6005489837265539,
-      "grad_norm": 0.46372899413108826,
-      "learning_rate": 9.188985944380457e-06,
-      "loss": 0.3733,
-      "step": 9189
-    },
-    {
-      "epoch": 0.6006143389320959,
-      "grad_norm": 0.45023807883262634,
-      "learning_rate": 9.188795282858307e-06,
-      "loss": 0.4154,
-      "step": 9190
-    },
-    {
-      "epoch": 0.6006796941376381,
-      "grad_norm": 0.39405637979507446,
-      "learning_rate": 9.188604600905952e-06,
-      "loss": 0.2964,
-      "step": 9191
-    },
-    {
-      "epoch": 0.6007450493431802,
-      "grad_norm": 0.49776163697242737,
-      "learning_rate": 9.18841389852432e-06,
-      "loss": 0.4051,
-      "step": 9192
-    },
-    {
-      "epoch": 0.6008104045487223,
-      "grad_norm": 0.4413900673389435,
-      "learning_rate": 9.188223175714343e-06,
-      "loss": 0.3814,
-      "step": 9193
-    },
-    {
-      "epoch": 0.6008757597542644,
-      "grad_norm": 0.45196112990379333,
-      "learning_rate": 9.18803243247695e-06,
-      "loss": 0.3959,
-      "step": 9194
-    },
-    {
-      "epoch": 0.6009411149598065,
-      "grad_norm": 0.44850677251815796,
-      "learning_rate": 9.187841668813074e-06,
-      "loss": 0.4009,
-      "step": 9195
-    },
-    {
-      "epoch": 0.6010064701653487,
-      "grad_norm": 0.4772990345954895,
-      "learning_rate": 9.187650884723642e-06,
-      "loss": 0.4528,
-      "step": 9196
-    },
-    {
-      "epoch": 0.6010718253708908,
-      "grad_norm": 0.43468621373176575,
-      "learning_rate": 9.187460080209585e-06,
-      "loss": 0.3953,
-      "step": 9197
-    },
-    {
-      "epoch": 0.601137180576433,
-      "grad_norm": 0.48254331946372986,
-      "learning_rate": 9.187269255271835e-06,
-      "loss": 0.4459,
-      "step": 9198
-    },
-    {
-      "epoch": 0.601202535781975,
-      "grad_norm": 0.45564985275268555,
-      "learning_rate": 9.187078409911322e-06,
-      "loss": 0.3997,
-      "step": 9199
-    },
-    {
-      "epoch": 0.6012678909875172,
-      "grad_norm": 0.4532981514930725,
-      "learning_rate": 9.18688754412898e-06,
-      "loss": 0.4067,
-      "step": 9200
-    },
-    {
-      "epoch": 0.6013332461930593,
-      "grad_norm": 0.430982381105423,
-      "learning_rate": 9.186696657925734e-06,
-      "loss": 0.3572,
-      "step": 9201
-    },
-    {
-      "epoch": 0.6013986013986014,
-      "grad_norm": 0.4672890603542328,
-      "learning_rate": 9.18650575130252e-06,
-      "loss": 0.4089,
-      "step": 9202
-    },
-    {
-      "epoch": 0.6014639566041435,
-      "grad_norm": 0.4574647545814514,
-      "learning_rate": 9.186314824260265e-06,
-      "loss": 0.3526,
-      "step": 9203
-    },
-    {
-      "epoch": 0.6015293118096856,
-      "grad_norm": 0.4146009683609009,
-      "learning_rate": 9.186123876799902e-06,
-      "loss": 0.3522,
-      "step": 9204
-    },
-    {
-      "epoch": 0.6015946670152278,
-      "grad_norm": 0.47112053632736206,
-      "learning_rate": 9.185932908922364e-06,
-      "loss": 0.3958,
-      "step": 9205
-    },
-    {
-      "epoch": 0.6016600222207699,
-      "grad_norm": 0.49419665336608887,
-      "learning_rate": 9.185741920628582e-06,
-      "loss": 0.4321,
-      "step": 9206
-    },
-    {
-      "epoch": 0.601725377426312,
-      "grad_norm": 0.44286206364631653,
-      "learning_rate": 9.185550911919485e-06,
-      "loss": 0.4233,
-      "step": 9207
-    },
-    {
-      "epoch": 0.6017907326318541,
-      "grad_norm": 0.4228411614894867,
-      "learning_rate": 9.185359882796006e-06,
-      "loss": 0.3486,
-      "step": 9208
-    },
-    {
-      "epoch": 0.6018560878373963,
-      "grad_norm": 0.4780553877353668,
-      "learning_rate": 9.185168833259077e-06,
-      "loss": 0.4282,
-      "step": 9209
-    },
-    {
-      "epoch": 0.6019214430429384,
-      "grad_norm": 0.4292789399623871,
-      "learning_rate": 9.18497776330963e-06,
-      "loss": 0.3883,
-      "step": 9210
-    },
-    {
-      "epoch": 0.6019867982484804,
-      "grad_norm": 0.4652314782142639,
-      "learning_rate": 9.184786672948599e-06,
-      "loss": 0.382,
-      "step": 9211
-    },
-    {
-      "epoch": 0.6020521534540226,
-      "grad_norm": 0.41927310824394226,
-      "learning_rate": 9.18459556217691e-06,
-      "loss": 0.3811,
-      "step": 9212
-    },
-    {
-      "epoch": 0.6021175086595647,
-      "grad_norm": 0.42894890904426575,
-      "learning_rate": 9.184404430995499e-06,
-      "loss": 0.3664,
-      "step": 9213
-    },
-    {
-      "epoch": 0.6021828638651069,
-      "grad_norm": 0.4503488540649414,
-      "learning_rate": 9.184213279405302e-06,
-      "loss": 0.3721,
-      "step": 9214
-    },
-    {
-      "epoch": 0.602248219070649,
-      "grad_norm": 0.4546520709991455,
-      "learning_rate": 9.184022107407243e-06,
-      "loss": 0.402,
-      "step": 9215
-    },
-    {
-      "epoch": 0.6023135742761911,
-      "grad_norm": 0.4758225381374359,
-      "learning_rate": 9.183830915002261e-06,
-      "loss": 0.411,
-      "step": 9216
-    },
-    {
-      "epoch": 0.6023789294817332,
-      "grad_norm": 0.4261079728603363,
-      "learning_rate": 9.183639702191285e-06,
-      "loss": 0.3331,
-      "step": 9217
-    },
-    {
-      "epoch": 0.6024442846872753,
-      "grad_norm": 0.4577682316303253,
-      "learning_rate": 9.183448468975248e-06,
-      "loss": 0.3785,
-      "step": 9218
-    },
-    {
-      "epoch": 0.6025096398928175,
-      "grad_norm": 0.49699866771698,
-      "learning_rate": 9.183257215355086e-06,
-      "loss": 0.4658,
-      "step": 9219
-    },
-    {
-      "epoch": 0.6025749950983595,
-      "grad_norm": 0.48968544602394104,
-      "learning_rate": 9.183065941331729e-06,
-      "loss": 0.3937,
-      "step": 9220
-    },
-    {
-      "epoch": 0.6026403503039017,
-      "grad_norm": 0.45355671644210815,
-      "learning_rate": 9.182874646906108e-06,
-      "loss": 0.4025,
-      "step": 9221
-    },
-    {
-      "epoch": 0.6027057055094438,
-      "grad_norm": 0.4412393271923065,
-      "learning_rate": 9.182683332079158e-06,
-      "loss": 0.3605,
-      "step": 9222
-    },
-    {
-      "epoch": 0.602771060714986,
-      "grad_norm": 0.45830315351486206,
-      "learning_rate": 9.182491996851816e-06,
-      "loss": 0.4112,
-      "step": 9223
-    },
-    {
-      "epoch": 0.6028364159205281,
-      "grad_norm": 0.46280384063720703,
-      "learning_rate": 9.182300641225009e-06,
-      "loss": 0.3903,
-      "step": 9224
-    },
-    {
-      "epoch": 0.6029017711260702,
-      "grad_norm": 0.47330477833747864,
-      "learning_rate": 9.182109265199674e-06,
-      "loss": 0.3751,
-      "step": 9225
-    },
-    {
-      "epoch": 0.6029671263316123,
-      "grad_norm": 0.41610467433929443,
-      "learning_rate": 9.181917868776741e-06,
-      "loss": 0.3447,
-      "step": 9226
-    },
-    {
-      "epoch": 0.6030324815371544,
-      "grad_norm": 0.4560438096523285,
-      "learning_rate": 9.18172645195715e-06,
-      "loss": 0.3833,
-      "step": 9227
-    },
-    {
-      "epoch": 0.6030978367426966,
-      "grad_norm": 0.5108293294906616,
-      "learning_rate": 9.181535014741827e-06,
-      "loss": 0.3996,
-      "step": 9228
-    },
-    {
-      "epoch": 0.6031631919482386,
-      "grad_norm": 0.4822651147842407,
-      "learning_rate": 9.18134355713171e-06,
-      "loss": 0.4373,
-      "step": 9229
-    },
-    {
-      "epoch": 0.6032285471537808,
-      "grad_norm": 0.43790221214294434,
-      "learning_rate": 9.18115207912773e-06,
-      "loss": 0.3831,
-      "step": 9230
-    },
-    {
-      "epoch": 0.6032939023593229,
-      "grad_norm": 0.4580710530281067,
-      "learning_rate": 9.180960580730826e-06,
-      "loss": 0.43,
-      "step": 9231
-    },
-    {
-      "epoch": 0.6033592575648651,
-      "grad_norm": 0.48753198981285095,
-      "learning_rate": 9.180769061941927e-06,
-      "loss": 0.4282,
-      "step": 9232
-    },
-    {
-      "epoch": 0.6034246127704072,
-      "grad_norm": 0.466320276260376,
-      "learning_rate": 9.18057752276197e-06,
-      "loss": 0.4073,
-      "step": 9233
-    },
-    {
-      "epoch": 0.6034899679759493,
-      "grad_norm": 0.45421427488327026,
-      "learning_rate": 9.180385963191888e-06,
-      "loss": 0.3648,
-      "step": 9234
-    },
-    {
-      "epoch": 0.6035553231814914,
-      "grad_norm": 0.4825940430164337,
-      "learning_rate": 9.180194383232614e-06,
-      "loss": 0.4729,
-      "step": 9235
-    },
-    {
-      "epoch": 0.6036206783870335,
-      "grad_norm": 0.4419664442539215,
-      "learning_rate": 9.180002782885086e-06,
-      "loss": 0.3844,
-      "step": 9236
-    },
-    {
-      "epoch": 0.6036860335925757,
-      "grad_norm": 0.4473826289176941,
-      "learning_rate": 9.179811162150234e-06,
-      "loss": 0.3653,
-      "step": 9237
-    },
-    {
-      "epoch": 0.6037513887981177,
-      "grad_norm": 0.46292537450790405,
-      "learning_rate": 9.179619521028997e-06,
-      "loss": 0.4106,
-      "step": 9238
-    },
-    {
-      "epoch": 0.6038167440036599,
-      "grad_norm": 0.4575139582157135,
-      "learning_rate": 9.179427859522307e-06,
-      "loss": 0.4048,
-      "step": 9239
-    },
-    {
-      "epoch": 0.603882099209202,
-      "grad_norm": 0.5099268555641174,
-      "learning_rate": 9.179236177631098e-06,
-      "loss": 0.4383,
-      "step": 9240
-    },
-    {
-      "epoch": 0.6039474544147442,
-      "grad_norm": 0.4492853879928589,
-      "learning_rate": 9.17904447535631e-06,
-      "loss": 0.4265,
-      "step": 9241
-    },
-    {
-      "epoch": 0.6040128096202863,
-      "grad_norm": 0.4846152365207672,
-      "learning_rate": 9.17885275269887e-06,
-      "loss": 0.3551,
-      "step": 9242
-    },
-    {
-      "epoch": 0.6040781648258283,
-      "grad_norm": 0.4736369550228119,
-      "learning_rate": 9.17866100965972e-06,
-      "loss": 0.3912,
-      "step": 9243
-    },
-    {
-      "epoch": 0.6041435200313705,
-      "grad_norm": 0.4351763129234314,
-      "learning_rate": 9.178469246239792e-06,
-      "loss": 0.3586,
-      "step": 9244
-    },
-    {
-      "epoch": 0.6042088752369126,
-      "grad_norm": 0.4278549253940582,
-      "learning_rate": 9.178277462440021e-06,
-      "loss": 0.3663,
-      "step": 9245
-    },
-    {
-      "epoch": 0.6042742304424548,
-      "grad_norm": 0.463532954454422,
-      "learning_rate": 9.178085658261345e-06,
-      "loss": 0.3654,
-      "step": 9246
-    },
-    {
-      "epoch": 0.6043395856479968,
-      "grad_norm": 0.48848146200180054,
-      "learning_rate": 9.177893833704697e-06,
-      "loss": 0.4193,
-      "step": 9247
-    },
-    {
-      "epoch": 0.604404940853539,
-      "grad_norm": 0.49999526143074036,
-      "learning_rate": 9.177701988771014e-06,
-      "loss": 0.4546,
-      "step": 9248
-    },
-    {
-      "epoch": 0.6044702960590811,
-      "grad_norm": 0.4398805499076843,
-      "learning_rate": 9.17751012346123e-06,
-      "loss": 0.3845,
-      "step": 9249
-    },
-    {
-      "epoch": 0.6045356512646233,
-      "grad_norm": 0.4630897045135498,
-      "learning_rate": 9.177318237776282e-06,
-      "loss": 0.3946,
-      "step": 9250
-    },
-    {
-      "epoch": 0.6046010064701653,
-      "grad_norm": 0.45941224694252014,
-      "learning_rate": 9.177126331717108e-06,
-      "loss": 0.3547,
-      "step": 9251
-    },
-    {
-      "epoch": 0.6046663616757074,
-      "grad_norm": 0.45796695351600647,
-      "learning_rate": 9.176934405284638e-06,
-      "loss": 0.4086,
-      "step": 9252
-    },
-    {
-      "epoch": 0.6047317168812496,
-      "grad_norm": 0.45740631222724915,
-      "learning_rate": 9.176742458479815e-06,
-      "loss": 0.379,
-      "step": 9253
-    },
-    {
-      "epoch": 0.6047970720867917,
-      "grad_norm": 0.45659109950065613,
-      "learning_rate": 9.176550491303571e-06,
-      "loss": 0.3924,
-      "step": 9254
-    },
-    {
-      "epoch": 0.6048624272923339,
-      "grad_norm": 0.4223993420600891,
-      "learning_rate": 9.176358503756844e-06,
-      "loss": 0.3316,
-      "step": 9255
-    },
-    {
-      "epoch": 0.6049277824978759,
-      "grad_norm": 0.47225451469421387,
-      "learning_rate": 9.176166495840569e-06,
-      "loss": 0.4003,
-      "step": 9256
-    },
-    {
-      "epoch": 0.6049931377034181,
-      "grad_norm": 0.4650261402130127,
-      "learning_rate": 9.175974467555682e-06,
-      "loss": 0.38,
-      "step": 9257
-    },
-    {
-      "epoch": 0.6050584929089602,
-      "grad_norm": 0.4409800171852112,
-      "learning_rate": 9.175782418903122e-06,
-      "loss": 0.3384,
-      "step": 9258
-    },
-    {
-      "epoch": 0.6051238481145024,
-      "grad_norm": 0.47074276208877563,
-      "learning_rate": 9.175590349883825e-06,
-      "loss": 0.4261,
-      "step": 9259
-    },
-    {
-      "epoch": 0.6051892033200444,
-      "grad_norm": 0.4849025309085846,
-      "learning_rate": 9.175398260498728e-06,
-      "loss": 0.4337,
-      "step": 9260
-    },
-    {
-      "epoch": 0.6052545585255865,
-      "grad_norm": 0.47750529646873474,
-      "learning_rate": 9.175206150748766e-06,
-      "loss": 0.4175,
-      "step": 9261
-    },
-    {
-      "epoch": 0.6053199137311287,
-      "grad_norm": 0.43006545305252075,
-      "learning_rate": 9.175014020634877e-06,
-      "loss": 0.3488,
-      "step": 9262
-    },
-    {
-      "epoch": 0.6053852689366708,
-      "grad_norm": 0.5229668617248535,
-      "learning_rate": 9.174821870158e-06,
-      "loss": 0.4438,
-      "step": 9263
-    },
-    {
-      "epoch": 0.605450624142213,
-      "grad_norm": 0.4007505476474762,
-      "learning_rate": 9.174629699319068e-06,
-      "loss": 0.3662,
-      "step": 9264
-    },
-    {
-      "epoch": 0.605515979347755,
-      "grad_norm": 0.4569067060947418,
-      "learning_rate": 9.174437508119022e-06,
-      "loss": 0.399,
-      "step": 9265
-    },
-    {
-      "epoch": 0.6055813345532972,
-      "grad_norm": 0.5018656849861145,
-      "learning_rate": 9.174245296558797e-06,
-      "loss": 0.4844,
-      "step": 9266
-    },
-    {
-      "epoch": 0.6056466897588393,
-      "grad_norm": 0.4632817506790161,
-      "learning_rate": 9.174053064639333e-06,
-      "loss": 0.4058,
-      "step": 9267
-    },
-    {
-      "epoch": 0.6057120449643814,
-      "grad_norm": 0.4557321071624756,
-      "learning_rate": 9.173860812361565e-06,
-      "loss": 0.3833,
-      "step": 9268
-    },
-    {
-      "epoch": 0.6057774001699235,
-      "grad_norm": 0.4242776930332184,
-      "learning_rate": 9.173668539726432e-06,
-      "loss": 0.3411,
-      "step": 9269
-    },
-    {
-      "epoch": 0.6058427553754656,
-      "grad_norm": 0.4838681221008301,
-      "learning_rate": 9.173476246734874e-06,
-      "loss": 0.4205,
-      "step": 9270
-    },
-    {
-      "epoch": 0.6059081105810078,
-      "grad_norm": 0.46381068229675293,
-      "learning_rate": 9.173283933387825e-06,
-      "loss": 0.3974,
-      "step": 9271
-    },
-    {
-      "epoch": 0.6059734657865499,
-      "grad_norm": 0.45779410004615784,
-      "learning_rate": 9.173091599686224e-06,
-      "loss": 0.397,
-      "step": 9272
-    },
-    {
-      "epoch": 0.606038820992092,
-      "grad_norm": 0.43464165925979614,
-      "learning_rate": 9.172899245631009e-06,
-      "loss": 0.3662,
-      "step": 9273
-    },
-    {
-      "epoch": 0.6061041761976341,
-      "grad_norm": 0.44983839988708496,
-      "learning_rate": 9.172706871223118e-06,
-      "loss": 0.3824,
-      "step": 9274
-    },
-    {
-      "epoch": 0.6061695314031763,
-      "grad_norm": 0.44361603260040283,
-      "learning_rate": 9.172514476463492e-06,
-      "loss": 0.3846,
-      "step": 9275
-    },
-    {
-      "epoch": 0.6062348866087184,
-      "grad_norm": 0.4228716790676117,
-      "learning_rate": 9.172322061353067e-06,
-      "loss": 0.3398,
-      "step": 9276
-    },
-    {
-      "epoch": 0.6063002418142605,
-      "grad_norm": 0.4249505400657654,
-      "learning_rate": 9.172129625892783e-06,
-      "loss": 0.3482,
-      "step": 9277
-    },
-    {
-      "epoch": 0.6063655970198026,
-      "grad_norm": 0.4705303907394409,
-      "learning_rate": 9.171937170083576e-06,
-      "loss": 0.3762,
-      "step": 9278
-    },
-    {
-      "epoch": 0.6064309522253447,
-      "grad_norm": 0.47127071022987366,
-      "learning_rate": 9.171744693926385e-06,
-      "loss": 0.4079,
-      "step": 9279
-    },
-    {
-      "epoch": 0.6064963074308869,
-      "grad_norm": 0.471608966588974,
-      "learning_rate": 9.171552197422152e-06,
-      "loss": 0.4513,
-      "step": 9280
-    },
-    {
-      "epoch": 0.606561662636429,
-      "grad_norm": 0.49204134941101074,
-      "learning_rate": 9.171359680571813e-06,
-      "loss": 0.4552,
-      "step": 9281
-    },
-    {
-      "epoch": 0.6066270178419712,
-      "grad_norm": 0.44507670402526855,
-      "learning_rate": 9.171167143376307e-06,
-      "loss": 0.3592,
-      "step": 9282
-    },
-    {
-      "epoch": 0.6066923730475132,
-      "grad_norm": 0.43678945302963257,
-      "learning_rate": 9.170974585836575e-06,
-      "loss": 0.3695,
-      "step": 9283
-    },
-    {
-      "epoch": 0.6067577282530554,
-      "grad_norm": 0.44916483759880066,
-      "learning_rate": 9.170782007953554e-06,
-      "loss": 0.399,
-      "step": 9284
-    },
-    {
-      "epoch": 0.6068230834585975,
-      "grad_norm": 0.4740779995918274,
-      "learning_rate": 9.170589409728185e-06,
-      "loss": 0.4306,
-      "step": 9285
-    },
-    {
-      "epoch": 0.6068884386641396,
-      "grad_norm": 0.481114000082016,
-      "learning_rate": 9.170396791161407e-06,
-      "loss": 0.4212,
-      "step": 9286
-    },
-    {
-      "epoch": 0.6069537938696817,
-      "grad_norm": 0.46107128262519836,
-      "learning_rate": 9.170204152254159e-06,
-      "loss": 0.4126,
-      "step": 9287
-    },
-    {
-      "epoch": 0.6070191490752238,
-      "grad_norm": 0.39378905296325684,
-      "learning_rate": 9.170011493007379e-06,
-      "loss": 0.3048,
-      "step": 9288
-    },
-    {
-      "epoch": 0.607084504280766,
-      "grad_norm": 0.5063765645027161,
-      "learning_rate": 9.16981881342201e-06,
-      "loss": 0.4835,
-      "step": 9289
-    },
-    {
-      "epoch": 0.6071498594863081,
-      "grad_norm": 0.45550718903541565,
-      "learning_rate": 9.16962611349899e-06,
-      "loss": 0.3817,
-      "step": 9290
-    },
-    {
-      "epoch": 0.6072152146918502,
-      "grad_norm": 0.5076361894607544,
-      "learning_rate": 9.169433393239258e-06,
-      "loss": 0.4454,
-      "step": 9291
-    },
-    {
-      "epoch": 0.6072805698973923,
-      "grad_norm": 0.43833643198013306,
-      "learning_rate": 9.169240652643756e-06,
-      "loss": 0.3643,
-      "step": 9292
-    },
-    {
-      "epoch": 0.6073459251029345,
-      "grad_norm": 0.40341854095458984,
-      "learning_rate": 9.169047891713422e-06,
-      "loss": 0.3335,
-      "step": 9293
-    },
-    {
-      "epoch": 0.6074112803084766,
-      "grad_norm": 0.4650060832500458,
-      "learning_rate": 9.168855110449198e-06,
-      "loss": 0.375,
-      "step": 9294
-    },
-    {
-      "epoch": 0.6074766355140186,
-      "grad_norm": 0.47409045696258545,
-      "learning_rate": 9.168662308852021e-06,
-      "loss": 0.3964,
-      "step": 9295
-    },
-    {
-      "epoch": 0.6075419907195608,
-      "grad_norm": 0.4652184247970581,
-      "learning_rate": 9.168469486922838e-06,
-      "loss": 0.3611,
-      "step": 9296
-    },
-    {
-      "epoch": 0.6076073459251029,
-      "grad_norm": 0.4325246810913086,
-      "learning_rate": 9.168276644662581e-06,
-      "loss": 0.3591,
-      "step": 9297
-    },
-    {
-      "epoch": 0.6076727011306451,
-      "grad_norm": 0.4370783567428589,
-      "learning_rate": 9.168083782072196e-06,
-      "loss": 0.334,
-      "step": 9298
-    },
-    {
-      "epoch": 0.6077380563361872,
-      "grad_norm": 0.46329495310783386,
-      "learning_rate": 9.167890899152624e-06,
-      "loss": 0.4051,
-      "step": 9299
-    },
-    {
-      "epoch": 0.6078034115417293,
-      "grad_norm": 0.43963465094566345,
-      "learning_rate": 9.167697995904802e-06,
-      "loss": 0.3603,
-      "step": 9300
-    },
-    {
-      "epoch": 0.6078687667472714,
-      "grad_norm": 0.4647124707698822,
-      "learning_rate": 9.167505072329677e-06,
-      "loss": 0.3787,
-      "step": 9301
-    },
-    {
-      "epoch": 0.6079341219528135,
-      "grad_norm": 0.5020391345024109,
-      "learning_rate": 9.167312128428181e-06,
-      "loss": 0.4252,
-      "step": 9302
-    },
-    {
-      "epoch": 0.6079994771583557,
-      "grad_norm": 0.45752403140068054,
-      "learning_rate": 9.167119164201263e-06,
-      "loss": 0.4102,
-      "step": 9303
-    },
-    {
-      "epoch": 0.6080648323638977,
-      "grad_norm": 0.44788506627082825,
-      "learning_rate": 9.16692617964986e-06,
-      "loss": 0.3645,
-      "step": 9304
-    },
-    {
-      "epoch": 0.6081301875694399,
-      "grad_norm": 0.42570140957832336,
-      "learning_rate": 9.166733174774915e-06,
-      "loss": 0.3683,
-      "step": 9305
-    },
-    {
-      "epoch": 0.608195542774982,
-      "grad_norm": 0.4321891963481903,
-      "learning_rate": 9.166540149577369e-06,
-      "loss": 0.3538,
-      "step": 9306
-    },
-    {
-      "epoch": 0.6082608979805242,
-      "grad_norm": 0.4512529671192169,
-      "learning_rate": 9.166347104058164e-06,
-      "loss": 0.3583,
-      "step": 9307
-    },
-    {
-      "epoch": 0.6083262531860663,
-      "grad_norm": 0.4835556149482727,
-      "learning_rate": 9.16615403821824e-06,
-      "loss": 0.378,
-      "step": 9308
-    },
-    {
-      "epoch": 0.6083916083916084,
-      "grad_norm": 0.48107215762138367,
-      "learning_rate": 9.165960952058538e-06,
-      "loss": 0.4056,
-      "step": 9309
-    },
-    {
-      "epoch": 0.6084569635971505,
-      "grad_norm": 0.48607125878334045,
-      "learning_rate": 9.165767845580004e-06,
-      "loss": 0.4121,
-      "step": 9310
-    },
-    {
-      "epoch": 0.6085223188026926,
-      "grad_norm": 0.453477144241333,
-      "learning_rate": 9.165574718783575e-06,
-      "loss": 0.4037,
-      "step": 9311
-    },
-    {
-      "epoch": 0.6085876740082348,
-      "grad_norm": 0.4463033676147461,
-      "learning_rate": 9.165381571670195e-06,
-      "loss": 0.3856,
-      "step": 9312
-    },
-    {
-      "epoch": 0.6086530292137768,
-      "grad_norm": 0.4055895209312439,
-      "learning_rate": 9.165188404240808e-06,
-      "loss": 0.3262,
-      "step": 9313
-    },
-    {
-      "epoch": 0.608718384419319,
-      "grad_norm": 0.4310225546360016,
-      "learning_rate": 9.164995216496354e-06,
-      "loss": 0.3695,
-      "step": 9314
-    },
-    {
-      "epoch": 0.6087837396248611,
-      "grad_norm": 0.48405686020851135,
-      "learning_rate": 9.164802008437772e-06,
-      "loss": 0.3959,
-      "step": 9315
-    },
-    {
-      "epoch": 0.6088490948304033,
-      "grad_norm": 0.4695260524749756,
-      "learning_rate": 9.164608780066011e-06,
-      "loss": 0.4043,
-      "step": 9316
-    },
-    {
-      "epoch": 0.6089144500359454,
-      "grad_norm": 0.4271462559700012,
-      "learning_rate": 9.164415531382009e-06,
-      "loss": 0.3703,
-      "step": 9317
-    },
-    {
-      "epoch": 0.6089798052414875,
-      "grad_norm": 0.4622548520565033,
-      "learning_rate": 9.16422226238671e-06,
-      "loss": 0.3775,
-      "step": 9318
-    },
-    {
-      "epoch": 0.6090451604470296,
-      "grad_norm": 0.4441933035850525,
-      "learning_rate": 9.164028973081057e-06,
-      "loss": 0.3782,
-      "step": 9319
-    },
-    {
-      "epoch": 0.6091105156525717,
-      "grad_norm": 0.45878997445106506,
-      "learning_rate": 9.163835663465992e-06,
-      "loss": 0.3747,
-      "step": 9320
-    },
-    {
-      "epoch": 0.6091758708581139,
-      "grad_norm": 0.45647621154785156,
-      "learning_rate": 9.163642333542457e-06,
-      "loss": 0.3813,
-      "step": 9321
-    },
-    {
-      "epoch": 0.6092412260636559,
-      "grad_norm": 0.43723830580711365,
-      "learning_rate": 9.163448983311396e-06,
-      "loss": 0.38,
-      "step": 9322
-    },
-    {
-      "epoch": 0.6093065812691981,
-      "grad_norm": 0.44617873430252075,
-      "learning_rate": 9.163255612773752e-06,
-      "loss": 0.3714,
-      "step": 9323
-    },
-    {
-      "epoch": 0.6093719364747402,
-      "grad_norm": 0.42884573340415955,
-      "learning_rate": 9.16306222193047e-06,
-      "loss": 0.3488,
-      "step": 9324
-    },
-    {
-      "epoch": 0.6094372916802824,
-      "grad_norm": 0.4828496277332306,
-      "learning_rate": 9.16286881078249e-06,
-      "loss": 0.4345,
-      "step": 9325
-    },
-    {
-      "epoch": 0.6095026468858245,
-      "grad_norm": 0.46828553080558777,
-      "learning_rate": 9.162675379330757e-06,
-      "loss": 0.4007,
-      "step": 9326
-    },
-    {
-      "epoch": 0.6095680020913665,
-      "grad_norm": 0.4563262164592743,
-      "learning_rate": 9.162481927576213e-06,
-      "loss": 0.3796,
-      "step": 9327
-    },
-    {
-      "epoch": 0.6096333572969087,
-      "grad_norm": 0.408500075340271,
-      "learning_rate": 9.162288455519803e-06,
-      "loss": 0.3276,
-      "step": 9328
-    },
-    {
-      "epoch": 0.6096987125024508,
-      "grad_norm": 0.5024822354316711,
-      "learning_rate": 9.16209496316247e-06,
-      "loss": 0.337,
-      "step": 9329
-    },
-    {
-      "epoch": 0.609764067707993,
-      "grad_norm": 0.47403833270072937,
-      "learning_rate": 9.16190145050516e-06,
-      "loss": 0.3705,
-      "step": 9330
-    },
-    {
-      "epoch": 0.609829422913535,
-      "grad_norm": 0.45196640491485596,
-      "learning_rate": 9.161707917548813e-06,
-      "loss": 0.385,
-      "step": 9331
-    },
-    {
-      "epoch": 0.6098947781190772,
-      "grad_norm": 0.47436442971229553,
-      "learning_rate": 9.161514364294373e-06,
-      "loss": 0.4057,
-      "step": 9332
-    },
-    {
-      "epoch": 0.6099601333246193,
-      "grad_norm": 0.4377812445163727,
-      "learning_rate": 9.16132079074279e-06,
-      "loss": 0.3866,
-      "step": 9333
-    },
-    {
-      "epoch": 0.6100254885301615,
-      "grad_norm": 0.45032066106796265,
-      "learning_rate": 9.161127196895e-06,
-      "loss": 0.401,
-      "step": 9334
-    },
-    {
-      "epoch": 0.6100908437357035,
-      "grad_norm": 0.46344923973083496,
-      "learning_rate": 9.160933582751953e-06,
-      "loss": 0.3959,
-      "step": 9335
-    },
-    {
-      "epoch": 0.6101561989412456,
-      "grad_norm": 0.4364248812198639,
-      "learning_rate": 9.160739948314591e-06,
-      "loss": 0.3634,
-      "step": 9336
-    },
-    {
-      "epoch": 0.6102215541467878,
-      "grad_norm": 0.46375197172164917,
-      "learning_rate": 9.160546293583858e-06,
-      "loss": 0.391,
-      "step": 9337
-    },
-    {
-      "epoch": 0.6102869093523299,
-      "grad_norm": 0.4995439946651459,
-      "learning_rate": 9.160352618560702e-06,
-      "loss": 0.4723,
-      "step": 9338
-    },
-    {
-      "epoch": 0.6103522645578721,
-      "grad_norm": 0.4590868651866913,
-      "learning_rate": 9.160158923246062e-06,
-      "loss": 0.3908,
-      "step": 9339
-    },
-    {
-      "epoch": 0.6104176197634141,
-      "grad_norm": 0.46573808789253235,
-      "learning_rate": 9.159965207640889e-06,
-      "loss": 0.3823,
-      "step": 9340
-    },
-    {
-      "epoch": 0.6104829749689563,
-      "grad_norm": 0.425341933965683,
-      "learning_rate": 9.159771471746122e-06,
-      "loss": 0.359,
-      "step": 9341
-    },
-    {
-      "epoch": 0.6105483301744984,
-      "grad_norm": 0.42857423424720764,
-      "learning_rate": 9.159577715562709e-06,
-      "loss": 0.374,
-      "step": 9342
-    },
-    {
-      "epoch": 0.6106136853800406,
-      "grad_norm": 0.41146010160446167,
-      "learning_rate": 9.159383939091594e-06,
-      "loss": 0.3417,
-      "step": 9343
-    },
-    {
-      "epoch": 0.6106790405855826,
-      "grad_norm": 0.4427056610584259,
-      "learning_rate": 9.159190142333724e-06,
-      "loss": 0.3579,
-      "step": 9344
-    },
-    {
-      "epoch": 0.6107443957911247,
-      "grad_norm": 0.38539740443229675,
-      "learning_rate": 9.158996325290041e-06,
-      "loss": 0.2726,
-      "step": 9345
-    },
-    {
-      "epoch": 0.6108097509966669,
-      "grad_norm": 0.44148820638656616,
-      "learning_rate": 9.158802487961493e-06,
-      "loss": 0.3861,
-      "step": 9346
-    },
-    {
-      "epoch": 0.610875106202209,
-      "grad_norm": 0.458450585603714,
-      "learning_rate": 9.158608630349025e-06,
-      "loss": 0.3775,
-      "step": 9347
-    },
-    {
-      "epoch": 0.6109404614077512,
-      "grad_norm": 0.4560156464576721,
-      "learning_rate": 9.158414752453582e-06,
-      "loss": 0.4279,
-      "step": 9348
-    },
-    {
-      "epoch": 0.6110058166132932,
-      "grad_norm": 0.44887226819992065,
-      "learning_rate": 9.158220854276108e-06,
-      "loss": 0.3703,
-      "step": 9349
-    },
-    {
-      "epoch": 0.6110711718188354,
-      "grad_norm": 0.4397490918636322,
-      "learning_rate": 9.158026935817552e-06,
-      "loss": 0.3885,
-      "step": 9350
-    },
-    {
-      "epoch": 0.6111365270243775,
-      "grad_norm": 0.4468318819999695,
-      "learning_rate": 9.157832997078859e-06,
-      "loss": 0.3997,
-      "step": 9351
-    },
-    {
-      "epoch": 0.6112018822299197,
-      "grad_norm": 0.4876280128955841,
-      "learning_rate": 9.157639038060972e-06,
-      "loss": 0.4132,
-      "step": 9352
-    },
-    {
-      "epoch": 0.6112672374354617,
-      "grad_norm": 0.4092002213001251,
-      "learning_rate": 9.15744505876484e-06,
-      "loss": 0.3486,
-      "step": 9353
-    },
-    {
-      "epoch": 0.6113325926410038,
-      "grad_norm": 0.4698246717453003,
-      "learning_rate": 9.157251059191408e-06,
-      "loss": 0.4555,
-      "step": 9354
-    },
-    {
-      "epoch": 0.611397947846546,
-      "grad_norm": 0.45148321986198425,
-      "learning_rate": 9.15705703934162e-06,
-      "loss": 0.4153,
-      "step": 9355
-    },
-    {
-      "epoch": 0.6114633030520881,
-      "grad_norm": 0.4464689791202545,
-      "learning_rate": 9.15686299921643e-06,
-      "loss": 0.4065,
-      "step": 9356
-    },
-    {
-      "epoch": 0.6115286582576303,
-      "grad_norm": 0.4503811001777649,
-      "learning_rate": 9.156668938816776e-06,
-      "loss": 0.4099,
-      "step": 9357
-    },
-    {
-      "epoch": 0.6115940134631723,
-      "grad_norm": 0.4475444555282593,
-      "learning_rate": 9.156474858143607e-06,
-      "loss": 0.3891,
-      "step": 9358
-    },
-    {
-      "epoch": 0.6116593686687145,
-      "grad_norm": 0.4001803398132324,
-      "learning_rate": 9.15628075719787e-06,
-      "loss": 0.2989,
-      "step": 9359
-    },
-    {
-      "epoch": 0.6117247238742566,
-      "grad_norm": 0.4261928200721741,
-      "learning_rate": 9.156086635980515e-06,
-      "loss": 0.3362,
-      "step": 9360
-    },
-    {
-      "epoch": 0.6117900790797987,
-      "grad_norm": 0.46899622678756714,
-      "learning_rate": 9.155892494492483e-06,
-      "loss": 0.3836,
-      "step": 9361
-    },
-    {
-      "epoch": 0.6118554342853408,
-      "grad_norm": 0.4666925370693207,
-      "learning_rate": 9.155698332734724e-06,
-      "loss": 0.3707,
-      "step": 9362
-    },
-    {
-      "epoch": 0.6119207894908829,
-      "grad_norm": 0.44752249121665955,
-      "learning_rate": 9.155504150708183e-06,
-      "loss": 0.3341,
-      "step": 9363
-    },
-    {
-      "epoch": 0.6119861446964251,
-      "grad_norm": 0.4533834159374237,
-      "learning_rate": 9.15530994841381e-06,
-      "loss": 0.3728,
-      "step": 9364
-    },
-    {
-      "epoch": 0.6120514999019672,
-      "grad_norm": 0.45615315437316895,
-      "learning_rate": 9.155115725852552e-06,
-      "loss": 0.4314,
-      "step": 9365
-    },
-    {
-      "epoch": 0.6121168551075094,
-      "grad_norm": 0.448549747467041,
-      "learning_rate": 9.154921483025355e-06,
-      "loss": 0.375,
-      "step": 9366
-    },
-    {
-      "epoch": 0.6121822103130514,
-      "grad_norm": 0.4391409158706665,
-      "learning_rate": 9.154727219933165e-06,
-      "loss": 0.3772,
-      "step": 9367
-    },
-    {
-      "epoch": 0.6122475655185936,
-      "grad_norm": 0.457964152097702,
-      "learning_rate": 9.154532936576931e-06,
-      "loss": 0.4139,
-      "step": 9368
-    },
-    {
-      "epoch": 0.6123129207241357,
-      "grad_norm": 0.4666399657726288,
-      "learning_rate": 9.154338632957603e-06,
-      "loss": 0.4676,
-      "step": 9369
-    },
-    {
-      "epoch": 0.6123782759296778,
-      "grad_norm": 0.4332391023635864,
-      "learning_rate": 9.154144309076124e-06,
-      "loss": 0.3582,
-      "step": 9370
-    },
-    {
-      "epoch": 0.6124436311352199,
-      "grad_norm": 0.450079083442688,
-      "learning_rate": 9.153949964933445e-06,
-      "loss": 0.3465,
-      "step": 9371
-    },
-    {
-      "epoch": 0.612508986340762,
-      "grad_norm": 0.46680015325546265,
-      "learning_rate": 9.153755600530512e-06,
-      "loss": 0.3781,
-      "step": 9372
-    },
-    {
-      "epoch": 0.6125743415463042,
-      "grad_norm": 0.5027528405189514,
-      "learning_rate": 9.153561215868274e-06,
-      "loss": 0.4332,
-      "step": 9373
-    },
-    {
-      "epoch": 0.6126396967518463,
-      "grad_norm": 0.4723973274230957,
-      "learning_rate": 9.15336681094768e-06,
-      "loss": 0.4151,
-      "step": 9374
-    },
-    {
-      "epoch": 0.6127050519573884,
-      "grad_norm": 0.5401503443717957,
-      "learning_rate": 9.153172385769678e-06,
-      "loss": 0.5215,
-      "step": 9375
-    },
-    {
-      "epoch": 0.6127704071629305,
-      "grad_norm": 0.46017375588417053,
-      "learning_rate": 9.152977940335213e-06,
-      "loss": 0.3931,
-      "step": 9376
-    },
-    {
-      "epoch": 0.6128357623684727,
-      "grad_norm": 0.46815547347068787,
-      "learning_rate": 9.152783474645237e-06,
-      "loss": 0.3997,
-      "step": 9377
-    },
-    {
-      "epoch": 0.6129011175740148,
-      "grad_norm": 0.4949282109737396,
-      "learning_rate": 9.152588988700697e-06,
-      "loss": 0.4594,
-      "step": 9378
-    },
-    {
-      "epoch": 0.6129664727795568,
-      "grad_norm": 0.4952649772167206,
-      "learning_rate": 9.152394482502543e-06,
-      "loss": 0.4142,
-      "step": 9379
-    },
-    {
-      "epoch": 0.613031827985099,
-      "grad_norm": 0.4407411217689514,
-      "learning_rate": 9.152199956051721e-06,
-      "loss": 0.3609,
-      "step": 9380
-    },
-    {
-      "epoch": 0.6130971831906411,
-      "grad_norm": 0.4774309992790222,
-      "learning_rate": 9.152005409349182e-06,
-      "loss": 0.4013,
-      "step": 9381
-    },
-    {
-      "epoch": 0.6131625383961833,
-      "grad_norm": 0.44647300243377686,
-      "learning_rate": 9.151810842395876e-06,
-      "loss": 0.3817,
-      "step": 9382
-    },
-    {
-      "epoch": 0.6132278936017254,
-      "grad_norm": 0.4410271644592285,
-      "learning_rate": 9.151616255192749e-06,
-      "loss": 0.3861,
-      "step": 9383
-    },
-    {
-      "epoch": 0.6132932488072675,
-      "grad_norm": 0.46410071849823,
-      "learning_rate": 9.151421647740751e-06,
-      "loss": 0.3684,
-      "step": 9384
-    },
-    {
-      "epoch": 0.6133586040128096,
-      "grad_norm": 0.4376247525215149,
-      "learning_rate": 9.151227020040832e-06,
-      "loss": 0.3688,
-      "step": 9385
-    },
-    {
-      "epoch": 0.6134239592183517,
-      "grad_norm": 0.4442978501319885,
-      "learning_rate": 9.15103237209394e-06,
-      "loss": 0.386,
-      "step": 9386
-    },
-    {
-      "epoch": 0.6134893144238939,
-      "grad_norm": 0.560441792011261,
-      "learning_rate": 9.150837703901025e-06,
-      "loss": 0.4521,
-      "step": 9387
-    },
-    {
-      "epoch": 0.613554669629436,
-      "grad_norm": 0.4365333616733551,
-      "learning_rate": 9.150643015463036e-06,
-      "loss": 0.3634,
-      "step": 9388
-    },
-    {
-      "epoch": 0.6136200248349781,
-      "grad_norm": 0.4355616867542267,
-      "learning_rate": 9.150448306780923e-06,
-      "loss": 0.3559,
-      "step": 9389
-    },
-    {
-      "epoch": 0.6136853800405202,
-      "grad_norm": 0.4717211127281189,
-      "learning_rate": 9.150253577855637e-06,
-      "loss": 0.426,
-      "step": 9390
-    },
-    {
-      "epoch": 0.6137507352460624,
-      "grad_norm": 0.451812207698822,
-      "learning_rate": 9.150058828688127e-06,
-      "loss": 0.395,
-      "step": 9391
-    },
-    {
-      "epoch": 0.6138160904516045,
-      "grad_norm": 0.4561154842376709,
-      "learning_rate": 9.14986405927934e-06,
-      "loss": 0.396,
-      "step": 9392
-    },
-    {
-      "epoch": 0.6138814456571466,
-      "grad_norm": 0.4951837956905365,
-      "learning_rate": 9.14966926963023e-06,
-      "loss": 0.455,
-      "step": 9393
-    },
-    {
-      "epoch": 0.6139468008626887,
-      "grad_norm": 0.4486224949359894,
-      "learning_rate": 9.149474459741747e-06,
-      "loss": 0.3446,
-      "step": 9394
-    },
-    {
-      "epoch": 0.6140121560682308,
-      "grad_norm": 0.4535962641239166,
-      "learning_rate": 9.149279629614836e-06,
-      "loss": 0.4157,
-      "step": 9395
-    },
-    {
-      "epoch": 0.614077511273773,
-      "grad_norm": 0.5602654814720154,
-      "learning_rate": 9.149084779250453e-06,
-      "loss": 0.3312,
-      "step": 9396
-    },
-    {
-      "epoch": 0.614142866479315,
-      "grad_norm": 0.4514555335044861,
-      "learning_rate": 9.148889908649546e-06,
-      "loss": 0.349,
-      "step": 9397
-    },
-    {
-      "epoch": 0.6142082216848572,
-      "grad_norm": 0.4832721948623657,
-      "learning_rate": 9.148695017813065e-06,
-      "loss": 0.4652,
-      "step": 9398
-    },
-    {
-      "epoch": 0.6142735768903993,
-      "grad_norm": 0.4278484284877777,
-      "learning_rate": 9.148500106741963e-06,
-      "loss": 0.3439,
-      "step": 9399
-    },
-    {
-      "epoch": 0.6143389320959415,
-      "grad_norm": 0.7185001969337463,
-      "learning_rate": 9.148305175437187e-06,
-      "loss": 0.4237,
-      "step": 9400
-    },
-    {
-      "epoch": 0.6144042873014836,
-      "grad_norm": 0.48224517703056335,
-      "learning_rate": 9.148110223899689e-06,
-      "loss": 0.4252,
-      "step": 9401
-    },
-    {
-      "epoch": 0.6144696425070257,
-      "grad_norm": 0.47076553106307983,
-      "learning_rate": 9.147915252130421e-06,
-      "loss": 0.3946,
-      "step": 9402
-    },
-    {
-      "epoch": 0.6145349977125678,
-      "grad_norm": 0.4633639454841614,
-      "learning_rate": 9.147720260130332e-06,
-      "loss": 0.3961,
-      "step": 9403
-    },
-    {
-      "epoch": 0.6146003529181099,
-      "grad_norm": 0.43805554509162903,
-      "learning_rate": 9.147525247900377e-06,
-      "loss": 0.3479,
-      "step": 9404
-    },
-    {
-      "epoch": 0.6146657081236521,
-      "grad_norm": 0.4406869411468506,
-      "learning_rate": 9.147330215441504e-06,
-      "loss": 0.3766,
-      "step": 9405
-    },
-    {
-      "epoch": 0.6147310633291941,
-      "grad_norm": 0.44196370244026184,
-      "learning_rate": 9.147135162754663e-06,
-      "loss": 0.3773,
-      "step": 9406
-    },
-    {
-      "epoch": 0.6147964185347363,
-      "grad_norm": 0.4745652675628662,
-      "learning_rate": 9.146940089840809e-06,
-      "loss": 0.353,
-      "step": 9407
-    },
-    {
-      "epoch": 0.6148617737402784,
-      "grad_norm": 0.41446709632873535,
-      "learning_rate": 9.146744996700891e-06,
-      "loss": 0.3251,
-      "step": 9408
-    },
-    {
-      "epoch": 0.6149271289458206,
-      "grad_norm": 0.45017939805984497,
-      "learning_rate": 9.14654988333586e-06,
-      "loss": 0.4431,
-      "step": 9409
-    },
-    {
-      "epoch": 0.6149924841513627,
-      "grad_norm": 0.45253807306289673,
-      "learning_rate": 9.146354749746672e-06,
-      "loss": 0.3608,
-      "step": 9410
-    },
-    {
-      "epoch": 0.6150578393569047,
-      "grad_norm": 0.4453875422477722,
-      "learning_rate": 9.146159595934272e-06,
-      "loss": 0.378,
-      "step": 9411
-    },
-    {
-      "epoch": 0.6151231945624469,
-      "grad_norm": 0.4556255638599396,
-      "learning_rate": 9.145964421899617e-06,
-      "loss": 0.3905,
-      "step": 9412
-    },
-    {
-      "epoch": 0.615188549767989,
-      "grad_norm": 0.4538557529449463,
-      "learning_rate": 9.145769227643655e-06,
-      "loss": 0.4013,
-      "step": 9413
-    },
-    {
-      "epoch": 0.6152539049735312,
-      "grad_norm": 0.43004491925239563,
-      "learning_rate": 9.145574013167342e-06,
-      "loss": 0.3288,
-      "step": 9414
-    },
-    {
-      "epoch": 0.6153192601790732,
-      "grad_norm": 0.4410993754863739,
-      "learning_rate": 9.14537877847163e-06,
-      "loss": 0.3944,
-      "step": 9415
-    },
-    {
-      "epoch": 0.6153846153846154,
-      "grad_norm": 0.4260723888874054,
-      "learning_rate": 9.145183523557465e-06,
-      "loss": 0.3568,
-      "step": 9416
-    },
-    {
-      "epoch": 0.6154499705901575,
-      "grad_norm": 0.478514164686203,
-      "learning_rate": 9.144988248425807e-06,
-      "loss": 0.4101,
-      "step": 9417
-    },
-    {
-      "epoch": 0.6155153257956997,
-      "grad_norm": 0.4262210428714752,
-      "learning_rate": 9.144792953077605e-06,
-      "loss": 0.3751,
-      "step": 9418
-    },
-    {
-      "epoch": 0.6155806810012417,
-      "grad_norm": 0.4362979829311371,
-      "learning_rate": 9.144597637513814e-06,
-      "loss": 0.3664,
-      "step": 9419
-    },
-    {
-      "epoch": 0.6156460362067838,
-      "grad_norm": 0.41937923431396484,
-      "learning_rate": 9.144402301735383e-06,
-      "loss": 0.3305,
-      "step": 9420
-    },
-    {
-      "epoch": 0.615711391412326,
-      "grad_norm": 0.4549425542354584,
-      "learning_rate": 9.144206945743264e-06,
-      "loss": 0.3817,
-      "step": 9421
-    },
-    {
-      "epoch": 0.6157767466178681,
-      "grad_norm": 0.43009620904922485,
-      "learning_rate": 9.144011569538414e-06,
-      "loss": 0.3465,
-      "step": 9422
-    },
-    {
-      "epoch": 0.6158421018234103,
-      "grad_norm": 0.5174602270126343,
-      "learning_rate": 9.143816173121785e-06,
-      "loss": 0.4855,
-      "step": 9423
-    },
-    {
-      "epoch": 0.6159074570289523,
-      "grad_norm": 0.44756612181663513,
-      "learning_rate": 9.143620756494327e-06,
-      "loss": 0.4159,
-      "step": 9424
-    },
-    {
-      "epoch": 0.6159728122344945,
-      "grad_norm": 0.47683805227279663,
-      "learning_rate": 9.143425319656995e-06,
-      "loss": 0.4467,
-      "step": 9425
-    },
-    {
-      "epoch": 0.6160381674400366,
-      "grad_norm": 0.4673953652381897,
-      "learning_rate": 9.143229862610742e-06,
-      "loss": 0.4009,
-      "step": 9426
-    },
-    {
-      "epoch": 0.6161035226455788,
-      "grad_norm": 0.45060044527053833,
-      "learning_rate": 9.143034385356525e-06,
-      "loss": 0.3755,
-      "step": 9427
-    },
-    {
-      "epoch": 0.6161688778511208,
-      "grad_norm": 0.4504525065422058,
-      "learning_rate": 9.14283888789529e-06,
-      "loss": 0.349,
-      "step": 9428
-    },
-    {
-      "epoch": 0.6162342330566629,
-      "grad_norm": 0.44463473558425903,
-      "learning_rate": 9.142643370227997e-06,
-      "loss": 0.3961,
-      "step": 9429
-    },
-    {
-      "epoch": 0.6162995882622051,
-      "grad_norm": 0.45240074396133423,
-      "learning_rate": 9.142447832355595e-06,
-      "loss": 0.4041,
-      "step": 9430
-    },
-    {
-      "epoch": 0.6163649434677472,
-      "grad_norm": 0.44422590732574463,
-      "learning_rate": 9.142252274279042e-06,
-      "loss": 0.3887,
-      "step": 9431
-    },
-    {
-      "epoch": 0.6164302986732894,
-      "grad_norm": 0.4349067211151123,
-      "learning_rate": 9.142056695999288e-06,
-      "loss": 0.3611,
-      "step": 9432
-    },
-    {
-      "epoch": 0.6164956538788314,
-      "grad_norm": 0.430692195892334,
-      "learning_rate": 9.14186109751729e-06,
-      "loss": 0.404,
-      "step": 9433
-    },
-    {
-      "epoch": 0.6165610090843736,
-      "grad_norm": 0.44676336646080017,
-      "learning_rate": 9.141665478834e-06,
-      "loss": 0.3599,
-      "step": 9434
-    },
-    {
-      "epoch": 0.6166263642899157,
-      "grad_norm": 0.4970923364162445,
-      "learning_rate": 9.141469839950372e-06,
-      "loss": 0.4883,
-      "step": 9435
-    },
-    {
-      "epoch": 0.6166917194954579,
-      "grad_norm": 0.4341626465320587,
-      "learning_rate": 9.141274180867361e-06,
-      "loss": 0.3818,
-      "step": 9436
-    },
-    {
-      "epoch": 0.6167570747009999,
-      "grad_norm": 0.4820190966129303,
-      "learning_rate": 9.141078501585921e-06,
-      "loss": 0.4028,
-      "step": 9437
-    },
-    {
-      "epoch": 0.616822429906542,
-      "grad_norm": 0.5133317112922668,
-      "learning_rate": 9.140882802107007e-06,
-      "loss": 0.447,
-      "step": 9438
-    },
-    {
-      "epoch": 0.6168877851120842,
-      "grad_norm": 0.42152732610702515,
-      "learning_rate": 9.140687082431574e-06,
-      "loss": 0.3504,
-      "step": 9439
-    },
-    {
-      "epoch": 0.6169531403176263,
-      "grad_norm": 0.4819971024990082,
-      "learning_rate": 9.140491342560575e-06,
-      "loss": 0.4251,
-      "step": 9440
-    },
-    {
-      "epoch": 0.6170184955231685,
-      "grad_norm": 0.4538039267063141,
-      "learning_rate": 9.140295582494965e-06,
-      "loss": 0.4082,
-      "step": 9441
-    },
-    {
-      "epoch": 0.6170838507287105,
-      "grad_norm": 0.42919978499412537,
-      "learning_rate": 9.140099802235699e-06,
-      "loss": 0.3378,
-      "step": 9442
-    },
-    {
-      "epoch": 0.6171492059342527,
-      "grad_norm": 0.4664906859397888,
-      "learning_rate": 9.139904001783732e-06,
-      "loss": 0.3922,
-      "step": 9443
-    },
-    {
-      "epoch": 0.6172145611397948,
-      "grad_norm": 0.47128826379776,
-      "learning_rate": 9.139708181140019e-06,
-      "loss": 0.3906,
-      "step": 9444
-    },
-    {
-      "epoch": 0.6172799163453369,
-      "grad_norm": 0.45355427265167236,
-      "learning_rate": 9.139512340305516e-06,
-      "loss": 0.4102,
-      "step": 9445
-    },
-    {
-      "epoch": 0.617345271550879,
-      "grad_norm": 0.4564480185508728,
-      "learning_rate": 9.139316479281175e-06,
-      "loss": 0.3898,
-      "step": 9446
-    },
-    {
-      "epoch": 0.6174106267564211,
-      "grad_norm": 0.4544818699359894,
-      "learning_rate": 9.139120598067955e-06,
-      "loss": 0.3651,
-      "step": 9447
-    },
-    {
-      "epoch": 0.6174759819619633,
-      "grad_norm": 0.41902658343315125,
-      "learning_rate": 9.13892469666681e-06,
-      "loss": 0.3753,
-      "step": 9448
-    },
-    {
-      "epoch": 0.6175413371675054,
-      "grad_norm": 0.4443846046924591,
-      "learning_rate": 9.138728775078695e-06,
-      "loss": 0.3739,
-      "step": 9449
-    },
-    {
-      "epoch": 0.6176066923730475,
-      "grad_norm": 0.42077818512916565,
-      "learning_rate": 9.138532833304567e-06,
-      "loss": 0.3499,
-      "step": 9450
-    },
-    {
-      "epoch": 0.6176720475785896,
-      "grad_norm": 0.4572470188140869,
-      "learning_rate": 9.13833687134538e-06,
-      "loss": 0.3819,
-      "step": 9451
-    },
-    {
-      "epoch": 0.6177374027841318,
-      "grad_norm": 0.5040394067764282,
-      "learning_rate": 9.138140889202089e-06,
-      "loss": 0.4459,
-      "step": 9452
-    },
-    {
-      "epoch": 0.6178027579896739,
-      "grad_norm": 0.44595399498939514,
-      "learning_rate": 9.137944886875654e-06,
-      "loss": 0.3474,
-      "step": 9453
-    },
-    {
-      "epoch": 0.617868113195216,
-      "grad_norm": 0.4859713315963745,
-      "learning_rate": 9.137748864367026e-06,
-      "loss": 0.3867,
-      "step": 9454
-    },
-    {
-      "epoch": 0.6179334684007581,
-      "grad_norm": 0.4658839702606201,
-      "learning_rate": 9.137552821677164e-06,
-      "loss": 0.3491,
-      "step": 9455
-    },
-    {
-      "epoch": 0.6179988236063002,
-      "grad_norm": 0.4738259017467499,
-      "learning_rate": 9.137356758807025e-06,
-      "loss": 0.4353,
-      "step": 9456
-    },
-    {
-      "epoch": 0.6180641788118424,
-      "grad_norm": 0.43269017338752747,
-      "learning_rate": 9.137160675757561e-06,
-      "loss": 0.3731,
-      "step": 9457
-    },
-    {
-      "epoch": 0.6181295340173845,
-      "grad_norm": 0.5084611177444458,
-      "learning_rate": 9.136964572529734e-06,
-      "loss": 0.4506,
-      "step": 9458
-    },
-    {
-      "epoch": 0.6181948892229266,
-      "grad_norm": 0.46275830268859863,
-      "learning_rate": 9.136768449124495e-06,
-      "loss": 0.4012,
-      "step": 9459
-    },
-    {
-      "epoch": 0.6182602444284687,
-      "grad_norm": 0.45646172761917114,
-      "learning_rate": 9.136572305542806e-06,
-      "loss": 0.4114,
-      "step": 9460
-    },
-    {
-      "epoch": 0.6183255996340109,
-      "grad_norm": 0.42141151428222656,
-      "learning_rate": 9.13637614178562e-06,
-      "loss": 0.3213,
-      "step": 9461
-    },
-    {
-      "epoch": 0.618390954839553,
-      "grad_norm": 0.4637605845928192,
-      "learning_rate": 9.136179957853893e-06,
-      "loss": 0.3814,
-      "step": 9462
-    },
-    {
-      "epoch": 0.618456310045095,
-      "grad_norm": 0.449079304933548,
-      "learning_rate": 9.135983753748582e-06,
-      "loss": 0.3794,
-      "step": 9463
-    },
-    {
-      "epoch": 0.6185216652506372,
-      "grad_norm": 0.4622590243816376,
-      "learning_rate": 9.135787529470649e-06,
-      "loss": 0.4105,
-      "step": 9464
-    },
-    {
-      "epoch": 0.6185870204561793,
-      "grad_norm": 0.43435537815093994,
-      "learning_rate": 9.135591285021045e-06,
-      "loss": 0.3599,
-      "step": 9465
-    },
-    {
-      "epoch": 0.6186523756617215,
-      "grad_norm": 0.47140899300575256,
-      "learning_rate": 9.135395020400733e-06,
-      "loss": 0.4003,
-      "step": 9466
-    },
-    {
-      "epoch": 0.6187177308672636,
-      "grad_norm": 0.4565064013004303,
-      "learning_rate": 9.135198735610664e-06,
-      "loss": 0.37,
-      "step": 9467
-    },
-    {
-      "epoch": 0.6187830860728057,
-      "grad_norm": 0.4229309558868408,
-      "learning_rate": 9.135002430651798e-06,
-      "loss": 0.3407,
-      "step": 9468
-    },
-    {
-      "epoch": 0.6188484412783478,
-      "grad_norm": 0.44590333104133606,
-      "learning_rate": 9.134806105525093e-06,
-      "loss": 0.3773,
-      "step": 9469
-    },
-    {
-      "epoch": 0.6189137964838899,
-      "grad_norm": 0.46590927243232727,
-      "learning_rate": 9.134609760231506e-06,
-      "loss": 0.3617,
-      "step": 9470
-    },
-    {
-      "epoch": 0.6189791516894321,
-      "grad_norm": 0.4947170913219452,
-      "learning_rate": 9.134413394771996e-06,
-      "loss": 0.4397,
-      "step": 9471
-    },
-    {
-      "epoch": 0.6190445068949741,
-      "grad_norm": 0.48017892241477966,
-      "learning_rate": 9.134217009147518e-06,
-      "loss": 0.3935,
-      "step": 9472
-    },
-    {
-      "epoch": 0.6191098621005163,
-      "grad_norm": 0.37933260202407837,
-      "learning_rate": 9.134020603359033e-06,
-      "loss": 0.287,
-      "step": 9473
-    },
-    {
-      "epoch": 0.6191752173060584,
-      "grad_norm": 0.4307333528995514,
-      "learning_rate": 9.133824177407496e-06,
-      "loss": 0.3453,
-      "step": 9474
-    },
-    {
-      "epoch": 0.6192405725116006,
-      "grad_norm": 0.4457012712955475,
-      "learning_rate": 9.133627731293868e-06,
-      "loss": 0.4216,
-      "step": 9475
-    },
-    {
-      "epoch": 0.6193059277171427,
-      "grad_norm": 0.4871172308921814,
-      "learning_rate": 9.133431265019106e-06,
-      "loss": 0.3958,
-      "step": 9476
-    },
-    {
-      "epoch": 0.6193712829226848,
-      "grad_norm": 0.43701526522636414,
-      "learning_rate": 9.133234778584166e-06,
-      "loss": 0.4003,
-      "step": 9477
-    },
-    {
-      "epoch": 0.6194366381282269,
-      "grad_norm": 0.4886641204357147,
-      "learning_rate": 9.133038271990007e-06,
-      "loss": 0.4187,
-      "step": 9478
-    },
-    {
-      "epoch": 0.619501993333769,
-      "grad_norm": 0.4610545039176941,
-      "learning_rate": 9.13284174523759e-06,
-      "loss": 0.4227,
-      "step": 9479
-    },
-    {
-      "epoch": 0.6195673485393112,
-      "grad_norm": 0.42697811126708984,
-      "learning_rate": 9.132645198327871e-06,
-      "loss": 0.3507,
-      "step": 9480
-    },
-    {
-      "epoch": 0.6196327037448532,
-      "grad_norm": 0.42546194791793823,
-      "learning_rate": 9.132448631261813e-06,
-      "loss": 0.3554,
-      "step": 9481
-    },
-    {
-      "epoch": 0.6196980589503954,
-      "grad_norm": 0.45359355211257935,
-      "learning_rate": 9.132252044040368e-06,
-      "loss": 0.4129,
-      "step": 9482
-    },
-    {
-      "epoch": 0.6197634141559375,
-      "grad_norm": 0.48987266421318054,
-      "learning_rate": 9.132055436664499e-06,
-      "loss": 0.4749,
-      "step": 9483
-    },
-    {
-      "epoch": 0.6198287693614797,
-      "grad_norm": 0.44682779908180237,
-      "learning_rate": 9.131858809135165e-06,
-      "loss": 0.3918,
-      "step": 9484
-    },
-    {
-      "epoch": 0.6198941245670218,
-      "grad_norm": 0.44950276613235474,
-      "learning_rate": 9.131662161453325e-06,
-      "loss": 0.393,
-      "step": 9485
-    },
-    {
-      "epoch": 0.6199594797725639,
-      "grad_norm": 0.4727022349834442,
-      "learning_rate": 9.131465493619936e-06,
-      "loss": 0.3808,
-      "step": 9486
-    },
-    {
-      "epoch": 0.620024834978106,
-      "grad_norm": 0.4463884234428406,
-      "learning_rate": 9.131268805635958e-06,
-      "loss": 0.4073,
-      "step": 9487
-    },
-    {
-      "epoch": 0.6200901901836481,
-      "grad_norm": 0.47201770544052124,
-      "learning_rate": 9.131072097502352e-06,
-      "loss": 0.4237,
-      "step": 9488
-    },
-    {
-      "epoch": 0.6201555453891903,
-      "grad_norm": 0.47684377431869507,
-      "learning_rate": 9.130875369220074e-06,
-      "loss": 0.4357,
-      "step": 9489
-    },
-    {
-      "epoch": 0.6202209005947323,
-      "grad_norm": 0.4322279095649719,
-      "learning_rate": 9.130678620790088e-06,
-      "loss": 0.3504,
-      "step": 9490
-    },
-    {
-      "epoch": 0.6202862558002745,
-      "grad_norm": 0.5034162998199463,
-      "learning_rate": 9.130481852213351e-06,
-      "loss": 0.4291,
-      "step": 9491
-    },
-    {
-      "epoch": 0.6203516110058166,
-      "grad_norm": 0.426683634519577,
-      "learning_rate": 9.130285063490822e-06,
-      "loss": 0.3221,
-      "step": 9492
-    },
-    {
-      "epoch": 0.6204169662113588,
-      "grad_norm": 0.4612686336040497,
-      "learning_rate": 9.130088254623462e-06,
-      "loss": 0.3957,
-      "step": 9493
-    },
-    {
-      "epoch": 0.6204823214169009,
-      "grad_norm": 0.4370593726634979,
-      "learning_rate": 9.129891425612232e-06,
-      "loss": 0.3698,
-      "step": 9494
-    },
-    {
-      "epoch": 0.6205476766224429,
-      "grad_norm": 0.43835359811782837,
-      "learning_rate": 9.12969457645809e-06,
-      "loss": 0.3693,
-      "step": 9495
-    },
-    {
-      "epoch": 0.6206130318279851,
-      "grad_norm": 0.4201563596725464,
-      "learning_rate": 9.129497707161998e-06,
-      "loss": 0.3282,
-      "step": 9496
-    },
-    {
-      "epoch": 0.6206783870335272,
-      "grad_norm": 0.48881956934928894,
-      "learning_rate": 9.129300817724914e-06,
-      "loss": 0.4248,
-      "step": 9497
-    },
-    {
-      "epoch": 0.6207437422390694,
-      "grad_norm": 0.46462857723236084,
-      "learning_rate": 9.129103908147798e-06,
-      "loss": 0.4152,
-      "step": 9498
-    },
-    {
-      "epoch": 0.6208090974446114,
-      "grad_norm": 0.39168354868888855,
-      "learning_rate": 9.128906978431615e-06,
-      "loss": 0.3081,
-      "step": 9499
-    },
-    {
-      "epoch": 0.6208744526501536,
-      "grad_norm": 0.4305526912212372,
-      "learning_rate": 9.12871002857732e-06,
-      "loss": 0.3577,
-      "step": 9500
-    },
-    {
-      "epoch": 0.6209398078556957,
-      "grad_norm": 0.49615317583084106,
-      "learning_rate": 9.128513058585877e-06,
-      "loss": 0.466,
-      "step": 9501
-    },
-    {
-      "epoch": 0.6210051630612379,
-      "grad_norm": 0.41772010922431946,
-      "learning_rate": 9.128316068458245e-06,
-      "loss": 0.3446,
-      "step": 9502
-    },
-    {
-      "epoch": 0.62107051826678,
-      "grad_norm": 0.455108106136322,
-      "learning_rate": 9.128119058195385e-06,
-      "loss": 0.3907,
-      "step": 9503
-    },
-    {
-      "epoch": 0.621135873472322,
-      "grad_norm": 0.4205496907234192,
-      "learning_rate": 9.127922027798259e-06,
-      "loss": 0.3337,
-      "step": 9504
-    },
-    {
-      "epoch": 0.6212012286778642,
-      "grad_norm": 0.45498114824295044,
-      "learning_rate": 9.127724977267827e-06,
-      "loss": 0.4106,
-      "step": 9505
-    },
-    {
-      "epoch": 0.6212665838834063,
-      "grad_norm": 0.43127432465553284,
-      "learning_rate": 9.12752790660505e-06,
-      "loss": 0.3621,
-      "step": 9506
-    },
-    {
-      "epoch": 0.6213319390889485,
-      "grad_norm": 0.42468681931495667,
-      "learning_rate": 9.127330815810888e-06,
-      "loss": 0.3383,
-      "step": 9507
-    },
-    {
-      "epoch": 0.6213972942944905,
-      "grad_norm": 0.41988471150398254,
-      "learning_rate": 9.127133704886307e-06,
-      "loss": 0.3574,
-      "step": 9508
-    },
-    {
-      "epoch": 0.6214626495000327,
-      "grad_norm": 0.4805356562137604,
-      "learning_rate": 9.126936573832264e-06,
-      "loss": 0.3758,
-      "step": 9509
-    },
-    {
-      "epoch": 0.6215280047055748,
-      "grad_norm": 0.4301708936691284,
-      "learning_rate": 9.12673942264972e-06,
-      "loss": 0.3779,
-      "step": 9510
-    },
-    {
-      "epoch": 0.621593359911117,
-      "grad_norm": 0.4447043836116791,
-      "learning_rate": 9.126542251339639e-06,
-      "loss": 0.364,
-      "step": 9511
-    },
-    {
-      "epoch": 0.621658715116659,
-      "grad_norm": 0.4358770549297333,
-      "learning_rate": 9.126345059902984e-06,
-      "loss": 0.3749,
-      "step": 9512
-    },
-    {
-      "epoch": 0.6217240703222011,
-      "grad_norm": 0.4476655125617981,
-      "learning_rate": 9.126147848340711e-06,
-      "loss": 0.3622,
-      "step": 9513
-    },
-    {
-      "epoch": 0.6217894255277433,
-      "grad_norm": 0.4278465807437897,
-      "learning_rate": 9.125950616653787e-06,
-      "loss": 0.3388,
-      "step": 9514
-    },
-    {
-      "epoch": 0.6218547807332854,
-      "grad_norm": 0.46168437600135803,
-      "learning_rate": 9.125753364843174e-06,
-      "loss": 0.3715,
-      "step": 9515
-    },
-    {
-      "epoch": 0.6219201359388276,
-      "grad_norm": 0.46428367495536804,
-      "learning_rate": 9.12555609290983e-06,
-      "loss": 0.3864,
-      "step": 9516
-    },
-    {
-      "epoch": 0.6219854911443696,
-      "grad_norm": 0.47808322310447693,
-      "learning_rate": 9.125358800854723e-06,
-      "loss": 0.3845,
-      "step": 9517
-    },
-    {
-      "epoch": 0.6220508463499118,
-      "grad_norm": 0.41962409019470215,
-      "learning_rate": 9.12516148867881e-06,
-      "loss": 0.3196,
-      "step": 9518
-    },
-    {
-      "epoch": 0.6221162015554539,
-      "grad_norm": 0.4695814549922943,
-      "learning_rate": 9.124964156383054e-06,
-      "loss": 0.4069,
-      "step": 9519
-    },
-    {
-      "epoch": 0.6221815567609961,
-      "grad_norm": 0.4293384850025177,
-      "learning_rate": 9.124766803968421e-06,
-      "loss": 0.3695,
-      "step": 9520
-    },
-    {
-      "epoch": 0.6222469119665381,
-      "grad_norm": 0.4725792109966278,
-      "learning_rate": 9.12456943143587e-06,
-      "loss": 0.4653,
-      "step": 9521
-    },
-    {
-      "epoch": 0.6223122671720802,
-      "grad_norm": 0.4451233446598053,
-      "learning_rate": 9.124372038786366e-06,
-      "loss": 0.36,
-      "step": 9522
-    },
-    {
-      "epoch": 0.6223776223776224,
-      "grad_norm": 0.46407151222229004,
-      "learning_rate": 9.124174626020869e-06,
-      "loss": 0.3954,
-      "step": 9523
-    },
-    {
-      "epoch": 0.6224429775831645,
-      "grad_norm": 0.4411452114582062,
-      "learning_rate": 9.123977193140346e-06,
-      "loss": 0.362,
-      "step": 9524
-    },
-    {
-      "epoch": 0.6225083327887067,
-      "grad_norm": 0.47463229298591614,
-      "learning_rate": 9.123779740145758e-06,
-      "loss": 0.3988,
-      "step": 9525
-    },
-    {
-      "epoch": 0.6225736879942487,
-      "grad_norm": 0.44011837244033813,
-      "learning_rate": 9.123582267038064e-06,
-      "loss": 0.3771,
-      "step": 9526
-    },
-    {
-      "epoch": 0.6226390431997909,
-      "grad_norm": 0.4925641119480133,
-      "learning_rate": 9.123384773818234e-06,
-      "loss": 0.3735,
-      "step": 9527
-    },
-    {
-      "epoch": 0.622704398405333,
-      "grad_norm": 0.4725680351257324,
-      "learning_rate": 9.123187260487226e-06,
-      "loss": 0.4013,
-      "step": 9528
-    },
-    {
-      "epoch": 0.622769753610875,
-      "grad_norm": 0.4580378532409668,
-      "learning_rate": 9.122989727046006e-06,
-      "loss": 0.4369,
-      "step": 9529
-    },
-    {
-      "epoch": 0.6228351088164172,
-      "grad_norm": 0.5389646291732788,
-      "learning_rate": 9.122792173495536e-06,
-      "loss": 0.3723,
-      "step": 9530
-    },
-    {
-      "epoch": 0.6229004640219593,
-      "grad_norm": 0.42884403467178345,
-      "learning_rate": 9.122594599836783e-06,
-      "loss": 0.3548,
-      "step": 9531
-    },
-    {
-      "epoch": 0.6229658192275015,
-      "grad_norm": 0.4398644268512726,
-      "learning_rate": 9.122397006070705e-06,
-      "loss": 0.3535,
-      "step": 9532
-    },
-    {
-      "epoch": 0.6230311744330436,
-      "grad_norm": 0.45770788192749023,
-      "learning_rate": 9.12219939219827e-06,
-      "loss": 0.3725,
-      "step": 9533
-    },
-    {
-      "epoch": 0.6230965296385857,
-      "grad_norm": 0.43530353903770447,
-      "learning_rate": 9.12200175822044e-06,
-      "loss": 0.3573,
-      "step": 9534
-    },
-    {
-      "epoch": 0.6231618848441278,
-      "grad_norm": 0.44287198781967163,
-      "learning_rate": 9.121804104138178e-06,
-      "loss": 0.3484,
-      "step": 9535
-    },
-    {
-      "epoch": 0.62322724004967,
-      "grad_norm": 0.43603357672691345,
-      "learning_rate": 9.121606429952453e-06,
-      "loss": 0.365,
-      "step": 9536
-    },
-    {
-      "epoch": 0.6232925952552121,
-      "grad_norm": 0.4030922055244446,
-      "learning_rate": 9.121408735664223e-06,
-      "loss": 0.3405,
-      "step": 9537
-    },
-    {
-      "epoch": 0.6233579504607542,
-      "grad_norm": 0.45667538046836853,
-      "learning_rate": 9.121211021274456e-06,
-      "loss": 0.3942,
-      "step": 9538
-    },
-    {
-      "epoch": 0.6234233056662963,
-      "grad_norm": 0.4265606999397278,
-      "learning_rate": 9.121013286784114e-06,
-      "loss": 0.372,
-      "step": 9539
-    },
-    {
-      "epoch": 0.6234886608718384,
-      "grad_norm": 0.4281396269798279,
-      "learning_rate": 9.120815532194162e-06,
-      "loss": 0.3541,
-      "step": 9540
-    },
-    {
-      "epoch": 0.6235540160773806,
-      "grad_norm": 0.439257949590683,
-      "learning_rate": 9.120617757505568e-06,
-      "loss": 0.3594,
-      "step": 9541
-    },
-    {
-      "epoch": 0.6236193712829227,
-      "grad_norm": 0.4113325774669647,
-      "learning_rate": 9.120419962719291e-06,
-      "loss": 0.3595,
-      "step": 9542
-    },
-    {
-      "epoch": 0.6236847264884648,
-      "grad_norm": 0.4289771616458893,
-      "learning_rate": 9.120222147836299e-06,
-      "loss": 0.36,
-      "step": 9543
-    },
-    {
-      "epoch": 0.6237500816940069,
-      "grad_norm": 0.47416773438453674,
-      "learning_rate": 9.120024312857557e-06,
-      "loss": 0.3907,
-      "step": 9544
-    },
-    {
-      "epoch": 0.6238154368995491,
-      "grad_norm": 0.42772915959358215,
-      "learning_rate": 9.119826457784028e-06,
-      "loss": 0.3318,
-      "step": 9545
-    },
-    {
-      "epoch": 0.6238807921050912,
-      "grad_norm": 0.46171560883522034,
-      "learning_rate": 9.11962858261668e-06,
-      "loss": 0.4351,
-      "step": 9546
-    },
-    {
-      "epoch": 0.6239461473106332,
-      "grad_norm": 0.4971867501735687,
-      "learning_rate": 9.119430687356474e-06,
-      "loss": 0.4852,
-      "step": 9547
-    },
-    {
-      "epoch": 0.6240115025161754,
-      "grad_norm": 0.46011096239089966,
-      "learning_rate": 9.119232772004378e-06,
-      "loss": 0.3876,
-      "step": 9548
-    },
-    {
-      "epoch": 0.6240768577217175,
-      "grad_norm": 0.4509351849555969,
-      "learning_rate": 9.119034836561358e-06,
-      "loss": 0.3718,
-      "step": 9549
-    },
-    {
-      "epoch": 0.6241422129272597,
-      "grad_norm": 0.43959373235702515,
-      "learning_rate": 9.118836881028377e-06,
-      "loss": 0.3949,
-      "step": 9550
-    },
-    {
-      "epoch": 0.6242075681328018,
-      "grad_norm": 0.43528345227241516,
-      "learning_rate": 9.118638905406402e-06,
-      "loss": 0.3672,
-      "step": 9551
-    },
-    {
-      "epoch": 0.6242729233383439,
-      "grad_norm": 0.5049495100975037,
-      "learning_rate": 9.118440909696397e-06,
-      "loss": 0.4554,
-      "step": 9552
-    },
-    {
-      "epoch": 0.624338278543886,
-      "grad_norm": 0.5229260325431824,
-      "learning_rate": 9.118242893899331e-06,
-      "loss": 0.4922,
-      "step": 9553
-    },
-    {
-      "epoch": 0.6244036337494281,
-      "grad_norm": 0.44667717814445496,
-      "learning_rate": 9.118044858016166e-06,
-      "loss": 0.3649,
-      "step": 9554
-    },
-    {
-      "epoch": 0.6244689889549703,
-      "grad_norm": 0.4446372091770172,
-      "learning_rate": 9.117846802047871e-06,
-      "loss": 0.3869,
-      "step": 9555
-    },
-    {
-      "epoch": 0.6245343441605123,
-      "grad_norm": 0.45658573508262634,
-      "learning_rate": 9.117648725995409e-06,
-      "loss": 0.3918,
-      "step": 9556
-    },
-    {
-      "epoch": 0.6245996993660545,
-      "grad_norm": 0.48933708667755127,
-      "learning_rate": 9.11745062985975e-06,
-      "loss": 0.4334,
-      "step": 9557
-    },
-    {
-      "epoch": 0.6246650545715966,
-      "grad_norm": 0.48574692010879517,
-      "learning_rate": 9.117252513641855e-06,
-      "loss": 0.4377,
-      "step": 9558
-    },
-    {
-      "epoch": 0.6247304097771388,
-      "grad_norm": 0.5007457733154297,
-      "learning_rate": 9.117054377342695e-06,
-      "loss": 0.4446,
-      "step": 9559
-    },
-    {
-      "epoch": 0.6247957649826809,
-      "grad_norm": 0.4362432658672333,
-      "learning_rate": 9.116856220963236e-06,
-      "loss": 0.3517,
-      "step": 9560
-    },
-    {
-      "epoch": 0.624861120188223,
-      "grad_norm": 0.4492630362510681,
-      "learning_rate": 9.11665804450444e-06,
-      "loss": 0.3935,
-      "step": 9561
-    },
-    {
-      "epoch": 0.6249264753937651,
-      "grad_norm": 0.45661497116088867,
-      "learning_rate": 9.116459847967276e-06,
-      "loss": 0.4001,
-      "step": 9562
-    },
-    {
-      "epoch": 0.6249918305993072,
-      "grad_norm": 0.44248563051223755,
-      "learning_rate": 9.116261631352714e-06,
-      "loss": 0.3819,
-      "step": 9563
-    },
-    {
-      "epoch": 0.6250571858048494,
-      "grad_norm": 0.4681680500507355,
-      "learning_rate": 9.116063394661716e-06,
-      "loss": 0.4456,
-      "step": 9564
-    },
-    {
-      "epoch": 0.6251225410103914,
-      "grad_norm": 0.45694154500961304,
-      "learning_rate": 9.115865137895252e-06,
-      "loss": 0.3714,
-      "step": 9565
-    },
-    {
-      "epoch": 0.6251878962159336,
-      "grad_norm": 0.46180787682533264,
-      "learning_rate": 9.115666861054289e-06,
-      "loss": 0.3567,
-      "step": 9566
-    },
-    {
-      "epoch": 0.6252532514214757,
-      "grad_norm": 0.4309927821159363,
-      "learning_rate": 9.115468564139791e-06,
-      "loss": 0.3564,
-      "step": 9567
-    },
-    {
-      "epoch": 0.6253186066270179,
-      "grad_norm": 0.48116418719291687,
-      "learning_rate": 9.115270247152728e-06,
-      "loss": 0.4245,
-      "step": 9568
-    },
-    {
-      "epoch": 0.62538396183256,
-      "grad_norm": 0.4717462956905365,
-      "learning_rate": 9.115071910094065e-06,
-      "loss": 0.4339,
-      "step": 9569
-    },
-    {
-      "epoch": 0.6254493170381021,
-      "grad_norm": 0.47250643372535706,
-      "learning_rate": 9.114873552964771e-06,
-      "loss": 0.3938,
-      "step": 9570
-    },
-    {
-      "epoch": 0.6255146722436442,
-      "grad_norm": 0.4773271977901459,
-      "learning_rate": 9.114675175765814e-06,
-      "loss": 0.3937,
-      "step": 9571
-    },
-    {
-      "epoch": 0.6255800274491863,
-      "grad_norm": 0.4597764313220978,
-      "learning_rate": 9.114476778498161e-06,
-      "loss": 0.4021,
-      "step": 9572
-    },
-    {
-      "epoch": 0.6256453826547285,
-      "grad_norm": 0.47342589497566223,
-      "learning_rate": 9.114278361162778e-06,
-      "loss": 0.4336,
-      "step": 9573
-    },
-    {
-      "epoch": 0.6257107378602705,
-      "grad_norm": 0.4563991129398346,
-      "learning_rate": 9.114079923760636e-06,
-      "loss": 0.3922,
-      "step": 9574
-    },
-    {
-      "epoch": 0.6257760930658127,
-      "grad_norm": 0.4252253472805023,
-      "learning_rate": 9.1138814662927e-06,
-      "loss": 0.3307,
-      "step": 9575
-    },
-    {
-      "epoch": 0.6258414482713548,
-      "grad_norm": 0.4634683430194855,
-      "learning_rate": 9.11368298875994e-06,
-      "loss": 0.3882,
-      "step": 9576
-    },
-    {
-      "epoch": 0.625906803476897,
-      "grad_norm": 0.4656892716884613,
-      "learning_rate": 9.11348449116332e-06,
-      "loss": 0.4308,
-      "step": 9577
-    },
-    {
-      "epoch": 0.625972158682439,
-      "grad_norm": 0.45494982600212097,
-      "learning_rate": 9.113285973503813e-06,
-      "loss": 0.4255,
-      "step": 9578
-    },
-    {
-      "epoch": 0.6260375138879811,
-      "grad_norm": 0.45242008566856384,
-      "learning_rate": 9.113087435782387e-06,
-      "loss": 0.3886,
-      "step": 9579
-    },
-    {
-      "epoch": 0.6261028690935233,
-      "grad_norm": 0.46740055084228516,
-      "learning_rate": 9.112888878000005e-06,
-      "loss": 0.3931,
-      "step": 9580
-    },
-    {
-      "epoch": 0.6261682242990654,
-      "grad_norm": 0.4570982754230499,
-      "learning_rate": 9.112690300157642e-06,
-      "loss": 0.4083,
-      "step": 9581
-    },
-    {
-      "epoch": 0.6262335795046076,
-      "grad_norm": 0.43806904554367065,
-      "learning_rate": 9.112491702256262e-06,
-      "loss": 0.3553,
-      "step": 9582
-    },
-    {
-      "epoch": 0.6262989347101496,
-      "grad_norm": 0.46506205201148987,
-      "learning_rate": 9.112293084296836e-06,
-      "loss": 0.3455,
-      "step": 9583
-    },
-    {
-      "epoch": 0.6263642899156918,
-      "grad_norm": 0.4456247389316559,
-      "learning_rate": 9.112094446280332e-06,
-      "loss": 0.3385,
-      "step": 9584
-    },
-    {
-      "epoch": 0.6264296451212339,
-      "grad_norm": 0.4501979649066925,
-      "learning_rate": 9.111895788207718e-06,
-      "loss": 0.3652,
-      "step": 9585
-    },
-    {
-      "epoch": 0.6264950003267761,
-      "grad_norm": 0.45896342396736145,
-      "learning_rate": 9.111697110079964e-06,
-      "loss": 0.3807,
-      "step": 9586
-    },
-    {
-      "epoch": 0.6265603555323181,
-      "grad_norm": 0.4563565254211426,
-      "learning_rate": 9.11149841189804e-06,
-      "loss": 0.3342,
-      "step": 9587
-    },
-    {
-      "epoch": 0.6266257107378602,
-      "grad_norm": 0.4744052588939667,
-      "learning_rate": 9.111299693662913e-06,
-      "loss": 0.3988,
-      "step": 9588
-    },
-    {
-      "epoch": 0.6266910659434024,
-      "grad_norm": 0.4680745303630829,
-      "learning_rate": 9.111100955375554e-06,
-      "loss": 0.4467,
-      "step": 9589
-    },
-    {
-      "epoch": 0.6267564211489445,
-      "grad_norm": 0.5005598068237305,
-      "learning_rate": 9.110902197036931e-06,
-      "loss": 0.4595,
-      "step": 9590
-    },
-    {
-      "epoch": 0.6268217763544867,
-      "grad_norm": 0.4681922197341919,
-      "learning_rate": 9.110703418648012e-06,
-      "loss": 0.4319,
-      "step": 9591
-    },
-    {
-      "epoch": 0.6268871315600287,
-      "grad_norm": 0.47896018624305725,
-      "learning_rate": 9.11050462020977e-06,
-      "loss": 0.4208,
-      "step": 9592
-    },
-    {
-      "epoch": 0.6269524867655709,
-      "grad_norm": 0.4624868333339691,
-      "learning_rate": 9.110305801723173e-06,
-      "loss": 0.3888,
-      "step": 9593
-    },
-    {
-      "epoch": 0.627017841971113,
-      "grad_norm": 0.4650159478187561,
-      "learning_rate": 9.11010696318919e-06,
-      "loss": 0.3707,
-      "step": 9594
-    },
-    {
-      "epoch": 0.6270831971766552,
-      "grad_norm": 0.4010483920574188,
-      "learning_rate": 9.109908104608792e-06,
-      "loss": 0.3541,
-      "step": 9595
-    },
-    {
-      "epoch": 0.6271485523821972,
-      "grad_norm": 0.44370007514953613,
-      "learning_rate": 9.109709225982947e-06,
-      "loss": 0.3813,
-      "step": 9596
-    },
-    {
-      "epoch": 0.6272139075877393,
-      "grad_norm": 0.4745508134365082,
-      "learning_rate": 9.109510327312628e-06,
-      "loss": 0.4053,
-      "step": 9597
-    },
-    {
-      "epoch": 0.6272792627932815,
-      "grad_norm": 0.4936065971851349,
-      "learning_rate": 9.109311408598805e-06,
-      "loss": 0.418,
-      "step": 9598
-    },
-    {
-      "epoch": 0.6273446179988236,
-      "grad_norm": 0.4746595621109009,
-      "learning_rate": 9.109112469842442e-06,
-      "loss": 0.4025,
-      "step": 9599
-    },
-    {
-      "epoch": 0.6274099732043658,
-      "grad_norm": 0.4503948986530304,
-      "learning_rate": 9.108913511044519e-06,
-      "loss": 0.3995,
-      "step": 9600
-    },
-    {
-      "epoch": 0.6274753284099078,
-      "grad_norm": 0.4455104470252991,
-      "learning_rate": 9.108714532205998e-06,
-      "loss": 0.358,
-      "step": 9601
-    },
-    {
-      "epoch": 0.62754068361545,
-      "grad_norm": 0.43602004647254944,
-      "learning_rate": 9.108515533327855e-06,
-      "loss": 0.3617,
-      "step": 9602
-    },
-    {
-      "epoch": 0.6276060388209921,
-      "grad_norm": 0.43457356095314026,
-      "learning_rate": 9.108316514411057e-06,
-      "loss": 0.3284,
-      "step": 9603
-    },
-    {
-      "epoch": 0.6276713940265343,
-      "grad_norm": 0.5141921639442444,
-      "learning_rate": 9.108117475456575e-06,
-      "loss": 0.3951,
-      "step": 9604
-    },
-    {
-      "epoch": 0.6277367492320763,
-      "grad_norm": 0.4688136577606201,
-      "learning_rate": 9.107918416465382e-06,
-      "loss": 0.3644,
-      "step": 9605
-    },
-    {
-      "epoch": 0.6278021044376184,
-      "grad_norm": 0.4773247241973877,
-      "learning_rate": 9.107719337438449e-06,
-      "loss": 0.4137,
-      "step": 9606
-    },
-    {
-      "epoch": 0.6278674596431606,
-      "grad_norm": 0.4520193934440613,
-      "learning_rate": 9.107520238376745e-06,
-      "loss": 0.402,
-      "step": 9607
-    },
-    {
-      "epoch": 0.6279328148487027,
-      "grad_norm": 0.4773165285587311,
-      "learning_rate": 9.10732111928124e-06,
-      "loss": 0.4258,
-      "step": 9608
-    },
-    {
-      "epoch": 0.6279981700542449,
-      "grad_norm": 0.47521254420280457,
-      "learning_rate": 9.107121980152908e-06,
-      "loss": 0.4111,
-      "step": 9609
-    },
-    {
-      "epoch": 0.6280635252597869,
-      "grad_norm": 0.5069689154624939,
-      "learning_rate": 9.106922820992721e-06,
-      "loss": 0.4194,
-      "step": 9610
-    },
-    {
-      "epoch": 0.6281288804653291,
-      "grad_norm": 0.460077166557312,
-      "learning_rate": 9.106723641801648e-06,
-      "loss": 0.3834,
-      "step": 9611
-    },
-    {
-      "epoch": 0.6281942356708712,
-      "grad_norm": 0.46278902888298035,
-      "learning_rate": 9.10652444258066e-06,
-      "loss": 0.4018,
-      "step": 9612
-    },
-    {
-      "epoch": 0.6282595908764133,
-      "grad_norm": 0.42567697167396545,
-      "learning_rate": 9.10632522333073e-06,
-      "loss": 0.3718,
-      "step": 9613
-    },
-    {
-      "epoch": 0.6283249460819554,
-      "grad_norm": 0.4435652494430542,
-      "learning_rate": 9.10612598405283e-06,
-      "loss": 0.3729,
-      "step": 9614
-    },
-    {
-      "epoch": 0.6283903012874975,
-      "grad_norm": 0.43816253542900085,
-      "learning_rate": 9.10592672474793e-06,
-      "loss": 0.3615,
-      "step": 9615
-    },
-    {
-      "epoch": 0.6284556564930397,
-      "grad_norm": 0.47340625524520874,
-      "learning_rate": 9.105727445417002e-06,
-      "loss": 0.4025,
-      "step": 9616
-    },
-    {
-      "epoch": 0.6285210116985818,
-      "grad_norm": 0.4708377420902252,
-      "learning_rate": 9.105528146061023e-06,
-      "loss": 0.3429,
-      "step": 9617
-    },
-    {
-      "epoch": 0.628586366904124,
-      "grad_norm": 0.48139289021492004,
-      "learning_rate": 9.105328826680957e-06,
-      "loss": 0.4622,
-      "step": 9618
-    },
-    {
-      "epoch": 0.628651722109666,
-      "grad_norm": 0.44649291038513184,
-      "learning_rate": 9.105129487277781e-06,
-      "loss": 0.4029,
-      "step": 9619
-    },
-    {
-      "epoch": 0.6287170773152082,
-      "grad_norm": 0.4421707093715668,
-      "learning_rate": 9.104930127852468e-06,
-      "loss": 0.355,
-      "step": 9620
-    },
-    {
-      "epoch": 0.6287824325207503,
-      "grad_norm": 0.46626222133636475,
-      "learning_rate": 9.104730748405988e-06,
-      "loss": 0.3937,
-      "step": 9621
-    },
-    {
-      "epoch": 0.6288477877262924,
-      "grad_norm": 0.479063481092453,
-      "learning_rate": 9.104531348939313e-06,
-      "loss": 0.4117,
-      "step": 9622
-    },
-    {
-      "epoch": 0.6289131429318345,
-      "grad_norm": 0.44687584042549133,
-      "learning_rate": 9.104331929453417e-06,
-      "loss": 0.4084,
-      "step": 9623
-    },
-    {
-      "epoch": 0.6289784981373766,
-      "grad_norm": 0.49097737669944763,
-      "learning_rate": 9.104132489949272e-06,
-      "loss": 0.3949,
-      "step": 9624
-    },
-    {
-      "epoch": 0.6290438533429188,
-      "grad_norm": 0.4492281973361969,
-      "learning_rate": 9.103933030427852e-06,
-      "loss": 0.4013,
-      "step": 9625
-    },
-    {
-      "epoch": 0.6291092085484609,
-      "grad_norm": 0.44036224484443665,
-      "learning_rate": 9.103733550890128e-06,
-      "loss": 0.3528,
-      "step": 9626
-    },
-    {
-      "epoch": 0.629174563754003,
-      "grad_norm": 0.41973841190338135,
-      "learning_rate": 9.103534051337074e-06,
-      "loss": 0.3228,
-      "step": 9627
-    },
-    {
-      "epoch": 0.6292399189595451,
-      "grad_norm": 0.46852245926856995,
-      "learning_rate": 9.103334531769664e-06,
-      "loss": 0.405,
-      "step": 9628
-    },
-    {
-      "epoch": 0.6293052741650873,
-      "grad_norm": 0.4788178503513336,
-      "learning_rate": 9.103134992188869e-06,
-      "loss": 0.3918,
-      "step": 9629
-    },
-    {
-      "epoch": 0.6293706293706294,
-      "grad_norm": 0.45014268159866333,
-      "learning_rate": 9.102935432595664e-06,
-      "loss": 0.4112,
-      "step": 9630
-    },
-    {
-      "epoch": 0.6294359845761714,
-      "grad_norm": 0.4370774030685425,
-      "learning_rate": 9.102735852991019e-06,
-      "loss": 0.3529,
-      "step": 9631
-    },
-    {
-      "epoch": 0.6295013397817136,
-      "grad_norm": 0.42277172207832336,
-      "learning_rate": 9.102536253375913e-06,
-      "loss": 0.3499,
-      "step": 9632
-    },
-    {
-      "epoch": 0.6295666949872557,
-      "grad_norm": 0.5183576941490173,
-      "learning_rate": 9.102336633751314e-06,
-      "loss": 0.4955,
-      "step": 9633
-    },
-    {
-      "epoch": 0.6296320501927979,
-      "grad_norm": 0.45432087779045105,
-      "learning_rate": 9.1021369941182e-06,
-      "loss": 0.3891,
-      "step": 9634
-    },
-    {
-      "epoch": 0.62969740539834,
-      "grad_norm": 0.4378865659236908,
-      "learning_rate": 9.101937334477542e-06,
-      "loss": 0.3496,
-      "step": 9635
-    },
-    {
-      "epoch": 0.6297627606038821,
-      "grad_norm": 0.4116170108318329,
-      "learning_rate": 9.101737654830313e-06,
-      "loss": 0.3077,
-      "step": 9636
-    },
-    {
-      "epoch": 0.6298281158094242,
-      "grad_norm": 0.5482897162437439,
-      "learning_rate": 9.101537955177491e-06,
-      "loss": 0.3867,
-      "step": 9637
-    },
-    {
-      "epoch": 0.6298934710149663,
-      "grad_norm": 0.4400225877761841,
-      "learning_rate": 9.101338235520046e-06,
-      "loss": 0.377,
-      "step": 9638
-    },
-    {
-      "epoch": 0.6299588262205085,
-      "grad_norm": 0.41989219188690186,
-      "learning_rate": 9.101138495858954e-06,
-      "loss": 0.3508,
-      "step": 9639
-    },
-    {
-      "epoch": 0.6300241814260505,
-      "grad_norm": 0.446665495634079,
-      "learning_rate": 9.100938736195188e-06,
-      "loss": 0.3776,
-      "step": 9640
-    },
-    {
-      "epoch": 0.6300895366315927,
-      "grad_norm": 0.47875264286994934,
-      "learning_rate": 9.100738956529724e-06,
-      "loss": 0.4348,
-      "step": 9641
-    },
-    {
-      "epoch": 0.6301548918371348,
-      "grad_norm": 0.4916093349456787,
-      "learning_rate": 9.100539156863536e-06,
-      "loss": 0.4538,
-      "step": 9642
-    },
-    {
-      "epoch": 0.630220247042677,
-      "grad_norm": 0.49365484714508057,
-      "learning_rate": 9.100339337197597e-06,
-      "loss": 0.421,
-      "step": 9643
-    },
-    {
-      "epoch": 0.6302856022482191,
-      "grad_norm": 0.41142141819000244,
-      "learning_rate": 9.100139497532882e-06,
-      "loss": 0.2936,
-      "step": 9644
-    },
-    {
-      "epoch": 0.6303509574537612,
-      "grad_norm": 0.5032975673675537,
-      "learning_rate": 9.099939637870369e-06,
-      "loss": 0.4358,
-      "step": 9645
-    },
-    {
-      "epoch": 0.6304163126593033,
-      "grad_norm": 0.4381714463233948,
-      "learning_rate": 9.099739758211028e-06,
-      "loss": 0.351,
-      "step": 9646
-    },
-    {
-      "epoch": 0.6304816678648454,
-      "grad_norm": 0.49288493394851685,
-      "learning_rate": 9.099539858555836e-06,
-      "loss": 0.4392,
-      "step": 9647
-    },
-    {
-      "epoch": 0.6305470230703876,
-      "grad_norm": 0.4541897177696228,
-      "learning_rate": 9.099339938905767e-06,
-      "loss": 0.3679,
-      "step": 9648
-    },
-    {
-      "epoch": 0.6306123782759296,
-      "grad_norm": 0.6171749830245972,
-      "learning_rate": 9.099139999261799e-06,
-      "loss": 0.3491,
-      "step": 9649
-    },
-    {
-      "epoch": 0.6306777334814718,
-      "grad_norm": 0.41822025179862976,
-      "learning_rate": 9.098940039624904e-06,
-      "loss": 0.3692,
-      "step": 9650
-    },
-    {
-      "epoch": 0.6307430886870139,
-      "grad_norm": 0.4923498034477234,
-      "learning_rate": 9.098740059996058e-06,
-      "loss": 0.4255,
-      "step": 9651
-    },
-    {
-      "epoch": 0.6308084438925561,
-      "grad_norm": 0.48193684220314026,
-      "learning_rate": 9.098540060376238e-06,
-      "loss": 0.3788,
-      "step": 9652
-    },
-    {
-      "epoch": 0.6308737990980982,
-      "grad_norm": 0.46530207991600037,
-      "learning_rate": 9.098340040766417e-06,
-      "loss": 0.3639,
-      "step": 9653
-    },
-    {
-      "epoch": 0.6309391543036403,
-      "grad_norm": 0.4175734221935272,
-      "learning_rate": 9.098140001167572e-06,
-      "loss": 0.3546,
-      "step": 9654
-    },
-    {
-      "epoch": 0.6310045095091824,
-      "grad_norm": 0.4892028570175171,
-      "learning_rate": 9.097939941580679e-06,
-      "loss": 0.3784,
-      "step": 9655
-    },
-    {
-      "epoch": 0.6310698647147245,
-      "grad_norm": 0.48101481795310974,
-      "learning_rate": 9.097739862006714e-06,
-      "loss": 0.4351,
-      "step": 9656
-    },
-    {
-      "epoch": 0.6311352199202667,
-      "grad_norm": 0.43300995230674744,
-      "learning_rate": 9.09753976244665e-06,
-      "loss": 0.3522,
-      "step": 9657
-    },
-    {
-      "epoch": 0.6312005751258087,
-      "grad_norm": 0.4683457612991333,
-      "learning_rate": 9.097339642901466e-06,
-      "loss": 0.3994,
-      "step": 9658
-    },
-    {
-      "epoch": 0.6312659303313509,
-      "grad_norm": 0.47194018959999084,
-      "learning_rate": 9.097139503372138e-06,
-      "loss": 0.3942,
-      "step": 9659
-    },
-    {
-      "epoch": 0.631331285536893,
-      "grad_norm": 0.4986540377140045,
-      "learning_rate": 9.096939343859641e-06,
-      "loss": 0.3847,
-      "step": 9660
-    },
-    {
-      "epoch": 0.6313966407424352,
-      "grad_norm": 0.6394871473312378,
-      "learning_rate": 9.09673916436495e-06,
-      "loss": 0.3852,
-      "step": 9661
-    },
-    {
-      "epoch": 0.6314619959479773,
-      "grad_norm": 0.44969215989112854,
-      "learning_rate": 9.096538964889043e-06,
-      "loss": 0.3612,
-      "step": 9662
-    },
-    {
-      "epoch": 0.6315273511535193,
-      "grad_norm": 0.4730345904827118,
-      "learning_rate": 9.096338745432899e-06,
-      "loss": 0.3715,
-      "step": 9663
-    },
-    {
-      "epoch": 0.6315927063590615,
-      "grad_norm": 0.456365168094635,
-      "learning_rate": 9.096138505997489e-06,
-      "loss": 0.4057,
-      "step": 9664
-    },
-    {
-      "epoch": 0.6316580615646036,
-      "grad_norm": 0.440775066614151,
-      "learning_rate": 9.095938246583796e-06,
-      "loss": 0.3604,
-      "step": 9665
-    },
-    {
-      "epoch": 0.6317234167701458,
-      "grad_norm": 0.4914442002773285,
-      "learning_rate": 9.09573796719279e-06,
-      "loss": 0.3982,
-      "step": 9666
-    },
-    {
-      "epoch": 0.6317887719756878,
-      "grad_norm": 0.49754732847213745,
-      "learning_rate": 9.095537667825452e-06,
-      "loss": 0.4219,
-      "step": 9667
-    },
-    {
-      "epoch": 0.63185412718123,
-      "grad_norm": 0.42917394638061523,
-      "learning_rate": 9.095337348482757e-06,
-      "loss": 0.3613,
-      "step": 9668
-    },
-    {
-      "epoch": 0.6319194823867721,
-      "grad_norm": 0.4514814615249634,
-      "learning_rate": 9.095137009165682e-06,
-      "loss": 0.4145,
-      "step": 9669
-    },
-    {
-      "epoch": 0.6319848375923143,
-      "grad_norm": 0.4623293876647949,
-      "learning_rate": 9.094936649875207e-06,
-      "loss": 0.3888,
-      "step": 9670
-    },
-    {
-      "epoch": 0.6320501927978563,
-      "grad_norm": 0.45264896750450134,
-      "learning_rate": 9.094736270612308e-06,
-      "loss": 0.3523,
-      "step": 9671
-    },
-    {
-      "epoch": 0.6321155480033984,
-      "grad_norm": 0.6229252815246582,
-      "learning_rate": 9.094535871377961e-06,
-      "loss": 0.4706,
-      "step": 9672
-    },
-    {
-      "epoch": 0.6321809032089406,
-      "grad_norm": 0.48615366220474243,
-      "learning_rate": 9.094335452173144e-06,
-      "loss": 0.4286,
-      "step": 9673
-    },
-    {
-      "epoch": 0.6322462584144827,
-      "grad_norm": 0.46924880146980286,
-      "learning_rate": 9.094135012998834e-06,
-      "loss": 0.4273,
-      "step": 9674
-    },
-    {
-      "epoch": 0.6323116136200249,
-      "grad_norm": 0.46026837825775146,
-      "learning_rate": 9.09393455385601e-06,
-      "loss": 0.3809,
-      "step": 9675
-    },
-    {
-      "epoch": 0.6323769688255669,
-      "grad_norm": 0.5779929161071777,
-      "learning_rate": 9.093734074745649e-06,
-      "loss": 0.3373,
-      "step": 9676
-    },
-    {
-      "epoch": 0.6324423240311091,
-      "grad_norm": 0.47199392318725586,
-      "learning_rate": 9.093533575668728e-06,
-      "loss": 0.3852,
-      "step": 9677
-    },
-    {
-      "epoch": 0.6325076792366512,
-      "grad_norm": 0.4600350856781006,
-      "learning_rate": 9.093333056626226e-06,
-      "loss": 0.3555,
-      "step": 9678
-    },
-    {
-      "epoch": 0.6325730344421934,
-      "grad_norm": 0.4776606857776642,
-      "learning_rate": 9.09313251761912e-06,
-      "loss": 0.4381,
-      "step": 9679
-    },
-    {
-      "epoch": 0.6326383896477354,
-      "grad_norm": 0.47482728958129883,
-      "learning_rate": 9.09293195864839e-06,
-      "loss": 0.3769,
-      "step": 9680
-    },
-    {
-      "epoch": 0.6327037448532775,
-      "grad_norm": 0.4603939354419708,
-      "learning_rate": 9.092731379715012e-06,
-      "loss": 0.3632,
-      "step": 9681
-    },
-    {
-      "epoch": 0.6327691000588197,
-      "grad_norm": 0.42856353521347046,
-      "learning_rate": 9.092530780819965e-06,
-      "loss": 0.3697,
-      "step": 9682
-    },
-    {
-      "epoch": 0.6328344552643618,
-      "grad_norm": 0.44450482726097107,
-      "learning_rate": 9.092330161964229e-06,
-      "loss": 0.3556,
-      "step": 9683
-    },
-    {
-      "epoch": 0.632899810469904,
-      "grad_norm": 0.4556718170642853,
-      "learning_rate": 9.09212952314878e-06,
-      "loss": 0.3703,
-      "step": 9684
-    },
-    {
-      "epoch": 0.632965165675446,
-      "grad_norm": 0.4377409815788269,
-      "learning_rate": 9.091928864374597e-06,
-      "loss": 0.3618,
-      "step": 9685
-    },
-    {
-      "epoch": 0.6330305208809882,
-      "grad_norm": 0.47488972544670105,
-      "learning_rate": 9.09172818564266e-06,
-      "loss": 0.401,
-      "step": 9686
-    },
-    {
-      "epoch": 0.6330958760865303,
-      "grad_norm": 0.4416325092315674,
-      "learning_rate": 9.091527486953947e-06,
-      "loss": 0.422,
-      "step": 9687
-    },
-    {
-      "epoch": 0.6331612312920725,
-      "grad_norm": 0.48478466272354126,
-      "learning_rate": 9.091326768309437e-06,
-      "loss": 0.4016,
-      "step": 9688
-    },
-    {
-      "epoch": 0.6332265864976145,
-      "grad_norm": 0.45282062888145447,
-      "learning_rate": 9.091126029710109e-06,
-      "loss": 0.4032,
-      "step": 9689
-    },
-    {
-      "epoch": 0.6332919417031566,
-      "grad_norm": 0.4405279755592346,
-      "learning_rate": 9.090925271156944e-06,
-      "loss": 0.3871,
-      "step": 9690
-    },
-    {
-      "epoch": 0.6333572969086988,
-      "grad_norm": 0.42671340703964233,
-      "learning_rate": 9.090724492650915e-06,
-      "loss": 0.3695,
-      "step": 9691
-    },
-    {
-      "epoch": 0.6334226521142409,
-      "grad_norm": 0.43295741081237793,
-      "learning_rate": 9.09052369419301e-06,
-      "loss": 0.3602,
-      "step": 9692
-    },
-    {
-      "epoch": 0.633488007319783,
-      "grad_norm": 0.47926750779151917,
-      "learning_rate": 9.090322875784202e-06,
-      "loss": 0.4064,
-      "step": 9693
-    },
-    {
-      "epoch": 0.6335533625253251,
-      "grad_norm": 0.40310317277908325,
-      "learning_rate": 9.090122037425471e-06,
-      "loss": 0.3284,
-      "step": 9694
-    },
-    {
-      "epoch": 0.6336187177308673,
-      "grad_norm": 0.46224668622016907,
-      "learning_rate": 9.089921179117798e-06,
-      "loss": 0.4115,
-      "step": 9695
-    },
-    {
-      "epoch": 0.6336840729364094,
-      "grad_norm": 0.4418547749519348,
-      "learning_rate": 9.089720300862164e-06,
-      "loss": 0.3853,
-      "step": 9696
-    },
-    {
-      "epoch": 0.6337494281419515,
-      "grad_norm": 0.4346838593482971,
-      "learning_rate": 9.089519402659548e-06,
-      "loss": 0.3575,
-      "step": 9697
-    },
-    {
-      "epoch": 0.6338147833474936,
-      "grad_norm": 0.4280930757522583,
-      "learning_rate": 9.089318484510927e-06,
-      "loss": 0.3947,
-      "step": 9698
-    },
-    {
-      "epoch": 0.6338801385530357,
-      "grad_norm": 0.4797299802303314,
-      "learning_rate": 9.089117546417284e-06,
-      "loss": 0.4431,
-      "step": 9699
-    },
-    {
-      "epoch": 0.6339454937585779,
-      "grad_norm": 0.4703907072544098,
-      "learning_rate": 9.088916588379598e-06,
-      "loss": 0.4059,
-      "step": 9700
-    },
-    {
-      "epoch": 0.63401084896412,
-      "grad_norm": 0.40756452083587646,
-      "learning_rate": 9.08871561039885e-06,
-      "loss": 0.3371,
-      "step": 9701
-    },
-    {
-      "epoch": 0.6340762041696621,
-      "grad_norm": 0.48780328035354614,
-      "learning_rate": 9.088514612476018e-06,
-      "loss": 0.4488,
-      "step": 9702
-    },
-    {
-      "epoch": 0.6341415593752042,
-      "grad_norm": 0.4277057945728302,
-      "learning_rate": 9.088313594612085e-06,
-      "loss": 0.3516,
-      "step": 9703
-    },
-    {
-      "epoch": 0.6342069145807464,
-      "grad_norm": 0.47396546602249146,
-      "learning_rate": 9.08811255680803e-06,
-      "loss": 0.4141,
-      "step": 9704
-    },
-    {
-      "epoch": 0.6342722697862885,
-      "grad_norm": 0.4338968098163605,
-      "learning_rate": 9.087911499064835e-06,
-      "loss": 0.3885,
-      "step": 9705
-    },
-    {
-      "epoch": 0.6343376249918306,
-      "grad_norm": 0.446594774723053,
-      "learning_rate": 9.087710421383477e-06,
-      "loss": 0.3842,
-      "step": 9706
-    },
-    {
-      "epoch": 0.6344029801973727,
-      "grad_norm": 0.4124334454536438,
-      "learning_rate": 9.087509323764941e-06,
-      "loss": 0.3193,
-      "step": 9707
-    },
-    {
-      "epoch": 0.6344683354029148,
-      "grad_norm": 0.45303478837013245,
-      "learning_rate": 9.087308206210204e-06,
-      "loss": 0.3565,
-      "step": 9708
-    },
-    {
-      "epoch": 0.634533690608457,
-      "grad_norm": 0.4626130759716034,
-      "learning_rate": 9.087107068720251e-06,
-      "loss": 0.3577,
-      "step": 9709
-    },
-    {
-      "epoch": 0.6345990458139991,
-      "grad_norm": 0.4484632611274719,
-      "learning_rate": 9.08690591129606e-06,
-      "loss": 0.3799,
-      "step": 9710
-    },
-    {
-      "epoch": 0.6346644010195412,
-      "grad_norm": 0.4217261075973511,
-      "learning_rate": 9.086704733938612e-06,
-      "loss": 0.3402,
-      "step": 9711
-    },
-    {
-      "epoch": 0.6347297562250833,
-      "grad_norm": 0.46096986532211304,
-      "learning_rate": 9.086503536648891e-06,
-      "loss": 0.4022,
-      "step": 9712
-    },
-    {
-      "epoch": 0.6347951114306255,
-      "grad_norm": 0.4396562874317169,
-      "learning_rate": 9.086302319427875e-06,
-      "loss": 0.3887,
-      "step": 9713
-    },
-    {
-      "epoch": 0.6348604666361676,
-      "grad_norm": 0.4368191063404083,
-      "learning_rate": 9.086101082276549e-06,
-      "loss": 0.4044,
-      "step": 9714
-    },
-    {
-      "epoch": 0.6349258218417096,
-      "grad_norm": 0.46362486481666565,
-      "learning_rate": 9.085899825195892e-06,
-      "loss": 0.3935,
-      "step": 9715
-    },
-    {
-      "epoch": 0.6349911770472518,
-      "grad_norm": 0.4828219413757324,
-      "learning_rate": 9.085698548186885e-06,
-      "loss": 0.4429,
-      "step": 9716
-    },
-    {
-      "epoch": 0.6350565322527939,
-      "grad_norm": 0.4345232844352722,
-      "learning_rate": 9.08549725125051e-06,
-      "loss": 0.3542,
-      "step": 9717
-    },
-    {
-      "epoch": 0.6351218874583361,
-      "grad_norm": 0.507585883140564,
-      "learning_rate": 9.085295934387752e-06,
-      "loss": 0.4885,
-      "step": 9718
-    },
-    {
-      "epoch": 0.6351872426638782,
-      "grad_norm": 0.4546085000038147,
-      "learning_rate": 9.085094597599589e-06,
-      "loss": 0.3786,
-      "step": 9719
-    },
-    {
-      "epoch": 0.6352525978694203,
-      "grad_norm": 0.43173351883888245,
-      "learning_rate": 9.084893240887005e-06,
-      "loss": 0.3803,
-      "step": 9720
-    },
-    {
-      "epoch": 0.6353179530749624,
-      "grad_norm": 0.4759143888950348,
-      "learning_rate": 9.08469186425098e-06,
-      "loss": 0.4048,
-      "step": 9721
-    },
-    {
-      "epoch": 0.6353833082805045,
-      "grad_norm": 0.45320361852645874,
-      "learning_rate": 9.0844904676925e-06,
-      "loss": 0.384,
-      "step": 9722
-    },
-    {
-      "epoch": 0.6354486634860467,
-      "grad_norm": 0.4534291923046112,
-      "learning_rate": 9.084289051212544e-06,
-      "loss": 0.3764,
-      "step": 9723
-    },
-    {
-      "epoch": 0.6355140186915887,
-      "grad_norm": 0.45015692710876465,
-      "learning_rate": 9.084087614812093e-06,
-      "loss": 0.3885,
-      "step": 9724
-    },
-    {
-      "epoch": 0.6355793738971309,
-      "grad_norm": 0.43396201729774475,
-      "learning_rate": 9.083886158492136e-06,
-      "loss": 0.3536,
-      "step": 9725
-    },
-    {
-      "epoch": 0.635644729102673,
-      "grad_norm": 0.4757612347602844,
-      "learning_rate": 9.08368468225365e-06,
-      "loss": 0.3941,
-      "step": 9726
-    },
-    {
-      "epoch": 0.6357100843082152,
-      "grad_norm": 0.4937000870704651,
-      "learning_rate": 9.083483186097616e-06,
-      "loss": 0.3941,
-      "step": 9727
-    },
-    {
-      "epoch": 0.6357754395137573,
-      "grad_norm": 0.47819656133651733,
-      "learning_rate": 9.083281670025024e-06,
-      "loss": 0.4073,
-      "step": 9728
-    },
-    {
-      "epoch": 0.6358407947192994,
-      "grad_norm": 0.4278254508972168,
-      "learning_rate": 9.083080134036851e-06,
-      "loss": 0.3591,
-      "step": 9729
-    },
-    {
-      "epoch": 0.6359061499248415,
-      "grad_norm": 0.4464288651943207,
-      "learning_rate": 9.082878578134082e-06,
-      "loss": 0.3979,
-      "step": 9730
-    },
-    {
-      "epoch": 0.6359715051303836,
-      "grad_norm": 0.45361876487731934,
-      "learning_rate": 9.0826770023177e-06,
-      "loss": 0.4091,
-      "step": 9731
-    },
-    {
-      "epoch": 0.6360368603359258,
-      "grad_norm": 0.4655058681964874,
-      "learning_rate": 9.082475406588686e-06,
-      "loss": 0.3877,
-      "step": 9732
-    },
-    {
-      "epoch": 0.6361022155414678,
-      "grad_norm": 0.44808417558670044,
-      "learning_rate": 9.082273790948027e-06,
-      "loss": 0.3874,
-      "step": 9733
-    },
-    {
-      "epoch": 0.63616757074701,
-      "grad_norm": 0.41832366585731506,
-      "learning_rate": 9.082072155396704e-06,
-      "loss": 0.3386,
-      "step": 9734
-    },
-    {
-      "epoch": 0.6362329259525521,
-      "grad_norm": 0.4112207889556885,
-      "learning_rate": 9.081870499935701e-06,
-      "loss": 0.3353,
-      "step": 9735
-    },
-    {
-      "epoch": 0.6362982811580943,
-      "grad_norm": 0.423401802778244,
-      "learning_rate": 9.081668824566002e-06,
-      "loss": 0.3663,
-      "step": 9736
-    },
-    {
-      "epoch": 0.6363636363636364,
-      "grad_norm": 0.46037980914115906,
-      "learning_rate": 9.081467129288589e-06,
-      "loss": 0.3953,
-      "step": 9737
-    },
-    {
-      "epoch": 0.6364289915691785,
-      "grad_norm": 0.43599933385849,
-      "learning_rate": 9.081265414104448e-06,
-      "loss": 0.3629,
-      "step": 9738
-    },
-    {
-      "epoch": 0.6364943467747206,
-      "grad_norm": 0.429604172706604,
-      "learning_rate": 9.08106367901456e-06,
-      "loss": 0.3367,
-      "step": 9739
-    },
-    {
-      "epoch": 0.6365597019802627,
-      "grad_norm": 0.42948323488235474,
-      "learning_rate": 9.080861924019912e-06,
-      "loss": 0.3685,
-      "step": 9740
-    },
-    {
-      "epoch": 0.6366250571858049,
-      "grad_norm": 0.4661257266998291,
-      "learning_rate": 9.080660149121487e-06,
-      "loss": 0.4667,
-      "step": 9741
-    },
-    {
-      "epoch": 0.6366904123913469,
-      "grad_norm": 0.4274749755859375,
-      "learning_rate": 9.080458354320267e-06,
-      "loss": 0.3483,
-      "step": 9742
-    },
-    {
-      "epoch": 0.6367557675968891,
-      "grad_norm": 0.46764200925827026,
-      "learning_rate": 9.08025653961724e-06,
-      "loss": 0.4217,
-      "step": 9743
-    },
-    {
-      "epoch": 0.6368211228024312,
-      "grad_norm": 0.4310452938079834,
-      "learning_rate": 9.080054705013387e-06,
-      "loss": 0.3702,
-      "step": 9744
-    },
-    {
-      "epoch": 0.6368864780079734,
-      "grad_norm": 0.4607413113117218,
-      "learning_rate": 9.079852850509694e-06,
-      "loss": 0.3794,
-      "step": 9745
-    },
-    {
-      "epoch": 0.6369518332135155,
-      "grad_norm": 0.46990257501602173,
-      "learning_rate": 9.079650976107147e-06,
-      "loss": 0.4215,
-      "step": 9746
-    },
-    {
-      "epoch": 0.6370171884190575,
-      "grad_norm": 0.46286189556121826,
-      "learning_rate": 9.079449081806726e-06,
-      "loss": 0.3654,
-      "step": 9747
-    },
-    {
-      "epoch": 0.6370825436245997,
-      "grad_norm": 0.41074416041374207,
-      "learning_rate": 9.079247167609419e-06,
-      "loss": 0.3235,
-      "step": 9748
-    },
-    {
-      "epoch": 0.6371478988301418,
-      "grad_norm": 0.4421272575855255,
-      "learning_rate": 9.079045233516213e-06,
-      "loss": 0.3509,
-      "step": 9749
-    },
-    {
-      "epoch": 0.637213254035684,
-      "grad_norm": 0.44864943623542786,
-      "learning_rate": 9.078843279528087e-06,
-      "loss": 0.3534,
-      "step": 9750
-    },
-    {
-      "epoch": 0.637278609241226,
-      "grad_norm": 0.45992445945739746,
-      "learning_rate": 9.078641305646032e-06,
-      "loss": 0.3895,
-      "step": 9751
-    },
-    {
-      "epoch": 0.6373439644467682,
-      "grad_norm": 0.432346373796463,
-      "learning_rate": 9.078439311871029e-06,
-      "loss": 0.3408,
-      "step": 9752
-    },
-    {
-      "epoch": 0.6374093196523103,
-      "grad_norm": 0.46255505084991455,
-      "learning_rate": 9.078237298204065e-06,
-      "loss": 0.3708,
-      "step": 9753
-    },
-    {
-      "epoch": 0.6374746748578525,
-      "grad_norm": 0.4515262246131897,
-      "learning_rate": 9.078035264646123e-06,
-      "loss": 0.383,
-      "step": 9754
-    },
-    {
-      "epoch": 0.6375400300633945,
-      "grad_norm": 0.4577791392803192,
-      "learning_rate": 9.077833211198192e-06,
-      "loss": 0.3924,
-      "step": 9755
-    },
-    {
-      "epoch": 0.6376053852689366,
-      "grad_norm": 0.4261878430843353,
-      "learning_rate": 9.077631137861255e-06,
-      "loss": 0.3724,
-      "step": 9756
-    },
-    {
-      "epoch": 0.6376707404744788,
-      "grad_norm": 0.4302404522895813,
-      "learning_rate": 9.0774290446363e-06,
-      "loss": 0.3362,
-      "step": 9757
-    },
-    {
-      "epoch": 0.6377360956800209,
-      "grad_norm": 0.47554245591163635,
-      "learning_rate": 9.07722693152431e-06,
-      "loss": 0.4322,
-      "step": 9758
-    },
-    {
-      "epoch": 0.6378014508855631,
-      "grad_norm": 0.4756726324558258,
-      "learning_rate": 9.077024798526273e-06,
-      "loss": 0.4354,
-      "step": 9759
-    },
-    {
-      "epoch": 0.6378668060911051,
-      "grad_norm": 0.44856563210487366,
-      "learning_rate": 9.07682264564317e-06,
-      "loss": 0.3674,
-      "step": 9760
-    },
-    {
-      "epoch": 0.6379321612966473,
-      "grad_norm": 0.4309118688106537,
-      "learning_rate": 9.076620472875994e-06,
-      "loss": 0.3612,
-      "step": 9761
-    },
-    {
-      "epoch": 0.6379975165021894,
-      "grad_norm": 0.4074327349662781,
-      "learning_rate": 9.076418280225727e-06,
-      "loss": 0.3181,
-      "step": 9762
-    },
-    {
-      "epoch": 0.6380628717077316,
-      "grad_norm": 0.46789079904556274,
-      "learning_rate": 9.076216067693355e-06,
-      "loss": 0.3931,
-      "step": 9763
-    },
-    {
-      "epoch": 0.6381282269132736,
-      "grad_norm": 0.423562616109848,
-      "learning_rate": 9.076013835279865e-06,
-      "loss": 0.3655,
-      "step": 9764
-    },
-    {
-      "epoch": 0.6381935821188157,
-      "grad_norm": 0.43467196822166443,
-      "learning_rate": 9.075811582986244e-06,
-      "loss": 0.3929,
-      "step": 9765
-    },
-    {
-      "epoch": 0.6382589373243579,
-      "grad_norm": 0.47196051478385925,
-      "learning_rate": 9.075609310813478e-06,
-      "loss": 0.3943,
-      "step": 9766
-    },
-    {
-      "epoch": 0.6383242925299,
-      "grad_norm": 0.4757579565048218,
-      "learning_rate": 9.075407018762554e-06,
-      "loss": 0.4466,
-      "step": 9767
-    },
-    {
-      "epoch": 0.6383896477354422,
-      "grad_norm": 0.42987725138664246,
-      "learning_rate": 9.075204706834458e-06,
-      "loss": 0.3845,
-      "step": 9768
-    },
-    {
-      "epoch": 0.6384550029409842,
-      "grad_norm": 0.4808485209941864,
-      "learning_rate": 9.075002375030176e-06,
-      "loss": 0.46,
-      "step": 9769
-    },
-    {
-      "epoch": 0.6385203581465264,
-      "grad_norm": 0.4296873211860657,
-      "learning_rate": 9.074800023350696e-06,
-      "loss": 0.3636,
-      "step": 9770
-    },
-    {
-      "epoch": 0.6385857133520685,
-      "grad_norm": 0.45422402024269104,
-      "learning_rate": 9.074597651797004e-06,
-      "loss": 0.3951,
-      "step": 9771
-    },
-    {
-      "epoch": 0.6386510685576107,
-      "grad_norm": 0.4010392725467682,
-      "learning_rate": 9.074395260370088e-06,
-      "loss": 0.3225,
-      "step": 9772
-    },
-    {
-      "epoch": 0.6387164237631527,
-      "grad_norm": 0.4246975779533386,
-      "learning_rate": 9.074192849070936e-06,
-      "loss": 0.3692,
-      "step": 9773
-    },
-    {
-      "epoch": 0.6387817789686948,
-      "grad_norm": 0.45356330275535583,
-      "learning_rate": 9.073990417900533e-06,
-      "loss": 0.3873,
-      "step": 9774
-    },
-    {
-      "epoch": 0.638847134174237,
-      "grad_norm": 0.45848318934440613,
-      "learning_rate": 9.073787966859866e-06,
-      "loss": 0.4449,
-      "step": 9775
-    },
-    {
-      "epoch": 0.6389124893797791,
-      "grad_norm": 0.43433430790901184,
-      "learning_rate": 9.073585495949927e-06,
-      "loss": 0.3788,
-      "step": 9776
-    },
-    {
-      "epoch": 0.6389778445853213,
-      "grad_norm": 0.47940051555633545,
-      "learning_rate": 9.073383005171699e-06,
-      "loss": 0.4517,
-      "step": 9777
-    },
-    {
-      "epoch": 0.6390431997908633,
-      "grad_norm": 0.416909784078598,
-      "learning_rate": 9.07318049452617e-06,
-      "loss": 0.3309,
-      "step": 9778
-    },
-    {
-      "epoch": 0.6391085549964055,
-      "grad_norm": 0.4129504859447479,
-      "learning_rate": 9.07297796401433e-06,
-      "loss": 0.3248,
-      "step": 9779
-    },
-    {
-      "epoch": 0.6391739102019476,
-      "grad_norm": 0.4612945318222046,
-      "learning_rate": 9.072775413637163e-06,
-      "loss": 0.3767,
-      "step": 9780
-    },
-    {
-      "epoch": 0.6392392654074897,
-      "grad_norm": 0.46480366587638855,
-      "learning_rate": 9.072572843395661e-06,
-      "loss": 0.4069,
-      "step": 9781
-    },
-    {
-      "epoch": 0.6393046206130318,
-      "grad_norm": 0.4225768744945526,
-      "learning_rate": 9.072370253290813e-06,
-      "loss": 0.3796,
-      "step": 9782
-    },
-    {
-      "epoch": 0.6393699758185739,
-      "grad_norm": 0.4254634380340576,
-      "learning_rate": 9.0721676433236e-06,
-      "loss": 0.3772,
-      "step": 9783
-    },
-    {
-      "epoch": 0.6394353310241161,
-      "grad_norm": 0.4734971225261688,
-      "learning_rate": 9.071965013495017e-06,
-      "loss": 0.3798,
-      "step": 9784
-    },
-    {
-      "epoch": 0.6395006862296582,
-      "grad_norm": 0.4534674286842346,
-      "learning_rate": 9.07176236380605e-06,
-      "loss": 0.4017,
-      "step": 9785
-    },
-    {
-      "epoch": 0.6395660414352003,
-      "grad_norm": 0.4039697051048279,
-      "learning_rate": 9.071559694257686e-06,
-      "loss": 0.3289,
-      "step": 9786
-    },
-    {
-      "epoch": 0.6396313966407424,
-      "grad_norm": 0.45211726427078247,
-      "learning_rate": 9.071357004850915e-06,
-      "loss": 0.3866,
-      "step": 9787
-    },
-    {
-      "epoch": 0.6396967518462846,
-      "grad_norm": 0.44150295853614807,
-      "learning_rate": 9.071154295586727e-06,
-      "loss": 0.3445,
-      "step": 9788
-    },
-    {
-      "epoch": 0.6397621070518267,
-      "grad_norm": 0.4525618553161621,
-      "learning_rate": 9.070951566466109e-06,
-      "loss": 0.3692,
-      "step": 9789
-    },
-    {
-      "epoch": 0.6398274622573688,
-      "grad_norm": 0.4716513156890869,
-      "learning_rate": 9.07074881749005e-06,
-      "loss": 0.387,
-      "step": 9790
-    },
-    {
-      "epoch": 0.6398928174629109,
-      "grad_norm": 0.45230627059936523,
-      "learning_rate": 9.070546048659537e-06,
-      "loss": 0.4024,
-      "step": 9791
-    },
-    {
-      "epoch": 0.639958172668453,
-      "grad_norm": 0.4527648687362671,
-      "learning_rate": 9.07034325997556e-06,
-      "loss": 0.3378,
-      "step": 9792
-    },
-    {
-      "epoch": 0.6400235278739952,
-      "grad_norm": 0.43456801772117615,
-      "learning_rate": 9.07014045143911e-06,
-      "loss": 0.3922,
-      "step": 9793
-    },
-    {
-      "epoch": 0.6400888830795373,
-      "grad_norm": 0.4388381242752075,
-      "learning_rate": 9.069937623051177e-06,
-      "loss": 0.3462,
-      "step": 9794
-    },
-    {
-      "epoch": 0.6401542382850794,
-      "grad_norm": 0.44833528995513916,
-      "learning_rate": 9.069734774812747e-06,
-      "loss": 0.3753,
-      "step": 9795
-    },
-    {
-      "epoch": 0.6402195934906215,
-      "grad_norm": 0.44104278087615967,
-      "learning_rate": 9.06953190672481e-06,
-      "loss": 0.3365,
-      "step": 9796
-    },
-    {
-      "epoch": 0.6402849486961637,
-      "grad_norm": 0.4822522699832916,
-      "learning_rate": 9.069329018788357e-06,
-      "loss": 0.4566,
-      "step": 9797
-    },
-    {
-      "epoch": 0.6403503039017058,
-      "grad_norm": 0.4667324125766754,
-      "learning_rate": 9.069126111004376e-06,
-      "loss": 0.4038,
-      "step": 9798
-    },
-    {
-      "epoch": 0.6404156591072478,
-      "grad_norm": 0.471744179725647,
-      "learning_rate": 9.068923183373856e-06,
-      "loss": 0.3891,
-      "step": 9799
-    },
-    {
-      "epoch": 0.64048101431279,
-      "grad_norm": 0.46452444791793823,
-      "learning_rate": 9.06872023589779e-06,
-      "loss": 0.3823,
-      "step": 9800
-    },
-    {
-      "epoch": 0.6405463695183321,
-      "grad_norm": 0.39168450236320496,
-      "learning_rate": 9.068517268577166e-06,
-      "loss": 0.2793,
-      "step": 9801
-    },
-    {
-      "epoch": 0.6406117247238743,
-      "grad_norm": 0.40817102789878845,
-      "learning_rate": 9.068314281412974e-06,
-      "loss": 0.3181,
-      "step": 9802
-    },
-    {
-      "epoch": 0.6406770799294164,
-      "grad_norm": 0.4927980601787567,
-      "learning_rate": 9.068111274406202e-06,
-      "loss": 0.4012,
-      "step": 9803
-    },
-    {
-      "epoch": 0.6407424351349585,
-      "grad_norm": 0.4919649660587311,
-      "learning_rate": 9.067908247557842e-06,
-      "loss": 0.443,
-      "step": 9804
-    },
-    {
-      "epoch": 0.6408077903405006,
-      "grad_norm": 0.4291025996208191,
-      "learning_rate": 9.067705200868886e-06,
-      "loss": 0.3336,
-      "step": 9805
-    },
-    {
-      "epoch": 0.6408731455460427,
-      "grad_norm": 0.46074026823043823,
-      "learning_rate": 9.067502134340321e-06,
-      "loss": 0.4087,
-      "step": 9806
-    },
-    {
-      "epoch": 0.6409385007515849,
-      "grad_norm": 0.45968097448349,
-      "learning_rate": 9.06729904797314e-06,
-      "loss": 0.4073,
-      "step": 9807
-    },
-    {
-      "epoch": 0.641003855957127,
-      "grad_norm": 0.4455585777759552,
-      "learning_rate": 9.067095941768332e-06,
-      "loss": 0.3662,
-      "step": 9808
-    },
-    {
-      "epoch": 0.6410692111626691,
-      "grad_norm": 0.442227303981781,
-      "learning_rate": 9.066892815726888e-06,
-      "loss": 0.3402,
-      "step": 9809
-    },
-    {
-      "epoch": 0.6411345663682112,
-      "grad_norm": 0.4429609179496765,
-      "learning_rate": 9.0666896698498e-06,
-      "loss": 0.3847,
-      "step": 9810
-    },
-    {
-      "epoch": 0.6411999215737534,
-      "grad_norm": 0.4293217957019806,
-      "learning_rate": 9.066486504138056e-06,
-      "loss": 0.3697,
-      "step": 9811
-    },
-    {
-      "epoch": 0.6412652767792955,
-      "grad_norm": 0.4350472092628479,
-      "learning_rate": 9.06628331859265e-06,
-      "loss": 0.372,
-      "step": 9812
-    },
-    {
-      "epoch": 0.6413306319848376,
-      "grad_norm": 0.4853617548942566,
-      "learning_rate": 9.066080113214571e-06,
-      "loss": 0.4415,
-      "step": 9813
-    },
-    {
-      "epoch": 0.6413959871903797,
-      "grad_norm": 0.4751304090023041,
-      "learning_rate": 9.06587688800481e-06,
-      "loss": 0.3688,
-      "step": 9814
-    },
-    {
-      "epoch": 0.6414613423959218,
-      "grad_norm": 0.44478118419647217,
-      "learning_rate": 9.065673642964358e-06,
-      "loss": 0.3377,
-      "step": 9815
-    },
-    {
-      "epoch": 0.641526697601464,
-      "grad_norm": 0.4865337312221527,
-      "learning_rate": 9.06547037809421e-06,
-      "loss": 0.4008,
-      "step": 9816
-    },
-    {
-      "epoch": 0.641592052807006,
-      "grad_norm": 0.4339168667793274,
-      "learning_rate": 9.065267093395353e-06,
-      "loss": 0.402,
-      "step": 9817
-    },
-    {
-      "epoch": 0.6416574080125482,
-      "grad_norm": 0.43428879976272583,
-      "learning_rate": 9.06506378886878e-06,
-      "loss": 0.3741,
-      "step": 9818
-    },
-    {
-      "epoch": 0.6417227632180903,
-      "grad_norm": 0.4526311159133911,
-      "learning_rate": 9.064860464515481e-06,
-      "loss": 0.4022,
-      "step": 9819
-    },
-    {
-      "epoch": 0.6417881184236325,
-      "grad_norm": 0.45775577425956726,
-      "learning_rate": 9.064657120336452e-06,
-      "loss": 0.3982,
-      "step": 9820
-    },
-    {
-      "epoch": 0.6418534736291746,
-      "grad_norm": 0.44582659006118774,
-      "learning_rate": 9.06445375633268e-06,
-      "loss": 0.3815,
-      "step": 9821
-    },
-    {
-      "epoch": 0.6419188288347167,
-      "grad_norm": 0.44306325912475586,
-      "learning_rate": 9.064250372505162e-06,
-      "loss": 0.3701,
-      "step": 9822
-    },
-    {
-      "epoch": 0.6419841840402588,
-      "grad_norm": 0.40925756096839905,
-      "learning_rate": 9.064046968854885e-06,
-      "loss": 0.3188,
-      "step": 9823
-    },
-    {
-      "epoch": 0.6420495392458009,
-      "grad_norm": 0.4516238570213318,
-      "learning_rate": 9.063843545382841e-06,
-      "loss": 0.4295,
-      "step": 9824
-    },
-    {
-      "epoch": 0.6421148944513431,
-      "grad_norm": 0.503147542476654,
-      "learning_rate": 9.063640102090029e-06,
-      "loss": 0.3328,
-      "step": 9825
-    },
-    {
-      "epoch": 0.6421802496568851,
-      "grad_norm": 0.4248664081096649,
-      "learning_rate": 9.063436638977432e-06,
-      "loss": 0.3851,
-      "step": 9826
-    },
-    {
-      "epoch": 0.6422456048624273,
-      "grad_norm": 0.46630939841270447,
-      "learning_rate": 9.06323315604605e-06,
-      "loss": 0.4135,
-      "step": 9827
-    },
-    {
-      "epoch": 0.6423109600679694,
-      "grad_norm": 0.4635048508644104,
-      "learning_rate": 9.063029653296868e-06,
-      "loss": 0.3847,
-      "step": 9828
-    },
-    {
-      "epoch": 0.6423763152735116,
-      "grad_norm": 0.4125978350639343,
-      "learning_rate": 9.062826130730886e-06,
-      "loss": 0.3321,
-      "step": 9829
-    },
-    {
-      "epoch": 0.6424416704790537,
-      "grad_norm": 0.4719914197921753,
-      "learning_rate": 9.062622588349094e-06,
-      "loss": 0.3317,
-      "step": 9830
-    },
-    {
-      "epoch": 0.6425070256845957,
-      "grad_norm": 0.4191725254058838,
-      "learning_rate": 9.062419026152483e-06,
-      "loss": 0.3445,
-      "step": 9831
-    },
-    {
-      "epoch": 0.6425723808901379,
-      "grad_norm": 0.4677099287509918,
-      "learning_rate": 9.062215444142047e-06,
-      "loss": 0.41,
-      "step": 9832
-    },
-    {
-      "epoch": 0.64263773609568,
-      "grad_norm": 0.4397822320461273,
-      "learning_rate": 9.06201184231878e-06,
-      "loss": 0.3687,
-      "step": 9833
-    },
-    {
-      "epoch": 0.6427030913012222,
-      "grad_norm": 0.4193592369556427,
-      "learning_rate": 9.061808220683672e-06,
-      "loss": 0.3517,
-      "step": 9834
-    },
-    {
-      "epoch": 0.6427684465067642,
-      "grad_norm": 0.4300253093242645,
-      "learning_rate": 9.06160457923772e-06,
-      "loss": 0.3616,
-      "step": 9835
-    },
-    {
-      "epoch": 0.6428338017123064,
-      "grad_norm": 0.42338287830352783,
-      "learning_rate": 9.061400917981915e-06,
-      "loss": 0.366,
-      "step": 9836
-    },
-    {
-      "epoch": 0.6428991569178485,
-      "grad_norm": 0.45557650923728943,
-      "learning_rate": 9.06119723691725e-06,
-      "loss": 0.3821,
-      "step": 9837
-    },
-    {
-      "epoch": 0.6429645121233907,
-      "grad_norm": 0.451945424079895,
-      "learning_rate": 9.06099353604472e-06,
-      "loss": 0.4168,
-      "step": 9838
-    },
-    {
-      "epoch": 0.6430298673289327,
-      "grad_norm": 0.4171278774738312,
-      "learning_rate": 9.060789815365317e-06,
-      "loss": 0.3309,
-      "step": 9839
-    },
-    {
-      "epoch": 0.6430952225344748,
-      "grad_norm": 0.45155543088912964,
-      "learning_rate": 9.060586074880036e-06,
-      "loss": 0.4161,
-      "step": 9840
-    },
-    {
-      "epoch": 0.643160577740017,
-      "grad_norm": 0.43141868710517883,
-      "learning_rate": 9.060382314589871e-06,
-      "loss": 0.3604,
-      "step": 9841
-    },
-    {
-      "epoch": 0.6432259329455591,
-      "grad_norm": 0.40812331438064575,
-      "learning_rate": 9.060178534495811e-06,
-      "loss": 0.3135,
-      "step": 9842
-    },
-    {
-      "epoch": 0.6432912881511013,
-      "grad_norm": 0.4390416145324707,
-      "learning_rate": 9.059974734598858e-06,
-      "loss": 0.3702,
-      "step": 9843
-    },
-    {
-      "epoch": 0.6433566433566433,
-      "grad_norm": 0.4729102849960327,
-      "learning_rate": 9.059770914899999e-06,
-      "loss": 0.3938,
-      "step": 9844
-    },
-    {
-      "epoch": 0.6434219985621855,
-      "grad_norm": 0.4624210596084595,
-      "learning_rate": 9.059567075400232e-06,
-      "loss": 0.434,
-      "step": 9845
-    },
-    {
-      "epoch": 0.6434873537677276,
-      "grad_norm": 0.42629072070121765,
-      "learning_rate": 9.05936321610055e-06,
-      "loss": 0.3571,
-      "step": 9846
-    },
-    {
-      "epoch": 0.6435527089732698,
-      "grad_norm": 0.46381500363349915,
-      "learning_rate": 9.059159337001945e-06,
-      "loss": 0.4149,
-      "step": 9847
-    },
-    {
-      "epoch": 0.6436180641788118,
-      "grad_norm": 0.46328607201576233,
-      "learning_rate": 9.058955438105416e-06,
-      "loss": 0.378,
-      "step": 9848
-    },
-    {
-      "epoch": 0.6436834193843539,
-      "grad_norm": 0.45271798968315125,
-      "learning_rate": 9.058751519411957e-06,
-      "loss": 0.4004,
-      "step": 9849
-    },
-    {
-      "epoch": 0.6437487745898961,
-      "grad_norm": 0.4705389738082886,
-      "learning_rate": 9.058547580922556e-06,
-      "loss": 0.4381,
-      "step": 9850
-    },
-    {
-      "epoch": 0.6438141297954382,
-      "grad_norm": 0.4463994801044464,
-      "learning_rate": 9.058343622638218e-06,
-      "loss": 0.3849,
-      "step": 9851
-    },
-    {
-      "epoch": 0.6438794850009804,
-      "grad_norm": 0.45492780208587646,
-      "learning_rate": 9.058139644559929e-06,
-      "loss": 0.4019,
-      "step": 9852
-    },
-    {
-      "epoch": 0.6439448402065224,
-      "grad_norm": 0.4657062590122223,
-      "learning_rate": 9.057935646688685e-06,
-      "loss": 0.3998,
-      "step": 9853
-    },
-    {
-      "epoch": 0.6440101954120646,
-      "grad_norm": 0.4456426799297333,
-      "learning_rate": 9.057731629025485e-06,
-      "loss": 0.426,
-      "step": 9854
-    },
-    {
-      "epoch": 0.6440755506176067,
-      "grad_norm": 0.44574275612831116,
-      "learning_rate": 9.057527591571325e-06,
-      "loss": 0.3418,
-      "step": 9855
-    },
-    {
-      "epoch": 0.6441409058231489,
-      "grad_norm": 0.4843095541000366,
-      "learning_rate": 9.057323534327194e-06,
-      "loss": 0.4494,
-      "step": 9856
-    },
-    {
-      "epoch": 0.6442062610286909,
-      "grad_norm": 0.45727428793907166,
-      "learning_rate": 9.05711945729409e-06,
-      "loss": 0.3891,
-      "step": 9857
-    },
-    {
-      "epoch": 0.644271616234233,
-      "grad_norm": 0.45042911171913147,
-      "learning_rate": 9.056915360473011e-06,
-      "loss": 0.3648,
-      "step": 9858
-    },
-    {
-      "epoch": 0.6443369714397752,
-      "grad_norm": 0.44673705101013184,
-      "learning_rate": 9.056711243864949e-06,
-      "loss": 0.3898,
-      "step": 9859
-    },
-    {
-      "epoch": 0.6444023266453173,
-      "grad_norm": 0.44384297728538513,
-      "learning_rate": 9.056507107470901e-06,
-      "loss": 0.3464,
-      "step": 9860
-    },
-    {
-      "epoch": 0.6444676818508595,
-      "grad_norm": 0.4594115912914276,
-      "learning_rate": 9.056302951291863e-06,
-      "loss": 0.3769,
-      "step": 9861
-    },
-    {
-      "epoch": 0.6445330370564015,
-      "grad_norm": 0.4539910554885864,
-      "learning_rate": 9.056098775328829e-06,
-      "loss": 0.3618,
-      "step": 9862
-    },
-    {
-      "epoch": 0.6445983922619437,
-      "grad_norm": 0.4160584509372711,
-      "learning_rate": 9.055894579582798e-06,
-      "loss": 0.2929,
-      "step": 9863
-    },
-    {
-      "epoch": 0.6446637474674858,
-      "grad_norm": 0.46673911809921265,
-      "learning_rate": 9.055690364054764e-06,
-      "loss": 0.3894,
-      "step": 9864
-    },
-    {
-      "epoch": 0.6447291026730279,
-      "grad_norm": 0.4321063756942749,
-      "learning_rate": 9.055486128745723e-06,
-      "loss": 0.3678,
-      "step": 9865
-    },
-    {
-      "epoch": 0.64479445787857,
-      "grad_norm": 0.446098268032074,
-      "learning_rate": 9.05528187365667e-06,
-      "loss": 0.336,
-      "step": 9866
-    },
-    {
-      "epoch": 0.6448598130841121,
-      "grad_norm": 0.43968868255615234,
-      "learning_rate": 9.055077598788603e-06,
-      "loss": 0.4202,
-      "step": 9867
-    },
-    {
-      "epoch": 0.6449251682896543,
-      "grad_norm": 0.47653335332870483,
-      "learning_rate": 9.054873304142518e-06,
-      "loss": 0.404,
-      "step": 9868
-    },
-    {
-      "epoch": 0.6449905234951964,
-      "grad_norm": 0.4718948006629944,
-      "learning_rate": 9.05466898971941e-06,
-      "loss": 0.4296,
-      "step": 9869
-    },
-    {
-      "epoch": 0.6450558787007385,
-      "grad_norm": 0.4796658754348755,
-      "learning_rate": 9.054464655520278e-06,
-      "loss": 0.451,
-      "step": 9870
-    },
-    {
-      "epoch": 0.6451212339062806,
-      "grad_norm": 0.45507287979125977,
-      "learning_rate": 9.054260301546116e-06,
-      "loss": 0.4176,
-      "step": 9871
-    },
-    {
-      "epoch": 0.6451865891118228,
-      "grad_norm": 0.4399968683719635,
-      "learning_rate": 9.054055927797924e-06,
-      "loss": 0.3779,
-      "step": 9872
-    },
-    {
-      "epoch": 0.6452519443173649,
-      "grad_norm": 0.4842342436313629,
-      "learning_rate": 9.053851534276695e-06,
-      "loss": 0.4452,
-      "step": 9873
-    },
-    {
-      "epoch": 0.645317299522907,
-      "grad_norm": 0.4472092092037201,
-      "learning_rate": 9.053647120983428e-06,
-      "loss": 0.3702,
-      "step": 9874
-    },
-    {
-      "epoch": 0.6453826547284491,
-      "grad_norm": 0.4460185170173645,
-      "learning_rate": 9.053442687919121e-06,
-      "loss": 0.4037,
-      "step": 9875
-    },
-    {
-      "epoch": 0.6454480099339912,
-      "grad_norm": 0.4718061685562134,
-      "learning_rate": 9.053238235084768e-06,
-      "loss": 0.4239,
-      "step": 9876
-    },
-    {
-      "epoch": 0.6455133651395334,
-      "grad_norm": 0.45335057377815247,
-      "learning_rate": 9.05303376248137e-06,
-      "loss": 0.4205,
-      "step": 9877
-    },
-    {
-      "epoch": 0.6455787203450755,
-      "grad_norm": 0.4372648596763611,
-      "learning_rate": 9.05282927010992e-06,
-      "loss": 0.3722,
-      "step": 9878
-    },
-    {
-      "epoch": 0.6456440755506176,
-      "grad_norm": 0.44156113266944885,
-      "learning_rate": 9.052624757971418e-06,
-      "loss": 0.3832,
-      "step": 9879
-    },
-    {
-      "epoch": 0.6457094307561597,
-      "grad_norm": 0.44538334012031555,
-      "learning_rate": 9.05242022606686e-06,
-      "loss": 0.3445,
-      "step": 9880
-    },
-    {
-      "epoch": 0.6457747859617019,
-      "grad_norm": 0.44533470273017883,
-      "learning_rate": 9.052215674397249e-06,
-      "loss": 0.3613,
-      "step": 9881
-    },
-    {
-      "epoch": 0.645840141167244,
-      "grad_norm": 0.41601648926734924,
-      "learning_rate": 9.052011102963574e-06,
-      "loss": 0.3331,
-      "step": 9882
-    },
-    {
-      "epoch": 0.645905496372786,
-      "grad_norm": 0.4672471284866333,
-      "learning_rate": 9.051806511766839e-06,
-      "loss": 0.3815,
-      "step": 9883
-    },
-    {
-      "epoch": 0.6459708515783282,
-      "grad_norm": 0.4555833637714386,
-      "learning_rate": 9.051601900808041e-06,
-      "loss": 0.414,
-      "step": 9884
-    },
-    {
-      "epoch": 0.6460362067838703,
-      "grad_norm": 0.4334162175655365,
-      "learning_rate": 9.051397270088174e-06,
-      "loss": 0.3493,
-      "step": 9885
-    },
-    {
-      "epoch": 0.6461015619894125,
-      "grad_norm": 0.4448656141757965,
-      "learning_rate": 9.05119261960824e-06,
-      "loss": 0.3652,
-      "step": 9886
-    },
-    {
-      "epoch": 0.6461669171949546,
-      "grad_norm": 0.495453804731369,
-      "learning_rate": 9.050987949369237e-06,
-      "loss": 0.4441,
-      "step": 9887
-    },
-    {
-      "epoch": 0.6462322724004967,
-      "grad_norm": 0.43579965829849243,
-      "learning_rate": 9.050783259372163e-06,
-      "loss": 0.3957,
-      "step": 9888
-    },
-    {
-      "epoch": 0.6462976276060388,
-      "grad_norm": 0.4534274637699127,
-      "learning_rate": 9.050578549618015e-06,
-      "loss": 0.4208,
-      "step": 9889
-    },
-    {
-      "epoch": 0.6463629828115809,
-      "grad_norm": 0.4602999687194824,
-      "learning_rate": 9.050373820107791e-06,
-      "loss": 0.395,
-      "step": 9890
-    },
-    {
-      "epoch": 0.6464283380171231,
-      "grad_norm": 0.4523613452911377,
-      "learning_rate": 9.050169070842492e-06,
-      "loss": 0.3936,
-      "step": 9891
-    },
-    {
-      "epoch": 0.6464936932226651,
-      "grad_norm": 0.4505181908607483,
-      "learning_rate": 9.049964301823114e-06,
-      "loss": 0.3851,
-      "step": 9892
-    },
-    {
-      "epoch": 0.6465590484282073,
-      "grad_norm": 0.47814926505088806,
-      "learning_rate": 9.049759513050657e-06,
-      "loss": 0.4121,
-      "step": 9893
-    },
-    {
-      "epoch": 0.6466244036337494,
-      "grad_norm": 0.43603309988975525,
-      "learning_rate": 9.049554704526122e-06,
-      "loss": 0.3961,
-      "step": 9894
-    },
-    {
-      "epoch": 0.6466897588392916,
-      "grad_norm": 0.4300692677497864,
-      "learning_rate": 9.049349876250506e-06,
-      "loss": 0.3836,
-      "step": 9895
-    },
-    {
-      "epoch": 0.6467551140448337,
-      "grad_norm": 0.4427059292793274,
-      "learning_rate": 9.049145028224806e-06,
-      "loss": 0.377,
-      "step": 9896
-    },
-    {
-      "epoch": 0.6468204692503758,
-      "grad_norm": 0.506506085395813,
-      "learning_rate": 9.048940160450023e-06,
-      "loss": 0.3877,
-      "step": 9897
-    },
-    {
-      "epoch": 0.6468858244559179,
-      "grad_norm": 0.45513319969177246,
-      "learning_rate": 9.048735272927156e-06,
-      "loss": 0.4026,
-      "step": 9898
-    },
-    {
-      "epoch": 0.64695117966146,
-      "grad_norm": 0.42491936683654785,
-      "learning_rate": 9.048530365657205e-06,
-      "loss": 0.3396,
-      "step": 9899
-    },
-    {
-      "epoch": 0.6470165348670022,
-      "grad_norm": 0.46710488200187683,
-      "learning_rate": 9.04832543864117e-06,
-      "loss": 0.3938,
-      "step": 9900
-    },
-    {
-      "epoch": 0.6470818900725442,
-      "grad_norm": 0.5009361505508423,
-      "learning_rate": 9.048120491880047e-06,
-      "loss": 0.5027,
-      "step": 9901
-    },
-    {
-      "epoch": 0.6471472452780864,
-      "grad_norm": 0.4301183521747589,
-      "learning_rate": 9.04791552537484e-06,
-      "loss": 0.3785,
-      "step": 9902
-    },
-    {
-      "epoch": 0.6472126004836285,
-      "grad_norm": 0.43310508131980896,
-      "learning_rate": 9.047710539126546e-06,
-      "loss": 0.3588,
-      "step": 9903
-    },
-    {
-      "epoch": 0.6472779556891707,
-      "grad_norm": 0.439488023519516,
-      "learning_rate": 9.047505533136165e-06,
-      "loss": 0.3706,
-      "step": 9904
-    },
-    {
-      "epoch": 0.6473433108947128,
-      "grad_norm": 0.43672558665275574,
-      "learning_rate": 9.047300507404698e-06,
-      "loss": 0.4033,
-      "step": 9905
-    },
-    {
-      "epoch": 0.6474086661002549,
-      "grad_norm": 0.4547431468963623,
-      "learning_rate": 9.047095461933145e-06,
-      "loss": 0.4193,
-      "step": 9906
-    },
-    {
-      "epoch": 0.647474021305797,
-      "grad_norm": 0.4387088716030121,
-      "learning_rate": 9.046890396722503e-06,
-      "loss": 0.3858,
-      "step": 9907
-    },
-    {
-      "epoch": 0.6475393765113391,
-      "grad_norm": 0.45879948139190674,
-      "learning_rate": 9.046685311773775e-06,
-      "loss": 0.3902,
-      "step": 9908
-    },
-    {
-      "epoch": 0.6476047317168813,
-      "grad_norm": 0.4278230369091034,
-      "learning_rate": 9.046480207087962e-06,
-      "loss": 0.349,
-      "step": 9909
-    },
-    {
-      "epoch": 0.6476700869224233,
-      "grad_norm": 0.46290960907936096,
-      "learning_rate": 9.046275082666064e-06,
-      "loss": 0.3722,
-      "step": 9910
-    },
-    {
-      "epoch": 0.6477354421279655,
-      "grad_norm": 0.4628424048423767,
-      "learning_rate": 9.046069938509078e-06,
-      "loss": 0.4117,
-      "step": 9911
-    },
-    {
-      "epoch": 0.6478007973335076,
-      "grad_norm": 0.44287049770355225,
-      "learning_rate": 9.04586477461801e-06,
-      "loss": 0.3894,
-      "step": 9912
-    },
-    {
-      "epoch": 0.6478661525390498,
-      "grad_norm": 0.4193098843097687,
-      "learning_rate": 9.045659590993856e-06,
-      "loss": 0.3577,
-      "step": 9913
-    },
-    {
-      "epoch": 0.6479315077445919,
-      "grad_norm": 0.4594117999076843,
-      "learning_rate": 9.04545438763762e-06,
-      "loss": 0.4194,
-      "step": 9914
-    },
-    {
-      "epoch": 0.6479968629501339,
-      "grad_norm": 0.4377639889717102,
-      "learning_rate": 9.0452491645503e-06,
-      "loss": 0.3881,
-      "step": 9915
-    },
-    {
-      "epoch": 0.6480622181556761,
-      "grad_norm": 0.4739892780780792,
-      "learning_rate": 9.0450439217329e-06,
-      "loss": 0.3911,
-      "step": 9916
-    },
-    {
-      "epoch": 0.6481275733612182,
-      "grad_norm": 0.4282979667186737,
-      "learning_rate": 9.044838659186417e-06,
-      "loss": 0.3675,
-      "step": 9917
-    },
-    {
-      "epoch": 0.6481929285667604,
-      "grad_norm": 0.4598449468612671,
-      "learning_rate": 9.044633376911857e-06,
-      "loss": 0.3787,
-      "step": 9918
-    },
-    {
-      "epoch": 0.6482582837723024,
-      "grad_norm": 0.4358254075050354,
-      "learning_rate": 9.04442807491022e-06,
-      "loss": 0.3695,
-      "step": 9919
-    },
-    {
-      "epoch": 0.6483236389778446,
-      "grad_norm": 0.4686878025531769,
-      "learning_rate": 9.044222753182502e-06,
-      "loss": 0.4344,
-      "step": 9920
-    },
-    {
-      "epoch": 0.6483889941833867,
-      "grad_norm": 0.4279601573944092,
-      "learning_rate": 9.04401741172971e-06,
-      "loss": 0.3576,
-      "step": 9921
-    },
-    {
-      "epoch": 0.6484543493889289,
-      "grad_norm": 0.4866029918193817,
-      "learning_rate": 9.043812050552847e-06,
-      "loss": 0.403,
-      "step": 9922
-    },
-    {
-      "epoch": 0.648519704594471,
-      "grad_norm": 0.4783462584018707,
-      "learning_rate": 9.043606669652909e-06,
-      "loss": 0.4194,
-      "step": 9923
-    },
-    {
-      "epoch": 0.648585059800013,
-      "grad_norm": 0.456325888633728,
-      "learning_rate": 9.0434012690309e-06,
-      "loss": 0.3844,
-      "step": 9924
-    },
-    {
-      "epoch": 0.6486504150055552,
-      "grad_norm": 0.4528539180755615,
-      "learning_rate": 9.043195848687824e-06,
-      "loss": 0.38,
-      "step": 9925
-    },
-    {
-      "epoch": 0.6487157702110973,
-      "grad_norm": 0.5295087695121765,
-      "learning_rate": 9.04299040862468e-06,
-      "loss": 0.4731,
-      "step": 9926
-    },
-    {
-      "epoch": 0.6487811254166395,
-      "grad_norm": 0.4058385193347931,
-      "learning_rate": 9.042784948842471e-06,
-      "loss": 0.3166,
-      "step": 9927
-    },
-    {
-      "epoch": 0.6488464806221815,
-      "grad_norm": 0.4409596920013428,
-      "learning_rate": 9.042579469342201e-06,
-      "loss": 0.3637,
-      "step": 9928
-    },
-    {
-      "epoch": 0.6489118358277237,
-      "grad_norm": 0.44786572456359863,
-      "learning_rate": 9.042373970124869e-06,
-      "loss": 0.3934,
-      "step": 9929
-    },
-    {
-      "epoch": 0.6489771910332658,
-      "grad_norm": 0.4519539177417755,
-      "learning_rate": 9.042168451191478e-06,
-      "loss": 0.3635,
-      "step": 9930
-    },
-    {
-      "epoch": 0.649042546238808,
-      "grad_norm": 0.42357999086380005,
-      "learning_rate": 9.041962912543033e-06,
-      "loss": 0.3735,
-      "step": 9931
-    },
-    {
-      "epoch": 0.64910790144435,
-      "grad_norm": 0.44480448961257935,
-      "learning_rate": 9.041757354180533e-06,
-      "loss": 0.3543,
-      "step": 9932
-    },
-    {
-      "epoch": 0.6491732566498921,
-      "grad_norm": 0.43090957403182983,
-      "learning_rate": 9.041551776104982e-06,
-      "loss": 0.3414,
-      "step": 9933
-    },
-    {
-      "epoch": 0.6492386118554343,
-      "grad_norm": 0.4563465714454651,
-      "learning_rate": 9.041346178317385e-06,
-      "loss": 0.4193,
-      "step": 9934
-    },
-    {
-      "epoch": 0.6493039670609764,
-      "grad_norm": 0.43380168080329895,
-      "learning_rate": 9.041140560818742e-06,
-      "loss": 0.3844,
-      "step": 9935
-    },
-    {
-      "epoch": 0.6493693222665186,
-      "grad_norm": 0.4181695878505707,
-      "learning_rate": 9.040934923610055e-06,
-      "loss": 0.3412,
-      "step": 9936
-    },
-    {
-      "epoch": 0.6494346774720606,
-      "grad_norm": 0.4257018566131592,
-      "learning_rate": 9.040729266692329e-06,
-      "loss": 0.3695,
-      "step": 9937
-    },
-    {
-      "epoch": 0.6495000326776028,
-      "grad_norm": 0.4463997185230255,
-      "learning_rate": 9.040523590066567e-06,
-      "loss": 0.3977,
-      "step": 9938
-    },
-    {
-      "epoch": 0.6495653878831449,
-      "grad_norm": 0.4287792444229126,
-      "learning_rate": 9.040317893733772e-06,
-      "loss": 0.3249,
-      "step": 9939
-    },
-    {
-      "epoch": 0.6496307430886871,
-      "grad_norm": 0.44113147258758545,
-      "learning_rate": 9.040112177694947e-06,
-      "loss": 0.3643,
-      "step": 9940
-    },
-    {
-      "epoch": 0.6496960982942291,
-      "grad_norm": 0.4458673596382141,
-      "learning_rate": 9.039906441951095e-06,
-      "loss": 0.3844,
-      "step": 9941
-    },
-    {
-      "epoch": 0.6497614534997712,
-      "grad_norm": 0.5157591700553894,
-      "learning_rate": 9.039700686503218e-06,
-      "loss": 0.4501,
-      "step": 9942
-    },
-    {
-      "epoch": 0.6498268087053134,
-      "grad_norm": 0.42506077885627747,
-      "learning_rate": 9.039494911352324e-06,
-      "loss": 0.372,
-      "step": 9943
-    },
-    {
-      "epoch": 0.6498921639108555,
-      "grad_norm": 0.44462844729423523,
-      "learning_rate": 9.039289116499412e-06,
-      "loss": 0.4034,
-      "step": 9944
-    },
-    {
-      "epoch": 0.6499575191163977,
-      "grad_norm": 0.4465774893760681,
-      "learning_rate": 9.039083301945489e-06,
-      "loss": 0.3769,
-      "step": 9945
-    },
-    {
-      "epoch": 0.6500228743219397,
-      "grad_norm": 0.5269215703010559,
-      "learning_rate": 9.038877467691555e-06,
-      "loss": 0.489,
-      "step": 9946
-    },
-    {
-      "epoch": 0.6500882295274819,
-      "grad_norm": 0.4372009336948395,
-      "learning_rate": 9.03867161373862e-06,
-      "loss": 0.3708,
-      "step": 9947
-    },
-    {
-      "epoch": 0.650153584733024,
-      "grad_norm": 0.40767350792884827,
-      "learning_rate": 9.038465740087683e-06,
-      "loss": 0.3304,
-      "step": 9948
-    },
-    {
-      "epoch": 0.650218939938566,
-      "grad_norm": 0.4977891445159912,
-      "learning_rate": 9.038259846739748e-06,
-      "loss": 0.3841,
-      "step": 9949
-    },
-    {
-      "epoch": 0.6502842951441082,
-      "grad_norm": 0.4250839650630951,
-      "learning_rate": 9.038053933695823e-06,
-      "loss": 0.3424,
-      "step": 9950
-    },
-    {
-      "epoch": 0.6503496503496503,
-      "grad_norm": 0.4356164038181305,
-      "learning_rate": 9.037848000956908e-06,
-      "loss": 0.3512,
-      "step": 9951
-    },
-    {
-      "epoch": 0.6504150055551925,
-      "grad_norm": 0.46105310320854187,
-      "learning_rate": 9.03764204852401e-06,
-      "loss": 0.4104,
-      "step": 9952
-    },
-    {
-      "epoch": 0.6504803607607346,
-      "grad_norm": 0.4970625340938568,
-      "learning_rate": 9.037436076398134e-06,
-      "loss": 0.4359,
-      "step": 9953
-    },
-    {
-      "epoch": 0.6505457159662767,
-      "grad_norm": 0.47230228781700134,
-      "learning_rate": 9.037230084580281e-06,
-      "loss": 0.3773,
-      "step": 9954
-    },
-    {
-      "epoch": 0.6506110711718188,
-      "grad_norm": 0.46225905418395996,
-      "learning_rate": 9.037024073071461e-06,
-      "loss": 0.405,
-      "step": 9955
-    },
-    {
-      "epoch": 0.650676426377361,
-      "grad_norm": 0.462067186832428,
-      "learning_rate": 9.036818041872674e-06,
-      "loss": 0.413,
-      "step": 9956
-    },
-    {
-      "epoch": 0.6507417815829031,
-      "grad_norm": 0.45248153805732727,
-      "learning_rate": 9.036611990984929e-06,
-      "loss": 0.4196,
-      "step": 9957
-    },
-    {
-      "epoch": 0.6508071367884452,
-      "grad_norm": 0.4767220914363861,
-      "learning_rate": 9.036405920409229e-06,
-      "loss": 0.3788,
-      "step": 9958
-    },
-    {
-      "epoch": 0.6508724919939873,
-      "grad_norm": 0.4543706178665161,
-      "learning_rate": 9.036199830146577e-06,
-      "loss": 0.3815,
-      "step": 9959
-    },
-    {
-      "epoch": 0.6509378471995294,
-      "grad_norm": 0.46349817514419556,
-      "learning_rate": 9.035993720197982e-06,
-      "loss": 0.3651,
-      "step": 9960
-    },
-    {
-      "epoch": 0.6510032024050716,
-      "grad_norm": 0.44254183769226074,
-      "learning_rate": 9.035787590564446e-06,
-      "loss": 0.3244,
-      "step": 9961
-    },
-    {
-      "epoch": 0.6510685576106137,
-      "grad_norm": 0.47205328941345215,
-      "learning_rate": 9.035581441246977e-06,
-      "loss": 0.3976,
-      "step": 9962
-    },
-    {
-      "epoch": 0.6511339128161558,
-      "grad_norm": 0.48458147048950195,
-      "learning_rate": 9.035375272246579e-06,
-      "loss": 0.4239,
-      "step": 9963
-    },
-    {
-      "epoch": 0.6511992680216979,
-      "grad_norm": 0.45239412784576416,
-      "learning_rate": 9.035169083564257e-06,
-      "loss": 0.3671,
-      "step": 9964
-    },
-    {
-      "epoch": 0.6512646232272401,
-      "grad_norm": 0.46970847249031067,
-      "learning_rate": 9.034962875201016e-06,
-      "loss": 0.3501,
-      "step": 9965
-    },
-    {
-      "epoch": 0.6513299784327822,
-      "grad_norm": 0.46872833371162415,
-      "learning_rate": 9.034756647157864e-06,
-      "loss": 0.3421,
-      "step": 9966
-    },
-    {
-      "epoch": 0.6513953336383242,
-      "grad_norm": 0.46653464436531067,
-      "learning_rate": 9.034550399435808e-06,
-      "loss": 0.3739,
-      "step": 9967
-    },
-    {
-      "epoch": 0.6514606888438664,
-      "grad_norm": 0.4715349078178406,
-      "learning_rate": 9.034344132035853e-06,
-      "loss": 0.4085,
-      "step": 9968
-    },
-    {
-      "epoch": 0.6515260440494085,
-      "grad_norm": 0.4499291479587555,
-      "learning_rate": 9.034137844959e-06,
-      "loss": 0.3884,
-      "step": 9969
-    },
-    {
-      "epoch": 0.6515913992549507,
-      "grad_norm": 0.49279195070266724,
-      "learning_rate": 9.033931538206263e-06,
-      "loss": 0.3394,
-      "step": 9970
-    },
-    {
-      "epoch": 0.6516567544604928,
-      "grad_norm": 0.4347681999206543,
-      "learning_rate": 9.033725211778641e-06,
-      "loss": 0.3516,
-      "step": 9971
-    },
-    {
-      "epoch": 0.6517221096660349,
-      "grad_norm": 0.4902048110961914,
-      "learning_rate": 9.033518865677147e-06,
-      "loss": 0.3936,
-      "step": 9972
-    },
-    {
-      "epoch": 0.651787464871577,
-      "grad_norm": 0.4415493607521057,
-      "learning_rate": 9.033312499902782e-06,
-      "loss": 0.3792,
-      "step": 9973
-    },
-    {
-      "epoch": 0.6518528200771191,
-      "grad_norm": 0.4481008052825928,
-      "learning_rate": 9.033106114456555e-06,
-      "loss": 0.342,
-      "step": 9974
-    },
-    {
-      "epoch": 0.6519181752826613,
-      "grad_norm": 0.42084819078445435,
-      "learning_rate": 9.032899709339473e-06,
-      "loss": 0.3781,
-      "step": 9975
-    },
-    {
-      "epoch": 0.6519835304882033,
-      "grad_norm": 0.4919080436229706,
-      "learning_rate": 9.032693284552541e-06,
-      "loss": 0.4113,
-      "step": 9976
-    },
-    {
-      "epoch": 0.6520488856937455,
-      "grad_norm": 0.5064375996589661,
-      "learning_rate": 9.032486840096768e-06,
-      "loss": 0.4421,
-      "step": 9977
-    },
-    {
-      "epoch": 0.6521142408992876,
-      "grad_norm": 0.48175880312919617,
-      "learning_rate": 9.03228037597316e-06,
-      "loss": 0.4486,
-      "step": 9978
-    },
-    {
-      "epoch": 0.6521795961048298,
-      "grad_norm": 0.42637887597084045,
-      "learning_rate": 9.032073892182721e-06,
-      "loss": 0.3972,
-      "step": 9979
-    },
-    {
-      "epoch": 0.6522449513103719,
-      "grad_norm": 0.44986969232559204,
-      "learning_rate": 9.031867388726463e-06,
-      "loss": 0.4007,
-      "step": 9980
-    },
-    {
-      "epoch": 0.652310306515914,
-      "grad_norm": 0.420212984085083,
-      "learning_rate": 9.031660865605389e-06,
-      "loss": 0.3474,
-      "step": 9981
-    },
-    {
-      "epoch": 0.6523756617214561,
-      "grad_norm": 0.4912068247795105,
-      "learning_rate": 9.031454322820511e-06,
-      "loss": 0.4405,
-      "step": 9982
-    },
-    {
-      "epoch": 0.6524410169269982,
-      "grad_norm": 0.46444404125213623,
-      "learning_rate": 9.031247760372831e-06,
-      "loss": 0.3987,
-      "step": 9983
-    },
-    {
-      "epoch": 0.6525063721325404,
-      "grad_norm": 0.4622892141342163,
-      "learning_rate": 9.031041178263362e-06,
-      "loss": 0.4178,
-      "step": 9984
-    },
-    {
-      "epoch": 0.6525717273380824,
-      "grad_norm": 0.4760500192642212,
-      "learning_rate": 9.030834576493105e-06,
-      "loss": 0.4224,
-      "step": 9985
-    },
-    {
-      "epoch": 0.6526370825436246,
-      "grad_norm": 0.44501349329948425,
-      "learning_rate": 9.030627955063075e-06,
-      "loss": 0.3686,
-      "step": 9986
-    },
-    {
-      "epoch": 0.6527024377491667,
-      "grad_norm": 0.4655097723007202,
-      "learning_rate": 9.030421313974275e-06,
-      "loss": 0.4026,
-      "step": 9987
-    },
-    {
-      "epoch": 0.6527677929547089,
-      "grad_norm": 0.5385804176330566,
-      "learning_rate": 9.030214653227713e-06,
-      "loss": 0.4118,
-      "step": 9988
-    },
-    {
-      "epoch": 0.652833148160251,
-      "grad_norm": 0.4324919581413269,
-      "learning_rate": 9.0300079728244e-06,
-      "loss": 0.3532,
-      "step": 9989
-    },
-    {
-      "epoch": 0.6528985033657931,
-      "grad_norm": 0.4297036826610565,
-      "learning_rate": 9.02980127276534e-06,
-      "loss": 0.3586,
-      "step": 9990
-    },
-    {
-      "epoch": 0.6529638585713352,
-      "grad_norm": 0.436282753944397,
-      "learning_rate": 9.029594553051543e-06,
-      "loss": 0.3716,
-      "step": 9991
-    },
-    {
-      "epoch": 0.6530292137768773,
-      "grad_norm": 0.46413522958755493,
-      "learning_rate": 9.029387813684018e-06,
-      "loss": 0.4013,
-      "step": 9992
-    },
-    {
-      "epoch": 0.6530945689824195,
-      "grad_norm": 0.48544174432754517,
-      "learning_rate": 9.029181054663772e-06,
-      "loss": 0.3904,
-      "step": 9993
-    },
-    {
-      "epoch": 0.6531599241879615,
-      "grad_norm": 0.43784505128860474,
-      "learning_rate": 9.028974275991815e-06,
-      "loss": 0.3558,
-      "step": 9994
-    },
-    {
-      "epoch": 0.6532252793935037,
-      "grad_norm": 0.4579785168170929,
-      "learning_rate": 9.028767477669156e-06,
-      "loss": 0.4014,
-      "step": 9995
-    },
-    {
-      "epoch": 0.6532906345990458,
-      "grad_norm": 0.5092059969902039,
-      "learning_rate": 9.0285606596968e-06,
-      "loss": 0.501,
-      "step": 9996
-    },
-    {
-      "epoch": 0.653355989804588,
-      "grad_norm": 0.4619104266166687,
-      "learning_rate": 9.02835382207576e-06,
-      "loss": 0.417,
-      "step": 9997
-    },
-    {
-      "epoch": 0.65342134501013,
-      "grad_norm": 0.4432462155818939,
-      "learning_rate": 9.02814696480704e-06,
-      "loss": 0.4044,
-      "step": 9998
-    },
-    {
-      "epoch": 0.6534867002156722,
-      "grad_norm": 0.45185190439224243,
-      "learning_rate": 9.027940087891655e-06,
-      "loss": 0.4071,
-      "step": 9999
-    },
-    {
-      "epoch": 0.6535520554212143,
-      "grad_norm": 0.42490679025650024,
-      "learning_rate": 9.02773319133061e-06,
-      "loss": 0.3441,
-      "step": 10000
-    },
-    {
-      "epoch": 0.6536174106267564,
-      "grad_norm": 0.4673260748386383,
-      "learning_rate": 9.027526275124913e-06,
-      "loss": 0.4092,
-      "step": 10001
-    },
-    {
-      "epoch": 0.6536827658322986,
-      "grad_norm": 0.4620700478553772,
-      "learning_rate": 9.027319339275577e-06,
-      "loss": 0.3714,
-      "step": 10002
-    },
-    {
-      "epoch": 0.6537481210378406,
-      "grad_norm": 0.45150327682495117,
-      "learning_rate": 9.027112383783608e-06,
-      "loss": 0.3714,
-      "step": 10003
-    },
-    {
-      "epoch": 0.6538134762433828,
-      "grad_norm": 0.47253942489624023,
-      "learning_rate": 9.026905408650017e-06,
-      "loss": 0.4131,
-      "step": 10004
-    },
-    {
-      "epoch": 0.6538788314489249,
-      "grad_norm": 0.47828763723373413,
-      "learning_rate": 9.026698413875815e-06,
-      "loss": 0.3979,
-      "step": 10005
-    },
-    {
-      "epoch": 0.6539441866544671,
-      "grad_norm": 0.4524450898170471,
-      "learning_rate": 9.026491399462008e-06,
-      "loss": 0.3672,
-      "step": 10006
-    },
-    {
-      "epoch": 0.6540095418600091,
-      "grad_norm": 0.5243582725524902,
-      "learning_rate": 9.026284365409608e-06,
-      "loss": 0.4326,
-      "step": 10007
-    },
-    {
-      "epoch": 0.6540748970655512,
-      "grad_norm": 0.44284766912460327,
-      "learning_rate": 9.026077311719622e-06,
-      "loss": 0.3493,
-      "step": 10008
-    },
-    {
-      "epoch": 0.6541402522710934,
-      "grad_norm": 0.4645557403564453,
-      "learning_rate": 9.025870238393067e-06,
-      "loss": 0.3862,
-      "step": 10009
-    },
-    {
-      "epoch": 0.6542056074766355,
-      "grad_norm": 0.4484542906284332,
-      "learning_rate": 9.025663145430945e-06,
-      "loss": 0.3679,
-      "step": 10010
-    },
-    {
-      "epoch": 0.6542709626821777,
-      "grad_norm": 0.47134894132614136,
-      "learning_rate": 9.02545603283427e-06,
-      "loss": 0.4028,
-      "step": 10011
-    },
-    {
-      "epoch": 0.6543363178877197,
-      "grad_norm": 0.46165215969085693,
-      "learning_rate": 9.025248900604052e-06,
-      "loss": 0.3893,
-      "step": 10012
-    },
-    {
-      "epoch": 0.6544016730932619,
-      "grad_norm": 0.4478997588157654,
-      "learning_rate": 9.0250417487413e-06,
-      "loss": 0.378,
-      "step": 10013
-    },
-    {
-      "epoch": 0.654467028298804,
-      "grad_norm": 0.4455048143863678,
-      "learning_rate": 9.024834577247024e-06,
-      "loss": 0.3596,
-      "step": 10014
-    },
-    {
-      "epoch": 0.6545323835043462,
-      "grad_norm": 0.45794814825057983,
-      "learning_rate": 9.024627386122238e-06,
-      "loss": 0.4211,
-      "step": 10015
-    },
-    {
-      "epoch": 0.6545977387098882,
-      "grad_norm": 0.543355405330658,
-      "learning_rate": 9.024420175367947e-06,
-      "loss": 0.3924,
-      "step": 10016
-    },
-    {
-      "epoch": 0.6546630939154303,
-      "grad_norm": 0.4654608368873596,
-      "learning_rate": 9.024212944985167e-06,
-      "loss": 0.4013,
-      "step": 10017
-    },
-    {
-      "epoch": 0.6547284491209725,
-      "grad_norm": 0.4858318865299225,
-      "learning_rate": 9.024005694974904e-06,
-      "loss": 0.416,
-      "step": 10018
-    },
-    {
-      "epoch": 0.6547938043265146,
-      "grad_norm": 0.4304412603378296,
-      "learning_rate": 9.023798425338173e-06,
-      "loss": 0.3342,
-      "step": 10019
-    },
-    {
-      "epoch": 0.6548591595320568,
-      "grad_norm": 0.4354506731033325,
-      "learning_rate": 9.023591136075982e-06,
-      "loss": 0.3341,
-      "step": 10020
-    },
-    {
-      "epoch": 0.6549245147375988,
-      "grad_norm": 0.4382390081882477,
-      "learning_rate": 9.023383827189345e-06,
-      "loss": 0.3692,
-      "step": 10021
-    },
-    {
-      "epoch": 0.654989869943141,
-      "grad_norm": 0.48136866092681885,
-      "learning_rate": 9.02317649867927e-06,
-      "loss": 0.4207,
-      "step": 10022
-    },
-    {
-      "epoch": 0.6550552251486831,
-      "grad_norm": 0.47718822956085205,
-      "learning_rate": 9.022969150546769e-06,
-      "loss": 0.3914,
-      "step": 10023
-    },
-    {
-      "epoch": 0.6551205803542253,
-      "grad_norm": 0.45839494466781616,
-      "learning_rate": 9.022761782792855e-06,
-      "loss": 0.388,
-      "step": 10024
-    },
-    {
-      "epoch": 0.6551859355597673,
-      "grad_norm": 0.40892085433006287,
-      "learning_rate": 9.022554395418537e-06,
-      "loss": 0.3322,
-      "step": 10025
-    },
-    {
-      "epoch": 0.6552512907653094,
-      "grad_norm": 0.4357730448246002,
-      "learning_rate": 9.022346988424827e-06,
-      "loss": 0.3414,
-      "step": 10026
-    },
-    {
-      "epoch": 0.6553166459708516,
-      "grad_norm": 0.41670113801956177,
-      "learning_rate": 9.02213956181274e-06,
-      "loss": 0.3221,
-      "step": 10027
-    },
-    {
-      "epoch": 0.6553820011763937,
-      "grad_norm": 0.4133478105068207,
-      "learning_rate": 9.021932115583282e-06,
-      "loss": 0.342,
-      "step": 10028
-    },
-    {
-      "epoch": 0.6554473563819359,
-      "grad_norm": 0.49278882145881653,
-      "learning_rate": 9.021724649737469e-06,
-      "loss": 0.3871,
-      "step": 10029
-    },
-    {
-      "epoch": 0.6555127115874779,
-      "grad_norm": 0.482332706451416,
-      "learning_rate": 9.021517164276312e-06,
-      "loss": 0.4091,
-      "step": 10030
-    },
-    {
-      "epoch": 0.6555780667930201,
-      "grad_norm": 0.448758602142334,
-      "learning_rate": 9.021309659200822e-06,
-      "loss": 0.3432,
-      "step": 10031
-    },
-    {
-      "epoch": 0.6556434219985622,
-      "grad_norm": 0.44994425773620605,
-      "learning_rate": 9.021102134512011e-06,
-      "loss": 0.3957,
-      "step": 10032
-    },
-    {
-      "epoch": 0.6557087772041043,
-      "grad_norm": 0.4392760097980499,
-      "learning_rate": 9.020894590210893e-06,
-      "loss": 0.3801,
-      "step": 10033
-    },
-    {
-      "epoch": 0.6557741324096464,
-      "grad_norm": 0.48466914892196655,
-      "learning_rate": 9.020687026298478e-06,
-      "loss": 0.441,
-      "step": 10034
-    },
-    {
-      "epoch": 0.6558394876151885,
-      "grad_norm": 0.4930497407913208,
-      "learning_rate": 9.02047944277578e-06,
-      "loss": 0.3783,
-      "step": 10035
-    },
-    {
-      "epoch": 0.6559048428207307,
-      "grad_norm": 0.4536300003528595,
-      "learning_rate": 9.020271839643813e-06,
-      "loss": 0.3825,
-      "step": 10036
-    },
-    {
-      "epoch": 0.6559701980262728,
-      "grad_norm": 0.44621482491493225,
-      "learning_rate": 9.020064216903586e-06,
-      "loss": 0.3624,
-      "step": 10037
-    },
-    {
-      "epoch": 0.656035553231815,
-      "grad_norm": 0.4715002477169037,
-      "learning_rate": 9.019856574556112e-06,
-      "loss": 0.4179,
-      "step": 10038
-    },
-    {
-      "epoch": 0.656100908437357,
-      "grad_norm": 0.4448879361152649,
-      "learning_rate": 9.019648912602405e-06,
-      "loss": 0.3903,
-      "step": 10039
-    },
-    {
-      "epoch": 0.6561662636428992,
-      "grad_norm": 0.43843865394592285,
-      "learning_rate": 9.01944123104348e-06,
-      "loss": 0.3967,
-      "step": 10040
-    },
-    {
-      "epoch": 0.6562316188484413,
-      "grad_norm": 0.4688895642757416,
-      "learning_rate": 9.019233529880346e-06,
-      "loss": 0.4029,
-      "step": 10041
-    },
-    {
-      "epoch": 0.6562969740539834,
-      "grad_norm": 0.433026522397995,
-      "learning_rate": 9.019025809114018e-06,
-      "loss": 0.3715,
-      "step": 10042
-    },
-    {
-      "epoch": 0.6563623292595255,
-      "grad_norm": 0.5355081558227539,
-      "learning_rate": 9.018818068745507e-06,
-      "loss": 0.4969,
-      "step": 10043
-    },
-    {
-      "epoch": 0.6564276844650676,
-      "grad_norm": 0.42214125394821167,
-      "learning_rate": 9.01861030877583e-06,
-      "loss": 0.3387,
-      "step": 10044
-    },
-    {
-      "epoch": 0.6564930396706098,
-      "grad_norm": 0.4066252112388611,
-      "learning_rate": 9.018402529205998e-06,
-      "loss": 0.3542,
-      "step": 10045
-    },
-    {
-      "epoch": 0.6565583948761519,
-      "grad_norm": 0.4637167453765869,
-      "learning_rate": 9.018194730037024e-06,
-      "loss": 0.4214,
-      "step": 10046
-    },
-    {
-      "epoch": 0.656623750081694,
-      "grad_norm": 0.4537128806114197,
-      "learning_rate": 9.017986911269924e-06,
-      "loss": 0.4039,
-      "step": 10047
-    },
-    {
-      "epoch": 0.6566891052872361,
-      "grad_norm": 0.468189001083374,
-      "learning_rate": 9.01777907290571e-06,
-      "loss": 0.4015,
-      "step": 10048
-    },
-    {
-      "epoch": 0.6567544604927783,
-      "grad_norm": 0.4477204382419586,
-      "learning_rate": 9.017571214945394e-06,
-      "loss": 0.3523,
-      "step": 10049
-    },
-    {
-      "epoch": 0.6568198156983204,
-      "grad_norm": 0.46103546023368835,
-      "learning_rate": 9.01736333738999e-06,
-      "loss": 0.4216,
-      "step": 10050
-    },
-    {
-      "epoch": 0.6568851709038624,
-      "grad_norm": 0.5244162082672119,
-      "learning_rate": 9.017155440240517e-06,
-      "loss": 0.4784,
-      "step": 10051
-    },
-    {
-      "epoch": 0.6569505261094046,
-      "grad_norm": 0.4482291340827942,
-      "learning_rate": 9.016947523497983e-06,
-      "loss": 0.3992,
-      "step": 10052
-    },
-    {
-      "epoch": 0.6570158813149467,
-      "grad_norm": 0.4626525342464447,
-      "learning_rate": 9.016739587163403e-06,
-      "loss": 0.3678,
-      "step": 10053
-    },
-    {
-      "epoch": 0.6570812365204889,
-      "grad_norm": 0.428783655166626,
-      "learning_rate": 9.016531631237794e-06,
-      "loss": 0.3786,
-      "step": 10054
-    },
-    {
-      "epoch": 0.657146591726031,
-      "grad_norm": 0.4373444616794586,
-      "learning_rate": 9.01632365572217e-06,
-      "loss": 0.3669,
-      "step": 10055
-    },
-    {
-      "epoch": 0.6572119469315731,
-      "grad_norm": 0.44966745376586914,
-      "learning_rate": 9.016115660617543e-06,
-      "loss": 0.3896,
-      "step": 10056
-    },
-    {
-      "epoch": 0.6572773021371152,
-      "grad_norm": 0.45996198058128357,
-      "learning_rate": 9.015907645924929e-06,
-      "loss": 0.4169,
-      "step": 10057
-    },
-    {
-      "epoch": 0.6573426573426573,
-      "grad_norm": 0.44895103573799133,
-      "learning_rate": 9.01569961164534e-06,
-      "loss": 0.3689,
-      "step": 10058
-    },
-    {
-      "epoch": 0.6574080125481995,
-      "grad_norm": 0.4172862470149994,
-      "learning_rate": 9.015491557779796e-06,
-      "loss": 0.3591,
-      "step": 10059
-    },
-    {
-      "epoch": 0.6574733677537415,
-      "grad_norm": 0.4612584710121155,
-      "learning_rate": 9.015283484329307e-06,
-      "loss": 0.4059,
-      "step": 10060
-    },
-    {
-      "epoch": 0.6575387229592837,
-      "grad_norm": 0.43840399384498596,
-      "learning_rate": 9.015075391294889e-06,
-      "loss": 0.3864,
-      "step": 10061
-    },
-    {
-      "epoch": 0.6576040781648258,
-      "grad_norm": 0.47902926802635193,
-      "learning_rate": 9.014867278677559e-06,
-      "loss": 0.431,
-      "step": 10062
-    },
-    {
-      "epoch": 0.657669433370368,
-      "grad_norm": 0.41883060336112976,
-      "learning_rate": 9.014659146478329e-06,
-      "loss": 0.3382,
-      "step": 10063
-    },
-    {
-      "epoch": 0.6577347885759101,
-      "grad_norm": 0.4539177715778351,
-      "learning_rate": 9.014450994698217e-06,
-      "loss": 0.3757,
-      "step": 10064
-    },
-    {
-      "epoch": 0.6578001437814522,
-      "grad_norm": 0.41955089569091797,
-      "learning_rate": 9.014242823338235e-06,
-      "loss": 0.3736,
-      "step": 10065
-    },
-    {
-      "epoch": 0.6578654989869943,
-      "grad_norm": 0.4466819167137146,
-      "learning_rate": 9.0140346323994e-06,
-      "loss": 0.3787,
-      "step": 10066
-    },
-    {
-      "epoch": 0.6579308541925364,
-      "grad_norm": 0.46580857038497925,
-      "learning_rate": 9.01382642188273e-06,
-      "loss": 0.4382,
-      "step": 10067
-    },
-    {
-      "epoch": 0.6579962093980786,
-      "grad_norm": 0.46903690695762634,
-      "learning_rate": 9.013618191789236e-06,
-      "loss": 0.4091,
-      "step": 10068
-    },
-    {
-      "epoch": 0.6580615646036206,
-      "grad_norm": 0.42461681365966797,
-      "learning_rate": 9.013409942119935e-06,
-      "loss": 0.3415,
-      "step": 10069
-    },
-    {
-      "epoch": 0.6581269198091628,
-      "grad_norm": 0.45153144001960754,
-      "learning_rate": 9.013201672875844e-06,
-      "loss": 0.3655,
-      "step": 10070
-    },
-    {
-      "epoch": 0.6581922750147049,
-      "grad_norm": 0.44758397340774536,
-      "learning_rate": 9.012993384057978e-06,
-      "loss": 0.4027,
-      "step": 10071
-    },
-    {
-      "epoch": 0.6582576302202471,
-      "grad_norm": 0.4283483624458313,
-      "learning_rate": 9.012785075667354e-06,
-      "loss": 0.356,
-      "step": 10072
-    },
-    {
-      "epoch": 0.6583229854257892,
-      "grad_norm": 0.4353366494178772,
-      "learning_rate": 9.012576747704987e-06,
-      "loss": 0.3689,
-      "step": 10073
-    },
-    {
-      "epoch": 0.6583883406313313,
-      "grad_norm": 0.5016412734985352,
-      "learning_rate": 9.012368400171891e-06,
-      "loss": 0.5056,
-      "step": 10074
-    },
-    {
-      "epoch": 0.6584536958368734,
-      "grad_norm": 0.44737544655799866,
-      "learning_rate": 9.012160033069087e-06,
-      "loss": 0.3622,
-      "step": 10075
-    },
-    {
-      "epoch": 0.6585190510424155,
-      "grad_norm": 0.440403014421463,
-      "learning_rate": 9.011951646397587e-06,
-      "loss": 0.3641,
-      "step": 10076
-    },
-    {
-      "epoch": 0.6585844062479577,
-      "grad_norm": 0.474662184715271,
-      "learning_rate": 9.01174324015841e-06,
-      "loss": 0.4777,
-      "step": 10077
-    },
-    {
-      "epoch": 0.6586497614534997,
-      "grad_norm": 0.45909667015075684,
-      "learning_rate": 9.01153481435257e-06,
-      "loss": 0.4237,
-      "step": 10078
-    },
-    {
-      "epoch": 0.6587151166590419,
-      "grad_norm": 0.4570311903953552,
-      "learning_rate": 9.011326368981086e-06,
-      "loss": 0.4152,
-      "step": 10079
-    },
-    {
-      "epoch": 0.658780471864584,
-      "grad_norm": 0.46616530418395996,
-      "learning_rate": 9.011117904044972e-06,
-      "loss": 0.4215,
-      "step": 10080
-    },
-    {
-      "epoch": 0.6588458270701262,
-      "grad_norm": 0.4463542103767395,
-      "learning_rate": 9.010909419545248e-06,
-      "loss": 0.3888,
-      "step": 10081
-    },
-    {
-      "epoch": 0.6589111822756683,
-      "grad_norm": 0.466397225856781,
-      "learning_rate": 9.01070091548293e-06,
-      "loss": 0.4124,
-      "step": 10082
-    },
-    {
-      "epoch": 0.6589765374812104,
-      "grad_norm": 0.4393870532512665,
-      "learning_rate": 9.010492391859033e-06,
-      "loss": 0.3644,
-      "step": 10083
-    },
-    {
-      "epoch": 0.6590418926867525,
-      "grad_norm": 0.5008548498153687,
-      "learning_rate": 9.010283848674574e-06,
-      "loss": 0.4393,
-      "step": 10084
-    },
-    {
-      "epoch": 0.6591072478922946,
-      "grad_norm": 0.4108535349369049,
-      "learning_rate": 9.010075285930574e-06,
-      "loss": 0.3062,
-      "step": 10085
-    },
-    {
-      "epoch": 0.6591726030978368,
-      "grad_norm": 0.46413442492485046,
-      "learning_rate": 9.009866703628047e-06,
-      "loss": 0.4246,
-      "step": 10086
-    },
-    {
-      "epoch": 0.6592379583033788,
-      "grad_norm": 0.45351120829582214,
-      "learning_rate": 9.009658101768011e-06,
-      "loss": 0.4035,
-      "step": 10087
-    },
-    {
-      "epoch": 0.659303313508921,
-      "grad_norm": 0.44074538350105286,
-      "learning_rate": 9.009449480351483e-06,
-      "loss": 0.3542,
-      "step": 10088
-    },
-    {
-      "epoch": 0.6593686687144631,
-      "grad_norm": 0.45235756039619446,
-      "learning_rate": 9.009240839379479e-06,
-      "loss": 0.3827,
-      "step": 10089
-    },
-    {
-      "epoch": 0.6594340239200053,
-      "grad_norm": 0.4972812831401825,
-      "learning_rate": 9.00903217885302e-06,
-      "loss": 0.4378,
-      "step": 10090
-    },
-    {
-      "epoch": 0.6594993791255473,
-      "grad_norm": 0.42305776476860046,
-      "learning_rate": 9.008823498773122e-06,
-      "loss": 0.3952,
-      "step": 10091
-    },
-    {
-      "epoch": 0.6595647343310894,
-      "grad_norm": 0.4234190583229065,
-      "learning_rate": 9.008614799140804e-06,
-      "loss": 0.3445,
-      "step": 10092
-    },
-    {
-      "epoch": 0.6596300895366316,
-      "grad_norm": 0.46396926045417786,
-      "learning_rate": 9.008406079957081e-06,
-      "loss": 0.3906,
-      "step": 10093
-    },
-    {
-      "epoch": 0.6596954447421737,
-      "grad_norm": 0.4174894392490387,
-      "learning_rate": 9.008197341222975e-06,
-      "loss": 0.3444,
-      "step": 10094
-    },
-    {
-      "epoch": 0.6597607999477159,
-      "grad_norm": 0.45253807306289673,
-      "learning_rate": 9.007988582939502e-06,
-      "loss": 0.4106,
-      "step": 10095
-    },
-    {
-      "epoch": 0.6598261551532579,
-      "grad_norm": 0.425199031829834,
-      "learning_rate": 9.007779805107679e-06,
-      "loss": 0.3645,
-      "step": 10096
-    },
-    {
-      "epoch": 0.6598915103588001,
-      "grad_norm": 0.45950135588645935,
-      "learning_rate": 9.007571007728526e-06,
-      "loss": 0.3749,
-      "step": 10097
-    },
-    {
-      "epoch": 0.6599568655643422,
-      "grad_norm": 0.4661332666873932,
-      "learning_rate": 9.007362190803059e-06,
-      "loss": 0.4126,
-      "step": 10098
-    },
-    {
-      "epoch": 0.6600222207698844,
-      "grad_norm": 0.427520215511322,
-      "learning_rate": 9.0071533543323e-06,
-      "loss": 0.3579,
-      "step": 10099
-    },
-    {
-      "epoch": 0.6600875759754264,
-      "grad_norm": 0.45677629113197327,
-      "learning_rate": 9.006944498317268e-06,
-      "loss": 0.3946,
-      "step": 10100
-    },
-    {
-      "epoch": 0.6601529311809685,
-      "grad_norm": 0.4137193262577057,
-      "learning_rate": 9.006735622758976e-06,
-      "loss": 0.3316,
-      "step": 10101
-    },
-    {
-      "epoch": 0.6602182863865107,
-      "grad_norm": 0.408157080411911,
-      "learning_rate": 9.006526727658446e-06,
-      "loss": 0.3489,
-      "step": 10102
-    },
-    {
-      "epoch": 0.6602836415920528,
-      "grad_norm": 0.4726017713546753,
-      "learning_rate": 9.0063178130167e-06,
-      "loss": 0.4008,
-      "step": 10103
-    },
-    {
-      "epoch": 0.660348996797595,
-      "grad_norm": 0.4968073070049286,
-      "learning_rate": 9.006108878834752e-06,
-      "loss": 0.4466,
-      "step": 10104
-    },
-    {
-      "epoch": 0.660414352003137,
-      "grad_norm": 0.5128054618835449,
-      "learning_rate": 9.005899925113625e-06,
-      "loss": 0.4228,
-      "step": 10105
-    },
-    {
-      "epoch": 0.6604797072086792,
-      "grad_norm": 0.45394113659858704,
-      "learning_rate": 9.005690951854335e-06,
-      "loss": 0.3922,
-      "step": 10106
-    },
-    {
-      "epoch": 0.6605450624142213,
-      "grad_norm": 0.4606708884239197,
-      "learning_rate": 9.005481959057903e-06,
-      "loss": 0.3878,
-      "step": 10107
-    },
-    {
-      "epoch": 0.6606104176197635,
-      "grad_norm": 0.4643462896347046,
-      "learning_rate": 9.005272946725347e-06,
-      "loss": 0.3779,
-      "step": 10108
-    },
-    {
-      "epoch": 0.6606757728253055,
-      "grad_norm": 0.4563404619693756,
-      "learning_rate": 9.00506391485769e-06,
-      "loss": 0.4075,
-      "step": 10109
-    },
-    {
-      "epoch": 0.6607411280308476,
-      "grad_norm": 0.41870981454849243,
-      "learning_rate": 9.004854863455945e-06,
-      "loss": 0.3354,
-      "step": 10110
-    },
-    {
-      "epoch": 0.6608064832363898,
-      "grad_norm": 0.43378958106040955,
-      "learning_rate": 9.004645792521139e-06,
-      "loss": 0.3719,
-      "step": 10111
-    },
-    {
-      "epoch": 0.6608718384419319,
-      "grad_norm": 0.45352861285209656,
-      "learning_rate": 9.004436702054285e-06,
-      "loss": 0.3937,
-      "step": 10112
-    },
-    {
-      "epoch": 0.660937193647474,
-      "grad_norm": 0.4332524538040161,
-      "learning_rate": 9.004227592056408e-06,
-      "loss": 0.3672,
-      "step": 10113
-    },
-    {
-      "epoch": 0.6610025488530161,
-      "grad_norm": 0.4357057511806488,
-      "learning_rate": 9.004018462528524e-06,
-      "loss": 0.363,
-      "step": 10114
-    },
-    {
-      "epoch": 0.6610679040585583,
-      "grad_norm": 0.44081050157546997,
-      "learning_rate": 9.003809313471657e-06,
-      "loss": 0.3888,
-      "step": 10115
-    },
-    {
-      "epoch": 0.6611332592641004,
-      "grad_norm": 0.4499940574169159,
-      "learning_rate": 9.003600144886823e-06,
-      "loss": 0.3809,
-      "step": 10116
-    },
-    {
-      "epoch": 0.6611986144696425,
-      "grad_norm": 0.5249917507171631,
-      "learning_rate": 9.003390956775046e-06,
-      "loss": 0.4402,
-      "step": 10117
-    },
-    {
-      "epoch": 0.6612639696751846,
-      "grad_norm": 0.4136514663696289,
-      "learning_rate": 9.003181749137342e-06,
-      "loss": 0.3356,
-      "step": 10118
-    },
-    {
-      "epoch": 0.6613293248807267,
-      "grad_norm": 0.44662079215049744,
-      "learning_rate": 9.002972521974735e-06,
-      "loss": 0.3668,
-      "step": 10119
-    },
-    {
-      "epoch": 0.6613946800862689,
-      "grad_norm": 0.45356494188308716,
-      "learning_rate": 9.002763275288244e-06,
-      "loss": 0.4114,
-      "step": 10120
-    },
-    {
-      "epoch": 0.661460035291811,
-      "grad_norm": 0.46098124980926514,
-      "learning_rate": 9.00255400907889e-06,
-      "loss": 0.4092,
-      "step": 10121
-    },
-    {
-      "epoch": 0.6615253904973531,
-      "grad_norm": 0.4753960371017456,
-      "learning_rate": 9.002344723347694e-06,
-      "loss": 0.415,
-      "step": 10122
-    },
-    {
-      "epoch": 0.6615907457028952,
-      "grad_norm": 0.46595531702041626,
-      "learning_rate": 9.002135418095677e-06,
-      "loss": 0.4204,
-      "step": 10123
-    },
-    {
-      "epoch": 0.6616561009084374,
-      "grad_norm": 0.4618552029132843,
-      "learning_rate": 9.001926093323858e-06,
-      "loss": 0.3964,
-      "step": 10124
-    },
-    {
-      "epoch": 0.6617214561139795,
-      "grad_norm": 0.4507905840873718,
-      "learning_rate": 9.001716749033259e-06,
-      "loss": 0.3689,
-      "step": 10125
-    },
-    {
-      "epoch": 0.6617868113195216,
-      "grad_norm": 0.4708883762359619,
-      "learning_rate": 9.001507385224902e-06,
-      "loss": 0.4309,
-      "step": 10126
-    },
-    {
-      "epoch": 0.6618521665250637,
-      "grad_norm": 0.4288370907306671,
-      "learning_rate": 9.001298001899806e-06,
-      "loss": 0.3147,
-      "step": 10127
-    },
-    {
-      "epoch": 0.6619175217306058,
-      "grad_norm": 0.4522435963153839,
-      "learning_rate": 9.001088599058993e-06,
-      "loss": 0.3859,
-      "step": 10128
-    },
-    {
-      "epoch": 0.661982876936148,
-      "grad_norm": 0.46277713775634766,
-      "learning_rate": 9.000879176703485e-06,
-      "loss": 0.4061,
-      "step": 10129
-    },
-    {
-      "epoch": 0.6620482321416901,
-      "grad_norm": 0.43465983867645264,
-      "learning_rate": 9.000669734834304e-06,
-      "loss": 0.343,
-      "step": 10130
-    },
-    {
-      "epoch": 0.6621135873472322,
-      "grad_norm": 0.44178035855293274,
-      "learning_rate": 9.000460273452471e-06,
-      "loss": 0.3689,
-      "step": 10131
-    },
-    {
-      "epoch": 0.6621789425527743,
-      "grad_norm": 0.4584408104419708,
-      "learning_rate": 9.000250792559007e-06,
-      "loss": 0.3767,
-      "step": 10132
-    },
-    {
-      "epoch": 0.6622442977583165,
-      "grad_norm": 0.43404266238212585,
-      "learning_rate": 9.000041292154934e-06,
-      "loss": 0.3772,
-      "step": 10133
-    },
-    {
-      "epoch": 0.6623096529638586,
-      "grad_norm": 0.4522802233695984,
-      "learning_rate": 8.999831772241274e-06,
-      "loss": 0.4029,
-      "step": 10134
-    },
-    {
-      "epoch": 0.6623750081694006,
-      "grad_norm": 0.43835920095443726,
-      "learning_rate": 8.999622232819048e-06,
-      "loss": 0.3624,
-      "step": 10135
-    },
-    {
-      "epoch": 0.6624403633749428,
-      "grad_norm": 0.45403867959976196,
-      "learning_rate": 8.99941267388928e-06,
-      "loss": 0.3808,
-      "step": 10136
-    },
-    {
-      "epoch": 0.6625057185804849,
-      "grad_norm": 0.46891212463378906,
-      "learning_rate": 8.999203095452992e-06,
-      "loss": 0.399,
-      "step": 10137
-    },
-    {
-      "epoch": 0.6625710737860271,
-      "grad_norm": 0.4498675465583801,
-      "learning_rate": 8.9989934975112e-06,
-      "loss": 0.3691,
-      "step": 10138
-    },
-    {
-      "epoch": 0.6626364289915692,
-      "grad_norm": 0.460763156414032,
-      "learning_rate": 8.998783880064936e-06,
-      "loss": 0.383,
-      "step": 10139
-    },
-    {
-      "epoch": 0.6627017841971113,
-      "grad_norm": 0.4195670783519745,
-      "learning_rate": 8.998574243115216e-06,
-      "loss": 0.3422,
-      "step": 10140
-    },
-    {
-      "epoch": 0.6627671394026534,
-      "grad_norm": 0.46796733140945435,
-      "learning_rate": 8.998364586663064e-06,
-      "loss": 0.3729,
-      "step": 10141
-    },
-    {
-      "epoch": 0.6628324946081955,
-      "grad_norm": 0.4123980700969696,
-      "learning_rate": 8.998154910709505e-06,
-      "loss": 0.3425,
-      "step": 10142
-    },
-    {
-      "epoch": 0.6628978498137377,
-      "grad_norm": 0.4784967005252838,
-      "learning_rate": 8.997945215255557e-06,
-      "loss": 0.4061,
-      "step": 10143
-    },
-    {
-      "epoch": 0.6629632050192797,
-      "grad_norm": 0.4615115821361542,
-      "learning_rate": 8.997735500302246e-06,
-      "loss": 0.3443,
-      "step": 10144
-    },
-    {
-      "epoch": 0.6630285602248219,
-      "grad_norm": 0.44597235321998596,
-      "learning_rate": 8.997525765850594e-06,
-      "loss": 0.4011,
-      "step": 10145
-    },
-    {
-      "epoch": 0.663093915430364,
-      "grad_norm": 0.44559240341186523,
-      "learning_rate": 8.997316011901624e-06,
-      "loss": 0.3581,
-      "step": 10146
-    },
-    {
-      "epoch": 0.6631592706359062,
-      "grad_norm": 0.44751402735710144,
-      "learning_rate": 8.997106238456358e-06,
-      "loss": 0.3693,
-      "step": 10147
-    },
-    {
-      "epoch": 0.6632246258414483,
-      "grad_norm": 0.42394378781318665,
-      "learning_rate": 8.996896445515821e-06,
-      "loss": 0.3066,
-      "step": 10148
-    },
-    {
-      "epoch": 0.6632899810469904,
-      "grad_norm": 0.47000759840011597,
-      "learning_rate": 8.996686633081036e-06,
-      "loss": 0.4365,
-      "step": 10149
-    },
-    {
-      "epoch": 0.6633553362525325,
-      "grad_norm": 0.4312097132205963,
-      "learning_rate": 8.996476801153025e-06,
-      "loss": 0.35,
-      "step": 10150
-    },
-    {
-      "epoch": 0.6634206914580746,
-      "grad_norm": 0.48415398597717285,
-      "learning_rate": 8.996266949732811e-06,
-      "loss": 0.4425,
-      "step": 10151
-    },
-    {
-      "epoch": 0.6634860466636168,
-      "grad_norm": 0.4553367793560028,
-      "learning_rate": 8.996057078821421e-06,
-      "loss": 0.3917,
-      "step": 10152
-    },
-    {
-      "epoch": 0.6635514018691588,
-      "grad_norm": 0.4832029938697815,
-      "learning_rate": 8.995847188419875e-06,
-      "loss": 0.3904,
-      "step": 10153
-    },
-    {
-      "epoch": 0.663616757074701,
-      "grad_norm": 0.4952784776687622,
-      "learning_rate": 8.995637278529197e-06,
-      "loss": 0.402,
-      "step": 10154
-    },
-    {
-      "epoch": 0.6636821122802431,
-      "grad_norm": 0.4514484405517578,
-      "learning_rate": 8.995427349150414e-06,
-      "loss": 0.4033,
-      "step": 10155
-    },
-    {
-      "epoch": 0.6637474674857853,
-      "grad_norm": 0.4274147152900696,
-      "learning_rate": 8.995217400284547e-06,
-      "loss": 0.3255,
-      "step": 10156
-    },
-    {
-      "epoch": 0.6638128226913274,
-      "grad_norm": 0.4665510356426239,
-      "learning_rate": 8.995007431932619e-06,
-      "loss": 0.3793,
-      "step": 10157
-    },
-    {
-      "epoch": 0.6638781778968695,
-      "grad_norm": 0.46595051884651184,
-      "learning_rate": 8.994797444095658e-06,
-      "loss": 0.3842,
-      "step": 10158
-    },
-    {
-      "epoch": 0.6639435331024116,
-      "grad_norm": 0.464510977268219,
-      "learning_rate": 8.994587436774684e-06,
-      "loss": 0.3673,
-      "step": 10159
-    },
-    {
-      "epoch": 0.6640088883079537,
-      "grad_norm": 0.4403513967990875,
-      "learning_rate": 8.994377409970723e-06,
-      "loss": 0.4024,
-      "step": 10160
-    },
-    {
-      "epoch": 0.6640742435134959,
-      "grad_norm": 0.49048399925231934,
-      "learning_rate": 8.994167363684803e-06,
-      "loss": 0.4465,
-      "step": 10161
-    },
-    {
-      "epoch": 0.6641395987190379,
-      "grad_norm": 0.44556862115859985,
-      "learning_rate": 8.993957297917942e-06,
-      "loss": 0.3659,
-      "step": 10162
-    },
-    {
-      "epoch": 0.6642049539245801,
-      "grad_norm": 0.44027361273765564,
-      "learning_rate": 8.993747212671167e-06,
-      "loss": 0.3457,
-      "step": 10163
-    },
-    {
-      "epoch": 0.6642703091301222,
-      "grad_norm": 0.46174749732017517,
-      "learning_rate": 8.993537107945505e-06,
-      "loss": 0.3912,
-      "step": 10164
-    },
-    {
-      "epoch": 0.6643356643356644,
-      "grad_norm": 0.43804579973220825,
-      "learning_rate": 8.99332698374198e-06,
-      "loss": 0.3369,
-      "step": 10165
-    },
-    {
-      "epoch": 0.6644010195412065,
-      "grad_norm": 0.42869675159454346,
-      "learning_rate": 8.993116840061613e-06,
-      "loss": 0.3668,
-      "step": 10166
-    },
-    {
-      "epoch": 0.6644663747467486,
-      "grad_norm": 0.43275701999664307,
-      "learning_rate": 8.992906676905432e-06,
-      "loss": 0.3489,
-      "step": 10167
-    },
-    {
-      "epoch": 0.6645317299522907,
-      "grad_norm": 0.46868565678596497,
-      "learning_rate": 8.992696494274464e-06,
-      "loss": 0.4167,
-      "step": 10168
-    },
-    {
-      "epoch": 0.6645970851578328,
-      "grad_norm": 0.45260968804359436,
-      "learning_rate": 8.992486292169732e-06,
-      "loss": 0.3629,
-      "step": 10169
-    },
-    {
-      "epoch": 0.664662440363375,
-      "grad_norm": 0.43209195137023926,
-      "learning_rate": 8.99227607059226e-06,
-      "loss": 0.3508,
-      "step": 10170
-    },
-    {
-      "epoch": 0.664727795568917,
-      "grad_norm": 0.4369298815727234,
-      "learning_rate": 8.992065829543075e-06,
-      "loss": 0.3673,
-      "step": 10171
-    },
-    {
-      "epoch": 0.6647931507744592,
-      "grad_norm": 0.4446268379688263,
-      "learning_rate": 8.991855569023203e-06,
-      "loss": 0.3434,
-      "step": 10172
-    },
-    {
-      "epoch": 0.6648585059800013,
-      "grad_norm": 0.4940401613712311,
-      "learning_rate": 8.991645289033666e-06,
-      "loss": 0.4306,
-      "step": 10173
-    },
-    {
-      "epoch": 0.6649238611855435,
-      "grad_norm": 0.45349887013435364,
-      "learning_rate": 8.991434989575493e-06,
-      "loss": 0.368,
-      "step": 10174
-    },
-    {
-      "epoch": 0.6649892163910855,
-      "grad_norm": 0.46481719613075256,
-      "learning_rate": 8.99122467064971e-06,
-      "loss": 0.4212,
-      "step": 10175
-    },
-    {
-      "epoch": 0.6650545715966276,
-      "grad_norm": 0.46498775482177734,
-      "learning_rate": 8.991014332257341e-06,
-      "loss": 0.401,
-      "step": 10176
-    },
-    {
-      "epoch": 0.6651199268021698,
-      "grad_norm": 0.42424118518829346,
-      "learning_rate": 8.990803974399413e-06,
-      "loss": 0.3369,
-      "step": 10177
-    },
-    {
-      "epoch": 0.6651852820077119,
-      "grad_norm": 0.4934605658054352,
-      "learning_rate": 8.99059359707695e-06,
-      "loss": 0.4262,
-      "step": 10178
-    },
-    {
-      "epoch": 0.6652506372132541,
-      "grad_norm": 0.46958664059638977,
-      "learning_rate": 8.99038320029098e-06,
-      "loss": 0.4016,
-      "step": 10179
-    },
-    {
-      "epoch": 0.6653159924187961,
-      "grad_norm": 0.49012336134910583,
-      "learning_rate": 8.99017278404253e-06,
-      "loss": 0.3984,
-      "step": 10180
-    },
-    {
-      "epoch": 0.6653813476243383,
-      "grad_norm": 0.47108718752861023,
-      "learning_rate": 8.989962348332624e-06,
-      "loss": 0.3778,
-      "step": 10181
-    },
-    {
-      "epoch": 0.6654467028298804,
-      "grad_norm": 0.44810351729393005,
-      "learning_rate": 8.98975189316229e-06,
-      "loss": 0.3931,
-      "step": 10182
-    },
-    {
-      "epoch": 0.6655120580354226,
-      "grad_norm": 0.4136602580547333,
-      "learning_rate": 8.989541418532552e-06,
-      "loss": 0.3269,
-      "step": 10183
-    },
-    {
-      "epoch": 0.6655774132409646,
-      "grad_norm": 0.4662415385246277,
-      "learning_rate": 8.989330924444441e-06,
-      "loss": 0.3793,
-      "step": 10184
-    },
-    {
-      "epoch": 0.6656427684465067,
-      "grad_norm": 0.4687197804450989,
-      "learning_rate": 8.989120410898979e-06,
-      "loss": 0.3868,
-      "step": 10185
-    },
-    {
-      "epoch": 0.6657081236520489,
-      "grad_norm": 0.45381027460098267,
-      "learning_rate": 8.988909877897196e-06,
-      "loss": 0.385,
-      "step": 10186
-    },
-    {
-      "epoch": 0.665773478857591,
-      "grad_norm": 0.44065558910369873,
-      "learning_rate": 8.988699325440117e-06,
-      "loss": 0.3872,
-      "step": 10187
-    },
-    {
-      "epoch": 0.6658388340631332,
-      "grad_norm": 0.43239185214042664,
-      "learning_rate": 8.98848875352877e-06,
-      "loss": 0.3974,
-      "step": 10188
-    },
-    {
-      "epoch": 0.6659041892686752,
-      "grad_norm": 0.4161105453968048,
-      "learning_rate": 8.988278162164181e-06,
-      "loss": 0.3458,
-      "step": 10189
-    },
-    {
-      "epoch": 0.6659695444742174,
-      "grad_norm": 0.46066081523895264,
-      "learning_rate": 8.988067551347378e-06,
-      "loss": 0.4135,
-      "step": 10190
-    },
-    {
-      "epoch": 0.6660348996797595,
-      "grad_norm": 0.42983028292655945,
-      "learning_rate": 8.987856921079387e-06,
-      "loss": 0.3636,
-      "step": 10191
-    },
-    {
-      "epoch": 0.6661002548853017,
-      "grad_norm": 0.45788678526878357,
-      "learning_rate": 8.98764627136124e-06,
-      "loss": 0.3747,
-      "step": 10192
-    },
-    {
-      "epoch": 0.6661656100908437,
-      "grad_norm": 0.4494451582431793,
-      "learning_rate": 8.987435602193956e-06,
-      "loss": 0.324,
-      "step": 10193
-    },
-    {
-      "epoch": 0.6662309652963858,
-      "grad_norm": 0.4337490499019623,
-      "learning_rate": 8.98722491357857e-06,
-      "loss": 0.38,
-      "step": 10194
-    },
-    {
-      "epoch": 0.666296320501928,
-      "grad_norm": 0.47315219044685364,
-      "learning_rate": 8.987014205516104e-06,
-      "loss": 0.3972,
-      "step": 10195
-    },
-    {
-      "epoch": 0.6663616757074701,
-      "grad_norm": 0.4512397348880768,
-      "learning_rate": 8.98680347800759e-06,
-      "loss": 0.4027,
-      "step": 10196
-    },
-    {
-      "epoch": 0.6664270309130123,
-      "grad_norm": 0.42925357818603516,
-      "learning_rate": 8.986592731054056e-06,
-      "loss": 0.3546,
-      "step": 10197
-    },
-    {
-      "epoch": 0.6664923861185543,
-      "grad_norm": 0.4276587963104248,
-      "learning_rate": 8.986381964656527e-06,
-      "loss": 0.3737,
-      "step": 10198
-    },
-    {
-      "epoch": 0.6665577413240965,
-      "grad_norm": 0.4193640649318695,
-      "learning_rate": 8.986171178816032e-06,
-      "loss": 0.3719,
-      "step": 10199
-    },
-    {
-      "epoch": 0.6666230965296386,
-      "grad_norm": 0.451722115278244,
-      "learning_rate": 8.9859603735336e-06,
-      "loss": 0.4072,
-      "step": 10200
-    },
-    {
-      "epoch": 0.6666884517351807,
-      "grad_norm": 0.4299343228340149,
-      "learning_rate": 8.985749548810256e-06,
-      "loss": 0.3142,
-      "step": 10201
-    },
-    {
-      "epoch": 0.6667538069407228,
-      "grad_norm": 0.4440017640590668,
-      "learning_rate": 8.985538704647034e-06,
-      "loss": 0.3948,
-      "step": 10202
-    },
-    {
-      "epoch": 0.6668191621462649,
-      "grad_norm": 0.479810893535614,
-      "learning_rate": 8.985327841044957e-06,
-      "loss": 0.4336,
-      "step": 10203
-    },
-    {
-      "epoch": 0.6668845173518071,
-      "grad_norm": 0.4192984402179718,
-      "learning_rate": 8.985116958005056e-06,
-      "loss": 0.3201,
-      "step": 10204
-    },
-    {
-      "epoch": 0.6669498725573492,
-      "grad_norm": 0.4778785705566406,
-      "learning_rate": 8.984906055528357e-06,
-      "loss": 0.3669,
-      "step": 10205
-    },
-    {
-      "epoch": 0.6670152277628913,
-      "grad_norm": 0.45121195912361145,
-      "learning_rate": 8.984695133615893e-06,
-      "loss": 0.3842,
-      "step": 10206
-    },
-    {
-      "epoch": 0.6670805829684334,
-      "grad_norm": 0.4319092631340027,
-      "learning_rate": 8.98448419226869e-06,
-      "loss": 0.3804,
-      "step": 10207
-    },
-    {
-      "epoch": 0.6671459381739756,
-      "grad_norm": 0.4295366108417511,
-      "learning_rate": 8.984273231487776e-06,
-      "loss": 0.3643,
-      "step": 10208
-    },
-    {
-      "epoch": 0.6672112933795177,
-      "grad_norm": 0.4476512670516968,
-      "learning_rate": 8.984062251274184e-06,
-      "loss": 0.3509,
-      "step": 10209
-    },
-    {
-      "epoch": 0.6672766485850598,
-      "grad_norm": 0.4452911615371704,
-      "learning_rate": 8.98385125162894e-06,
-      "loss": 0.363,
-      "step": 10210
-    },
-    {
-      "epoch": 0.6673420037906019,
-      "grad_norm": 0.498222678899765,
-      "learning_rate": 8.983640232553071e-06,
-      "loss": 0.4138,
-      "step": 10211
-    },
-    {
-      "epoch": 0.667407358996144,
-      "grad_norm": 0.4494498670101166,
-      "learning_rate": 8.983429194047608e-06,
-      "loss": 0.3839,
-      "step": 10212
-    },
-    {
-      "epoch": 0.6674727142016862,
-      "grad_norm": 0.47864049673080444,
-      "learning_rate": 8.983218136113583e-06,
-      "loss": 0.3595,
-      "step": 10213
-    },
-    {
-      "epoch": 0.6675380694072283,
-      "grad_norm": 0.4909796714782715,
-      "learning_rate": 8.98300705875202e-06,
-      "loss": 0.4401,
-      "step": 10214
-    },
-    {
-      "epoch": 0.6676034246127704,
-      "grad_norm": 0.46578124165534973,
-      "learning_rate": 8.982795961963956e-06,
-      "loss": 0.3778,
-      "step": 10215
-    },
-    {
-      "epoch": 0.6676687798183125,
-      "grad_norm": 0.45613938570022583,
-      "learning_rate": 8.982584845750415e-06,
-      "loss": 0.4082,
-      "step": 10216
-    },
-    {
-      "epoch": 0.6677341350238547,
-      "grad_norm": 0.46102020144462585,
-      "learning_rate": 8.982373710112426e-06,
-      "loss": 0.3641,
-      "step": 10217
-    },
-    {
-      "epoch": 0.6677994902293968,
-      "grad_norm": 0.4523315131664276,
-      "learning_rate": 8.982162555051024e-06,
-      "loss": 0.4057,
-      "step": 10218
-    },
-    {
-      "epoch": 0.6678648454349388,
-      "grad_norm": 0.44271978735923767,
-      "learning_rate": 8.981951380567233e-06,
-      "loss": 0.3704,
-      "step": 10219
-    },
-    {
-      "epoch": 0.667930200640481,
-      "grad_norm": 0.44290855526924133,
-      "learning_rate": 8.981740186662087e-06,
-      "loss": 0.3823,
-      "step": 10220
-    },
-    {
-      "epoch": 0.6679955558460231,
-      "grad_norm": 0.4370899200439453,
-      "learning_rate": 8.981528973336614e-06,
-      "loss": 0.3811,
-      "step": 10221
-    },
-    {
-      "epoch": 0.6680609110515653,
-      "grad_norm": 0.4476775825023651,
-      "learning_rate": 8.981317740591844e-06,
-      "loss": 0.3949,
-      "step": 10222
-    },
-    {
-      "epoch": 0.6681262662571074,
-      "grad_norm": 0.4386078715324402,
-      "learning_rate": 8.981106488428809e-06,
-      "loss": 0.3702,
-      "step": 10223
-    },
-    {
-      "epoch": 0.6681916214626495,
-      "grad_norm": 0.44177818298339844,
-      "learning_rate": 8.98089521684854e-06,
-      "loss": 0.3688,
-      "step": 10224
-    },
-    {
-      "epoch": 0.6682569766681916,
-      "grad_norm": 0.43435531854629517,
-      "learning_rate": 8.980683925852062e-06,
-      "loss": 0.3591,
-      "step": 10225
-    },
-    {
-      "epoch": 0.6683223318737337,
-      "grad_norm": 0.471281498670578,
-      "learning_rate": 8.980472615440412e-06,
-      "loss": 0.4274,
-      "step": 10226
-    },
-    {
-      "epoch": 0.6683876870792759,
-      "grad_norm": 0.44629737734794617,
-      "learning_rate": 8.98026128561462e-06,
-      "loss": 0.3701,
-      "step": 10227
-    },
-    {
-      "epoch": 0.6684530422848179,
-      "grad_norm": 0.43073129653930664,
-      "learning_rate": 8.980049936375712e-06,
-      "loss": 0.3695,
-      "step": 10228
-    },
-    {
-      "epoch": 0.6685183974903601,
-      "grad_norm": 0.43494707345962524,
-      "learning_rate": 8.979838567724723e-06,
-      "loss": 0.401,
-      "step": 10229
-    },
-    {
-      "epoch": 0.6685837526959022,
-      "grad_norm": 0.4401400685310364,
-      "learning_rate": 8.979627179662683e-06,
-      "loss": 0.3818,
-      "step": 10230
-    },
-    {
-      "epoch": 0.6686491079014444,
-      "grad_norm": 0.4596883952617645,
-      "learning_rate": 8.97941577219062e-06,
-      "loss": 0.3776,
-      "step": 10231
-    },
-    {
-      "epoch": 0.6687144631069865,
-      "grad_norm": 0.4445232152938843,
-      "learning_rate": 8.97920434530957e-06,
-      "loss": 0.369,
-      "step": 10232
-    },
-    {
-      "epoch": 0.6687798183125286,
-      "grad_norm": 0.48138898611068726,
-      "learning_rate": 8.978992899020561e-06,
-      "loss": 0.4236,
-      "step": 10233
-    },
-    {
-      "epoch": 0.6688451735180707,
-      "grad_norm": 0.4211460053920746,
-      "learning_rate": 8.978781433324626e-06,
-      "loss": 0.3362,
-      "step": 10234
-    },
-    {
-      "epoch": 0.6689105287236128,
-      "grad_norm": 0.4986729621887207,
-      "learning_rate": 8.978569948222796e-06,
-      "loss": 0.4129,
-      "step": 10235
-    },
-    {
-      "epoch": 0.668975883929155,
-      "grad_norm": 0.42038753628730774,
-      "learning_rate": 8.978358443716099e-06,
-      "loss": 0.3466,
-      "step": 10236
-    },
-    {
-      "epoch": 0.669041239134697,
-      "grad_norm": 0.43704989552497864,
-      "learning_rate": 8.978146919805573e-06,
-      "loss": 0.371,
-      "step": 10237
-    },
-    {
-      "epoch": 0.6691065943402392,
-      "grad_norm": 0.46122992038726807,
-      "learning_rate": 8.977935376492244e-06,
-      "loss": 0.4032,
-      "step": 10238
-    },
-    {
-      "epoch": 0.6691719495457813,
-      "grad_norm": 0.43309664726257324,
-      "learning_rate": 8.977723813777148e-06,
-      "loss": 0.3422,
-      "step": 10239
-    },
-    {
-      "epoch": 0.6692373047513235,
-      "grad_norm": 0.4502905309200287,
-      "learning_rate": 8.977512231661313e-06,
-      "loss": 0.3827,
-      "step": 10240
-    },
-    {
-      "epoch": 0.6693026599568656,
-      "grad_norm": 0.5038543343544006,
-      "learning_rate": 8.977300630145773e-06,
-      "loss": 0.4451,
-      "step": 10241
-    },
-    {
-      "epoch": 0.6693680151624077,
-      "grad_norm": 0.45643964409828186,
-      "learning_rate": 8.97708900923156e-06,
-      "loss": 0.376,
-      "step": 10242
-    },
-    {
-      "epoch": 0.6694333703679498,
-      "grad_norm": 0.43658164143562317,
-      "learning_rate": 8.976877368919709e-06,
-      "loss": 0.3633,
-      "step": 10243
-    },
-    {
-      "epoch": 0.6694987255734919,
-      "grad_norm": 0.6266259551048279,
-      "learning_rate": 8.976665709211248e-06,
-      "loss": 0.3804,
-      "step": 10244
-    },
-    {
-      "epoch": 0.6695640807790341,
-      "grad_norm": 0.4313041865825653,
-      "learning_rate": 8.976454030107209e-06,
-      "loss": 0.3749,
-      "step": 10245
-    },
-    {
-      "epoch": 0.6696294359845761,
-      "grad_norm": 0.4237399101257324,
-      "learning_rate": 8.976242331608627e-06,
-      "loss": 0.3571,
-      "step": 10246
-    },
-    {
-      "epoch": 0.6696947911901183,
-      "grad_norm": 0.4492327868938446,
-      "learning_rate": 8.976030613716533e-06,
-      "loss": 0.3879,
-      "step": 10247
-    },
-    {
-      "epoch": 0.6697601463956604,
-      "grad_norm": 0.4370712637901306,
-      "learning_rate": 8.97581887643196e-06,
-      "loss": 0.3185,
-      "step": 10248
-    },
-    {
-      "epoch": 0.6698255016012026,
-      "grad_norm": 0.4478670060634613,
-      "learning_rate": 8.975607119755944e-06,
-      "loss": 0.366,
-      "step": 10249
-    },
-    {
-      "epoch": 0.6698908568067447,
-      "grad_norm": 0.4454340636730194,
-      "learning_rate": 8.975395343689512e-06,
-      "loss": 0.3365,
-      "step": 10250
-    },
-    {
-      "epoch": 0.6699562120122868,
-      "grad_norm": 0.47780612111091614,
-      "learning_rate": 8.975183548233702e-06,
-      "loss": 0.4427,
-      "step": 10251
-    },
-    {
-      "epoch": 0.6700215672178289,
-      "grad_norm": 0.45170411467552185,
-      "learning_rate": 8.974971733389542e-06,
-      "loss": 0.3855,
-      "step": 10252
-    },
-    {
-      "epoch": 0.670086922423371,
-      "grad_norm": 0.4412384629249573,
-      "learning_rate": 8.97475989915807e-06,
-      "loss": 0.3835,
-      "step": 10253
-    },
-    {
-      "epoch": 0.6701522776289132,
-      "grad_norm": 0.44694584608078003,
-      "learning_rate": 8.974548045540315e-06,
-      "loss": 0.3484,
-      "step": 10254
-    },
-    {
-      "epoch": 0.6702176328344552,
-      "grad_norm": 0.43447887897491455,
-      "learning_rate": 8.974336172537313e-06,
-      "loss": 0.394,
-      "step": 10255
-    },
-    {
-      "epoch": 0.6702829880399974,
-      "grad_norm": 0.43939846754074097,
-      "learning_rate": 8.974124280150098e-06,
-      "loss": 0.3757,
-      "step": 10256
-    },
-    {
-      "epoch": 0.6703483432455395,
-      "grad_norm": 0.4591800570487976,
-      "learning_rate": 8.973912368379701e-06,
-      "loss": 0.3886,
-      "step": 10257
-    },
-    {
-      "epoch": 0.6704136984510817,
-      "grad_norm": 0.44207674264907837,
-      "learning_rate": 8.973700437227156e-06,
-      "loss": 0.3891,
-      "step": 10258
-    },
-    {
-      "epoch": 0.6704790536566237,
-      "grad_norm": 0.5039398074150085,
-      "learning_rate": 8.973488486693499e-06,
-      "loss": 0.4825,
-      "step": 10259
-    },
-    {
-      "epoch": 0.6705444088621658,
-      "grad_norm": 0.4677008390426636,
-      "learning_rate": 8.97327651677976e-06,
-      "loss": 0.4448,
-      "step": 10260
-    },
-    {
-      "epoch": 0.670609764067708,
-      "grad_norm": 0.45407986640930176,
-      "learning_rate": 8.973064527486976e-06,
-      "loss": 0.3861,
-      "step": 10261
-    },
-    {
-      "epoch": 0.6706751192732501,
-      "grad_norm": 0.43893203139305115,
-      "learning_rate": 8.97285251881618e-06,
-      "loss": 0.3145,
-      "step": 10262
-    },
-    {
-      "epoch": 0.6707404744787923,
-      "grad_norm": 0.4304412305355072,
-      "learning_rate": 8.972640490768407e-06,
-      "loss": 0.3881,
-      "step": 10263
-    },
-    {
-      "epoch": 0.6708058296843343,
-      "grad_norm": 0.4039929211139679,
-      "learning_rate": 8.972428443344687e-06,
-      "loss": 0.3233,
-      "step": 10264
-    },
-    {
-      "epoch": 0.6708711848898765,
-      "grad_norm": 0.4385417103767395,
-      "learning_rate": 8.972216376546061e-06,
-      "loss": 0.3776,
-      "step": 10265
-    },
-    {
-      "epoch": 0.6709365400954186,
-      "grad_norm": 0.46117353439331055,
-      "learning_rate": 8.972004290373558e-06,
-      "loss": 0.3903,
-      "step": 10266
-    },
-    {
-      "epoch": 0.6710018953009608,
-      "grad_norm": 0.432743102312088,
-      "learning_rate": 8.971792184828213e-06,
-      "loss": 0.3508,
-      "step": 10267
-    },
-    {
-      "epoch": 0.6710672505065028,
-      "grad_norm": 0.8923166990280151,
-      "learning_rate": 8.971580059911063e-06,
-      "loss": 0.4817,
-      "step": 10268
-    },
-    {
-      "epoch": 0.6711326057120449,
-      "grad_norm": 0.45550790429115295,
-      "learning_rate": 8.971367915623141e-06,
-      "loss": 0.3702,
-      "step": 10269
-    },
-    {
-      "epoch": 0.6711979609175871,
-      "grad_norm": 0.45742085576057434,
-      "learning_rate": 8.971155751965481e-06,
-      "loss": 0.3985,
-      "step": 10270
-    },
-    {
-      "epoch": 0.6712633161231292,
-      "grad_norm": 0.4713760316371918,
-      "learning_rate": 8.970943568939119e-06,
-      "loss": 0.3894,
-      "step": 10271
-    },
-    {
-      "epoch": 0.6713286713286714,
-      "grad_norm": 0.43509918451309204,
-      "learning_rate": 8.970731366545089e-06,
-      "loss": 0.3609,
-      "step": 10272
-    },
-    {
-      "epoch": 0.6713940265342134,
-      "grad_norm": 0.45157599449157715,
-      "learning_rate": 8.970519144784428e-06,
-      "loss": 0.3755,
-      "step": 10273
-    },
-    {
-      "epoch": 0.6714593817397556,
-      "grad_norm": 0.45870983600616455,
-      "learning_rate": 8.97030690365817e-06,
-      "loss": 0.3668,
-      "step": 10274
-    },
-    {
-      "epoch": 0.6715247369452977,
-      "grad_norm": 0.4608015716075897,
-      "learning_rate": 8.970094643167347e-06,
-      "loss": 0.3638,
-      "step": 10275
-    },
-    {
-      "epoch": 0.6715900921508399,
-      "grad_norm": 0.48547616600990295,
-      "learning_rate": 8.969882363313e-06,
-      "loss": 0.4158,
-      "step": 10276
-    },
-    {
-      "epoch": 0.6716554473563819,
-      "grad_norm": 0.4456138610839844,
-      "learning_rate": 8.969670064096158e-06,
-      "loss": 0.3733,
-      "step": 10277
-    },
-    {
-      "epoch": 0.671720802561924,
-      "grad_norm": 0.4796470105648041,
-      "learning_rate": 8.969457745517863e-06,
-      "loss": 0.3957,
-      "step": 10278
-    },
-    {
-      "epoch": 0.6717861577674662,
-      "grad_norm": 0.43504852056503296,
-      "learning_rate": 8.969245407579147e-06,
-      "loss": 0.4011,
-      "step": 10279
-    },
-    {
-      "epoch": 0.6718515129730083,
-      "grad_norm": 0.41327905654907227,
-      "learning_rate": 8.969033050281045e-06,
-      "loss": 0.3547,
-      "step": 10280
-    },
-    {
-      "epoch": 0.6719168681785505,
-      "grad_norm": 0.4182872176170349,
-      "learning_rate": 8.968820673624594e-06,
-      "loss": 0.3735,
-      "step": 10281
-    },
-    {
-      "epoch": 0.6719822233840925,
-      "grad_norm": 0.44985899329185486,
-      "learning_rate": 8.968608277610831e-06,
-      "loss": 0.3691,
-      "step": 10282
-    },
-    {
-      "epoch": 0.6720475785896347,
-      "grad_norm": 0.43563640117645264,
-      "learning_rate": 8.968395862240789e-06,
-      "loss": 0.3708,
-      "step": 10283
-    },
-    {
-      "epoch": 0.6721129337951768,
-      "grad_norm": 0.4224267899990082,
-      "learning_rate": 8.968183427515506e-06,
-      "loss": 0.376,
-      "step": 10284
-    },
-    {
-      "epoch": 0.6721782890007189,
-      "grad_norm": 0.4384387135505676,
-      "learning_rate": 8.967970973436017e-06,
-      "loss": 0.3929,
-      "step": 10285
-    },
-    {
-      "epoch": 0.672243644206261,
-      "grad_norm": 0.42206332087516785,
-      "learning_rate": 8.96775850000336e-06,
-      "loss": 0.3545,
-      "step": 10286
-    },
-    {
-      "epoch": 0.6723089994118031,
-      "grad_norm": 0.42261624336242676,
-      "learning_rate": 8.96754600721857e-06,
-      "loss": 0.3598,
-      "step": 10287
-    },
-    {
-      "epoch": 0.6723743546173453,
-      "grad_norm": 0.39704015851020813,
-      "learning_rate": 8.967333495082684e-06,
-      "loss": 0.3231,
-      "step": 10288
-    },
-    {
-      "epoch": 0.6724397098228874,
-      "grad_norm": 0.4982104003429413,
-      "learning_rate": 8.967120963596738e-06,
-      "loss": 0.4331,
-      "step": 10289
-    },
-    {
-      "epoch": 0.6725050650284295,
-      "grad_norm": 0.4854123294353485,
-      "learning_rate": 8.966908412761768e-06,
-      "loss": 0.3927,
-      "step": 10290
-    },
-    {
-      "epoch": 0.6725704202339716,
-      "grad_norm": 0.44044816493988037,
-      "learning_rate": 8.966695842578812e-06,
-      "loss": 0.38,
-      "step": 10291
-    },
-    {
-      "epoch": 0.6726357754395138,
-      "grad_norm": 0.4597327411174774,
-      "learning_rate": 8.966483253048906e-06,
-      "loss": 0.4095,
-      "step": 10292
-    },
-    {
-      "epoch": 0.6727011306450559,
-      "grad_norm": 0.4429006278514862,
-      "learning_rate": 8.966270644173087e-06,
-      "loss": 0.3505,
-      "step": 10293
-    },
-    {
-      "epoch": 0.672766485850598,
-      "grad_norm": 0.4318757653236389,
-      "learning_rate": 8.966058015952392e-06,
-      "loss": 0.3857,
-      "step": 10294
-    },
-    {
-      "epoch": 0.6728318410561401,
-      "grad_norm": 0.4408669173717499,
-      "learning_rate": 8.965845368387859e-06,
-      "loss": 0.392,
-      "step": 10295
-    },
-    {
-      "epoch": 0.6728971962616822,
-      "grad_norm": 0.42073020339012146,
-      "learning_rate": 8.965632701480524e-06,
-      "loss": 0.3516,
-      "step": 10296
-    },
-    {
-      "epoch": 0.6729625514672244,
-      "grad_norm": 0.4711887538433075,
-      "learning_rate": 8.965420015231423e-06,
-      "loss": 0.4263,
-      "step": 10297
-    },
-    {
-      "epoch": 0.6730279066727665,
-      "grad_norm": 0.4669957756996155,
-      "learning_rate": 8.965207309641596e-06,
-      "loss": 0.4359,
-      "step": 10298
-    },
-    {
-      "epoch": 0.6730932618783086,
-      "grad_norm": 0.4311840236186981,
-      "learning_rate": 8.964994584712078e-06,
-      "loss": 0.3683,
-      "step": 10299
-    },
-    {
-      "epoch": 0.6731586170838507,
-      "grad_norm": 0.4037550985813141,
-      "learning_rate": 8.96478184044391e-06,
-      "loss": 0.3208,
-      "step": 10300
-    },
-    {
-      "epoch": 0.6732239722893929,
-      "grad_norm": 0.4380420744419098,
-      "learning_rate": 8.964569076838125e-06,
-      "loss": 0.3864,
-      "step": 10301
-    },
-    {
-      "epoch": 0.673289327494935,
-      "grad_norm": 0.45357048511505127,
-      "learning_rate": 8.964356293895765e-06,
-      "loss": 0.3947,
-      "step": 10302
-    },
-    {
-      "epoch": 0.673354682700477,
-      "grad_norm": 0.46027740836143494,
-      "learning_rate": 8.964143491617865e-06,
-      "loss": 0.3709,
-      "step": 10303
-    },
-    {
-      "epoch": 0.6734200379060192,
-      "grad_norm": 0.4550207555294037,
-      "learning_rate": 8.963930670005465e-06,
-      "loss": 0.3782,
-      "step": 10304
-    },
-    {
-      "epoch": 0.6734853931115613,
-      "grad_norm": 0.44852039217948914,
-      "learning_rate": 8.963717829059601e-06,
-      "loss": 0.4004,
-      "step": 10305
-    },
-    {
-      "epoch": 0.6735507483171035,
-      "grad_norm": 0.43903085589408875,
-      "learning_rate": 8.963504968781312e-06,
-      "loss": 0.3443,
-      "step": 10306
-    },
-    {
-      "epoch": 0.6736161035226456,
-      "grad_norm": 0.4645479619503021,
-      "learning_rate": 8.963292089171637e-06,
-      "loss": 0.3752,
-      "step": 10307
-    },
-    {
-      "epoch": 0.6736814587281877,
-      "grad_norm": 0.44583413004875183,
-      "learning_rate": 8.963079190231611e-06,
-      "loss": 0.3689,
-      "step": 10308
-    },
-    {
-      "epoch": 0.6737468139337298,
-      "grad_norm": 0.4629260003566742,
-      "learning_rate": 8.962866271962278e-06,
-      "loss": 0.3696,
-      "step": 10309
-    },
-    {
-      "epoch": 0.6738121691392719,
-      "grad_norm": 0.449081689119339,
-      "learning_rate": 8.962653334364671e-06,
-      "loss": 0.3862,
-      "step": 10310
-    },
-    {
-      "epoch": 0.6738775243448141,
-      "grad_norm": 0.4531875550746918,
-      "learning_rate": 8.962440377439833e-06,
-      "loss": 0.3647,
-      "step": 10311
-    },
-    {
-      "epoch": 0.6739428795503561,
-      "grad_norm": 0.46704763174057007,
-      "learning_rate": 8.9622274011888e-06,
-      "loss": 0.4292,
-      "step": 10312
-    },
-    {
-      "epoch": 0.6740082347558983,
-      "grad_norm": 0.4282761812210083,
-      "learning_rate": 8.96201440561261e-06,
-      "loss": 0.3708,
-      "step": 10313
-    },
-    {
-      "epoch": 0.6740735899614404,
-      "grad_norm": 0.48998647928237915,
-      "learning_rate": 8.961801390712304e-06,
-      "loss": 0.3607,
-      "step": 10314
-    },
-    {
-      "epoch": 0.6741389451669826,
-      "grad_norm": 0.45318475365638733,
-      "learning_rate": 8.96158835648892e-06,
-      "loss": 0.3827,
-      "step": 10315
-    },
-    {
-      "epoch": 0.6742043003725247,
-      "grad_norm": 0.49559691548347473,
-      "learning_rate": 8.961375302943497e-06,
-      "loss": 0.4332,
-      "step": 10316
-    },
-    {
-      "epoch": 0.6742696555780668,
-      "grad_norm": 0.43280029296875,
-      "learning_rate": 8.961162230077073e-06,
-      "loss": 0.361,
-      "step": 10317
-    },
-    {
-      "epoch": 0.6743350107836089,
-      "grad_norm": 0.41879433393478394,
-      "learning_rate": 8.96094913789069e-06,
-      "loss": 0.3611,
-      "step": 10318
-    },
-    {
-      "epoch": 0.674400365989151,
-      "grad_norm": 0.4279939830303192,
-      "learning_rate": 8.960736026385387e-06,
-      "loss": 0.3486,
-      "step": 10319
-    },
-    {
-      "epoch": 0.6744657211946932,
-      "grad_norm": 0.4643529951572418,
-      "learning_rate": 8.960522895562201e-06,
-      "loss": 0.417,
-      "step": 10320
-    },
-    {
-      "epoch": 0.6745310764002352,
-      "grad_norm": 0.43617352843284607,
-      "learning_rate": 8.960309745422173e-06,
-      "loss": 0.4038,
-      "step": 10321
-    },
-    {
-      "epoch": 0.6745964316057774,
-      "grad_norm": 0.4094836413860321,
-      "learning_rate": 8.960096575966341e-06,
-      "loss": 0.3373,
-      "step": 10322
-    },
-    {
-      "epoch": 0.6746617868113195,
-      "grad_norm": 0.4339154064655304,
-      "learning_rate": 8.959883387195749e-06,
-      "loss": 0.3267,
-      "step": 10323
-    },
-    {
-      "epoch": 0.6747271420168617,
-      "grad_norm": 0.45063844323158264,
-      "learning_rate": 8.959670179111433e-06,
-      "loss": 0.4068,
-      "step": 10324
-    },
-    {
-      "epoch": 0.6747924972224038,
-      "grad_norm": 0.43268445134162903,
-      "learning_rate": 8.959456951714431e-06,
-      "loss": 0.3635,
-      "step": 10325
-    },
-    {
-      "epoch": 0.6748578524279459,
-      "grad_norm": 0.41768670082092285,
-      "learning_rate": 8.95924370500579e-06,
-      "loss": 0.3458,
-      "step": 10326
-    },
-    {
-      "epoch": 0.674923207633488,
-      "grad_norm": 0.4855414628982544,
-      "learning_rate": 8.959030438986542e-06,
-      "loss": 0.4194,
-      "step": 10327
-    },
-    {
-      "epoch": 0.6749885628390301,
-      "grad_norm": 0.4134969711303711,
-      "learning_rate": 8.958817153657732e-06,
-      "loss": 0.3513,
-      "step": 10328
-    },
-    {
-      "epoch": 0.6750539180445723,
-      "grad_norm": 0.4402543008327484,
-      "learning_rate": 8.9586038490204e-06,
-      "loss": 0.3819,
-      "step": 10329
-    },
-    {
-      "epoch": 0.6751192732501143,
-      "grad_norm": 0.4257301986217499,
-      "learning_rate": 8.958390525075587e-06,
-      "loss": 0.348,
-      "step": 10330
-    },
-    {
-      "epoch": 0.6751846284556565,
-      "grad_norm": 0.44497087597846985,
-      "learning_rate": 8.95817718182433e-06,
-      "loss": 0.356,
-      "step": 10331
-    },
-    {
-      "epoch": 0.6752499836611986,
-      "grad_norm": 0.8429772257804871,
-      "learning_rate": 8.957963819267671e-06,
-      "loss": 0.4179,
-      "step": 10332
-    },
-    {
-      "epoch": 0.6753153388667408,
-      "grad_norm": 0.4714336097240448,
-      "learning_rate": 8.957750437406654e-06,
-      "loss": 0.4201,
-      "step": 10333
-    },
-    {
-      "epoch": 0.6753806940722829,
-      "grad_norm": 0.4567609429359436,
-      "learning_rate": 8.957537036242315e-06,
-      "loss": 0.383,
-      "step": 10334
-    },
-    {
-      "epoch": 0.675446049277825,
-      "grad_norm": 0.459941565990448,
-      "learning_rate": 8.957323615775698e-06,
-      "loss": 0.4143,
-      "step": 10335
-    },
-    {
-      "epoch": 0.6755114044833671,
-      "grad_norm": 0.4573533236980438,
-      "learning_rate": 8.95711017600784e-06,
-      "loss": 0.3875,
-      "step": 10336
-    },
-    {
-      "epoch": 0.6755767596889092,
-      "grad_norm": 0.49541574716567993,
-      "learning_rate": 8.956896716939789e-06,
-      "loss": 0.4481,
-      "step": 10337
-    },
-    {
-      "epoch": 0.6756421148944514,
-      "grad_norm": 0.43781739473342896,
-      "learning_rate": 8.95668323857258e-06,
-      "loss": 0.3754,
-      "step": 10338
-    },
-    {
-      "epoch": 0.6757074700999934,
-      "grad_norm": 0.4408060312271118,
-      "learning_rate": 8.956469740907256e-06,
-      "loss": 0.4003,
-      "step": 10339
-    },
-    {
-      "epoch": 0.6757728253055356,
-      "grad_norm": 0.4413546025753021,
-      "learning_rate": 8.95625622394486e-06,
-      "loss": 0.3688,
-      "step": 10340
-    },
-    {
-      "epoch": 0.6758381805110777,
-      "grad_norm": 0.43218860030174255,
-      "learning_rate": 8.956042687686428e-06,
-      "loss": 0.3789,
-      "step": 10341
-    },
-    {
-      "epoch": 0.6759035357166199,
-      "grad_norm": 0.44783079624176025,
-      "learning_rate": 8.955829132133009e-06,
-      "loss": 0.4132,
-      "step": 10342
-    },
-    {
-      "epoch": 0.675968890922162,
-      "grad_norm": 0.4339519143104553,
-      "learning_rate": 8.955615557285638e-06,
-      "loss": 0.389,
-      "step": 10343
-    },
-    {
-      "epoch": 0.676034246127704,
-      "grad_norm": 0.4588874280452728,
-      "learning_rate": 8.955401963145362e-06,
-      "loss": 0.3935,
-      "step": 10344
-    },
-    {
-      "epoch": 0.6760996013332462,
-      "grad_norm": 0.43356090784072876,
-      "learning_rate": 8.95518834971322e-06,
-      "loss": 0.3769,
-      "step": 10345
-    },
-    {
-      "epoch": 0.6761649565387883,
-      "grad_norm": 0.4621839225292206,
-      "learning_rate": 8.954974716990253e-06,
-      "loss": 0.3769,
-      "step": 10346
-    },
-    {
-      "epoch": 0.6762303117443305,
-      "grad_norm": 0.4320388436317444,
-      "learning_rate": 8.954761064977504e-06,
-      "loss": 0.3976,
-      "step": 10347
-    },
-    {
-      "epoch": 0.6762956669498725,
-      "grad_norm": 0.4683499038219452,
-      "learning_rate": 8.954547393676017e-06,
-      "loss": 0.3769,
-      "step": 10348
-    },
-    {
-      "epoch": 0.6763610221554147,
-      "grad_norm": 0.4357524812221527,
-      "learning_rate": 8.95433370308683e-06,
-      "loss": 0.3384,
-      "step": 10349
-    },
-    {
-      "epoch": 0.6764263773609568,
-      "grad_norm": 0.4497321546077728,
-      "learning_rate": 8.95411999321099e-06,
-      "loss": 0.4055,
-      "step": 10350
-    },
-    {
-      "epoch": 0.676491732566499,
-      "grad_norm": 0.45245251059532166,
-      "learning_rate": 8.953906264049537e-06,
-      "loss": 0.4032,
-      "step": 10351
-    },
-    {
-      "epoch": 0.676557087772041,
-      "grad_norm": 0.469921737909317,
-      "learning_rate": 8.953692515603512e-06,
-      "loss": 0.3844,
-      "step": 10352
-    },
-    {
-      "epoch": 0.6766224429775831,
-      "grad_norm": 0.4377079904079437,
-      "learning_rate": 8.953478747873958e-06,
-      "loss": 0.3858,
-      "step": 10353
-    },
-    {
-      "epoch": 0.6766877981831253,
-      "grad_norm": 0.4820059835910797,
-      "learning_rate": 8.95326496086192e-06,
-      "loss": 0.4289,
-      "step": 10354
-    },
-    {
-      "epoch": 0.6767531533886674,
-      "grad_norm": 0.4293309152126312,
-      "learning_rate": 8.953051154568439e-06,
-      "loss": 0.3383,
-      "step": 10355
-    },
-    {
-      "epoch": 0.6768185085942096,
-      "grad_norm": 0.48739153146743774,
-      "learning_rate": 8.952837328994557e-06,
-      "loss": 0.4239,
-      "step": 10356
-    },
-    {
-      "epoch": 0.6768838637997516,
-      "grad_norm": 0.45735102891921997,
-      "learning_rate": 8.952623484141321e-06,
-      "loss": 0.3719,
-      "step": 10357
-    },
-    {
-      "epoch": 0.6769492190052938,
-      "grad_norm": 0.4400736093521118,
-      "learning_rate": 8.952409620009768e-06,
-      "loss": 0.3507,
-      "step": 10358
-    },
-    {
-      "epoch": 0.6770145742108359,
-      "grad_norm": 0.45328009128570557,
-      "learning_rate": 8.952195736600946e-06,
-      "loss": 0.4131,
-      "step": 10359
-    },
-    {
-      "epoch": 0.6770799294163781,
-      "grad_norm": 0.49128541350364685,
-      "learning_rate": 8.951981833915895e-06,
-      "loss": 0.4185,
-      "step": 10360
-    },
-    {
-      "epoch": 0.6771452846219201,
-      "grad_norm": 0.49053069949150085,
-      "learning_rate": 8.951767911955659e-06,
-      "loss": 0.405,
-      "step": 10361
-    },
-    {
-      "epoch": 0.6772106398274622,
-      "grad_norm": 0.496187299489975,
-      "learning_rate": 8.951553970721283e-06,
-      "loss": 0.39,
-      "step": 10362
-    },
-    {
-      "epoch": 0.6772759950330044,
-      "grad_norm": 0.43509534001350403,
-      "learning_rate": 8.951340010213807e-06,
-      "loss": 0.3868,
-      "step": 10363
-    },
-    {
-      "epoch": 0.6773413502385465,
-      "grad_norm": 0.47046181559562683,
-      "learning_rate": 8.951126030434281e-06,
-      "loss": 0.3827,
-      "step": 10364
-    },
-    {
-      "epoch": 0.6774067054440887,
-      "grad_norm": 0.4119715094566345,
-      "learning_rate": 8.950912031383742e-06,
-      "loss": 0.3062,
-      "step": 10365
-    },
-    {
-      "epoch": 0.6774720606496307,
-      "grad_norm": 0.42869341373443604,
-      "learning_rate": 8.950698013063237e-06,
-      "loss": 0.3603,
-      "step": 10366
-    },
-    {
-      "epoch": 0.6775374158551729,
-      "grad_norm": 0.5217639207839966,
-      "learning_rate": 8.950483975473808e-06,
-      "loss": 0.4298,
-      "step": 10367
-    },
-    {
-      "epoch": 0.677602771060715,
-      "grad_norm": 0.5211995840072632,
-      "learning_rate": 8.950269918616501e-06,
-      "loss": 0.4159,
-      "step": 10368
-    },
-    {
-      "epoch": 0.677668126266257,
-      "grad_norm": 0.4355018436908722,
-      "learning_rate": 8.950055842492359e-06,
-      "loss": 0.3571,
-      "step": 10369
-    },
-    {
-      "epoch": 0.6777334814717992,
-      "grad_norm": 0.4326300621032715,
-      "learning_rate": 8.949841747102425e-06,
-      "loss": 0.3681,
-      "step": 10370
-    },
-    {
-      "epoch": 0.6777988366773413,
-      "grad_norm": 0.4292009174823761,
-      "learning_rate": 8.949627632447747e-06,
-      "loss": 0.3575,
-      "step": 10371
-    },
-    {
-      "epoch": 0.6778641918828835,
-      "grad_norm": 0.42942360043525696,
-      "learning_rate": 8.949413498529364e-06,
-      "loss": 0.3786,
-      "step": 10372
-    },
-    {
-      "epoch": 0.6779295470884256,
-      "grad_norm": 0.4630257189273834,
-      "learning_rate": 8.949199345348326e-06,
-      "loss": 0.3831,
-      "step": 10373
-    },
-    {
-      "epoch": 0.6779949022939677,
-      "grad_norm": 0.45308834314346313,
-      "learning_rate": 8.948985172905673e-06,
-      "loss": 0.3735,
-      "step": 10374
-    },
-    {
-      "epoch": 0.6780602574995098,
-      "grad_norm": 0.46515166759490967,
-      "learning_rate": 8.94877098120245e-06,
-      "loss": 0.4155,
-      "step": 10375
-    },
-    {
-      "epoch": 0.678125612705052,
-      "grad_norm": 0.42242324352264404,
-      "learning_rate": 8.948556770239706e-06,
-      "loss": 0.3537,
-      "step": 10376
-    },
-    {
-      "epoch": 0.6781909679105941,
-      "grad_norm": 0.4392739236354828,
-      "learning_rate": 8.948342540018482e-06,
-      "loss": 0.3618,
-      "step": 10377
-    },
-    {
-      "epoch": 0.6782563231161362,
-      "grad_norm": 0.459957093000412,
-      "learning_rate": 8.94812829053982e-06,
-      "loss": 0.3923,
-      "step": 10378
-    },
-    {
-      "epoch": 0.6783216783216783,
-      "grad_norm": 0.42192912101745605,
-      "learning_rate": 8.947914021804774e-06,
-      "loss": 0.3502,
-      "step": 10379
-    },
-    {
-      "epoch": 0.6783870335272204,
-      "grad_norm": 0.45790332555770874,
-      "learning_rate": 8.94769973381438e-06,
-      "loss": 0.372,
-      "step": 10380
-    },
-    {
-      "epoch": 0.6784523887327626,
-      "grad_norm": 0.47560402750968933,
-      "learning_rate": 8.947485426569688e-06,
-      "loss": 0.3781,
-      "step": 10381
-    },
-    {
-      "epoch": 0.6785177439383047,
-      "grad_norm": 0.42340466380119324,
-      "learning_rate": 8.94727110007174e-06,
-      "loss": 0.3238,
-      "step": 10382
-    },
-    {
-      "epoch": 0.6785830991438468,
-      "grad_norm": 0.46564173698425293,
-      "learning_rate": 8.947056754321585e-06,
-      "loss": 0.4187,
-      "step": 10383
-    },
-    {
-      "epoch": 0.6786484543493889,
-      "grad_norm": 0.4427543580532074,
-      "learning_rate": 8.946842389320267e-06,
-      "loss": 0.3505,
-      "step": 10384
-    },
-    {
-      "epoch": 0.6787138095549311,
-      "grad_norm": 0.45892271399497986,
-      "learning_rate": 8.946628005068831e-06,
-      "loss": 0.3271,
-      "step": 10385
-    },
-    {
-      "epoch": 0.6787791647604732,
-      "grad_norm": 0.4201567769050598,
-      "learning_rate": 8.946413601568325e-06,
-      "loss": 0.3062,
-      "step": 10386
-    },
-    {
-      "epoch": 0.6788445199660152,
-      "grad_norm": 0.4735974371433258,
-      "learning_rate": 8.94619917881979e-06,
-      "loss": 0.4426,
-      "step": 10387
-    },
-    {
-      "epoch": 0.6789098751715574,
-      "grad_norm": 0.4726061224937439,
-      "learning_rate": 8.945984736824276e-06,
-      "loss": 0.4126,
-      "step": 10388
-    },
-    {
-      "epoch": 0.6789752303770995,
-      "grad_norm": 0.4313628375530243,
-      "learning_rate": 8.945770275582826e-06,
-      "loss": 0.3585,
-      "step": 10389
-    },
-    {
-      "epoch": 0.6790405855826417,
-      "grad_norm": 0.45539984107017517,
-      "learning_rate": 8.94555579509649e-06,
-      "loss": 0.3647,
-      "step": 10390
-    },
-    {
-      "epoch": 0.6791059407881838,
-      "grad_norm": 0.4311884045600891,
-      "learning_rate": 8.945341295366309e-06,
-      "loss": 0.3227,
-      "step": 10391
-    },
-    {
-      "epoch": 0.6791712959937259,
-      "grad_norm": 0.40638798475265503,
-      "learning_rate": 8.945126776393333e-06,
-      "loss": 0.3343,
-      "step": 10392
-    },
-    {
-      "epoch": 0.679236651199268,
-      "grad_norm": 0.427643358707428,
-      "learning_rate": 8.944912238178606e-06,
-      "loss": 0.3345,
-      "step": 10393
-    },
-    {
-      "epoch": 0.6793020064048101,
-      "grad_norm": 0.4728054702281952,
-      "learning_rate": 8.944697680723176e-06,
-      "loss": 0.4271,
-      "step": 10394
-    },
-    {
-      "epoch": 0.6793673616103523,
-      "grad_norm": 0.4511755108833313,
-      "learning_rate": 8.944483104028088e-06,
-      "loss": 0.3649,
-      "step": 10395
-    },
-    {
-      "epoch": 0.6794327168158943,
-      "grad_norm": 0.4559018909931183,
-      "learning_rate": 8.94426850809439e-06,
-      "loss": 0.3862,
-      "step": 10396
-    },
-    {
-      "epoch": 0.6794980720214365,
-      "grad_norm": 0.4411986470222473,
-      "learning_rate": 8.944053892923128e-06,
-      "loss": 0.3485,
-      "step": 10397
-    },
-    {
-      "epoch": 0.6795634272269786,
-      "grad_norm": 0.44940850138664246,
-      "learning_rate": 8.94383925851535e-06,
-      "loss": 0.3835,
-      "step": 10398
-    },
-    {
-      "epoch": 0.6796287824325208,
-      "grad_norm": 0.44346609711647034,
-      "learning_rate": 8.9436246048721e-06,
-      "loss": 0.3672,
-      "step": 10399
-    },
-    {
-      "epoch": 0.6796941376380629,
-      "grad_norm": 0.4554852843284607,
-      "learning_rate": 8.943409931994427e-06,
-      "loss": 0.3846,
-      "step": 10400
-    },
-    {
-      "epoch": 0.679759492843605,
-      "grad_norm": 0.4254249036312103,
-      "learning_rate": 8.943195239883377e-06,
-      "loss": 0.3456,
-      "step": 10401
-    },
-    {
-      "epoch": 0.6798248480491471,
-      "grad_norm": 0.4523789584636688,
-      "learning_rate": 8.94298052854e-06,
-      "loss": 0.3877,
-      "step": 10402
-    },
-    {
-      "epoch": 0.6798902032546892,
-      "grad_norm": 0.4845171570777893,
-      "learning_rate": 8.94276579796534e-06,
-      "loss": 0.4727,
-      "step": 10403
-    },
-    {
-      "epoch": 0.6799555584602314,
-      "grad_norm": 0.46508994698524475,
-      "learning_rate": 8.942551048160444e-06,
-      "loss": 0.3713,
-      "step": 10404
-    },
-    {
-      "epoch": 0.6800209136657734,
-      "grad_norm": 0.4565008580684662,
-      "learning_rate": 8.942336279126363e-06,
-      "loss": 0.38,
-      "step": 10405
-    },
-    {
-      "epoch": 0.6800862688713156,
-      "grad_norm": 0.4528326392173767,
-      "learning_rate": 8.942121490864139e-06,
-      "loss": 0.3936,
-      "step": 10406
-    },
-    {
-      "epoch": 0.6801516240768577,
-      "grad_norm": 0.4814760386943817,
-      "learning_rate": 8.941906683374826e-06,
-      "loss": 0.4181,
-      "step": 10407
-    },
-    {
-      "epoch": 0.6802169792823999,
-      "grad_norm": 0.4270453155040741,
-      "learning_rate": 8.941691856659466e-06,
-      "loss": 0.3847,
-      "step": 10408
-    },
-    {
-      "epoch": 0.680282334487942,
-      "grad_norm": 0.46762701869010925,
-      "learning_rate": 8.94147701071911e-06,
-      "loss": 0.387,
-      "step": 10409
-    },
-    {
-      "epoch": 0.6803476896934841,
-      "grad_norm": 0.4895259141921997,
-      "learning_rate": 8.941262145554807e-06,
-      "loss": 0.327,
-      "step": 10410
-    },
-    {
-      "epoch": 0.6804130448990262,
-      "grad_norm": 0.45179054141044617,
-      "learning_rate": 8.941047261167601e-06,
-      "loss": 0.3695,
-      "step": 10411
-    },
-    {
-      "epoch": 0.6804784001045683,
-      "grad_norm": 0.4389207661151886,
-      "learning_rate": 8.940832357558543e-06,
-      "loss": 0.3548,
-      "step": 10412
-    },
-    {
-      "epoch": 0.6805437553101105,
-      "grad_norm": 0.44351649284362793,
-      "learning_rate": 8.94061743472868e-06,
-      "loss": 0.3724,
-      "step": 10413
-    },
-    {
-      "epoch": 0.6806091105156525,
-      "grad_norm": 0.4505099654197693,
-      "learning_rate": 8.94040249267906e-06,
-      "loss": 0.3614,
-      "step": 10414
-    },
-    {
-      "epoch": 0.6806744657211947,
-      "grad_norm": 0.4455447494983673,
-      "learning_rate": 8.940187531410735e-06,
-      "loss": 0.3888,
-      "step": 10415
-    },
-    {
-      "epoch": 0.6807398209267368,
-      "grad_norm": 0.437529981136322,
-      "learning_rate": 8.939972550924746e-06,
-      "loss": 0.3555,
-      "step": 10416
-    },
-    {
-      "epoch": 0.680805176132279,
-      "grad_norm": 0.43726852536201477,
-      "learning_rate": 8.93975755122215e-06,
-      "loss": 0.3418,
-      "step": 10417
-    },
-    {
-      "epoch": 0.680870531337821,
-      "grad_norm": 0.45562511682510376,
-      "learning_rate": 8.93954253230399e-06,
-      "loss": 0.4094,
-      "step": 10418
-    },
-    {
-      "epoch": 0.6809358865433632,
-      "grad_norm": 0.42359310388565063,
-      "learning_rate": 8.939327494171315e-06,
-      "loss": 0.4072,
-      "step": 10419
-    },
-    {
-      "epoch": 0.6810012417489053,
-      "grad_norm": 0.46157050132751465,
-      "learning_rate": 8.939112436825177e-06,
-      "loss": 0.4001,
-      "step": 10420
-    },
-    {
-      "epoch": 0.6810665969544474,
-      "grad_norm": 0.5074307322502136,
-      "learning_rate": 8.938897360266621e-06,
-      "loss": 0.4772,
-      "step": 10421
-    },
-    {
-      "epoch": 0.6811319521599896,
-      "grad_norm": 0.44071605801582336,
-      "learning_rate": 8.938682264496699e-06,
-      "loss": 0.3764,
-      "step": 10422
-    },
-    {
-      "epoch": 0.6811973073655316,
-      "grad_norm": 0.6126153469085693,
-      "learning_rate": 8.938467149516459e-06,
-      "loss": 0.339,
-      "step": 10423
-    },
-    {
-      "epoch": 0.6812626625710738,
-      "grad_norm": 0.44011783599853516,
-      "learning_rate": 8.938252015326952e-06,
-      "loss": 0.3559,
-      "step": 10424
-    },
-    {
-      "epoch": 0.6813280177766159,
-      "grad_norm": 0.4675701856613159,
-      "learning_rate": 8.938036861929223e-06,
-      "loss": 0.3983,
-      "step": 10425
-    },
-    {
-      "epoch": 0.6813933729821581,
-      "grad_norm": 0.42904967069625854,
-      "learning_rate": 8.937821689324325e-06,
-      "loss": 0.3597,
-      "step": 10426
-    },
-    {
-      "epoch": 0.6814587281877001,
-      "grad_norm": 0.4446243643760681,
-      "learning_rate": 8.937606497513308e-06,
-      "loss": 0.3716,
-      "step": 10427
-    },
-    {
-      "epoch": 0.6815240833932422,
-      "grad_norm": 0.4483039081096649,
-      "learning_rate": 8.937391286497216e-06,
-      "loss": 0.3966,
-      "step": 10428
-    },
-    {
-      "epoch": 0.6815894385987844,
-      "grad_norm": 0.46698901057243347,
-      "learning_rate": 8.937176056277105e-06,
-      "loss": 0.4258,
-      "step": 10429
-    },
-    {
-      "epoch": 0.6816547938043265,
-      "grad_norm": 0.47981882095336914,
-      "learning_rate": 8.936960806854024e-06,
-      "loss": 0.3886,
-      "step": 10430
-    },
-    {
-      "epoch": 0.6817201490098687,
-      "grad_norm": 0.4871785044670105,
-      "learning_rate": 8.93674553822902e-06,
-      "loss": 0.4052,
-      "step": 10431
-    },
-    {
-      "epoch": 0.6817855042154107,
-      "grad_norm": 0.46038419008255005,
-      "learning_rate": 8.936530250403143e-06,
-      "loss": 0.392,
-      "step": 10432
-    },
-    {
-      "epoch": 0.6818508594209529,
-      "grad_norm": 0.47744181752204895,
-      "learning_rate": 8.936314943377447e-06,
-      "loss": 0.37,
-      "step": 10433
-    },
-    {
-      "epoch": 0.681916214626495,
-      "grad_norm": 0.4498889446258545,
-      "learning_rate": 8.936099617152977e-06,
-      "loss": 0.4048,
-      "step": 10434
-    },
-    {
-      "epoch": 0.6819815698320372,
-      "grad_norm": 0.4357393682003021,
-      "learning_rate": 8.935884271730787e-06,
-      "loss": 0.3686,
-      "step": 10435
-    },
-    {
-      "epoch": 0.6820469250375792,
-      "grad_norm": 0.44549882411956787,
-      "learning_rate": 8.935668907111923e-06,
-      "loss": 0.3881,
-      "step": 10436
-    },
-    {
-      "epoch": 0.6821122802431213,
-      "grad_norm": 0.44550469517707825,
-      "learning_rate": 8.935453523297442e-06,
-      "loss": 0.3674,
-      "step": 10437
-    },
-    {
-      "epoch": 0.6821776354486635,
-      "grad_norm": 0.4659557044506073,
-      "learning_rate": 8.935238120288388e-06,
-      "loss": 0.4223,
-      "step": 10438
-    },
-    {
-      "epoch": 0.6822429906542056,
-      "grad_norm": 0.443538099527359,
-      "learning_rate": 8.935022698085815e-06,
-      "loss": 0.3767,
-      "step": 10439
-    },
-    {
-      "epoch": 0.6823083458597478,
-      "grad_norm": 0.4619375169277191,
-      "learning_rate": 8.934807256690774e-06,
-      "loss": 0.402,
-      "step": 10440
-    },
-    {
-      "epoch": 0.6823737010652898,
-      "grad_norm": 0.4323079288005829,
-      "learning_rate": 8.934591796104315e-06,
-      "loss": 0.3486,
-      "step": 10441
-    },
-    {
-      "epoch": 0.682439056270832,
-      "grad_norm": 0.4612639844417572,
-      "learning_rate": 8.934376316327486e-06,
-      "loss": 0.4015,
-      "step": 10442
-    },
-    {
-      "epoch": 0.6825044114763741,
-      "grad_norm": 0.45452800393104553,
-      "learning_rate": 8.934160817361345e-06,
-      "loss": 0.4516,
-      "step": 10443
-    },
-    {
-      "epoch": 0.6825697666819163,
-      "grad_norm": 0.4550628960132599,
-      "learning_rate": 8.933945299206937e-06,
-      "loss": 0.4084,
-      "step": 10444
-    },
-    {
-      "epoch": 0.6826351218874583,
-      "grad_norm": 0.43161723017692566,
-      "learning_rate": 8.933729761865312e-06,
-      "loss": 0.3897,
-      "step": 10445
-    },
-    {
-      "epoch": 0.6827004770930004,
-      "grad_norm": 0.49983176589012146,
-      "learning_rate": 8.933514205337527e-06,
-      "loss": 0.3792,
-      "step": 10446
-    },
-    {
-      "epoch": 0.6827658322985426,
-      "grad_norm": 0.4445848762989044,
-      "learning_rate": 8.933298629624632e-06,
-      "loss": 0.3764,
-      "step": 10447
-    },
-    {
-      "epoch": 0.6828311875040847,
-      "grad_norm": 0.43100154399871826,
-      "learning_rate": 8.933083034727674e-06,
-      "loss": 0.3877,
-      "step": 10448
-    },
-    {
-      "epoch": 0.6828965427096269,
-      "grad_norm": 0.442611426115036,
-      "learning_rate": 8.932867420647709e-06,
-      "loss": 0.359,
-      "step": 10449
-    },
-    {
-      "epoch": 0.6829618979151689,
-      "grad_norm": 0.45714277029037476,
-      "learning_rate": 8.932651787385786e-06,
-      "loss": 0.3883,
-      "step": 10450
-    },
-    {
-      "epoch": 0.6830272531207111,
-      "grad_norm": 0.427101194858551,
-      "learning_rate": 8.93243613494296e-06,
-      "loss": 0.3387,
-      "step": 10451
-    },
-    {
-      "epoch": 0.6830926083262532,
-      "grad_norm": 0.4358592927455902,
-      "learning_rate": 8.932220463320278e-06,
-      "loss": 0.3739,
-      "step": 10452
-    },
-    {
-      "epoch": 0.6831579635317953,
-      "grad_norm": 0.4458143413066864,
-      "learning_rate": 8.932004772518796e-06,
-      "loss": 0.4092,
-      "step": 10453
-    },
-    {
-      "epoch": 0.6832233187373374,
-      "grad_norm": 0.47435426712036133,
-      "learning_rate": 8.931789062539566e-06,
-      "loss": 0.4197,
-      "step": 10454
-    },
-    {
-      "epoch": 0.6832886739428795,
-      "grad_norm": 0.4375659227371216,
-      "learning_rate": 8.931573333383636e-06,
-      "loss": 0.4025,
-      "step": 10455
-    },
-    {
-      "epoch": 0.6833540291484217,
-      "grad_norm": 0.42265719175338745,
-      "learning_rate": 8.931357585052063e-06,
-      "loss": 0.3746,
-      "step": 10456
-    },
-    {
-      "epoch": 0.6834193843539638,
-      "grad_norm": 0.44514164328575134,
-      "learning_rate": 8.931141817545896e-06,
-      "loss": 0.3928,
-      "step": 10457
-    },
-    {
-      "epoch": 0.683484739559506,
-      "grad_norm": 0.4276603162288666,
-      "learning_rate": 8.930926030866188e-06,
-      "loss": 0.3629,
-      "step": 10458
-    },
-    {
-      "epoch": 0.683550094765048,
-      "grad_norm": 0.4545811712741852,
-      "learning_rate": 8.930710225013992e-06,
-      "loss": 0.3797,
-      "step": 10459
-    },
-    {
-      "epoch": 0.6836154499705902,
-      "grad_norm": 0.4180017113685608,
-      "learning_rate": 8.930494399990361e-06,
-      "loss": 0.3488,
-      "step": 10460
-    },
-    {
-      "epoch": 0.6836808051761323,
-      "grad_norm": 0.43427959084510803,
-      "learning_rate": 8.930278555796347e-06,
-      "loss": 0.3388,
-      "step": 10461
-    },
-    {
-      "epoch": 0.6837461603816744,
-      "grad_norm": 0.4505934715270996,
-      "learning_rate": 8.930062692433004e-06,
-      "loss": 0.3623,
-      "step": 10462
-    },
-    {
-      "epoch": 0.6838115155872165,
-      "grad_norm": 0.4257461726665497,
-      "learning_rate": 8.92984680990138e-06,
-      "loss": 0.3729,
-      "step": 10463
-    },
-    {
-      "epoch": 0.6838768707927586,
-      "grad_norm": 0.457512766122818,
-      "learning_rate": 8.929630908202535e-06,
-      "loss": 0.3984,
-      "step": 10464
-    },
-    {
-      "epoch": 0.6839422259983008,
-      "grad_norm": 0.45155709981918335,
-      "learning_rate": 8.929414987337519e-06,
-      "loss": 0.3871,
-      "step": 10465
-    },
-    {
-      "epoch": 0.6840075812038429,
-      "grad_norm": 0.45669347047805786,
-      "learning_rate": 8.929199047307384e-06,
-      "loss": 0.3828,
-      "step": 10466
-    },
-    {
-      "epoch": 0.684072936409385,
-      "grad_norm": 0.4452977180480957,
-      "learning_rate": 8.928983088113184e-06,
-      "loss": 0.3853,
-      "step": 10467
-    },
-    {
-      "epoch": 0.6841382916149271,
-      "grad_norm": 0.46452596783638,
-      "learning_rate": 8.928767109755971e-06,
-      "loss": 0.4117,
-      "step": 10468
-    },
-    {
-      "epoch": 0.6842036468204693,
-      "grad_norm": 0.4296901524066925,
-      "learning_rate": 8.928551112236803e-06,
-      "loss": 0.3625,
-      "step": 10469
-    },
-    {
-      "epoch": 0.6842690020260114,
-      "grad_norm": 0.4388759732246399,
-      "learning_rate": 8.928335095556727e-06,
-      "loss": 0.3551,
-      "step": 10470
-    },
-    {
-      "epoch": 0.6843343572315534,
-      "grad_norm": 0.44955843687057495,
-      "learning_rate": 8.928119059716802e-06,
-      "loss": 0.3911,
-      "step": 10471
-    },
-    {
-      "epoch": 0.6843997124370956,
-      "grad_norm": 0.47844985127449036,
-      "learning_rate": 8.927903004718078e-06,
-      "loss": 0.4492,
-      "step": 10472
-    },
-    {
-      "epoch": 0.6844650676426377,
-      "grad_norm": 0.4134272336959839,
-      "learning_rate": 8.927686930561612e-06,
-      "loss": 0.3117,
-      "step": 10473
-    },
-    {
-      "epoch": 0.6845304228481799,
-      "grad_norm": 0.43954187631607056,
-      "learning_rate": 8.927470837248455e-06,
-      "loss": 0.3459,
-      "step": 10474
-    },
-    {
-      "epoch": 0.684595778053722,
-      "grad_norm": 0.5005001425743103,
-      "learning_rate": 8.927254724779661e-06,
-      "loss": 0.4935,
-      "step": 10475
-    },
-    {
-      "epoch": 0.6846611332592641,
-      "grad_norm": 0.42396217584609985,
-      "learning_rate": 8.927038593156287e-06,
-      "loss": 0.3381,
-      "step": 10476
-    },
-    {
-      "epoch": 0.6847264884648062,
-      "grad_norm": 0.46023210883140564,
-      "learning_rate": 8.926822442379383e-06,
-      "loss": 0.3611,
-      "step": 10477
-    },
-    {
-      "epoch": 0.6847918436703483,
-      "grad_norm": 0.4212297201156616,
-      "learning_rate": 8.92660627245001e-06,
-      "loss": 0.3165,
-      "step": 10478
-    },
-    {
-      "epoch": 0.6848571988758905,
-      "grad_norm": 0.6105642318725586,
-      "learning_rate": 8.926390083369214e-06,
-      "loss": 0.4126,
-      "step": 10479
-    },
-    {
-      "epoch": 0.6849225540814325,
-      "grad_norm": 0.38947272300720215,
-      "learning_rate": 8.926173875138053e-06,
-      "loss": 0.3039,
-      "step": 10480
-    },
-    {
-      "epoch": 0.6849879092869747,
-      "grad_norm": 0.4735604524612427,
-      "learning_rate": 8.925957647757584e-06,
-      "loss": 0.4305,
-      "step": 10481
-    },
-    {
-      "epoch": 0.6850532644925168,
-      "grad_norm": 0.4116586148738861,
-      "learning_rate": 8.92574140122886e-06,
-      "loss": 0.3562,
-      "step": 10482
-    },
-    {
-      "epoch": 0.685118619698059,
-      "grad_norm": 0.4315769374370575,
-      "learning_rate": 8.925525135552932e-06,
-      "loss": 0.373,
-      "step": 10483
-    },
-    {
-      "epoch": 0.6851839749036011,
-      "grad_norm": 0.4615154266357422,
-      "learning_rate": 8.925308850730862e-06,
-      "loss": 0.404,
-      "step": 10484
-    },
-    {
-      "epoch": 0.6852493301091432,
-      "grad_norm": 0.4887750446796417,
-      "learning_rate": 8.925092546763698e-06,
-      "loss": 0.3828,
-      "step": 10485
-    },
-    {
-      "epoch": 0.6853146853146853,
-      "grad_norm": 0.4732937812805176,
-      "learning_rate": 8.9248762236525e-06,
-      "loss": 0.4293,
-      "step": 10486
-    },
-    {
-      "epoch": 0.6853800405202274,
-      "grad_norm": 0.44955384731292725,
-      "learning_rate": 8.924659881398318e-06,
-      "loss": 0.4109,
-      "step": 10487
-    },
-    {
-      "epoch": 0.6854453957257696,
-      "grad_norm": 0.5000100135803223,
-      "learning_rate": 8.924443520002213e-06,
-      "loss": 0.4199,
-      "step": 10488
-    },
-    {
-      "epoch": 0.6855107509313116,
-      "grad_norm": 0.4586191475391388,
-      "learning_rate": 8.924227139465236e-06,
-      "loss": 0.4165,
-      "step": 10489
-    },
-    {
-      "epoch": 0.6855761061368538,
-      "grad_norm": 0.43072134256362915,
-      "learning_rate": 8.924010739788444e-06,
-      "loss": 0.3507,
-      "step": 10490
-    },
-    {
-      "epoch": 0.6856414613423959,
-      "grad_norm": 0.4799150824546814,
-      "learning_rate": 8.923794320972892e-06,
-      "loss": 0.4124,
-      "step": 10491
-    },
-    {
-      "epoch": 0.6857068165479381,
-      "grad_norm": 0.40612852573394775,
-      "learning_rate": 8.923577883019636e-06,
-      "loss": 0.3159,
-      "step": 10492
-    },
-    {
-      "epoch": 0.6857721717534802,
-      "grad_norm": 0.4481818675994873,
-      "learning_rate": 8.923361425929731e-06,
-      "loss": 0.3713,
-      "step": 10493
-    },
-    {
-      "epoch": 0.6858375269590223,
-      "grad_norm": 0.43309223651885986,
-      "learning_rate": 8.923144949704233e-06,
-      "loss": 0.3704,
-      "step": 10494
-    },
-    {
-      "epoch": 0.6859028821645644,
-      "grad_norm": 0.4391562342643738,
-      "learning_rate": 8.9229284543442e-06,
-      "loss": 0.375,
-      "step": 10495
-    },
-    {
-      "epoch": 0.6859682373701065,
-      "grad_norm": 0.4453137218952179,
-      "learning_rate": 8.922711939850684e-06,
-      "loss": 0.3949,
-      "step": 10496
-    },
-    {
-      "epoch": 0.6860335925756487,
-      "grad_norm": 0.4381144642829895,
-      "learning_rate": 8.922495406224743e-06,
-      "loss": 0.3782,
-      "step": 10497
-    },
-    {
-      "epoch": 0.6860989477811907,
-      "grad_norm": 0.45431220531463623,
-      "learning_rate": 8.922278853467432e-06,
-      "loss": 0.3895,
-      "step": 10498
-    },
-    {
-      "epoch": 0.6861643029867329,
-      "grad_norm": 0.4399212598800659,
-      "learning_rate": 8.922062281579811e-06,
-      "loss": 0.3682,
-      "step": 10499
-    },
-    {
-      "epoch": 0.686229658192275,
-      "grad_norm": 0.4298236072063446,
-      "learning_rate": 8.92184569056293e-06,
-      "loss": 0.3811,
-      "step": 10500
-    },
-    {
-      "epoch": 0.6862950133978172,
-      "grad_norm": 0.4866264760494232,
-      "learning_rate": 8.921629080417852e-06,
-      "loss": 0.4035,
-      "step": 10501
-    },
-    {
-      "epoch": 0.6863603686033592,
-      "grad_norm": 0.4533084034919739,
-      "learning_rate": 8.92141245114563e-06,
-      "loss": 0.3838,
-      "step": 10502
-    },
-    {
-      "epoch": 0.6864257238089014,
-      "grad_norm": 0.4726482331752777,
-      "learning_rate": 8.92119580274732e-06,
-      "loss": 0.4461,
-      "step": 10503
-    },
-    {
-      "epoch": 0.6864910790144435,
-      "grad_norm": 0.4625649154186249,
-      "learning_rate": 8.92097913522398e-06,
-      "loss": 0.3746,
-      "step": 10504
-    },
-    {
-      "epoch": 0.6865564342199856,
-      "grad_norm": 0.4844793975353241,
-      "learning_rate": 8.920762448576665e-06,
-      "loss": 0.414,
-      "step": 10505
-    },
-    {
-      "epoch": 0.6866217894255278,
-      "grad_norm": 0.4288451075553894,
-      "learning_rate": 8.920545742806436e-06,
-      "loss": 0.3974,
-      "step": 10506
-    },
-    {
-      "epoch": 0.6866871446310698,
-      "grad_norm": 0.43219995498657227,
-      "learning_rate": 8.920329017914345e-06,
-      "loss": 0.3614,
-      "step": 10507
-    },
-    {
-      "epoch": 0.686752499836612,
-      "grad_norm": 0.46927735209465027,
-      "learning_rate": 8.920112273901452e-06,
-      "loss": 0.3791,
-      "step": 10508
-    },
-    {
-      "epoch": 0.6868178550421541,
-      "grad_norm": 0.48191893100738525,
-      "learning_rate": 8.919895510768814e-06,
-      "loss": 0.426,
-      "step": 10509
-    },
-    {
-      "epoch": 0.6868832102476963,
-      "grad_norm": 0.4131309390068054,
-      "learning_rate": 8.919678728517487e-06,
-      "loss": 0.3243,
-      "step": 10510
-    },
-    {
-      "epoch": 0.6869485654532383,
-      "grad_norm": 0.4218220114707947,
-      "learning_rate": 8.91946192714853e-06,
-      "loss": 0.2949,
-      "step": 10511
-    },
-    {
-      "epoch": 0.6870139206587804,
-      "grad_norm": 0.4162497818470001,
-      "learning_rate": 8.919245106662997e-06,
-      "loss": 0.3278,
-      "step": 10512
-    },
-    {
-      "epoch": 0.6870792758643226,
-      "grad_norm": 0.425112247467041,
-      "learning_rate": 8.919028267061948e-06,
-      "loss": 0.3554,
-      "step": 10513
-    },
-    {
-      "epoch": 0.6871446310698647,
-      "grad_norm": 0.4841451942920685,
-      "learning_rate": 8.918811408346442e-06,
-      "loss": 0.4142,
-      "step": 10514
-    },
-    {
-      "epoch": 0.6872099862754069,
-      "grad_norm": 0.47394266724586487,
-      "learning_rate": 8.918594530517536e-06,
-      "loss": 0.4179,
-      "step": 10515
-    },
-    {
-      "epoch": 0.6872753414809489,
-      "grad_norm": 0.4626839756965637,
-      "learning_rate": 8.918377633576285e-06,
-      "loss": 0.4002,
-      "step": 10516
-    },
-    {
-      "epoch": 0.6873406966864911,
-      "grad_norm": 0.42641302943229675,
-      "learning_rate": 8.91816071752375e-06,
-      "loss": 0.3356,
-      "step": 10517
-    },
-    {
-      "epoch": 0.6874060518920332,
-      "grad_norm": 0.4462212920188904,
-      "learning_rate": 8.917943782360986e-06,
-      "loss": 0.3899,
-      "step": 10518
-    },
-    {
-      "epoch": 0.6874714070975754,
-      "grad_norm": 0.46592387557029724,
-      "learning_rate": 8.917726828089054e-06,
-      "loss": 0.3808,
-      "step": 10519
-    },
-    {
-      "epoch": 0.6875367623031174,
-      "grad_norm": 0.46447572112083435,
-      "learning_rate": 8.917509854709012e-06,
-      "loss": 0.4059,
-      "step": 10520
-    },
-    {
-      "epoch": 0.6876021175086595,
-      "grad_norm": 0.4563732445240021,
-      "learning_rate": 8.917292862221918e-06,
-      "loss": 0.365,
-      "step": 10521
-    },
-    {
-      "epoch": 0.6876674727142017,
-      "grad_norm": 0.4624747931957245,
-      "learning_rate": 8.917075850628827e-06,
-      "loss": 0.3966,
-      "step": 10522
-    },
-    {
-      "epoch": 0.6877328279197438,
-      "grad_norm": 0.48354753851890564,
-      "learning_rate": 8.916858819930801e-06,
-      "loss": 0.3962,
-      "step": 10523
-    },
-    {
-      "epoch": 0.687798183125286,
-      "grad_norm": 0.43341320753097534,
-      "learning_rate": 8.916641770128899e-06,
-      "loss": 0.3535,
-      "step": 10524
-    },
-    {
-      "epoch": 0.687863538330828,
-      "grad_norm": 0.3981284201145172,
-      "learning_rate": 8.916424701224176e-06,
-      "loss": 0.3221,
-      "step": 10525
-    },
-    {
-      "epoch": 0.6879288935363702,
-      "grad_norm": 0.45291438698768616,
-      "learning_rate": 8.916207613217695e-06,
-      "loss": 0.3689,
-      "step": 10526
-    },
-    {
-      "epoch": 0.6879942487419123,
-      "grad_norm": 0.4812171161174774,
-      "learning_rate": 8.91599050611051e-06,
-      "loss": 0.4138,
-      "step": 10527
-    },
-    {
-      "epoch": 0.6880596039474545,
-      "grad_norm": 0.42696821689605713,
-      "learning_rate": 8.915773379903685e-06,
-      "loss": 0.3286,
-      "step": 10528
-    },
-    {
-      "epoch": 0.6881249591529965,
-      "grad_norm": 0.46794548630714417,
-      "learning_rate": 8.915556234598276e-06,
-      "loss": 0.4165,
-      "step": 10529
-    },
-    {
-      "epoch": 0.6881903143585386,
-      "grad_norm": 0.46507856249809265,
-      "learning_rate": 8.915339070195344e-06,
-      "loss": 0.3388,
-      "step": 10530
-    },
-    {
-      "epoch": 0.6882556695640808,
-      "grad_norm": 0.434635728597641,
-      "learning_rate": 8.915121886695946e-06,
-      "loss": 0.385,
-      "step": 10531
-    },
-    {
-      "epoch": 0.6883210247696229,
-      "grad_norm": 0.4649295508861542,
-      "learning_rate": 8.914904684101143e-06,
-      "loss": 0.4223,
-      "step": 10532
-    },
-    {
-      "epoch": 0.688386379975165,
-      "grad_norm": 0.4520520567893982,
-      "learning_rate": 8.91468746241199e-06,
-      "loss": 0.3779,
-      "step": 10533
-    },
-    {
-      "epoch": 0.6884517351807071,
-      "grad_norm": 0.463011234998703,
-      "learning_rate": 8.914470221629554e-06,
-      "loss": 0.3743,
-      "step": 10534
-    },
-    {
-      "epoch": 0.6885170903862493,
-      "grad_norm": 0.4743647873401642,
-      "learning_rate": 8.91425296175489e-06,
-      "loss": 0.4088,
-      "step": 10535
-    },
-    {
-      "epoch": 0.6885824455917914,
-      "grad_norm": 0.42367467284202576,
-      "learning_rate": 8.914035682789058e-06,
-      "loss": 0.3493,
-      "step": 10536
-    },
-    {
-      "epoch": 0.6886478007973335,
-      "grad_norm": 0.48695218563079834,
-      "learning_rate": 8.913818384733117e-06,
-      "loss": 0.4029,
-      "step": 10537
-    },
-    {
-      "epoch": 0.6887131560028756,
-      "grad_norm": 0.4481523334980011,
-      "learning_rate": 8.913601067588128e-06,
-      "loss": 0.3826,
-      "step": 10538
-    },
-    {
-      "epoch": 0.6887785112084177,
-      "grad_norm": 0.48517176508903503,
-      "learning_rate": 8.913383731355152e-06,
-      "loss": 0.402,
-      "step": 10539
-    },
-    {
-      "epoch": 0.6888438664139599,
-      "grad_norm": 0.4523976445198059,
-      "learning_rate": 8.913166376035247e-06,
-      "loss": 0.403,
-      "step": 10540
-    },
-    {
-      "epoch": 0.688909221619502,
-      "grad_norm": 0.44856175780296326,
-      "learning_rate": 8.912949001629474e-06,
-      "loss": 0.3922,
-      "step": 10541
-    },
-    {
-      "epoch": 0.6889745768250441,
-      "grad_norm": 0.44000181555747986,
-      "learning_rate": 8.912731608138894e-06,
-      "loss": 0.3479,
-      "step": 10542
-    },
-    {
-      "epoch": 0.6890399320305862,
-      "grad_norm": 0.4488789439201355,
-      "learning_rate": 8.912514195564566e-06,
-      "loss": 0.3213,
-      "step": 10543
-    },
-    {
-      "epoch": 0.6891052872361284,
-      "grad_norm": 0.4425378739833832,
-      "learning_rate": 8.912296763907548e-06,
-      "loss": 0.3203,
-      "step": 10544
-    },
-    {
-      "epoch": 0.6891706424416705,
-      "grad_norm": 0.46558132767677307,
-      "learning_rate": 8.912079313168907e-06,
-      "loss": 0.4087,
-      "step": 10545
-    },
-    {
-      "epoch": 0.6892359976472126,
-      "grad_norm": 0.437294602394104,
-      "learning_rate": 8.9118618433497e-06,
-      "loss": 0.3789,
-      "step": 10546
-    },
-    {
-      "epoch": 0.6893013528527547,
-      "grad_norm": 0.4735149145126343,
-      "learning_rate": 8.911644354450986e-06,
-      "loss": 0.366,
-      "step": 10547
-    },
-    {
-      "epoch": 0.6893667080582968,
-      "grad_norm": 0.48468002676963806,
-      "learning_rate": 8.911426846473828e-06,
-      "loss": 0.4307,
-      "step": 10548
-    },
-    {
-      "epoch": 0.689432063263839,
-      "grad_norm": 0.425825297832489,
-      "learning_rate": 8.911209319419285e-06,
-      "loss": 0.3594,
-      "step": 10549
-    },
-    {
-      "epoch": 0.6894974184693811,
-      "grad_norm": 0.45564308762550354,
-      "learning_rate": 8.91099177328842e-06,
-      "loss": 0.3853,
-      "step": 10550
-    },
-    {
-      "epoch": 0.6895627736749232,
-      "grad_norm": 0.44014772772789,
-      "learning_rate": 8.910774208082293e-06,
-      "loss": 0.3767,
-      "step": 10551
-    },
-    {
-      "epoch": 0.6896281288804653,
-      "grad_norm": 0.47790876030921936,
-      "learning_rate": 8.910556623801966e-06,
-      "loss": 0.4091,
-      "step": 10552
-    },
-    {
-      "epoch": 0.6896934840860075,
-      "grad_norm": 0.4382498562335968,
-      "learning_rate": 8.9103390204485e-06,
-      "loss": 0.3735,
-      "step": 10553
-    },
-    {
-      "epoch": 0.6897588392915496,
-      "grad_norm": 0.4551611542701721,
-      "learning_rate": 8.910121398022956e-06,
-      "loss": 0.4214,
-      "step": 10554
-    },
-    {
-      "epoch": 0.6898241944970916,
-      "grad_norm": 0.4938529133796692,
-      "learning_rate": 8.909903756526393e-06,
-      "loss": 0.4592,
-      "step": 10555
-    },
-    {
-      "epoch": 0.6898895497026338,
-      "grad_norm": 0.44927355647087097,
-      "learning_rate": 8.909686095959877e-06,
-      "loss": 0.3901,
-      "step": 10556
-    },
-    {
-      "epoch": 0.6899549049081759,
-      "grad_norm": 0.47456619143486023,
-      "learning_rate": 8.909468416324467e-06,
-      "loss": 0.4062,
-      "step": 10557
-    },
-    {
-      "epoch": 0.6900202601137181,
-      "grad_norm": 0.44300612807273865,
-      "learning_rate": 8.909250717621225e-06,
-      "loss": 0.4017,
-      "step": 10558
-    },
-    {
-      "epoch": 0.6900856153192602,
-      "grad_norm": 0.45416632294654846,
-      "learning_rate": 8.909032999851213e-06,
-      "loss": 0.3796,
-      "step": 10559
-    },
-    {
-      "epoch": 0.6901509705248023,
-      "grad_norm": 0.4438626766204834,
-      "learning_rate": 8.908815263015492e-06,
-      "loss": 0.4059,
-      "step": 10560
-    },
-    {
-      "epoch": 0.6902163257303444,
-      "grad_norm": 0.41670164465904236,
-      "learning_rate": 8.908597507115127e-06,
-      "loss": 0.3566,
-      "step": 10561
-    },
-    {
-      "epoch": 0.6902816809358865,
-      "grad_norm": 0.47049680352211,
-      "learning_rate": 8.908379732151175e-06,
-      "loss": 0.3999,
-      "step": 10562
-    },
-    {
-      "epoch": 0.6903470361414287,
-      "grad_norm": 0.4610985219478607,
-      "learning_rate": 8.908161938124704e-06,
-      "loss": 0.3784,
-      "step": 10563
-    },
-    {
-      "epoch": 0.6904123913469707,
-      "grad_norm": 0.4252164363861084,
-      "learning_rate": 8.907944125036771e-06,
-      "loss": 0.3729,
-      "step": 10564
-    },
-    {
-      "epoch": 0.6904777465525129,
-      "grad_norm": 0.4724777340888977,
-      "learning_rate": 8.90772629288844e-06,
-      "loss": 0.3858,
-      "step": 10565
-    },
-    {
-      "epoch": 0.690543101758055,
-      "grad_norm": 0.44275957345962524,
-      "learning_rate": 8.907508441680778e-06,
-      "loss": 0.3927,
-      "step": 10566
-    },
-    {
-      "epoch": 0.6906084569635972,
-      "grad_norm": 0.4292038679122925,
-      "learning_rate": 8.907290571414841e-06,
-      "loss": 0.3836,
-      "step": 10567
-    },
-    {
-      "epoch": 0.6906738121691393,
-      "grad_norm": 0.4941498637199402,
-      "learning_rate": 8.907072682091695e-06,
-      "loss": 0.3412,
-      "step": 10568
-    },
-    {
-      "epoch": 0.6907391673746814,
-      "grad_norm": 0.4277919828891754,
-      "learning_rate": 8.906854773712402e-06,
-      "loss": 0.3665,
-      "step": 10569
-    },
-    {
-      "epoch": 0.6908045225802235,
-      "grad_norm": 0.4419439733028412,
-      "learning_rate": 8.906636846278023e-06,
-      "loss": 0.4015,
-      "step": 10570
-    },
-    {
-      "epoch": 0.6908698777857656,
-      "grad_norm": 0.39750900864601135,
-      "learning_rate": 8.906418899789624e-06,
-      "loss": 0.3012,
-      "step": 10571
-    },
-    {
-      "epoch": 0.6909352329913078,
-      "grad_norm": 0.5598298907279968,
-      "learning_rate": 8.906200934248267e-06,
-      "loss": 0.4767,
-      "step": 10572
-    },
-    {
-      "epoch": 0.6910005881968498,
-      "grad_norm": 0.4465446472167969,
-      "learning_rate": 8.905982949655014e-06,
-      "loss": 0.3541,
-      "step": 10573
-    },
-    {
-      "epoch": 0.691065943402392,
-      "grad_norm": 0.4126950204372406,
-      "learning_rate": 8.905764946010931e-06,
-      "loss": 0.3185,
-      "step": 10574
-    },
-    {
-      "epoch": 0.6911312986079341,
-      "grad_norm": 0.4510418772697449,
-      "learning_rate": 8.905546923317077e-06,
-      "loss": 0.3961,
-      "step": 10575
-    },
-    {
-      "epoch": 0.6911966538134763,
-      "grad_norm": 0.43746069073677063,
-      "learning_rate": 8.90532888157452e-06,
-      "loss": 0.404,
-      "step": 10576
-    },
-    {
-      "epoch": 0.6912620090190184,
-      "grad_norm": 0.49543002247810364,
-      "learning_rate": 8.905110820784319e-06,
-      "loss": 0.4365,
-      "step": 10577
-    },
-    {
-      "epoch": 0.6913273642245605,
-      "grad_norm": 0.48675885796546936,
-      "learning_rate": 8.904892740947539e-06,
-      "loss": 0.4306,
-      "step": 10578
-    },
-    {
-      "epoch": 0.6913927194301026,
-      "grad_norm": 0.4443257749080658,
-      "learning_rate": 8.904674642065247e-06,
-      "loss": 0.3956,
-      "step": 10579
-    },
-    {
-      "epoch": 0.6914580746356447,
-      "grad_norm": 0.4747105836868286,
-      "learning_rate": 8.904456524138503e-06,
-      "loss": 0.3717,
-      "step": 10580
-    },
-    {
-      "epoch": 0.6915234298411869,
-      "grad_norm": 0.48398444056510925,
-      "learning_rate": 8.904238387168371e-06,
-      "loss": 0.4199,
-      "step": 10581
-    },
-    {
-      "epoch": 0.6915887850467289,
-      "grad_norm": 0.46877458691596985,
-      "learning_rate": 8.904020231155916e-06,
-      "loss": 0.4041,
-      "step": 10582
-    },
-    {
-      "epoch": 0.6916541402522711,
-      "grad_norm": 0.4279765188694,
-      "learning_rate": 8.903802056102202e-06,
-      "loss": 0.3479,
-      "step": 10583
-    },
-    {
-      "epoch": 0.6917194954578132,
-      "grad_norm": 0.4481217563152313,
-      "learning_rate": 8.903583862008294e-06,
-      "loss": 0.3472,
-      "step": 10584
-    },
-    {
-      "epoch": 0.6917848506633554,
-      "grad_norm": 0.43060222268104553,
-      "learning_rate": 8.903365648875254e-06,
-      "loss": 0.3549,
-      "step": 10585
-    },
-    {
-      "epoch": 0.6918502058688974,
-      "grad_norm": 0.44700443744659424,
-      "learning_rate": 8.903147416704148e-06,
-      "loss": 0.4061,
-      "step": 10586
-    },
-    {
-      "epoch": 0.6919155610744396,
-      "grad_norm": 0.4763336777687073,
-      "learning_rate": 8.90292916549604e-06,
-      "loss": 0.4159,
-      "step": 10587
-    },
-    {
-      "epoch": 0.6919809162799817,
-      "grad_norm": 0.44735389947891235,
-      "learning_rate": 8.902710895251993e-06,
-      "loss": 0.383,
-      "step": 10588
-    },
-    {
-      "epoch": 0.6920462714855238,
-      "grad_norm": 0.4600410759449005,
-      "learning_rate": 8.902492605973074e-06,
-      "loss": 0.4224,
-      "step": 10589
-    },
-    {
-      "epoch": 0.692111626691066,
-      "grad_norm": 0.4634614586830139,
-      "learning_rate": 8.902274297660347e-06,
-      "loss": 0.3581,
-      "step": 10590
-    },
-    {
-      "epoch": 0.692176981896608,
-      "grad_norm": 0.4893110990524292,
-      "learning_rate": 8.902055970314875e-06,
-      "loss": 0.4458,
-      "step": 10591
-    },
-    {
-      "epoch": 0.6922423371021502,
-      "grad_norm": 0.4451541304588318,
-      "learning_rate": 8.901837623937726e-06,
-      "loss": 0.3626,
-      "step": 10592
-    },
-    {
-      "epoch": 0.6923076923076923,
-      "grad_norm": 0.45439523458480835,
-      "learning_rate": 8.901619258529963e-06,
-      "loss": 0.3615,
-      "step": 10593
-    },
-    {
-      "epoch": 0.6923730475132345,
-      "grad_norm": 0.4473332464694977,
-      "learning_rate": 8.90140087409265e-06,
-      "loss": 0.3609,
-      "step": 10594
-    },
-    {
-      "epoch": 0.6924384027187765,
-      "grad_norm": 0.4544123113155365,
-      "learning_rate": 8.901182470626855e-06,
-      "loss": 0.4165,
-      "step": 10595
-    },
-    {
-      "epoch": 0.6925037579243186,
-      "grad_norm": 0.4130313992500305,
-      "learning_rate": 8.90096404813364e-06,
-      "loss": 0.3646,
-      "step": 10596
-    },
-    {
-      "epoch": 0.6925691131298608,
-      "grad_norm": 0.4376696050167084,
-      "learning_rate": 8.900745606614072e-06,
-      "loss": 0.3931,
-      "step": 10597
-    },
-    {
-      "epoch": 0.6926344683354029,
-      "grad_norm": 0.4156436324119568,
-      "learning_rate": 8.900527146069218e-06,
-      "loss": 0.3595,
-      "step": 10598
-    },
-    {
-      "epoch": 0.6926998235409451,
-      "grad_norm": 0.4557853937149048,
-      "learning_rate": 8.90030866650014e-06,
-      "loss": 0.418,
-      "step": 10599
-    },
-    {
-      "epoch": 0.6927651787464871,
-      "grad_norm": 0.4538869559764862,
-      "learning_rate": 8.900090167907906e-06,
-      "loss": 0.3923,
-      "step": 10600
-    },
-    {
-      "epoch": 0.6928305339520293,
-      "grad_norm": 0.43243837356567383,
-      "learning_rate": 8.89987165029358e-06,
-      "loss": 0.369,
-      "step": 10601
-    },
-    {
-      "epoch": 0.6928958891575714,
-      "grad_norm": 0.4296075701713562,
-      "learning_rate": 8.899653113658232e-06,
-      "loss": 0.3524,
-      "step": 10602
-    },
-    {
-      "epoch": 0.6929612443631136,
-      "grad_norm": 0.4318280518054962,
-      "learning_rate": 8.899434558002923e-06,
-      "loss": 0.3608,
-      "step": 10603
-    },
-    {
-      "epoch": 0.6930265995686556,
-      "grad_norm": 0.46237117052078247,
-      "learning_rate": 8.899215983328721e-06,
-      "loss": 0.3822,
-      "step": 10604
-    },
-    {
-      "epoch": 0.6930919547741977,
-      "grad_norm": 0.46481773257255554,
-      "learning_rate": 8.898997389636691e-06,
-      "loss": 0.3839,
-      "step": 10605
-    },
-    {
-      "epoch": 0.6931573099797399,
-      "grad_norm": 0.4405609965324402,
-      "learning_rate": 8.898778776927901e-06,
-      "loss": 0.3792,
-      "step": 10606
-    },
-    {
-      "epoch": 0.693222665185282,
-      "grad_norm": 0.4056107997894287,
-      "learning_rate": 8.898560145203416e-06,
-      "loss": 0.3491,
-      "step": 10607
-    },
-    {
-      "epoch": 0.6932880203908242,
-      "grad_norm": 0.46750712394714355,
-      "learning_rate": 8.898341494464302e-06,
-      "loss": 0.4107,
-      "step": 10608
-    },
-    {
-      "epoch": 0.6933533755963662,
-      "grad_norm": 0.4560029208660126,
-      "learning_rate": 8.898122824711626e-06,
-      "loss": 0.3831,
-      "step": 10609
-    },
-    {
-      "epoch": 0.6934187308019084,
-      "grad_norm": 0.4878546893596649,
-      "learning_rate": 8.897904135946456e-06,
-      "loss": 0.445,
-      "step": 10610
-    },
-    {
-      "epoch": 0.6934840860074505,
-      "grad_norm": 0.4243931174278259,
-      "learning_rate": 8.897685428169856e-06,
-      "loss": 0.3499,
-      "step": 10611
-    },
-    {
-      "epoch": 0.6935494412129927,
-      "grad_norm": 0.4623061716556549,
-      "learning_rate": 8.897466701382894e-06,
-      "loss": 0.3608,
-      "step": 10612
-    },
-    {
-      "epoch": 0.6936147964185347,
-      "grad_norm": 0.4468550682067871,
-      "learning_rate": 8.897247955586637e-06,
-      "loss": 0.3618,
-      "step": 10613
-    },
-    {
-      "epoch": 0.6936801516240768,
-      "grad_norm": 0.49978509545326233,
-      "learning_rate": 8.89702919078215e-06,
-      "loss": 0.4836,
-      "step": 10614
-    },
-    {
-      "epoch": 0.693745506829619,
-      "grad_norm": 0.4840392768383026,
-      "learning_rate": 8.896810406970503e-06,
-      "loss": 0.4928,
-      "step": 10615
-    },
-    {
-      "epoch": 0.6938108620351611,
-      "grad_norm": 0.46464452147483826,
-      "learning_rate": 8.89659160415276e-06,
-      "loss": 0.3697,
-      "step": 10616
-    },
-    {
-      "epoch": 0.6938762172407033,
-      "grad_norm": 0.4350418448448181,
-      "learning_rate": 8.896372782329993e-06,
-      "loss": 0.3573,
-      "step": 10617
-    },
-    {
-      "epoch": 0.6939415724462453,
-      "grad_norm": 0.423578679561615,
-      "learning_rate": 8.896153941503265e-06,
-      "loss": 0.3606,
-      "step": 10618
-    },
-    {
-      "epoch": 0.6940069276517875,
-      "grad_norm": 0.407560259103775,
-      "learning_rate": 8.895935081673644e-06,
-      "loss": 0.3463,
-      "step": 10619
-    },
-    {
-      "epoch": 0.6940722828573296,
-      "grad_norm": 0.4485580325126648,
-      "learning_rate": 8.895716202842198e-06,
-      "loss": 0.3953,
-      "step": 10620
-    },
-    {
-      "epoch": 0.6941376380628717,
-      "grad_norm": 0.49691659212112427,
-      "learning_rate": 8.895497305009993e-06,
-      "loss": 0.3891,
-      "step": 10621
-    },
-    {
-      "epoch": 0.6942029932684138,
-      "grad_norm": 0.43800756335258484,
-      "learning_rate": 8.895278388178099e-06,
-      "loss": 0.4062,
-      "step": 10622
-    },
-    {
-      "epoch": 0.6942683484739559,
-      "grad_norm": 0.4745366871356964,
-      "learning_rate": 8.895059452347583e-06,
-      "loss": 0.4419,
-      "step": 10623
-    },
-    {
-      "epoch": 0.6943337036794981,
-      "grad_norm": 0.46037518978118896,
-      "learning_rate": 8.894840497519514e-06,
-      "loss": 0.3688,
-      "step": 10624
-    },
-    {
-      "epoch": 0.6943990588850402,
-      "grad_norm": 0.45265913009643555,
-      "learning_rate": 8.894621523694957e-06,
-      "loss": 0.4068,
-      "step": 10625
-    },
-    {
-      "epoch": 0.6944644140905823,
-      "grad_norm": 0.44792941212654114,
-      "learning_rate": 8.894402530874982e-06,
-      "loss": 0.3871,
-      "step": 10626
-    },
-    {
-      "epoch": 0.6945297692961244,
-      "grad_norm": 0.4481916129589081,
-      "learning_rate": 8.894183519060657e-06,
-      "loss": 0.3744,
-      "step": 10627
-    },
-    {
-      "epoch": 0.6945951245016666,
-      "grad_norm": 0.47108253836631775,
-      "learning_rate": 8.89396448825305e-06,
-      "loss": 0.4096,
-      "step": 10628
-    },
-    {
-      "epoch": 0.6946604797072087,
-      "grad_norm": 0.47146904468536377,
-      "learning_rate": 8.89374543845323e-06,
-      "loss": 0.3721,
-      "step": 10629
-    },
-    {
-      "epoch": 0.6947258349127508,
-      "grad_norm": 0.44914835691452026,
-      "learning_rate": 8.893526369662263e-06,
-      "loss": 0.3933,
-      "step": 10630
-    },
-    {
-      "epoch": 0.6947911901182929,
-      "grad_norm": 0.4471861720085144,
-      "learning_rate": 8.893307281881219e-06,
-      "loss": 0.3742,
-      "step": 10631
-    },
-    {
-      "epoch": 0.694856545323835,
-      "grad_norm": 0.4269276261329651,
-      "learning_rate": 8.893088175111167e-06,
-      "loss": 0.3649,
-      "step": 10632
-    },
-    {
-      "epoch": 0.6949219005293772,
-      "grad_norm": 0.45189711451530457,
-      "learning_rate": 8.892869049353175e-06,
-      "loss": 0.3628,
-      "step": 10633
-    },
-    {
-      "epoch": 0.6949872557349193,
-      "grad_norm": 0.45149579644203186,
-      "learning_rate": 8.892649904608312e-06,
-      "loss": 0.3919,
-      "step": 10634
-    },
-    {
-      "epoch": 0.6950526109404614,
-      "grad_norm": 0.4281890392303467,
-      "learning_rate": 8.892430740877649e-06,
-      "loss": 0.3673,
-      "step": 10635
-    },
-    {
-      "epoch": 0.6951179661460035,
-      "grad_norm": 0.42137566208839417,
-      "learning_rate": 8.89221155816225e-06,
-      "loss": 0.3714,
-      "step": 10636
-    },
-    {
-      "epoch": 0.6951833213515457,
-      "grad_norm": 0.43112459778785706,
-      "learning_rate": 8.891992356463188e-06,
-      "loss": 0.3598,
-      "step": 10637
-    },
-    {
-      "epoch": 0.6952486765570878,
-      "grad_norm": 0.4220387041568756,
-      "learning_rate": 8.891773135781533e-06,
-      "loss": 0.3343,
-      "step": 10638
-    },
-    {
-      "epoch": 0.6953140317626298,
-      "grad_norm": 0.4726293385028839,
-      "learning_rate": 8.891553896118348e-06,
-      "loss": 0.416,
-      "step": 10639
-    },
-    {
-      "epoch": 0.695379386968172,
-      "grad_norm": 0.4971306025981903,
-      "learning_rate": 8.89133463747471e-06,
-      "loss": 0.4577,
-      "step": 10640
-    },
-    {
-      "epoch": 0.6954447421737141,
-      "grad_norm": 0.4709372818470001,
-      "learning_rate": 8.891115359851683e-06,
-      "loss": 0.4255,
-      "step": 10641
-    },
-    {
-      "epoch": 0.6955100973792563,
-      "grad_norm": 0.4117278456687927,
-      "learning_rate": 8.890896063250338e-06,
-      "loss": 0.3245,
-      "step": 10642
-    },
-    {
-      "epoch": 0.6955754525847984,
-      "grad_norm": 0.4369412064552307,
-      "learning_rate": 8.890676747671746e-06,
-      "loss": 0.3713,
-      "step": 10643
-    },
-    {
-      "epoch": 0.6956408077903405,
-      "grad_norm": 0.4706905484199524,
-      "learning_rate": 8.890457413116976e-06,
-      "loss": 0.3993,
-      "step": 10644
-    },
-    {
-      "epoch": 0.6957061629958826,
-      "grad_norm": 0.4339410662651062,
-      "learning_rate": 8.890238059587096e-06,
-      "loss": 0.3546,
-      "step": 10645
-    },
-    {
-      "epoch": 0.6957715182014248,
-      "grad_norm": 0.44695916771888733,
-      "learning_rate": 8.890018687083178e-06,
-      "loss": 0.3152,
-      "step": 10646
-    },
-    {
-      "epoch": 0.6958368734069669,
-      "grad_norm": 0.4529878497123718,
-      "learning_rate": 8.889799295606292e-06,
-      "loss": 0.3884,
-      "step": 10647
-    },
-    {
-      "epoch": 0.6959022286125089,
-      "grad_norm": 0.4456702470779419,
-      "learning_rate": 8.889579885157507e-06,
-      "loss": 0.3849,
-      "step": 10648
-    },
-    {
-      "epoch": 0.6959675838180511,
-      "grad_norm": 0.4578702449798584,
-      "learning_rate": 8.889360455737892e-06,
-      "loss": 0.3821,
-      "step": 10649
-    },
-    {
-      "epoch": 0.6960329390235932,
-      "grad_norm": 0.44034576416015625,
-      "learning_rate": 8.88914100734852e-06,
-      "loss": 0.3437,
-      "step": 10650
-    },
-    {
-      "epoch": 0.6960982942291354,
-      "grad_norm": 0.4882797598838806,
-      "learning_rate": 8.888921539990461e-06,
-      "loss": 0.3948,
-      "step": 10651
-    },
-    {
-      "epoch": 0.6961636494346775,
-      "grad_norm": 0.44824862480163574,
-      "learning_rate": 8.888702053664781e-06,
-      "loss": 0.3996,
-      "step": 10652
-    },
-    {
-      "epoch": 0.6962290046402196,
-      "grad_norm": 0.46059808135032654,
-      "learning_rate": 8.888482548372556e-06,
-      "loss": 0.3877,
-      "step": 10653
-    },
-    {
-      "epoch": 0.6962943598457617,
-      "grad_norm": 0.41871094703674316,
-      "learning_rate": 8.888263024114857e-06,
-      "loss": 0.3669,
-      "step": 10654
-    },
-    {
-      "epoch": 0.6963597150513038,
-      "grad_norm": 0.4514888823032379,
-      "learning_rate": 8.888043480892749e-06,
-      "loss": 0.3934,
-      "step": 10655
-    },
-    {
-      "epoch": 0.696425070256846,
-      "grad_norm": 0.45794615149497986,
-      "learning_rate": 8.887823918707306e-06,
-      "loss": 0.3726,
-      "step": 10656
-    },
-    {
-      "epoch": 0.696490425462388,
-      "grad_norm": 0.4819653332233429,
-      "learning_rate": 8.8876043375596e-06,
-      "loss": 0.4225,
-      "step": 10657
-    },
-    {
-      "epoch": 0.6965557806679302,
-      "grad_norm": 0.4350774586200714,
-      "learning_rate": 8.887384737450701e-06,
-      "loss": 0.3375,
-      "step": 10658
-    },
-    {
-      "epoch": 0.6966211358734723,
-      "grad_norm": 0.5108208060264587,
-      "learning_rate": 8.887165118381678e-06,
-      "loss": 0.459,
-      "step": 10659
-    },
-    {
-      "epoch": 0.6966864910790145,
-      "grad_norm": 0.4641295075416565,
-      "learning_rate": 8.886945480353605e-06,
-      "loss": 0.4061,
-      "step": 10660
-    },
-    {
-      "epoch": 0.6967518462845566,
-      "grad_norm": 0.44970735907554626,
-      "learning_rate": 8.886725823367554e-06,
-      "loss": 0.3566,
-      "step": 10661
-    },
-    {
-      "epoch": 0.6968172014900987,
-      "grad_norm": 0.4390512704849243,
-      "learning_rate": 8.886506147424594e-06,
-      "loss": 0.3557,
-      "step": 10662
-    },
-    {
-      "epoch": 0.6968825566956408,
-      "grad_norm": 0.4361428916454315,
-      "learning_rate": 8.886286452525797e-06,
-      "loss": 0.3534,
-      "step": 10663
-    },
-    {
-      "epoch": 0.6969479119011829,
-      "grad_norm": 0.44793328642845154,
-      "learning_rate": 8.886066738672234e-06,
-      "loss": 0.3942,
-      "step": 10664
-    },
-    {
-      "epoch": 0.6970132671067251,
-      "grad_norm": 0.4622913599014282,
-      "learning_rate": 8.885847005864975e-06,
-      "loss": 0.4141,
-      "step": 10665
-    },
-    {
-      "epoch": 0.6970786223122671,
-      "grad_norm": 0.4830019772052765,
-      "learning_rate": 8.885627254105097e-06,
-      "loss": 0.4287,
-      "step": 10666
-    },
-    {
-      "epoch": 0.6971439775178093,
-      "grad_norm": 0.4581737816333771,
-      "learning_rate": 8.885407483393669e-06,
-      "loss": 0.3737,
-      "step": 10667
-    },
-    {
-      "epoch": 0.6972093327233514,
-      "grad_norm": 0.43531617522239685,
-      "learning_rate": 8.885187693731763e-06,
-      "loss": 0.3911,
-      "step": 10668
-    },
-    {
-      "epoch": 0.6972746879288936,
-      "grad_norm": 0.4650973081588745,
-      "learning_rate": 8.884967885120448e-06,
-      "loss": 0.388,
-      "step": 10669
-    },
-    {
-      "epoch": 0.6973400431344356,
-      "grad_norm": 1.0223784446716309,
-      "learning_rate": 8.884748057560801e-06,
-      "loss": 0.4184,
-      "step": 10670
-    },
-    {
-      "epoch": 0.6974053983399778,
-      "grad_norm": 0.47753241658210754,
-      "learning_rate": 8.884528211053891e-06,
-      "loss": 0.4055,
-      "step": 10671
-    },
-    {
-      "epoch": 0.6974707535455199,
-      "grad_norm": 0.4437274634838104,
-      "learning_rate": 8.884308345600792e-06,
-      "loss": 0.3933,
-      "step": 10672
-    },
-    {
-      "epoch": 0.697536108751062,
-      "grad_norm": 0.46149584650993347,
-      "learning_rate": 8.884088461202574e-06,
-      "loss": 0.3898,
-      "step": 10673
-    },
-    {
-      "epoch": 0.6976014639566042,
-      "grad_norm": 0.4530940353870392,
-      "learning_rate": 8.883868557860313e-06,
-      "loss": 0.4197,
-      "step": 10674
-    },
-    {
-      "epoch": 0.6976668191621462,
-      "grad_norm": 0.4554557502269745,
-      "learning_rate": 8.883648635575077e-06,
-      "loss": 0.3742,
-      "step": 10675
-    },
-    {
-      "epoch": 0.6977321743676884,
-      "grad_norm": 0.4334377348423004,
-      "learning_rate": 8.883428694347944e-06,
-      "loss": 0.3337,
-      "step": 10676
-    },
-    {
-      "epoch": 0.6977975295732305,
-      "grad_norm": 0.4683746099472046,
-      "learning_rate": 8.883208734179981e-06,
-      "loss": 0.4032,
-      "step": 10677
-    },
-    {
-      "epoch": 0.6978628847787727,
-      "grad_norm": 0.38879600167274475,
-      "learning_rate": 8.882988755072266e-06,
-      "loss": 0.3101,
-      "step": 10678
-    },
-    {
-      "epoch": 0.6979282399843147,
-      "grad_norm": 0.5270335674285889,
-      "learning_rate": 8.88276875702587e-06,
-      "loss": 0.4585,
-      "step": 10679
-    },
-    {
-      "epoch": 0.6979935951898568,
-      "grad_norm": 0.4615952968597412,
-      "learning_rate": 8.882548740041862e-06,
-      "loss": 0.3749,
-      "step": 10680
-    },
-    {
-      "epoch": 0.698058950395399,
-      "grad_norm": 0.43616750836372375,
-      "learning_rate": 8.882328704121324e-06,
-      "loss": 0.3485,
-      "step": 10681
-    },
-    {
-      "epoch": 0.6981243056009411,
-      "grad_norm": 0.44347694516181946,
-      "learning_rate": 8.882108649265319e-06,
-      "loss": 0.3613,
-      "step": 10682
-    },
-    {
-      "epoch": 0.6981896608064833,
-      "grad_norm": 0.4720376133918762,
-      "learning_rate": 8.881888575474929e-06,
-      "loss": 0.4117,
-      "step": 10683
-    },
-    {
-      "epoch": 0.6982550160120253,
-      "grad_norm": 0.43622103333473206,
-      "learning_rate": 8.881668482751221e-06,
-      "loss": 0.3714,
-      "step": 10684
-    },
-    {
-      "epoch": 0.6983203712175675,
-      "grad_norm": 0.41271963715553284,
-      "learning_rate": 8.881448371095272e-06,
-      "loss": 0.3646,
-      "step": 10685
-    },
-    {
-      "epoch": 0.6983857264231096,
-      "grad_norm": 0.4676041603088379,
-      "learning_rate": 8.881228240508155e-06,
-      "loss": 0.4016,
-      "step": 10686
-    },
-    {
-      "epoch": 0.6984510816286518,
-      "grad_norm": 0.4412485361099243,
-      "learning_rate": 8.881008090990944e-06,
-      "loss": 0.3905,
-      "step": 10687
-    },
-    {
-      "epoch": 0.6985164368341938,
-      "grad_norm": 0.459614098072052,
-      "learning_rate": 8.88078792254471e-06,
-      "loss": 0.3864,
-      "step": 10688
-    },
-    {
-      "epoch": 0.6985817920397359,
-      "grad_norm": 0.41537952423095703,
-      "learning_rate": 8.880567735170531e-06,
-      "loss": 0.3406,
-      "step": 10689
-    },
-    {
-      "epoch": 0.6986471472452781,
-      "grad_norm": 0.4189286231994629,
-      "learning_rate": 8.880347528869477e-06,
-      "loss": 0.3227,
-      "step": 10690
-    },
-    {
-      "epoch": 0.6987125024508202,
-      "grad_norm": 0.4343949258327484,
-      "learning_rate": 8.880127303642625e-06,
-      "loss": 0.371,
-      "step": 10691
-    },
-    {
-      "epoch": 0.6987778576563624,
-      "grad_norm": 0.43783894181251526,
-      "learning_rate": 8.879907059491048e-06,
-      "loss": 0.3517,
-      "step": 10692
-    },
-    {
-      "epoch": 0.6988432128619044,
-      "grad_norm": 0.4129463732242584,
-      "learning_rate": 8.879686796415818e-06,
-      "loss": 0.3378,
-      "step": 10693
-    },
-    {
-      "epoch": 0.6989085680674466,
-      "grad_norm": 0.45865529775619507,
-      "learning_rate": 8.879466514418014e-06,
-      "loss": 0.3631,
-      "step": 10694
-    },
-    {
-      "epoch": 0.6989739232729887,
-      "grad_norm": 0.4360487163066864,
-      "learning_rate": 8.879246213498707e-06,
-      "loss": 0.3628,
-      "step": 10695
-    },
-    {
-      "epoch": 0.6990392784785309,
-      "grad_norm": 0.46427857875823975,
-      "learning_rate": 8.879025893658973e-06,
-      "loss": 0.3751,
-      "step": 10696
-    },
-    {
-      "epoch": 0.6991046336840729,
-      "grad_norm": 0.5423113703727722,
-      "learning_rate": 8.878805554899885e-06,
-      "loss": 0.4671,
-      "step": 10697
-    },
-    {
-      "epoch": 0.699169988889615,
-      "grad_norm": 0.47656580805778503,
-      "learning_rate": 8.878585197222519e-06,
-      "loss": 0.4071,
-      "step": 10698
-    },
-    {
-      "epoch": 0.6992353440951572,
-      "grad_norm": 0.41309502720832825,
-      "learning_rate": 8.878364820627948e-06,
-      "loss": 0.3586,
-      "step": 10699
-    },
-    {
-      "epoch": 0.6993006993006993,
-      "grad_norm": 0.43951088190078735,
-      "learning_rate": 8.87814442511725e-06,
-      "loss": 0.3331,
-      "step": 10700
-    },
-    {
-      "epoch": 0.6993660545062415,
-      "grad_norm": 0.45432737469673157,
-      "learning_rate": 8.877924010691496e-06,
-      "loss": 0.3692,
-      "step": 10701
-    },
-    {
-      "epoch": 0.6994314097117835,
-      "grad_norm": 0.4661157727241516,
-      "learning_rate": 8.877703577351766e-06,
-      "loss": 0.4104,
-      "step": 10702
-    },
-    {
-      "epoch": 0.6994967649173257,
-      "grad_norm": 0.4366321861743927,
-      "learning_rate": 8.87748312509913e-06,
-      "loss": 0.369,
-      "step": 10703
-    },
-    {
-      "epoch": 0.6995621201228678,
-      "grad_norm": 0.45107367634773254,
-      "learning_rate": 8.877262653934667e-06,
-      "loss": 0.3703,
-      "step": 10704
-    },
-    {
-      "epoch": 0.6996274753284099,
-      "grad_norm": 0.41421547532081604,
-      "learning_rate": 8.87704216385945e-06,
-      "loss": 0.3259,
-      "step": 10705
-    },
-    {
-      "epoch": 0.699692830533952,
-      "grad_norm": 0.5069754719734192,
-      "learning_rate": 8.876821654874555e-06,
-      "loss": 0.4366,
-      "step": 10706
-    },
-    {
-      "epoch": 0.6997581857394941,
-      "grad_norm": 0.46444544196128845,
-      "learning_rate": 8.876601126981059e-06,
-      "loss": 0.4085,
-      "step": 10707
-    },
-    {
-      "epoch": 0.6998235409450363,
-      "grad_norm": 0.44071781635284424,
-      "learning_rate": 8.876380580180034e-06,
-      "loss": 0.3671,
-      "step": 10708
-    },
-    {
-      "epoch": 0.6998888961505784,
-      "grad_norm": 0.4483624994754791,
-      "learning_rate": 8.87616001447256e-06,
-      "loss": 0.4174,
-      "step": 10709
-    },
-    {
-      "epoch": 0.6999542513561205,
-      "grad_norm": 0.4800783693790436,
-      "learning_rate": 8.87593942985971e-06,
-      "loss": 0.3825,
-      "step": 10710
-    },
-    {
-      "epoch": 0.7000196065616626,
-      "grad_norm": 0.4607797861099243,
-      "learning_rate": 8.875718826342561e-06,
-      "loss": 0.3604,
-      "step": 10711
-    },
-    {
-      "epoch": 0.7000849617672048,
-      "grad_norm": 0.4988437294960022,
-      "learning_rate": 8.875498203922189e-06,
-      "loss": 0.4535,
-      "step": 10712
-    },
-    {
-      "epoch": 0.7001503169727469,
-      "grad_norm": 0.49454623460769653,
-      "learning_rate": 8.875277562599668e-06,
-      "loss": 0.4754,
-      "step": 10713
-    },
-    {
-      "epoch": 0.700215672178289,
-      "grad_norm": 0.4355059266090393,
-      "learning_rate": 8.875056902376076e-06,
-      "loss": 0.398,
-      "step": 10714
-    },
-    {
-      "epoch": 0.7002810273838311,
-      "grad_norm": 0.4247625470161438,
-      "learning_rate": 8.874836223252491e-06,
-      "loss": 0.3552,
-      "step": 10715
-    },
-    {
-      "epoch": 0.7003463825893732,
-      "grad_norm": 0.45553550124168396,
-      "learning_rate": 8.874615525229986e-06,
-      "loss": 0.4185,
-      "step": 10716
-    },
-    {
-      "epoch": 0.7004117377949154,
-      "grad_norm": 0.4440927505493164,
-      "learning_rate": 8.874394808309638e-06,
-      "loss": 0.3783,
-      "step": 10717
-    },
-    {
-      "epoch": 0.7004770930004575,
-      "grad_norm": 0.47648704051971436,
-      "learning_rate": 8.874174072492528e-06,
-      "loss": 0.4023,
-      "step": 10718
-    },
-    {
-      "epoch": 0.7005424482059996,
-      "grad_norm": 0.4460158944129944,
-      "learning_rate": 8.873953317779724e-06,
-      "loss": 0.3869,
-      "step": 10719
-    },
-    {
-      "epoch": 0.7006078034115417,
-      "grad_norm": 0.4699411690235138,
-      "learning_rate": 8.87373254417231e-06,
-      "loss": 0.4209,
-      "step": 10720
-    },
-    {
-      "epoch": 0.7006731586170839,
-      "grad_norm": 0.5043027997016907,
-      "learning_rate": 8.873511751671361e-06,
-      "loss": 0.3493,
-      "step": 10721
-    },
-    {
-      "epoch": 0.700738513822626,
-      "grad_norm": 0.46910256147384644,
-      "learning_rate": 8.873290940277952e-06,
-      "loss": 0.4264,
-      "step": 10722
-    },
-    {
-      "epoch": 0.700803869028168,
-      "grad_norm": 0.43721339106559753,
-      "learning_rate": 8.873070109993162e-06,
-      "loss": 0.351,
-      "step": 10723
-    },
-    {
-      "epoch": 0.7008692242337102,
-      "grad_norm": 0.45217421650886536,
-      "learning_rate": 8.872849260818066e-06,
-      "loss": 0.3642,
-      "step": 10724
-    },
-    {
-      "epoch": 0.7009345794392523,
-      "grad_norm": 0.42875149846076965,
-      "learning_rate": 8.872628392753747e-06,
-      "loss": 0.3758,
-      "step": 10725
-    },
-    {
-      "epoch": 0.7009999346447945,
-      "grad_norm": 0.45782139897346497,
-      "learning_rate": 8.872407505801273e-06,
-      "loss": 0.4035,
-      "step": 10726
-    },
-    {
-      "epoch": 0.7010652898503366,
-      "grad_norm": 0.45796889066696167,
-      "learning_rate": 8.872186599961727e-06,
-      "loss": 0.3799,
-      "step": 10727
-    },
-    {
-      "epoch": 0.7011306450558787,
-      "grad_norm": 0.4108940660953522,
-      "learning_rate": 8.871965675236185e-06,
-      "loss": 0.3738,
-      "step": 10728
-    },
-    {
-      "epoch": 0.7011960002614208,
-      "grad_norm": 0.46935567259788513,
-      "learning_rate": 8.871744731625727e-06,
-      "loss": 0.3723,
-      "step": 10729
-    },
-    {
-      "epoch": 0.701261355466963,
-      "grad_norm": 0.44690999388694763,
-      "learning_rate": 8.871523769131426e-06,
-      "loss": 0.393,
-      "step": 10730
-    },
-    {
-      "epoch": 0.7013267106725051,
-      "grad_norm": 0.46299657225608826,
-      "learning_rate": 8.871302787754364e-06,
-      "loss": 0.3847,
-      "step": 10731
-    },
-    {
-      "epoch": 0.7013920658780471,
-      "grad_norm": 0.44792985916137695,
-      "learning_rate": 8.871081787495617e-06,
-      "loss": 0.361,
-      "step": 10732
-    },
-    {
-      "epoch": 0.7014574210835893,
-      "grad_norm": 0.4198303520679474,
-      "learning_rate": 8.870860768356264e-06,
-      "loss": 0.3334,
-      "step": 10733
-    },
-    {
-      "epoch": 0.7015227762891314,
-      "grad_norm": 0.45148947834968567,
-      "learning_rate": 8.87063973033738e-06,
-      "loss": 0.3715,
-      "step": 10734
-    },
-    {
-      "epoch": 0.7015881314946736,
-      "grad_norm": 0.44864290952682495,
-      "learning_rate": 8.870418673440048e-06,
-      "loss": 0.3615,
-      "step": 10735
-    },
-    {
-      "epoch": 0.7016534867002157,
-      "grad_norm": 0.4621596932411194,
-      "learning_rate": 8.870197597665342e-06,
-      "loss": 0.4117,
-      "step": 10736
-    },
-    {
-      "epoch": 0.7017188419057578,
-      "grad_norm": 0.49969062209129333,
-      "learning_rate": 8.86997650301434e-06,
-      "loss": 0.4514,
-      "step": 10737
-    },
-    {
-      "epoch": 0.7017841971112999,
-      "grad_norm": 0.4572676420211792,
-      "learning_rate": 8.869755389488122e-06,
-      "loss": 0.3911,
-      "step": 10738
-    },
-    {
-      "epoch": 0.701849552316842,
-      "grad_norm": 0.4414505362510681,
-      "learning_rate": 8.86953425708777e-06,
-      "loss": 0.4044,
-      "step": 10739
-    },
-    {
-      "epoch": 0.7019149075223842,
-      "grad_norm": 0.4709620177745819,
-      "learning_rate": 8.869313105814355e-06,
-      "loss": 0.4538,
-      "step": 10740
-    },
-    {
-      "epoch": 0.7019802627279262,
-      "grad_norm": 0.4614220857620239,
-      "learning_rate": 8.86909193566896e-06,
-      "loss": 0.4106,
-      "step": 10741
-    },
-    {
-      "epoch": 0.7020456179334684,
-      "grad_norm": 0.4344942569732666,
-      "learning_rate": 8.868870746652664e-06,
-      "loss": 0.3755,
-      "step": 10742
-    },
-    {
-      "epoch": 0.7021109731390105,
-      "grad_norm": 0.44991180300712585,
-      "learning_rate": 8.868649538766545e-06,
-      "loss": 0.3968,
-      "step": 10743
-    },
-    {
-      "epoch": 0.7021763283445527,
-      "grad_norm": 0.5085800886154175,
-      "learning_rate": 8.868428312011684e-06,
-      "loss": 0.4734,
-      "step": 10744
-    },
-    {
-      "epoch": 0.7022416835500948,
-      "grad_norm": 0.4519469439983368,
-      "learning_rate": 8.868207066389153e-06,
-      "loss": 0.4017,
-      "step": 10745
-    },
-    {
-      "epoch": 0.7023070387556369,
-      "grad_norm": 0.4577236771583557,
-      "learning_rate": 8.867985801900041e-06,
-      "loss": 0.3542,
-      "step": 10746
-    },
-    {
-      "epoch": 0.702372393961179,
-      "grad_norm": 0.47221171855926514,
-      "learning_rate": 8.86776451854542e-06,
-      "loss": 0.4036,
-      "step": 10747
-    },
-    {
-      "epoch": 0.7024377491667211,
-      "grad_norm": 0.4412024915218353,
-      "learning_rate": 8.86754321632637e-06,
-      "loss": 0.3753,
-      "step": 10748
-    },
-    {
-      "epoch": 0.7025031043722633,
-      "grad_norm": 0.43691593408584595,
-      "learning_rate": 8.867321895243974e-06,
-      "loss": 0.328,
-      "step": 10749
-    },
-    {
-      "epoch": 0.7025684595778053,
-      "grad_norm": 0.47266703844070435,
-      "learning_rate": 8.86710055529931e-06,
-      "loss": 0.4345,
-      "step": 10750
-    },
-    {
-      "epoch": 0.7026338147833475,
-      "grad_norm": 0.43213972449302673,
-      "learning_rate": 8.866879196493457e-06,
-      "loss": 0.3555,
-      "step": 10751
-    },
-    {
-      "epoch": 0.7026991699888896,
-      "grad_norm": 0.4508489668369293,
-      "learning_rate": 8.866657818827494e-06,
-      "loss": 0.4101,
-      "step": 10752
-    },
-    {
-      "epoch": 0.7027645251944318,
-      "grad_norm": 0.4328119456768036,
-      "learning_rate": 8.8664364223025e-06,
-      "loss": 0.3716,
-      "step": 10753
-    },
-    {
-      "epoch": 0.7028298803999738,
-      "grad_norm": 0.45387157797813416,
-      "learning_rate": 8.866215006919557e-06,
-      "loss": 0.3587,
-      "step": 10754
-    },
-    {
-      "epoch": 0.702895235605516,
-      "grad_norm": 0.4766331613063812,
-      "learning_rate": 8.865993572679743e-06,
-      "loss": 0.4045,
-      "step": 10755
-    },
-    {
-      "epoch": 0.7029605908110581,
-      "grad_norm": 0.4131259620189667,
-      "learning_rate": 8.865772119584141e-06,
-      "loss": 0.3334,
-      "step": 10756
-    },
-    {
-      "epoch": 0.7030259460166002,
-      "grad_norm": 0.41277629137039185,
-      "learning_rate": 8.865550647633828e-06,
-      "loss": 0.3157,
-      "step": 10757
-    },
-    {
-      "epoch": 0.7030913012221424,
-      "grad_norm": 0.4751417636871338,
-      "learning_rate": 8.865329156829886e-06,
-      "loss": 0.3785,
-      "step": 10758
-    },
-    {
-      "epoch": 0.7031566564276844,
-      "grad_norm": 0.469399094581604,
-      "learning_rate": 8.865107647173392e-06,
-      "loss": 0.4298,
-      "step": 10759
-    },
-    {
-      "epoch": 0.7032220116332266,
-      "grad_norm": 0.4965677261352539,
-      "learning_rate": 8.864886118665432e-06,
-      "loss": 0.3875,
-      "step": 10760
-    },
-    {
-      "epoch": 0.7032873668387687,
-      "grad_norm": 0.4735341966152191,
-      "learning_rate": 8.864664571307082e-06,
-      "loss": 0.3821,
-      "step": 10761
-    },
-    {
-      "epoch": 0.7033527220443109,
-      "grad_norm": 0.4623195230960846,
-      "learning_rate": 8.864443005099425e-06,
-      "loss": 0.3912,
-      "step": 10762
-    },
-    {
-      "epoch": 0.703418077249853,
-      "grad_norm": 0.46555569767951965,
-      "learning_rate": 8.86422142004354e-06,
-      "loss": 0.4143,
-      "step": 10763
-    },
-    {
-      "epoch": 0.703483432455395,
-      "grad_norm": 0.49839991331100464,
-      "learning_rate": 8.86399981614051e-06,
-      "loss": 0.4137,
-      "step": 10764
-    },
-    {
-      "epoch": 0.7035487876609372,
-      "grad_norm": 0.5141803622245789,
-      "learning_rate": 8.863778193391413e-06,
-      "loss": 0.4206,
-      "step": 10765
-    },
-    {
-      "epoch": 0.7036141428664793,
-      "grad_norm": 0.45348691940307617,
-      "learning_rate": 8.86355655179733e-06,
-      "loss": 0.3676,
-      "step": 10766
-    },
-    {
-      "epoch": 0.7036794980720215,
-      "grad_norm": 0.4878885746002197,
-      "learning_rate": 8.863334891359345e-06,
-      "loss": 0.4447,
-      "step": 10767
-    },
-    {
-      "epoch": 0.7037448532775635,
-      "grad_norm": 0.44455888867378235,
-      "learning_rate": 8.863113212078536e-06,
-      "loss": 0.3908,
-      "step": 10768
-    },
-    {
-      "epoch": 0.7038102084831057,
-      "grad_norm": 0.47392991185188293,
-      "learning_rate": 8.862891513955987e-06,
-      "loss": 0.3833,
-      "step": 10769
-    },
-    {
-      "epoch": 0.7038755636886478,
-      "grad_norm": 0.4671472907066345,
-      "learning_rate": 8.862669796992776e-06,
-      "loss": 0.377,
-      "step": 10770
-    },
-    {
-      "epoch": 0.70394091889419,
-      "grad_norm": 0.4605567455291748,
-      "learning_rate": 8.862448061189988e-06,
-      "loss": 0.4077,
-      "step": 10771
-    },
-    {
-      "epoch": 0.704006274099732,
-      "grad_norm": 0.426021009683609,
-      "learning_rate": 8.862226306548702e-06,
-      "loss": 0.336,
-      "step": 10772
-    },
-    {
-      "epoch": 0.7040716293052741,
-      "grad_norm": 0.47108277678489685,
-      "learning_rate": 8.86200453307e-06,
-      "loss": 0.4316,
-      "step": 10773
-    },
-    {
-      "epoch": 0.7041369845108163,
-      "grad_norm": 0.45935899019241333,
-      "learning_rate": 8.861782740754966e-06,
-      "loss": 0.3574,
-      "step": 10774
-    },
-    {
-      "epoch": 0.7042023397163584,
-      "grad_norm": 0.4728596806526184,
-      "learning_rate": 8.861560929604677e-06,
-      "loss": 0.3752,
-      "step": 10775
-    },
-    {
-      "epoch": 0.7042676949219006,
-      "grad_norm": 0.4555656313896179,
-      "learning_rate": 8.861339099620219e-06,
-      "loss": 0.3752,
-      "step": 10776
-    },
-    {
-      "epoch": 0.7043330501274426,
-      "grad_norm": 0.4627569913864136,
-      "learning_rate": 8.861117250802672e-06,
-      "loss": 0.435,
-      "step": 10777
-    },
-    {
-      "epoch": 0.7043984053329848,
-      "grad_norm": 0.44748517870903015,
-      "learning_rate": 8.860895383153119e-06,
-      "loss": 0.366,
-      "step": 10778
-    },
-    {
-      "epoch": 0.7044637605385269,
-      "grad_norm": 0.42327216267585754,
-      "learning_rate": 8.860673496672642e-06,
-      "loss": 0.3302,
-      "step": 10779
-    },
-    {
-      "epoch": 0.7045291157440691,
-      "grad_norm": 0.4513353109359741,
-      "learning_rate": 8.860451591362322e-06,
-      "loss": 0.4067,
-      "step": 10780
-    },
-    {
-      "epoch": 0.7045944709496111,
-      "grad_norm": 0.41613486409187317,
-      "learning_rate": 8.860229667223243e-06,
-      "loss": 0.3562,
-      "step": 10781
-    },
-    {
-      "epoch": 0.7046598261551532,
-      "grad_norm": 0.5639072060585022,
-      "learning_rate": 8.860007724256484e-06,
-      "loss": 0.4195,
-      "step": 10782
-    },
-    {
-      "epoch": 0.7047251813606954,
-      "grad_norm": 0.45426589250564575,
-      "learning_rate": 8.859785762463133e-06,
-      "loss": 0.3876,
-      "step": 10783
-    },
-    {
-      "epoch": 0.7047905365662375,
-      "grad_norm": 0.4279865324497223,
-      "learning_rate": 8.85956378184427e-06,
-      "loss": 0.3242,
-      "step": 10784
-    },
-    {
-      "epoch": 0.7048558917717797,
-      "grad_norm": 0.4605855345726013,
-      "learning_rate": 8.859341782400976e-06,
-      "loss": 0.3947,
-      "step": 10785
-    },
-    {
-      "epoch": 0.7049212469773217,
-      "grad_norm": 0.4505729675292969,
-      "learning_rate": 8.859119764134336e-06,
-      "loss": 0.4176,
-      "step": 10786
-    },
-    {
-      "epoch": 0.7049866021828639,
-      "grad_norm": 0.47500374913215637,
-      "learning_rate": 8.85889772704543e-06,
-      "loss": 0.4212,
-      "step": 10787
-    },
-    {
-      "epoch": 0.705051957388406,
-      "grad_norm": 0.41787174344062805,
-      "learning_rate": 8.858675671135345e-06,
-      "loss": 0.339,
-      "step": 10788
-    },
-    {
-      "epoch": 0.705117312593948,
-      "grad_norm": 0.4474586546421051,
-      "learning_rate": 8.858453596405161e-06,
-      "loss": 0.3584,
-      "step": 10789
-    },
-    {
-      "epoch": 0.7051826677994902,
-      "grad_norm": 0.44654515385627747,
-      "learning_rate": 8.858231502855964e-06,
-      "loss": 0.4001,
-      "step": 10790
-    },
-    {
-      "epoch": 0.7052480230050323,
-      "grad_norm": 0.45959019660949707,
-      "learning_rate": 8.858009390488833e-06,
-      "loss": 0.3826,
-      "step": 10791
-    },
-    {
-      "epoch": 0.7053133782105745,
-      "grad_norm": 0.47081857919692993,
-      "learning_rate": 8.857787259304854e-06,
-      "loss": 0.3235,
-      "step": 10792
-    },
-    {
-      "epoch": 0.7053787334161166,
-      "grad_norm": 0.4579770267009735,
-      "learning_rate": 8.857565109305112e-06,
-      "loss": 0.338,
-      "step": 10793
-    },
-    {
-      "epoch": 0.7054440886216587,
-      "grad_norm": 0.4581635594367981,
-      "learning_rate": 8.857342940490686e-06,
-      "loss": 0.3985,
-      "step": 10794
-    },
-    {
-      "epoch": 0.7055094438272008,
-      "grad_norm": 0.4913211166858673,
-      "learning_rate": 8.857120752862662e-06,
-      "loss": 0.4794,
-      "step": 10795
-    },
-    {
-      "epoch": 0.705574799032743,
-      "grad_norm": 0.4294610917568207,
-      "learning_rate": 8.856898546422126e-06,
-      "loss": 0.3445,
-      "step": 10796
-    },
-    {
-      "epoch": 0.7056401542382851,
-      "grad_norm": 0.459158331155777,
-      "learning_rate": 8.856676321170159e-06,
-      "loss": 0.3882,
-      "step": 10797
-    },
-    {
-      "epoch": 0.7057055094438272,
-      "grad_norm": 0.44121789932250977,
-      "learning_rate": 8.856454077107845e-06,
-      "loss": 0.3523,
-      "step": 10798
-    },
-    {
-      "epoch": 0.7057708646493693,
-      "grad_norm": 0.4617336392402649,
-      "learning_rate": 8.856231814236268e-06,
-      "loss": 0.4026,
-      "step": 10799
-    },
-    {
-      "epoch": 0.7058362198549114,
-      "grad_norm": 0.46041619777679443,
-      "learning_rate": 8.856009532556513e-06,
-      "loss": 0.3947,
-      "step": 10800
-    },
-    {
-      "epoch": 0.7059015750604536,
-      "grad_norm": 0.42607614398002625,
-      "learning_rate": 8.855787232069664e-06,
-      "loss": 0.373,
-      "step": 10801
-    },
-    {
-      "epoch": 0.7059669302659957,
-      "grad_norm": 0.42943528294563293,
-      "learning_rate": 8.855564912776806e-06,
-      "loss": 0.3641,
-      "step": 10802
-    },
-    {
-      "epoch": 0.7060322854715378,
-      "grad_norm": 0.479065477848053,
-      "learning_rate": 8.85534257467902e-06,
-      "loss": 0.4223,
-      "step": 10803
-    },
-    {
-      "epoch": 0.7060976406770799,
-      "grad_norm": 0.43014097213745117,
-      "learning_rate": 8.855120217777393e-06,
-      "loss": 0.3647,
-      "step": 10804
-    },
-    {
-      "epoch": 0.7061629958826221,
-      "grad_norm": 0.43675094842910767,
-      "learning_rate": 8.854897842073011e-06,
-      "loss": 0.3742,
-      "step": 10805
-    },
-    {
-      "epoch": 0.7062283510881642,
-      "grad_norm": 0.46332669258117676,
-      "learning_rate": 8.854675447566954e-06,
-      "loss": 0.3773,
-      "step": 10806
-    },
-    {
-      "epoch": 0.7062937062937062,
-      "grad_norm": 0.43585094809532166,
-      "learning_rate": 8.854453034260312e-06,
-      "loss": 0.3562,
-      "step": 10807
-    },
-    {
-      "epoch": 0.7063590614992484,
-      "grad_norm": 0.4546062648296356,
-      "learning_rate": 8.854230602154166e-06,
-      "loss": 0.4249,
-      "step": 10808
-    },
-    {
-      "epoch": 0.7064244167047905,
-      "grad_norm": 0.42176294326782227,
-      "learning_rate": 8.854008151249602e-06,
-      "loss": 0.3444,
-      "step": 10809
-    },
-    {
-      "epoch": 0.7064897719103327,
-      "grad_norm": 0.4762844443321228,
-      "learning_rate": 8.853785681547706e-06,
-      "loss": 0.4243,
-      "step": 10810
-    },
-    {
-      "epoch": 0.7065551271158748,
-      "grad_norm": 0.4549350142478943,
-      "learning_rate": 8.85356319304956e-06,
-      "loss": 0.3679,
-      "step": 10811
-    },
-    {
-      "epoch": 0.7066204823214169,
-      "grad_norm": 0.44891974329948425,
-      "learning_rate": 8.853340685756254e-06,
-      "loss": 0.4075,
-      "step": 10812
-    },
-    {
-      "epoch": 0.706685837526959,
-      "grad_norm": 0.45329487323760986,
-      "learning_rate": 8.85311815966887e-06,
-      "loss": 0.3636,
-      "step": 10813
-    },
-    {
-      "epoch": 0.7067511927325012,
-      "grad_norm": 0.44027382135391235,
-      "learning_rate": 8.852895614788493e-06,
-      "loss": 0.3424,
-      "step": 10814
-    },
-    {
-      "epoch": 0.7068165479380433,
-      "grad_norm": 0.4336678683757782,
-      "learning_rate": 8.85267305111621e-06,
-      "loss": 0.3408,
-      "step": 10815
-    },
-    {
-      "epoch": 0.7068819031435853,
-      "grad_norm": 0.4381355345249176,
-      "learning_rate": 8.852450468653105e-06,
-      "loss": 0.4034,
-      "step": 10816
-    },
-    {
-      "epoch": 0.7069472583491275,
-      "grad_norm": 0.4556027948856354,
-      "learning_rate": 8.852227867400265e-06,
-      "loss": 0.4002,
-      "step": 10817
-    },
-    {
-      "epoch": 0.7070126135546696,
-      "grad_norm": 0.4505718946456909,
-      "learning_rate": 8.852005247358775e-06,
-      "loss": 0.3931,
-      "step": 10818
-    },
-    {
-      "epoch": 0.7070779687602118,
-      "grad_norm": 0.4421040117740631,
-      "learning_rate": 8.851782608529721e-06,
-      "loss": 0.3873,
-      "step": 10819
-    },
-    {
-      "epoch": 0.7071433239657539,
-      "grad_norm": 0.4468221068382263,
-      "learning_rate": 8.851559950914189e-06,
-      "loss": 0.4217,
-      "step": 10820
-    },
-    {
-      "epoch": 0.707208679171296,
-      "grad_norm": 0.4986788332462311,
-      "learning_rate": 8.851337274513265e-06,
-      "loss": 0.3857,
-      "step": 10821
-    },
-    {
-      "epoch": 0.7072740343768381,
-      "grad_norm": 0.4252099096775055,
-      "learning_rate": 8.851114579328034e-06,
-      "loss": 0.3546,
-      "step": 10822
-    },
-    {
-      "epoch": 0.7073393895823802,
-      "grad_norm": 0.4833270311355591,
-      "learning_rate": 8.850891865359583e-06,
-      "loss": 0.4502,
-      "step": 10823
-    },
-    {
-      "epoch": 0.7074047447879224,
-      "grad_norm": 0.42057788372039795,
-      "learning_rate": 8.850669132609e-06,
-      "loss": 0.3267,
-      "step": 10824
-    },
-    {
-      "epoch": 0.7074700999934644,
-      "grad_norm": 0.45054686069488525,
-      "learning_rate": 8.850446381077368e-06,
-      "loss": 0.4101,
-      "step": 10825
-    },
-    {
-      "epoch": 0.7075354551990066,
-      "grad_norm": 0.4439249634742737,
-      "learning_rate": 8.850223610765777e-06,
-      "loss": 0.3795,
-      "step": 10826
-    },
-    {
-      "epoch": 0.7076008104045487,
-      "grad_norm": 0.45448774099349976,
-      "learning_rate": 8.850000821675308e-06,
-      "loss": 0.3883,
-      "step": 10827
-    },
-    {
-      "epoch": 0.7076661656100909,
-      "grad_norm": 0.45995283126831055,
-      "learning_rate": 8.849778013807053e-06,
-      "loss": 0.396,
-      "step": 10828
-    },
-    {
-      "epoch": 0.707731520815633,
-      "grad_norm": 0.4646743834018707,
-      "learning_rate": 8.849555187162099e-06,
-      "loss": 0.4227,
-      "step": 10829
-    },
-    {
-      "epoch": 0.7077968760211751,
-      "grad_norm": 0.4693724811077118,
-      "learning_rate": 8.849332341741529e-06,
-      "loss": 0.406,
-      "step": 10830
-    },
-    {
-      "epoch": 0.7078622312267172,
-      "grad_norm": 0.4972003102302551,
-      "learning_rate": 8.849109477546431e-06,
-      "loss": 0.4384,
-      "step": 10831
-    },
-    {
-      "epoch": 0.7079275864322593,
-      "grad_norm": 0.44735318422317505,
-      "learning_rate": 8.848886594577894e-06,
-      "loss": 0.4097,
-      "step": 10832
-    },
-    {
-      "epoch": 0.7079929416378015,
-      "grad_norm": 0.41161873936653137,
-      "learning_rate": 8.848663692837002e-06,
-      "loss": 0.3219,
-      "step": 10833
-    },
-    {
-      "epoch": 0.7080582968433435,
-      "grad_norm": 0.4374428391456604,
-      "learning_rate": 8.848440772324846e-06,
-      "loss": 0.3727,
-      "step": 10834
-    },
-    {
-      "epoch": 0.7081236520488857,
-      "grad_norm": 0.45904046297073364,
-      "learning_rate": 8.84821783304251e-06,
-      "loss": 0.3954,
-      "step": 10835
-    },
-    {
-      "epoch": 0.7081890072544278,
-      "grad_norm": 0.4530632793903351,
-      "learning_rate": 8.847994874991082e-06,
-      "loss": 0.3746,
-      "step": 10836
-    },
-    {
-      "epoch": 0.70825436245997,
-      "grad_norm": 0.4787178635597229,
-      "learning_rate": 8.847771898171651e-06,
-      "loss": 0.3831,
-      "step": 10837
-    },
-    {
-      "epoch": 0.708319717665512,
-      "grad_norm": 0.4304639995098114,
-      "learning_rate": 8.847548902585304e-06,
-      "loss": 0.361,
-      "step": 10838
-    },
-    {
-      "epoch": 0.7083850728710542,
-      "grad_norm": 0.46742579340934753,
-      "learning_rate": 8.847325888233125e-06,
-      "loss": 0.3653,
-      "step": 10839
-    },
-    {
-      "epoch": 0.7084504280765963,
-      "grad_norm": 0.46168291568756104,
-      "learning_rate": 8.847102855116207e-06,
-      "loss": 0.3687,
-      "step": 10840
-    },
-    {
-      "epoch": 0.7085157832821384,
-      "grad_norm": 0.4948406517505646,
-      "learning_rate": 8.846879803235636e-06,
-      "loss": 0.4293,
-      "step": 10841
-    },
-    {
-      "epoch": 0.7085811384876806,
-      "grad_norm": 0.46802011132240295,
-      "learning_rate": 8.846656732592498e-06,
-      "loss": 0.4366,
-      "step": 10842
-    },
-    {
-      "epoch": 0.7086464936932226,
-      "grad_norm": 0.4439346492290497,
-      "learning_rate": 8.846433643187884e-06,
-      "loss": 0.3525,
-      "step": 10843
-    },
-    {
-      "epoch": 0.7087118488987648,
-      "grad_norm": 0.43533918261528015,
-      "learning_rate": 8.846210535022878e-06,
-      "loss": 0.348,
-      "step": 10844
-    },
-    {
-      "epoch": 0.7087772041043069,
-      "grad_norm": 0.43249449133872986,
-      "learning_rate": 8.845987408098574e-06,
-      "loss": 0.3338,
-      "step": 10845
-    },
-    {
-      "epoch": 0.7088425593098491,
-      "grad_norm": 0.4447890520095825,
-      "learning_rate": 8.845764262416056e-06,
-      "loss": 0.3561,
-      "step": 10846
-    },
-    {
-      "epoch": 0.7089079145153911,
-      "grad_norm": 0.46714460849761963,
-      "learning_rate": 8.845541097976414e-06,
-      "loss": 0.4183,
-      "step": 10847
-    },
-    {
-      "epoch": 0.7089732697209332,
-      "grad_norm": 0.4190499782562256,
-      "learning_rate": 8.845317914780734e-06,
-      "loss": 0.3403,
-      "step": 10848
-    },
-    {
-      "epoch": 0.7090386249264754,
-      "grad_norm": 0.45065832138061523,
-      "learning_rate": 8.845094712830107e-06,
-      "loss": 0.3823,
-      "step": 10849
-    },
-    {
-      "epoch": 0.7091039801320175,
-      "grad_norm": 0.4573337137699127,
-      "learning_rate": 8.844871492125622e-06,
-      "loss": 0.4012,
-      "step": 10850
-    },
-    {
-      "epoch": 0.7091693353375597,
-      "grad_norm": 0.4844611585140228,
-      "learning_rate": 8.844648252668366e-06,
-      "loss": 0.3647,
-      "step": 10851
-    },
-    {
-      "epoch": 0.7092346905431017,
-      "grad_norm": 0.41030028462409973,
-      "learning_rate": 8.844424994459428e-06,
-      "loss": 0.3293,
-      "step": 10852
-    },
-    {
-      "epoch": 0.7093000457486439,
-      "grad_norm": 0.4440174996852875,
-      "learning_rate": 8.8442017174999e-06,
-      "loss": 0.3452,
-      "step": 10853
-    },
-    {
-      "epoch": 0.709365400954186,
-      "grad_norm": 0.4625392258167267,
-      "learning_rate": 8.843978421790866e-06,
-      "loss": 0.4057,
-      "step": 10854
-    },
-    {
-      "epoch": 0.7094307561597282,
-      "grad_norm": 0.4622591733932495,
-      "learning_rate": 8.843755107333418e-06,
-      "loss": 0.4069,
-      "step": 10855
-    },
-    {
-      "epoch": 0.7094961113652702,
-      "grad_norm": 0.4473128616809845,
-      "learning_rate": 8.843531774128646e-06,
-      "loss": 0.371,
-      "step": 10856
-    },
-    {
-      "epoch": 0.7095614665708123,
-      "grad_norm": 0.43539655208587646,
-      "learning_rate": 8.843308422177637e-06,
-      "loss": 0.342,
-      "step": 10857
-    },
-    {
-      "epoch": 0.7096268217763545,
-      "grad_norm": 0.48266685009002686,
-      "learning_rate": 8.84308505148148e-06,
-      "loss": 0.3869,
-      "step": 10858
-    },
-    {
-      "epoch": 0.7096921769818966,
-      "grad_norm": 0.43436679244041443,
-      "learning_rate": 8.842861662041268e-06,
-      "loss": 0.3357,
-      "step": 10859
-    },
-    {
-      "epoch": 0.7097575321874388,
-      "grad_norm": 0.4314224421977997,
-      "learning_rate": 8.842638253858086e-06,
-      "loss": 0.3478,
-      "step": 10860
-    },
-    {
-      "epoch": 0.7098228873929808,
-      "grad_norm": 0.478385329246521,
-      "learning_rate": 8.842414826933028e-06,
-      "loss": 0.4209,
-      "step": 10861
-    },
-    {
-      "epoch": 0.709888242598523,
-      "grad_norm": 0.42900168895721436,
-      "learning_rate": 8.842191381267182e-06,
-      "loss": 0.3746,
-      "step": 10862
-    },
-    {
-      "epoch": 0.7099535978040651,
-      "grad_norm": 0.47261881828308105,
-      "learning_rate": 8.841967916861636e-06,
-      "loss": 0.41,
-      "step": 10863
-    },
-    {
-      "epoch": 0.7100189530096073,
-      "grad_norm": 0.45555639266967773,
-      "learning_rate": 8.841744433717484e-06,
-      "loss": 0.3696,
-      "step": 10864
-    },
-    {
-      "epoch": 0.7100843082151493,
-      "grad_norm": 0.41827529668807983,
-      "learning_rate": 8.841520931835812e-06,
-      "loss": 0.3498,
-      "step": 10865
-    },
-    {
-      "epoch": 0.7101496634206914,
-      "grad_norm": 0.4690840542316437,
-      "learning_rate": 8.84129741121771e-06,
-      "loss": 0.4432,
-      "step": 10866
-    },
-    {
-      "epoch": 0.7102150186262336,
-      "grad_norm": 0.45258966088294983,
-      "learning_rate": 8.841073871864272e-06,
-      "loss": 0.3969,
-      "step": 10867
-    },
-    {
-      "epoch": 0.7102803738317757,
-      "grad_norm": 0.4109076261520386,
-      "learning_rate": 8.840850313776586e-06,
-      "loss": 0.3324,
-      "step": 10868
-    },
-    {
-      "epoch": 0.7103457290373179,
-      "grad_norm": 0.430026113986969,
-      "learning_rate": 8.84062673695574e-06,
-      "loss": 0.3603,
-      "step": 10869
-    },
-    {
-      "epoch": 0.7104110842428599,
-      "grad_norm": 0.4078415036201477,
-      "learning_rate": 8.840403141402829e-06,
-      "loss": 0.3253,
-      "step": 10870
-    },
-    {
-      "epoch": 0.7104764394484021,
-      "grad_norm": 0.43680548667907715,
-      "learning_rate": 8.840179527118942e-06,
-      "loss": 0.3311,
-      "step": 10871
-    },
-    {
-      "epoch": 0.7105417946539442,
-      "grad_norm": 0.5105580687522888,
-      "learning_rate": 8.839955894105167e-06,
-      "loss": 0.4159,
-      "step": 10872
-    },
-    {
-      "epoch": 0.7106071498594863,
-      "grad_norm": 0.5044821500778198,
-      "learning_rate": 8.839732242362598e-06,
-      "loss": 0.4195,
-      "step": 10873
-    },
-    {
-      "epoch": 0.7106725050650284,
-      "grad_norm": 0.4528474807739258,
-      "learning_rate": 8.839508571892325e-06,
-      "loss": 0.3536,
-      "step": 10874
-    },
-    {
-      "epoch": 0.7107378602705705,
-      "grad_norm": 0.40835288166999817,
-      "learning_rate": 8.839284882695438e-06,
-      "loss": 0.3191,
-      "step": 10875
-    },
-    {
-      "epoch": 0.7108032154761127,
-      "grad_norm": 0.5332349538803101,
-      "learning_rate": 8.839061174773029e-06,
-      "loss": 0.4065,
-      "step": 10876
-    },
-    {
-      "epoch": 0.7108685706816548,
-      "grad_norm": 0.44312018156051636,
-      "learning_rate": 8.838837448126188e-06,
-      "loss": 0.3668,
-      "step": 10877
-    },
-    {
-      "epoch": 0.710933925887197,
-      "grad_norm": 0.48014044761657715,
-      "learning_rate": 8.838613702756007e-06,
-      "loss": 0.409,
-      "step": 10878
-    },
-    {
-      "epoch": 0.710999281092739,
-      "grad_norm": 0.46533316373825073,
-      "learning_rate": 8.838389938663577e-06,
-      "loss": 0.3773,
-      "step": 10879
-    },
-    {
-      "epoch": 0.7110646362982812,
-      "grad_norm": 0.42617470026016235,
-      "learning_rate": 8.83816615584999e-06,
-      "loss": 0.3422,
-      "step": 10880
-    },
-    {
-      "epoch": 0.7111299915038233,
-      "grad_norm": 0.42829635739326477,
-      "learning_rate": 8.837942354316339e-06,
-      "loss": 0.3169,
-      "step": 10881
-    },
-    {
-      "epoch": 0.7111953467093654,
-      "grad_norm": 0.4599711298942566,
-      "learning_rate": 8.83771853406371e-06,
-      "loss": 0.3753,
-      "step": 10882
-    },
-    {
-      "epoch": 0.7112607019149075,
-      "grad_norm": 0.4736286997795105,
-      "learning_rate": 8.837494695093199e-06,
-      "loss": 0.4014,
-      "step": 10883
-    },
-    {
-      "epoch": 0.7113260571204496,
-      "grad_norm": 0.4122152328491211,
-      "learning_rate": 8.837270837405898e-06,
-      "loss": 0.3421,
-      "step": 10884
-    },
-    {
-      "epoch": 0.7113914123259918,
-      "grad_norm": 0.4581123888492584,
-      "learning_rate": 8.837046961002897e-06,
-      "loss": 0.3795,
-      "step": 10885
-    },
-    {
-      "epoch": 0.7114567675315339,
-      "grad_norm": 0.45057132840156555,
-      "learning_rate": 8.83682306588529e-06,
-      "loss": 0.4064,
-      "step": 10886
-    },
-    {
-      "epoch": 0.711522122737076,
-      "grad_norm": 0.4579734802246094,
-      "learning_rate": 8.836599152054166e-06,
-      "loss": 0.3461,
-      "step": 10887
-    },
-    {
-      "epoch": 0.7115874779426181,
-      "grad_norm": 0.4159832000732422,
-      "learning_rate": 8.836375219510618e-06,
-      "loss": 0.3296,
-      "step": 10888
-    },
-    {
-      "epoch": 0.7116528331481603,
-      "grad_norm": 0.4370971620082855,
-      "learning_rate": 8.836151268255742e-06,
-      "loss": 0.3648,
-      "step": 10889
-    },
-    {
-      "epoch": 0.7117181883537024,
-      "grad_norm": 0.45852354168891907,
-      "learning_rate": 8.835927298290625e-06,
-      "loss": 0.4075,
-      "step": 10890
-    },
-    {
-      "epoch": 0.7117835435592444,
-      "grad_norm": 0.4565734565258026,
-      "learning_rate": 8.835703309616364e-06,
-      "loss": 0.3773,
-      "step": 10891
-    },
-    {
-      "epoch": 0.7118488987647866,
-      "grad_norm": 0.43849247694015503,
-      "learning_rate": 8.835479302234047e-06,
-      "loss": 0.3948,
-      "step": 10892
-    },
-    {
-      "epoch": 0.7119142539703287,
-      "grad_norm": 0.4963937997817993,
-      "learning_rate": 8.835255276144769e-06,
-      "loss": 0.4058,
-      "step": 10893
-    },
-    {
-      "epoch": 0.7119796091758709,
-      "grad_norm": 0.4483307898044586,
-      "learning_rate": 8.835031231349622e-06,
-      "loss": 0.3813,
-      "step": 10894
-    },
-    {
-      "epoch": 0.712044964381413,
-      "grad_norm": 0.4734781086444855,
-      "learning_rate": 8.8348071678497e-06,
-      "loss": 0.394,
-      "step": 10895
-    },
-    {
-      "epoch": 0.7121103195869551,
-      "grad_norm": 0.4316871464252472,
-      "learning_rate": 8.834583085646095e-06,
-      "loss": 0.3194,
-      "step": 10896
-    },
-    {
-      "epoch": 0.7121756747924972,
-      "grad_norm": 0.48908916115760803,
-      "learning_rate": 8.8343589847399e-06,
-      "loss": 0.4083,
-      "step": 10897
-    },
-    {
-      "epoch": 0.7122410299980394,
-      "grad_norm": 0.4344671368598938,
-      "learning_rate": 8.834134865132207e-06,
-      "loss": 0.3855,
-      "step": 10898
-    },
-    {
-      "epoch": 0.7123063852035815,
-      "grad_norm": 0.4473215341567993,
-      "learning_rate": 8.833910726824111e-06,
-      "loss": 0.3448,
-      "step": 10899
-    },
-    {
-      "epoch": 0.7123717404091235,
-      "grad_norm": 0.4628905653953552,
-      "learning_rate": 8.833686569816702e-06,
-      "loss": 0.4,
-      "step": 10900
-    },
-    {
-      "epoch": 0.7124370956146657,
-      "grad_norm": 0.48413145542144775,
-      "learning_rate": 8.833462394111078e-06,
-      "loss": 0.3781,
-      "step": 10901
-    },
-    {
-      "epoch": 0.7125024508202078,
-      "grad_norm": 0.45187827944755554,
-      "learning_rate": 8.83323819970833e-06,
-      "loss": 0.3468,
-      "step": 10902
-    },
-    {
-      "epoch": 0.71256780602575,
-      "grad_norm": 0.44939520955085754,
-      "learning_rate": 8.833013986609549e-06,
-      "loss": 0.3865,
-      "step": 10903
-    },
-    {
-      "epoch": 0.7126331612312921,
-      "grad_norm": 0.4265718460083008,
-      "learning_rate": 8.832789754815834e-06,
-      "loss": 0.3789,
-      "step": 10904
-    },
-    {
-      "epoch": 0.7126985164368342,
-      "grad_norm": 0.4289838373661041,
-      "learning_rate": 8.832565504328274e-06,
-      "loss": 0.351,
-      "step": 10905
-    },
-    {
-      "epoch": 0.7127638716423763,
-      "grad_norm": 0.46329349279403687,
-      "learning_rate": 8.832341235147963e-06,
-      "loss": 0.4032,
-      "step": 10906
-    },
-    {
-      "epoch": 0.7128292268479184,
-      "grad_norm": 0.48146799206733704,
-      "learning_rate": 8.832116947275997e-06,
-      "loss": 0.3834,
-      "step": 10907
-    },
-    {
-      "epoch": 0.7128945820534606,
-      "grad_norm": 0.45707768201828003,
-      "learning_rate": 8.831892640713469e-06,
-      "loss": 0.3914,
-      "step": 10908
-    },
-    {
-      "epoch": 0.7129599372590026,
-      "grad_norm": 0.48288625478744507,
-      "learning_rate": 8.831668315461475e-06,
-      "loss": 0.4414,
-      "step": 10909
-    },
-    {
-      "epoch": 0.7130252924645448,
-      "grad_norm": 0.44983968138694763,
-      "learning_rate": 8.831443971521106e-06,
-      "loss": 0.4292,
-      "step": 10910
-    },
-    {
-      "epoch": 0.7130906476700869,
-      "grad_norm": 0.44615569710731506,
-      "learning_rate": 8.831219608893456e-06,
-      "loss": 0.3608,
-      "step": 10911
-    },
-    {
-      "epoch": 0.7131560028756291,
-      "grad_norm": 0.46797874569892883,
-      "learning_rate": 8.830995227579622e-06,
-      "loss": 0.4129,
-      "step": 10912
-    },
-    {
-      "epoch": 0.7132213580811712,
-      "grad_norm": 0.4359447956085205,
-      "learning_rate": 8.830770827580697e-06,
-      "loss": 0.3618,
-      "step": 10913
-    },
-    {
-      "epoch": 0.7132867132867133,
-      "grad_norm": 0.44202476739883423,
-      "learning_rate": 8.830546408897775e-06,
-      "loss": 0.3496,
-      "step": 10914
-    },
-    {
-      "epoch": 0.7133520684922554,
-      "grad_norm": 0.4439420998096466,
-      "learning_rate": 8.830321971531952e-06,
-      "loss": 0.3795,
-      "step": 10915
-    },
-    {
-      "epoch": 0.7134174236977975,
-      "grad_norm": 0.47986286878585815,
-      "learning_rate": 8.830097515484322e-06,
-      "loss": 0.3911,
-      "step": 10916
-    },
-    {
-      "epoch": 0.7134827789033397,
-      "grad_norm": 0.6670001149177551,
-      "learning_rate": 8.829873040755979e-06,
-      "loss": 0.4094,
-      "step": 10917
-    },
-    {
-      "epoch": 0.7135481341088817,
-      "grad_norm": 0.4551219642162323,
-      "learning_rate": 8.829648547348017e-06,
-      "loss": 0.3808,
-      "step": 10918
-    },
-    {
-      "epoch": 0.7136134893144239,
-      "grad_norm": 0.4279273748397827,
-      "learning_rate": 8.829424035261534e-06,
-      "loss": 0.349,
-      "step": 10919
-    },
-    {
-      "epoch": 0.713678844519966,
-      "grad_norm": 0.4488070011138916,
-      "learning_rate": 8.829199504497623e-06,
-      "loss": 0.3762,
-      "step": 10920
-    },
-    {
-      "epoch": 0.7137441997255082,
-      "grad_norm": 0.46613001823425293,
-      "learning_rate": 8.828974955057378e-06,
-      "loss": 0.3977,
-      "step": 10921
-    },
-    {
-      "epoch": 0.7138095549310502,
-      "grad_norm": 0.5064405202865601,
-      "learning_rate": 8.828750386941898e-06,
-      "loss": 0.4216,
-      "step": 10922
-    },
-    {
-      "epoch": 0.7138749101365924,
-      "grad_norm": 0.4371580183506012,
-      "learning_rate": 8.828525800152276e-06,
-      "loss": 0.3806,
-      "step": 10923
-    },
-    {
-      "epoch": 0.7139402653421345,
-      "grad_norm": 0.4593411684036255,
-      "learning_rate": 8.828301194689606e-06,
-      "loss": 0.3727,
-      "step": 10924
-    },
-    {
-      "epoch": 0.7140056205476766,
-      "grad_norm": 0.4481455087661743,
-      "learning_rate": 8.828076570554985e-06,
-      "loss": 0.4087,
-      "step": 10925
-    },
-    {
-      "epoch": 0.7140709757532188,
-      "grad_norm": 0.44991838932037354,
-      "learning_rate": 8.827851927749509e-06,
-      "loss": 0.3683,
-      "step": 10926
-    },
-    {
-      "epoch": 0.7141363309587608,
-      "grad_norm": 0.43994560837745667,
-      "learning_rate": 8.827627266274272e-06,
-      "loss": 0.377,
-      "step": 10927
-    },
-    {
-      "epoch": 0.714201686164303,
-      "grad_norm": 0.4755774438381195,
-      "learning_rate": 8.827402586130372e-06,
-      "loss": 0.3778,
-      "step": 10928
-    },
-    {
-      "epoch": 0.7142670413698451,
-      "grad_norm": 0.4460085332393646,
-      "learning_rate": 8.827177887318903e-06,
-      "loss": 0.4226,
-      "step": 10929
-    },
-    {
-      "epoch": 0.7143323965753873,
-      "grad_norm": 0.464880108833313,
-      "learning_rate": 8.82695316984096e-06,
-      "loss": 0.4037,
-      "step": 10930
-    },
-    {
-      "epoch": 0.7143977517809293,
-      "grad_norm": 0.4421575963497162,
-      "learning_rate": 8.826728433697643e-06,
-      "loss": 0.3343,
-      "step": 10931
-    },
-    {
-      "epoch": 0.7144631069864714,
-      "grad_norm": 0.4808433949947357,
-      "learning_rate": 8.826503678890045e-06,
-      "loss": 0.3898,
-      "step": 10932
-    },
-    {
-      "epoch": 0.7145284621920136,
-      "grad_norm": 0.42213189601898193,
-      "learning_rate": 8.826278905419264e-06,
-      "loss": 0.3359,
-      "step": 10933
-    },
-    {
-      "epoch": 0.7145938173975557,
-      "grad_norm": 0.4662804901599884,
-      "learning_rate": 8.826054113286394e-06,
-      "loss": 0.4158,
-      "step": 10934
-    },
-    {
-      "epoch": 0.7146591726030979,
-      "grad_norm": 0.4736062288284302,
-      "learning_rate": 8.825829302492532e-06,
-      "loss": 0.4586,
-      "step": 10935
-    },
-    {
-      "epoch": 0.7147245278086399,
-      "grad_norm": 0.4594738483428955,
-      "learning_rate": 8.825604473038777e-06,
-      "loss": 0.4074,
-      "step": 10936
-    },
-    {
-      "epoch": 0.7147898830141821,
-      "grad_norm": 0.4078019857406616,
-      "learning_rate": 8.825379624926222e-06,
-      "loss": 0.3007,
-      "step": 10937
-    },
-    {
-      "epoch": 0.7148552382197242,
-      "grad_norm": 0.5145253539085388,
-      "learning_rate": 8.825154758155964e-06,
-      "loss": 0.3368,
-      "step": 10938
-    },
-    {
-      "epoch": 0.7149205934252664,
-      "grad_norm": 0.4444692134857178,
-      "learning_rate": 8.824929872729105e-06,
-      "loss": 0.3356,
-      "step": 10939
-    },
-    {
-      "epoch": 0.7149859486308084,
-      "grad_norm": 0.4701055586338043,
-      "learning_rate": 8.824704968646736e-06,
-      "loss": 0.4099,
-      "step": 10940
-    },
-    {
-      "epoch": 0.7150513038363505,
-      "grad_norm": 0.6250319480895996,
-      "learning_rate": 8.824480045909955e-06,
-      "loss": 0.3978,
-      "step": 10941
-    },
-    {
-      "epoch": 0.7151166590418927,
-      "grad_norm": 0.42224377393722534,
-      "learning_rate": 8.82425510451986e-06,
-      "loss": 0.3229,
-      "step": 10942
-    },
-    {
-      "epoch": 0.7151820142474348,
-      "grad_norm": 0.44007593393325806,
-      "learning_rate": 8.824030144477548e-06,
-      "loss": 0.3425,
-      "step": 10943
-    },
-    {
-      "epoch": 0.715247369452977,
-      "grad_norm": 0.48423925042152405,
-      "learning_rate": 8.823805165784118e-06,
-      "loss": 0.449,
-      "step": 10944
-    },
-    {
-      "epoch": 0.715312724658519,
-      "grad_norm": 0.49116432666778564,
-      "learning_rate": 8.823580168440664e-06,
-      "loss": 0.439,
-      "step": 10945
-    },
-    {
-      "epoch": 0.7153780798640612,
-      "grad_norm": 0.4677906334400177,
-      "learning_rate": 8.823355152448285e-06,
-      "loss": 0.4068,
-      "step": 10946
-    },
-    {
-      "epoch": 0.7154434350696033,
-      "grad_norm": 0.4655870199203491,
-      "learning_rate": 8.823130117808079e-06,
-      "loss": 0.4314,
-      "step": 10947
-    },
-    {
-      "epoch": 0.7155087902751455,
-      "grad_norm": 0.4523985981941223,
-      "learning_rate": 8.822905064521143e-06,
-      "loss": 0.3642,
-      "step": 10948
-    },
-    {
-      "epoch": 0.7155741454806875,
-      "grad_norm": 0.4735022187232971,
-      "learning_rate": 8.822679992588575e-06,
-      "loss": 0.3843,
-      "step": 10949
-    },
-    {
-      "epoch": 0.7156395006862296,
-      "grad_norm": 0.4337882399559021,
-      "learning_rate": 8.82245490201147e-06,
-      "loss": 0.3373,
-      "step": 10950
-    },
-    {
-      "epoch": 0.7157048558917718,
-      "grad_norm": 0.4603765308856964,
-      "learning_rate": 8.82222979279093e-06,
-      "loss": 0.4014,
-      "step": 10951
-    },
-    {
-      "epoch": 0.7157702110973139,
-      "grad_norm": 0.421307772397995,
-      "learning_rate": 8.82200466492805e-06,
-      "loss": 0.3556,
-      "step": 10952
-    },
-    {
-      "epoch": 0.715835566302856,
-      "grad_norm": 0.44478708505630493,
-      "learning_rate": 8.821779518423932e-06,
-      "loss": 0.3485,
-      "step": 10953
-    },
-    {
-      "epoch": 0.7159009215083981,
-      "grad_norm": 0.44603121280670166,
-      "learning_rate": 8.82155435327967e-06,
-      "loss": 0.381,
-      "step": 10954
-    },
-    {
-      "epoch": 0.7159662767139403,
-      "grad_norm": 0.43115556240081787,
-      "learning_rate": 8.821329169496362e-06,
-      "loss": 0.3724,
-      "step": 10955
-    },
-    {
-      "epoch": 0.7160316319194824,
-      "grad_norm": 0.4249560832977295,
-      "learning_rate": 8.821103967075108e-06,
-      "loss": 0.3255,
-      "step": 10956
-    },
-    {
-      "epoch": 0.7160969871250245,
-      "grad_norm": 0.4381152093410492,
-      "learning_rate": 8.820878746017008e-06,
-      "loss": 0.3402,
-      "step": 10957
-    },
-    {
-      "epoch": 0.7161623423305666,
-      "grad_norm": 0.46246418356895447,
-      "learning_rate": 8.820653506323156e-06,
-      "loss": 0.3846,
-      "step": 10958
-    },
-    {
-      "epoch": 0.7162276975361087,
-      "grad_norm": 0.49482643604278564,
-      "learning_rate": 8.820428247994656e-06,
-      "loss": 0.3852,
-      "step": 10959
-    },
-    {
-      "epoch": 0.7162930527416509,
-      "grad_norm": 0.4686237573623657,
-      "learning_rate": 8.820202971032604e-06,
-      "loss": 0.3347,
-      "step": 10960
-    },
-    {
-      "epoch": 0.716358407947193,
-      "grad_norm": 0.4484345316886902,
-      "learning_rate": 8.819977675438096e-06,
-      "loss": 0.3881,
-      "step": 10961
-    },
-    {
-      "epoch": 0.7164237631527351,
-      "grad_norm": 0.43936142325401306,
-      "learning_rate": 8.819752361212235e-06,
-      "loss": 0.3689,
-      "step": 10962
-    },
-    {
-      "epoch": 0.7164891183582772,
-      "grad_norm": 0.4533930718898773,
-      "learning_rate": 8.819527028356118e-06,
-      "loss": 0.3805,
-      "step": 10963
-    },
-    {
-      "epoch": 0.7165544735638194,
-      "grad_norm": 0.43360522389411926,
-      "learning_rate": 8.819301676870847e-06,
-      "loss": 0.3692,
-      "step": 10964
-    },
-    {
-      "epoch": 0.7166198287693615,
-      "grad_norm": 0.4650508165359497,
-      "learning_rate": 8.819076306757514e-06,
-      "loss": 0.3679,
-      "step": 10965
-    },
-    {
-      "epoch": 0.7166851839749036,
-      "grad_norm": 0.4472239315509796,
-      "learning_rate": 8.818850918017225e-06,
-      "loss": 0.3899,
-      "step": 10966
-    },
-    {
-      "epoch": 0.7167505391804457,
-      "grad_norm": 0.4909222424030304,
-      "learning_rate": 8.818625510651077e-06,
-      "loss": 0.397,
-      "step": 10967
-    },
-    {
-      "epoch": 0.7168158943859878,
-      "grad_norm": 0.4445730149745941,
-      "learning_rate": 8.81840008466017e-06,
-      "loss": 0.3811,
-      "step": 10968
-    },
-    {
-      "epoch": 0.71688124959153,
-      "grad_norm": 0.4273037612438202,
-      "learning_rate": 8.818174640045605e-06,
-      "loss": 0.3553,
-      "step": 10969
-    },
-    {
-      "epoch": 0.7169466047970721,
-      "grad_norm": 0.4292222261428833,
-      "learning_rate": 8.817949176808476e-06,
-      "loss": 0.3503,
-      "step": 10970
-    },
-    {
-      "epoch": 0.7170119600026142,
-      "grad_norm": 0.45991745591163635,
-      "learning_rate": 8.817723694949887e-06,
-      "loss": 0.3717,
-      "step": 10971
-    },
-    {
-      "epoch": 0.7170773152081563,
-      "grad_norm": 0.44813650846481323,
-      "learning_rate": 8.81749819447094e-06,
-      "loss": 0.3981,
-      "step": 10972
-    },
-    {
-      "epoch": 0.7171426704136985,
-      "grad_norm": 0.49451524019241333,
-      "learning_rate": 8.817272675372728e-06,
-      "loss": 0.4176,
-      "step": 10973
-    },
-    {
-      "epoch": 0.7172080256192406,
-      "grad_norm": 0.47646835446357727,
-      "learning_rate": 8.817047137656356e-06,
-      "loss": 0.426,
-      "step": 10974
-    },
-    {
-      "epoch": 0.7172733808247826,
-      "grad_norm": 0.4439930021762848,
-      "learning_rate": 8.816821581322922e-06,
-      "loss": 0.3521,
-      "step": 10975
-    },
-    {
-      "epoch": 0.7173387360303248,
-      "grad_norm": 0.43444332480430603,
-      "learning_rate": 8.816596006373529e-06,
-      "loss": 0.3492,
-      "step": 10976
-    },
-    {
-      "epoch": 0.7174040912358669,
-      "grad_norm": 0.4854743778705597,
-      "learning_rate": 8.816370412809273e-06,
-      "loss": 0.4523,
-      "step": 10977
-    },
-    {
-      "epoch": 0.7174694464414091,
-      "grad_norm": 0.4733608663082123,
-      "learning_rate": 8.816144800631256e-06,
-      "loss": 0.4003,
-      "step": 10978
-    },
-    {
-      "epoch": 0.7175348016469512,
-      "grad_norm": 0.43836793303489685,
-      "learning_rate": 8.81591916984058e-06,
-      "loss": 0.3709,
-      "step": 10979
-    },
-    {
-      "epoch": 0.7176001568524933,
-      "grad_norm": 0.45413318276405334,
-      "learning_rate": 8.815693520438346e-06,
-      "loss": 0.3645,
-      "step": 10980
-    },
-    {
-      "epoch": 0.7176655120580354,
-      "grad_norm": 0.45094919204711914,
-      "learning_rate": 8.81546785242565e-06,
-      "loss": 0.3656,
-      "step": 10981
-    },
-    {
-      "epoch": 0.7177308672635776,
-      "grad_norm": 0.5256023406982422,
-      "learning_rate": 8.815242165803597e-06,
-      "loss": 0.4817,
-      "step": 10982
-    },
-    {
-      "epoch": 0.7177962224691197,
-      "grad_norm": 0.4483894407749176,
-      "learning_rate": 8.815016460573284e-06,
-      "loss": 0.3846,
-      "step": 10983
-    },
-    {
-      "epoch": 0.7178615776746617,
-      "grad_norm": 0.5038370490074158,
-      "learning_rate": 8.814790736735817e-06,
-      "loss": 0.4298,
-      "step": 10984
-    },
-    {
-      "epoch": 0.7179269328802039,
-      "grad_norm": 0.44803082942962646,
-      "learning_rate": 8.814564994292293e-06,
-      "loss": 0.3324,
-      "step": 10985
-    },
-    {
-      "epoch": 0.717992288085746,
-      "grad_norm": 0.4388437867164612,
-      "learning_rate": 8.814339233243813e-06,
-      "loss": 0.3726,
-      "step": 10986
-    },
-    {
-      "epoch": 0.7180576432912882,
-      "grad_norm": 0.46978771686553955,
-      "learning_rate": 8.81411345359148e-06,
-      "loss": 0.4075,
-      "step": 10987
-    },
-    {
-      "epoch": 0.7181229984968303,
-      "grad_norm": 0.429244726896286,
-      "learning_rate": 8.813887655336394e-06,
-      "loss": 0.3628,
-      "step": 10988
-    },
-    {
-      "epoch": 0.7181883537023724,
-      "grad_norm": 0.43495580554008484,
-      "learning_rate": 8.813661838479658e-06,
-      "loss": 0.3758,
-      "step": 10989
-    },
-    {
-      "epoch": 0.7182537089079145,
-      "grad_norm": 0.48604434728622437,
-      "learning_rate": 8.81343600302237e-06,
-      "loss": 0.3737,
-      "step": 10990
-    },
-    {
-      "epoch": 0.7183190641134566,
-      "grad_norm": 0.4476868510246277,
-      "learning_rate": 8.813210148965634e-06,
-      "loss": 0.3622,
-      "step": 10991
-    },
-    {
-      "epoch": 0.7183844193189988,
-      "grad_norm": 0.45974016189575195,
-      "learning_rate": 8.812984276310551e-06,
-      "loss": 0.3763,
-      "step": 10992
-    },
-    {
-      "epoch": 0.7184497745245408,
-      "grad_norm": 0.428249329328537,
-      "learning_rate": 8.812758385058225e-06,
-      "loss": 0.3667,
-      "step": 10993
-    },
-    {
-      "epoch": 0.718515129730083,
-      "grad_norm": 0.43195784091949463,
-      "learning_rate": 8.812532475209754e-06,
-      "loss": 0.3286,
-      "step": 10994
-    },
-    {
-      "epoch": 0.7185804849356251,
-      "grad_norm": 0.450755774974823,
-      "learning_rate": 8.81230654676624e-06,
-      "loss": 0.3568,
-      "step": 10995
-    },
-    {
-      "epoch": 0.7186458401411673,
-      "grad_norm": 0.47830522060394287,
-      "learning_rate": 8.812080599728787e-06,
-      "loss": 0.3891,
-      "step": 10996
-    },
-    {
-      "epoch": 0.7187111953467094,
-      "grad_norm": 0.42130833864212036,
-      "learning_rate": 8.811854634098497e-06,
-      "loss": 0.3639,
-      "step": 10997
-    },
-    {
-      "epoch": 0.7187765505522515,
-      "grad_norm": 0.4895265996456146,
-      "learning_rate": 8.811628649876471e-06,
-      "loss": 0.4119,
-      "step": 10998
-    },
-    {
-      "epoch": 0.7188419057577936,
-      "grad_norm": 0.5014936923980713,
-      "learning_rate": 8.811402647063812e-06,
-      "loss": 0.3907,
-      "step": 10999
-    },
-    {
-      "epoch": 0.7189072609633357,
-      "grad_norm": 0.4365863502025604,
-      "learning_rate": 8.811176625661622e-06,
-      "loss": 0.3865,
-      "step": 11000
-    },
-    {
-      "epoch": 0.7189726161688779,
-      "grad_norm": 0.4857371747493744,
-      "learning_rate": 8.810950585671003e-06,
-      "loss": 0.4793,
-      "step": 11001
-    },
-    {
-      "epoch": 0.7190379713744199,
-      "grad_norm": 0.4356517791748047,
-      "learning_rate": 8.810724527093057e-06,
-      "loss": 0.3624,
-      "step": 11002
-    },
-    {
-      "epoch": 0.7191033265799621,
-      "grad_norm": 0.4523765444755554,
-      "learning_rate": 8.810498449928888e-06,
-      "loss": 0.4028,
-      "step": 11003
-    },
-    {
-      "epoch": 0.7191686817855042,
-      "grad_norm": 0.47722652554512024,
-      "learning_rate": 8.810272354179598e-06,
-      "loss": 0.4262,
-      "step": 11004
-    },
-    {
-      "epoch": 0.7192340369910464,
-      "grad_norm": 0.43598616123199463,
-      "learning_rate": 8.81004623984629e-06,
-      "loss": 0.3193,
-      "step": 11005
-    },
-    {
-      "epoch": 0.7192993921965884,
-      "grad_norm": 0.4548710286617279,
-      "learning_rate": 8.809820106930066e-06,
-      "loss": 0.3794,
-      "step": 11006
-    },
-    {
-      "epoch": 0.7193647474021306,
-      "grad_norm": 0.4733636677265167,
-      "learning_rate": 8.80959395543203e-06,
-      "loss": 0.4245,
-      "step": 11007
-    },
-    {
-      "epoch": 0.7194301026076727,
-      "grad_norm": 0.46189406514167786,
-      "learning_rate": 8.809367785353284e-06,
-      "loss": 0.4062,
-      "step": 11008
-    },
-    {
-      "epoch": 0.7194954578132148,
-      "grad_norm": 0.48950931429862976,
-      "learning_rate": 8.809141596694932e-06,
-      "loss": 0.4353,
-      "step": 11009
-    },
-    {
-      "epoch": 0.719560813018757,
-      "grad_norm": 0.45719993114471436,
-      "learning_rate": 8.808915389458076e-06,
-      "loss": 0.3878,
-      "step": 11010
-    },
-    {
-      "epoch": 0.719626168224299,
-      "grad_norm": 0.42908379435539246,
-      "learning_rate": 8.80868916364382e-06,
-      "loss": 0.3694,
-      "step": 11011
-    },
-    {
-      "epoch": 0.7196915234298412,
-      "grad_norm": 0.4180840849876404,
-      "learning_rate": 8.808462919253268e-06,
-      "loss": 0.3811,
-      "step": 11012
-    },
-    {
-      "epoch": 0.7197568786353833,
-      "grad_norm": 0.4631643295288086,
-      "learning_rate": 8.808236656287523e-06,
-      "loss": 0.3788,
-      "step": 11013
-    },
-    {
-      "epoch": 0.7198222338409255,
-      "grad_norm": 0.5063216090202332,
-      "learning_rate": 8.808010374747688e-06,
-      "loss": 0.4949,
-      "step": 11014
-    },
-    {
-      "epoch": 0.7198875890464675,
-      "grad_norm": 0.47948136925697327,
-      "learning_rate": 8.807784074634868e-06,
-      "loss": 0.4339,
-      "step": 11015
-    },
-    {
-      "epoch": 0.7199529442520096,
-      "grad_norm": 0.45391443371772766,
-      "learning_rate": 8.807557755950167e-06,
-      "loss": 0.3784,
-      "step": 11016
-    },
-    {
-      "epoch": 0.7200182994575518,
-      "grad_norm": 0.43957698345184326,
-      "learning_rate": 8.807331418694687e-06,
-      "loss": 0.3866,
-      "step": 11017
-    },
-    {
-      "epoch": 0.7200836546630939,
-      "grad_norm": 0.4413336217403412,
-      "learning_rate": 8.807105062869533e-06,
-      "loss": 0.3682,
-      "step": 11018
-    },
-    {
-      "epoch": 0.7201490098686361,
-      "grad_norm": 0.4598570466041565,
-      "learning_rate": 8.806878688475806e-06,
-      "loss": 0.4059,
-      "step": 11019
-    },
-    {
-      "epoch": 0.7202143650741781,
-      "grad_norm": 0.4452202320098877,
-      "learning_rate": 8.806652295514615e-06,
-      "loss": 0.3558,
-      "step": 11020
-    },
-    {
-      "epoch": 0.7202797202797203,
-      "grad_norm": 0.4020130932331085,
-      "learning_rate": 8.806425883987063e-06,
-      "loss": 0.2888,
-      "step": 11021
-    },
-    {
-      "epoch": 0.7203450754852624,
-      "grad_norm": 0.4242806136608124,
-      "learning_rate": 8.806199453894251e-06,
-      "loss": 0.3682,
-      "step": 11022
-    },
-    {
-      "epoch": 0.7204104306908046,
-      "grad_norm": 0.4291136562824249,
-      "learning_rate": 8.805973005237287e-06,
-      "loss": 0.3582,
-      "step": 11023
-    },
-    {
-      "epoch": 0.7204757858963466,
-      "grad_norm": 0.455759733915329,
-      "learning_rate": 8.805746538017275e-06,
-      "loss": 0.3962,
-      "step": 11024
-    },
-    {
-      "epoch": 0.7205411411018887,
-      "grad_norm": 0.4600485563278198,
-      "learning_rate": 8.805520052235316e-06,
-      "loss": 0.4046,
-      "step": 11025
-    },
-    {
-      "epoch": 0.7206064963074309,
-      "grad_norm": 0.45278605818748474,
-      "learning_rate": 8.80529354789252e-06,
-      "loss": 0.377,
-      "step": 11026
-    },
-    {
-      "epoch": 0.720671851512973,
-      "grad_norm": 0.4613986313343048,
-      "learning_rate": 8.805067024989989e-06,
-      "loss": 0.44,
-      "step": 11027
-    },
-    {
-      "epoch": 0.7207372067185152,
-      "grad_norm": 0.4480716586112976,
-      "learning_rate": 8.804840483528824e-06,
-      "loss": 0.3674,
-      "step": 11028
-    },
-    {
-      "epoch": 0.7208025619240572,
-      "grad_norm": 0.43719980120658875,
-      "learning_rate": 8.804613923510138e-06,
-      "loss": 0.3594,
-      "step": 11029
-    },
-    {
-      "epoch": 0.7208679171295994,
-      "grad_norm": 0.4525455832481384,
-      "learning_rate": 8.804387344935031e-06,
-      "loss": 0.3968,
-      "step": 11030
-    },
-    {
-      "epoch": 0.7209332723351415,
-      "grad_norm": 0.43577420711517334,
-      "learning_rate": 8.804160747804608e-06,
-      "loss": 0.3375,
-      "step": 11031
-    },
-    {
-      "epoch": 0.7209986275406837,
-      "grad_norm": 0.4039052128791809,
-      "learning_rate": 8.803934132119976e-06,
-      "loss": 0.3682,
-      "step": 11032
-    },
-    {
-      "epoch": 0.7210639827462257,
-      "grad_norm": 0.4610409736633301,
-      "learning_rate": 8.803707497882239e-06,
-      "loss": 0.3992,
-      "step": 11033
-    },
-    {
-      "epoch": 0.7211293379517678,
-      "grad_norm": 0.5072036981582642,
-      "learning_rate": 8.803480845092503e-06,
-      "loss": 0.4154,
-      "step": 11034
-    },
-    {
-      "epoch": 0.72119469315731,
-      "grad_norm": 0.44230931997299194,
-      "learning_rate": 8.803254173751874e-06,
-      "loss": 0.3602,
-      "step": 11035
-    },
-    {
-      "epoch": 0.7212600483628521,
-      "grad_norm": 0.43132802844047546,
-      "learning_rate": 8.803027483861455e-06,
-      "loss": 0.3406,
-      "step": 11036
-    },
-    {
-      "epoch": 0.7213254035683943,
-      "grad_norm": 0.4197312593460083,
-      "learning_rate": 8.802800775422354e-06,
-      "loss": 0.3151,
-      "step": 11037
-    },
-    {
-      "epoch": 0.7213907587739363,
-      "grad_norm": 0.43612998723983765,
-      "learning_rate": 8.802574048435677e-06,
-      "loss": 0.3549,
-      "step": 11038
-    },
-    {
-      "epoch": 0.7214561139794785,
-      "grad_norm": 0.450834721326828,
-      "learning_rate": 8.802347302902528e-06,
-      "loss": 0.3917,
-      "step": 11039
-    },
-    {
-      "epoch": 0.7215214691850206,
-      "grad_norm": 0.44693103432655334,
-      "learning_rate": 8.802120538824015e-06,
-      "loss": 0.3489,
-      "step": 11040
-    },
-    {
-      "epoch": 0.7215868243905627,
-      "grad_norm": 0.43119072914123535,
-      "learning_rate": 8.801893756201242e-06,
-      "loss": 0.3489,
-      "step": 11041
-    },
-    {
-      "epoch": 0.7216521795961048,
-      "grad_norm": 0.4801838994026184,
-      "learning_rate": 8.801666955035317e-06,
-      "loss": 0.352,
-      "step": 11042
-    },
-    {
-      "epoch": 0.7217175348016469,
-      "grad_norm": 0.42381563782691956,
-      "learning_rate": 8.801440135327347e-06,
-      "loss": 0.3085,
-      "step": 11043
-    },
-    {
-      "epoch": 0.7217828900071891,
-      "grad_norm": 0.4279800355434418,
-      "learning_rate": 8.801213297078433e-06,
-      "loss": 0.2879,
-      "step": 11044
-    },
-    {
-      "epoch": 0.7218482452127312,
-      "grad_norm": 0.4466213285923004,
-      "learning_rate": 8.800986440289685e-06,
-      "loss": 0.4095,
-      "step": 11045
-    },
-    {
-      "epoch": 0.7219136004182733,
-      "grad_norm": 0.43376627564430237,
-      "learning_rate": 8.80075956496221e-06,
-      "loss": 0.3627,
-      "step": 11046
-    },
-    {
-      "epoch": 0.7219789556238154,
-      "grad_norm": 0.43545806407928467,
-      "learning_rate": 8.800532671097117e-06,
-      "loss": 0.3329,
-      "step": 11047
-    },
-    {
-      "epoch": 0.7220443108293576,
-      "grad_norm": 0.4581034481525421,
-      "learning_rate": 8.800305758695507e-06,
-      "loss": 0.395,
-      "step": 11048
-    },
-    {
-      "epoch": 0.7221096660348997,
-      "grad_norm": 0.44845613837242126,
-      "learning_rate": 8.80007882775849e-06,
-      "loss": 0.4128,
-      "step": 11049
-    },
-    {
-      "epoch": 0.7221750212404418,
-      "grad_norm": 0.4453403949737549,
-      "learning_rate": 8.79985187828717e-06,
-      "loss": 0.3796,
-      "step": 11050
-    },
-    {
-      "epoch": 0.7222403764459839,
-      "grad_norm": 0.42938685417175293,
-      "learning_rate": 8.799624910282658e-06,
-      "loss": 0.3521,
-      "step": 11051
-    },
-    {
-      "epoch": 0.722305731651526,
-      "grad_norm": 0.46365949511528015,
-      "learning_rate": 8.799397923746057e-06,
-      "loss": 0.4298,
-      "step": 11052
-    },
-    {
-      "epoch": 0.7223710868570682,
-      "grad_norm": 0.420899897813797,
-      "learning_rate": 8.799170918678479e-06,
-      "loss": 0.3483,
-      "step": 11053
-    },
-    {
-      "epoch": 0.7224364420626103,
-      "grad_norm": 0.4637686312198639,
-      "learning_rate": 8.798943895081026e-06,
-      "loss": 0.425,
-      "step": 11054
-    },
-    {
-      "epoch": 0.7225017972681524,
-      "grad_norm": 0.4429585635662079,
-      "learning_rate": 8.798716852954807e-06,
-      "loss": 0.3456,
-      "step": 11055
-    },
-    {
-      "epoch": 0.7225671524736945,
-      "grad_norm": 0.45941469073295593,
-      "learning_rate": 8.798489792300933e-06,
-      "loss": 0.3958,
-      "step": 11056
-    },
-    {
-      "epoch": 0.7226325076792367,
-      "grad_norm": 0.48170894384384155,
-      "learning_rate": 8.798262713120506e-06,
-      "loss": 0.4519,
-      "step": 11057
-    },
-    {
-      "epoch": 0.7226978628847788,
-      "grad_norm": 0.4276023209095001,
-      "learning_rate": 8.798035615414636e-06,
-      "loss": 0.3398,
-      "step": 11058
-    },
-    {
-      "epoch": 0.7227632180903208,
-      "grad_norm": 0.4634799361228943,
-      "learning_rate": 8.797808499184432e-06,
-      "loss": 0.3692,
-      "step": 11059
-    },
-    {
-      "epoch": 0.722828573295863,
-      "grad_norm": 0.4455621540546417,
-      "learning_rate": 8.797581364430997e-06,
-      "loss": 0.356,
-      "step": 11060
-    },
-    {
-      "epoch": 0.7228939285014051,
-      "grad_norm": 0.4130583107471466,
-      "learning_rate": 8.797354211155445e-06,
-      "loss": 0.3258,
-      "step": 11061
-    },
-    {
-      "epoch": 0.7229592837069473,
-      "grad_norm": 0.4679669737815857,
-      "learning_rate": 8.797127039358881e-06,
-      "loss": 0.3873,
-      "step": 11062
-    },
-    {
-      "epoch": 0.7230246389124894,
-      "grad_norm": 0.4282200038433075,
-      "learning_rate": 8.796899849042414e-06,
-      "loss": 0.3655,
-      "step": 11063
-    },
-    {
-      "epoch": 0.7230899941180315,
-      "grad_norm": 0.4634385108947754,
-      "learning_rate": 8.796672640207148e-06,
-      "loss": 0.3765,
-      "step": 11064
-    },
-    {
-      "epoch": 0.7231553493235736,
-      "grad_norm": 0.4974842667579651,
-      "learning_rate": 8.796445412854195e-06,
-      "loss": 0.4121,
-      "step": 11065
-    },
-    {
-      "epoch": 0.7232207045291158,
-      "grad_norm": 0.42820289731025696,
-      "learning_rate": 8.796218166984663e-06,
-      "loss": 0.3593,
-      "step": 11066
-    },
-    {
-      "epoch": 0.7232860597346579,
-      "grad_norm": 0.42884114384651184,
-      "learning_rate": 8.79599090259966e-06,
-      "loss": 0.3554,
-      "step": 11067
-    },
-    {
-      "epoch": 0.7233514149401999,
-      "grad_norm": 0.46795573830604553,
-      "learning_rate": 8.795763619700295e-06,
-      "loss": 0.409,
-      "step": 11068
-    },
-    {
-      "epoch": 0.7234167701457421,
-      "grad_norm": 0.48327428102493286,
-      "learning_rate": 8.795536318287674e-06,
-      "loss": 0.4499,
-      "step": 11069
-    },
-    {
-      "epoch": 0.7234821253512842,
-      "grad_norm": 0.4357799291610718,
-      "learning_rate": 8.795308998362909e-06,
-      "loss": 0.3592,
-      "step": 11070
-    },
-    {
-      "epoch": 0.7235474805568264,
-      "grad_norm": 0.45475080609321594,
-      "learning_rate": 8.795081659927108e-06,
-      "loss": 0.3872,
-      "step": 11071
-    },
-    {
-      "epoch": 0.7236128357623685,
-      "grad_norm": 0.43749937415122986,
-      "learning_rate": 8.794854302981376e-06,
-      "loss": 0.3966,
-      "step": 11072
-    },
-    {
-      "epoch": 0.7236781909679106,
-      "grad_norm": 0.4310551881790161,
-      "learning_rate": 8.794626927526829e-06,
-      "loss": 0.3256,
-      "step": 11073
-    },
-    {
-      "epoch": 0.7237435461734527,
-      "grad_norm": 0.42545050382614136,
-      "learning_rate": 8.794399533564569e-06,
-      "loss": 0.3265,
-      "step": 11074
-    },
-    {
-      "epoch": 0.7238089013789948,
-      "grad_norm": 0.48232147097587585,
-      "learning_rate": 8.794172121095708e-06,
-      "loss": 0.4247,
-      "step": 11075
-    },
-    {
-      "epoch": 0.723874256584537,
-      "grad_norm": 0.4450146555900574,
-      "learning_rate": 8.793944690121355e-06,
-      "loss": 0.3738,
-      "step": 11076
-    },
-    {
-      "epoch": 0.723939611790079,
-      "grad_norm": 0.4868414103984833,
-      "learning_rate": 8.793717240642621e-06,
-      "loss": 0.4165,
-      "step": 11077
-    },
-    {
-      "epoch": 0.7240049669956212,
-      "grad_norm": 0.4174163341522217,
-      "learning_rate": 8.793489772660613e-06,
-      "loss": 0.2987,
-      "step": 11078
-    },
-    {
-      "epoch": 0.7240703222011633,
-      "grad_norm": 0.4786052703857422,
-      "learning_rate": 8.793262286176441e-06,
-      "loss": 0.376,
-      "step": 11079
-    },
-    {
-      "epoch": 0.7241356774067055,
-      "grad_norm": 0.4679298996925354,
-      "learning_rate": 8.793034781191215e-06,
-      "loss": 0.3844,
-      "step": 11080
-    },
-    {
-      "epoch": 0.7242010326122476,
-      "grad_norm": 0.42517128586769104,
-      "learning_rate": 8.792807257706043e-06,
-      "loss": 0.3547,
-      "step": 11081
-    },
-    {
-      "epoch": 0.7242663878177897,
-      "grad_norm": 0.4264257848262787,
-      "learning_rate": 8.792579715722038e-06,
-      "loss": 0.3504,
-      "step": 11082
-    },
-    {
-      "epoch": 0.7243317430233318,
-      "grad_norm": 0.44104328751564026,
-      "learning_rate": 8.792352155240307e-06,
-      "loss": 0.3861,
-      "step": 11083
-    },
-    {
-      "epoch": 0.7243970982288739,
-      "grad_norm": 0.46385055780410767,
-      "learning_rate": 8.79212457626196e-06,
-      "loss": 0.387,
-      "step": 11084
-    },
-    {
-      "epoch": 0.7244624534344161,
-      "grad_norm": 0.4661731421947479,
-      "learning_rate": 8.791896978788108e-06,
-      "loss": 0.3996,
-      "step": 11085
-    },
-    {
-      "epoch": 0.7245278086399581,
-      "grad_norm": 0.4316199719905853,
-      "learning_rate": 8.79166936281986e-06,
-      "loss": 0.3463,
-      "step": 11086
-    },
-    {
-      "epoch": 0.7245931638455003,
-      "grad_norm": 0.43848204612731934,
-      "learning_rate": 8.79144172835833e-06,
-      "loss": 0.3674,
-      "step": 11087
-    },
-    {
-      "epoch": 0.7246585190510424,
-      "grad_norm": 0.4132622182369232,
-      "learning_rate": 8.791214075404623e-06,
-      "loss": 0.377,
-      "step": 11088
-    },
-    {
-      "epoch": 0.7247238742565846,
-      "grad_norm": 0.44539204239845276,
-      "learning_rate": 8.790986403959851e-06,
-      "loss": 0.3695,
-      "step": 11089
-    },
-    {
-      "epoch": 0.7247892294621266,
-      "grad_norm": 0.4502166211605072,
-      "learning_rate": 8.790758714025128e-06,
-      "loss": 0.3776,
-      "step": 11090
-    },
-    {
-      "epoch": 0.7248545846676688,
-      "grad_norm": 0.4646752178668976,
-      "learning_rate": 8.790531005601559e-06,
-      "loss": 0.4114,
-      "step": 11091
-    },
-    {
-      "epoch": 0.7249199398732109,
-      "grad_norm": 0.46941882371902466,
-      "learning_rate": 8.790303278690258e-06,
-      "loss": 0.4163,
-      "step": 11092
-    },
-    {
-      "epoch": 0.724985295078753,
-      "grad_norm": 0.45404377579689026,
-      "learning_rate": 8.790075533292332e-06,
-      "loss": 0.402,
-      "step": 11093
-    },
-    {
-      "epoch": 0.7250506502842952,
-      "grad_norm": 0.5291372537612915,
-      "learning_rate": 8.789847769408898e-06,
-      "loss": 0.4224,
-      "step": 11094
-    },
-    {
-      "epoch": 0.7251160054898372,
-      "grad_norm": 0.43661803007125854,
-      "learning_rate": 8.789619987041063e-06,
-      "loss": 0.3662,
-      "step": 11095
-    },
-    {
-      "epoch": 0.7251813606953794,
-      "grad_norm": 0.43203383684158325,
-      "learning_rate": 8.789392186189938e-06,
-      "loss": 0.3522,
-      "step": 11096
-    },
-    {
-      "epoch": 0.7252467159009215,
-      "grad_norm": 0.45868775248527527,
-      "learning_rate": 8.789164366856634e-06,
-      "loss": 0.3845,
-      "step": 11097
-    },
-    {
-      "epoch": 0.7253120711064637,
-      "grad_norm": 0.445512980222702,
-      "learning_rate": 8.788936529042264e-06,
-      "loss": 0.349,
-      "step": 11098
-    },
-    {
-      "epoch": 0.7253774263120057,
-      "grad_norm": 0.41102826595306396,
-      "learning_rate": 8.788708672747937e-06,
-      "loss": 0.3229,
-      "step": 11099
-    },
-    {
-      "epoch": 0.7254427815175478,
-      "grad_norm": 0.47712117433547974,
-      "learning_rate": 8.788480797974765e-06,
-      "loss": 0.383,
-      "step": 11100
-    },
-    {
-      "epoch": 0.72550813672309,
-      "grad_norm": 0.4242284297943115,
-      "learning_rate": 8.78825290472386e-06,
-      "loss": 0.3781,
-      "step": 11101
-    },
-    {
-      "epoch": 0.7255734919286321,
-      "grad_norm": 0.47264206409454346,
-      "learning_rate": 8.788024992996333e-06,
-      "loss": 0.4067,
-      "step": 11102
-    },
-    {
-      "epoch": 0.7256388471341743,
-      "grad_norm": 0.44981124997138977,
-      "learning_rate": 8.787797062793298e-06,
-      "loss": 0.3509,
-      "step": 11103
-    },
-    {
-      "epoch": 0.7257042023397163,
-      "grad_norm": 0.47995010018348694,
-      "learning_rate": 8.787569114115862e-06,
-      "loss": 0.4037,
-      "step": 11104
-    },
-    {
-      "epoch": 0.7257695575452585,
-      "grad_norm": 0.433910995721817,
-      "learning_rate": 8.78734114696514e-06,
-      "loss": 0.3509,
-      "step": 11105
-    },
-    {
-      "epoch": 0.7258349127508006,
-      "grad_norm": 0.44411203265190125,
-      "learning_rate": 8.787113161342243e-06,
-      "loss": 0.3583,
-      "step": 11106
-    },
-    {
-      "epoch": 0.7259002679563428,
-      "grad_norm": 0.45076310634613037,
-      "learning_rate": 8.786885157248281e-06,
-      "loss": 0.4022,
-      "step": 11107
-    },
-    {
-      "epoch": 0.7259656231618848,
-      "grad_norm": 0.48805731534957886,
-      "learning_rate": 8.78665713468437e-06,
-      "loss": 0.4192,
-      "step": 11108
-    },
-    {
-      "epoch": 0.7260309783674269,
-      "grad_norm": 0.4489264190196991,
-      "learning_rate": 8.786429093651622e-06,
-      "loss": 0.3725,
-      "step": 11109
-    },
-    {
-      "epoch": 0.7260963335729691,
-      "grad_norm": 0.46248388290405273,
-      "learning_rate": 8.786201034151147e-06,
-      "loss": 0.4336,
-      "step": 11110
-    },
-    {
-      "epoch": 0.7261616887785112,
-      "grad_norm": 0.45819878578186035,
-      "learning_rate": 8.785972956184056e-06,
-      "loss": 0.4106,
-      "step": 11111
-    },
-    {
-      "epoch": 0.7262270439840534,
-      "grad_norm": 0.4444325566291809,
-      "learning_rate": 8.785744859751465e-06,
-      "loss": 0.3858,
-      "step": 11112
-    },
-    {
-      "epoch": 0.7262923991895954,
-      "grad_norm": 0.4306245744228363,
-      "learning_rate": 8.785516744854485e-06,
-      "loss": 0.321,
-      "step": 11113
-    },
-    {
-      "epoch": 0.7263577543951376,
-      "grad_norm": 0.44880321621894836,
-      "learning_rate": 8.785288611494227e-06,
-      "loss": 0.3481,
-      "step": 11114
-    },
-    {
-      "epoch": 0.7264231096006797,
-      "grad_norm": 0.44093531370162964,
-      "learning_rate": 8.785060459671806e-06,
-      "loss": 0.3624,
-      "step": 11115
-    },
-    {
-      "epoch": 0.7264884648062219,
-      "grad_norm": 0.43183866143226624,
-      "learning_rate": 8.784832289388334e-06,
-      "loss": 0.356,
-      "step": 11116
-    },
-    {
-      "epoch": 0.7265538200117639,
-      "grad_norm": 0.4692407548427582,
-      "learning_rate": 8.784604100644922e-06,
-      "loss": 0.3844,
-      "step": 11117
-    },
-    {
-      "epoch": 0.726619175217306,
-      "grad_norm": 0.420350044965744,
-      "learning_rate": 8.784375893442687e-06,
-      "loss": 0.3425,
-      "step": 11118
-    },
-    {
-      "epoch": 0.7266845304228482,
-      "grad_norm": 0.44382867217063904,
-      "learning_rate": 8.784147667782739e-06,
-      "loss": 0.3661,
-      "step": 11119
-    },
-    {
-      "epoch": 0.7267498856283903,
-      "grad_norm": 0.45902663469314575,
-      "learning_rate": 8.783919423666191e-06,
-      "loss": 0.381,
-      "step": 11120
-    },
-    {
-      "epoch": 0.7268152408339325,
-      "grad_norm": 0.42323535680770874,
-      "learning_rate": 8.78369116109416e-06,
-      "loss": 0.3179,
-      "step": 11121
-    },
-    {
-      "epoch": 0.7268805960394745,
-      "grad_norm": 0.43687260150909424,
-      "learning_rate": 8.783462880067753e-06,
-      "loss": 0.386,
-      "step": 11122
-    },
-    {
-      "epoch": 0.7269459512450167,
-      "grad_norm": 0.4666963815689087,
-      "learning_rate": 8.783234580588089e-06,
-      "loss": 0.4866,
-      "step": 11123
-    },
-    {
-      "epoch": 0.7270113064505588,
-      "grad_norm": 0.4805375635623932,
-      "learning_rate": 8.783006262656277e-06,
-      "loss": 0.3845,
-      "step": 11124
-    },
-    {
-      "epoch": 0.7270766616561009,
-      "grad_norm": 0.444644033908844,
-      "learning_rate": 8.782777926273434e-06,
-      "loss": 0.4082,
-      "step": 11125
-    },
-    {
-      "epoch": 0.727142016861643,
-      "grad_norm": 0.4642644226551056,
-      "learning_rate": 8.782549571440673e-06,
-      "loss": 0.4384,
-      "step": 11126
-    },
-    {
-      "epoch": 0.7272073720671851,
-      "grad_norm": 0.4263255298137665,
-      "learning_rate": 8.782321198159107e-06,
-      "loss": 0.3425,
-      "step": 11127
-    },
-    {
-      "epoch": 0.7272727272727273,
-      "grad_norm": 0.41924771666526794,
-      "learning_rate": 8.78209280642985e-06,
-      "loss": 0.3759,
-      "step": 11128
-    },
-    {
-      "epoch": 0.7273380824782694,
-      "grad_norm": 0.46761152148246765,
-      "learning_rate": 8.781864396254016e-06,
-      "loss": 0.3966,
-      "step": 11129
-    },
-    {
-      "epoch": 0.7274034376838115,
-      "grad_norm": 0.4359188377857208,
-      "learning_rate": 8.78163596763272e-06,
-      "loss": 0.3722,
-      "step": 11130
-    },
-    {
-      "epoch": 0.7274687928893536,
-      "grad_norm": 0.4260074496269226,
-      "learning_rate": 8.781407520567076e-06,
-      "loss": 0.3709,
-      "step": 11131
-    },
-    {
-      "epoch": 0.7275341480948958,
-      "grad_norm": 0.43425846099853516,
-      "learning_rate": 8.781179055058196e-06,
-      "loss": 0.3514,
-      "step": 11132
-    },
-    {
-      "epoch": 0.7275995033004379,
-      "grad_norm": 0.45331141352653503,
-      "learning_rate": 8.780950571107197e-06,
-      "loss": 0.3464,
-      "step": 11133
-    },
-    {
-      "epoch": 0.72766485850598,
-      "grad_norm": 0.45970654487609863,
-      "learning_rate": 8.780722068715191e-06,
-      "loss": 0.3853,
-      "step": 11134
-    },
-    {
-      "epoch": 0.7277302137115221,
-      "grad_norm": 0.46538349986076355,
-      "learning_rate": 8.780493547883293e-06,
-      "loss": 0.4237,
-      "step": 11135
-    },
-    {
-      "epoch": 0.7277955689170642,
-      "grad_norm": 0.5452269911766052,
-      "learning_rate": 8.780265008612621e-06,
-      "loss": 0.3933,
-      "step": 11136
-    },
-    {
-      "epoch": 0.7278609241226064,
-      "grad_norm": 0.42780470848083496,
-      "learning_rate": 8.780036450904285e-06,
-      "loss": 0.3543,
-      "step": 11137
-    },
-    {
-      "epoch": 0.7279262793281485,
-      "grad_norm": 0.4398547410964966,
-      "learning_rate": 8.779807874759403e-06,
-      "loss": 0.3589,
-      "step": 11138
-    },
-    {
-      "epoch": 0.7279916345336906,
-      "grad_norm": 0.43431010842323303,
-      "learning_rate": 8.779579280179087e-06,
-      "loss": 0.3569,
-      "step": 11139
-    },
-    {
-      "epoch": 0.7280569897392327,
-      "grad_norm": 0.4678983986377716,
-      "learning_rate": 8.779350667164453e-06,
-      "loss": 0.4278,
-      "step": 11140
-    },
-    {
-      "epoch": 0.7281223449447749,
-      "grad_norm": 0.41606494784355164,
-      "learning_rate": 8.779122035716619e-06,
-      "loss": 0.3293,
-      "step": 11141
-    },
-    {
-      "epoch": 0.728187700150317,
-      "grad_norm": 0.437587708234787,
-      "learning_rate": 8.778893385836695e-06,
-      "loss": 0.3377,
-      "step": 11142
-    },
-    {
-      "epoch": 0.728253055355859,
-      "grad_norm": 0.4146372973918915,
-      "learning_rate": 8.778664717525802e-06,
-      "loss": 0.3291,
-      "step": 11143
-    },
-    {
-      "epoch": 0.7283184105614012,
-      "grad_norm": 0.4122779965400696,
-      "learning_rate": 8.778436030785049e-06,
-      "loss": 0.3035,
-      "step": 11144
-    },
-    {
-      "epoch": 0.7283837657669433,
-      "grad_norm": 0.4226057231426239,
-      "learning_rate": 8.778207325615556e-06,
-      "loss": 0.3606,
-      "step": 11145
-    },
-    {
-      "epoch": 0.7284491209724855,
-      "grad_norm": 0.46493902802467346,
-      "learning_rate": 8.777978602018436e-06,
-      "loss": 0.4174,
-      "step": 11146
-    },
-    {
-      "epoch": 0.7285144761780276,
-      "grad_norm": 0.4412344992160797,
-      "learning_rate": 8.777749859994806e-06,
-      "loss": 0.3742,
-      "step": 11147
-    },
-    {
-      "epoch": 0.7285798313835697,
-      "grad_norm": 0.4301016330718994,
-      "learning_rate": 8.777521099545783e-06,
-      "loss": 0.3946,
-      "step": 11148
-    },
-    {
-      "epoch": 0.7286451865891118,
-      "grad_norm": 0.43630465865135193,
-      "learning_rate": 8.777292320672479e-06,
-      "loss": 0.3496,
-      "step": 11149
-    },
-    {
-      "epoch": 0.728710541794654,
-      "grad_norm": 0.46967384219169617,
-      "learning_rate": 8.777063523376012e-06,
-      "loss": 0.4097,
-      "step": 11150
-    },
-    {
-      "epoch": 0.7287758970001961,
-      "grad_norm": 0.46150636672973633,
-      "learning_rate": 8.776834707657498e-06,
-      "loss": 0.4559,
-      "step": 11151
-    },
-    {
-      "epoch": 0.7288412522057381,
-      "grad_norm": 0.4512837529182434,
-      "learning_rate": 8.776605873518052e-06,
-      "loss": 0.3979,
-      "step": 11152
-    },
-    {
-      "epoch": 0.7289066074112803,
-      "grad_norm": 0.4495929777622223,
-      "learning_rate": 8.77637702095879e-06,
-      "loss": 0.3379,
-      "step": 11153
-    },
-    {
-      "epoch": 0.7289719626168224,
-      "grad_norm": 0.4882166087627411,
-      "learning_rate": 8.776148149980833e-06,
-      "loss": 0.4665,
-      "step": 11154
-    },
-    {
-      "epoch": 0.7290373178223646,
-      "grad_norm": 0.4099055230617523,
-      "learning_rate": 8.775919260585289e-06,
-      "loss": 0.3413,
-      "step": 11155
-    },
-    {
-      "epoch": 0.7291026730279067,
-      "grad_norm": 0.465614914894104,
-      "learning_rate": 8.77569035277328e-06,
-      "loss": 0.4076,
-      "step": 11156
-    },
-    {
-      "epoch": 0.7291680282334488,
-      "grad_norm": 0.42142704129219055,
-      "learning_rate": 8.775461426545922e-06,
-      "loss": 0.3596,
-      "step": 11157
-    },
-    {
-      "epoch": 0.7292333834389909,
-      "grad_norm": 0.4787689447402954,
-      "learning_rate": 8.77523248190433e-06,
-      "loss": 0.4263,
-      "step": 11158
-    },
-    {
-      "epoch": 0.729298738644533,
-      "grad_norm": 0.4259145259857178,
-      "learning_rate": 8.775003518849622e-06,
-      "loss": 0.3228,
-      "step": 11159
-    },
-    {
-      "epoch": 0.7293640938500752,
-      "grad_norm": 0.4332257807254791,
-      "learning_rate": 8.774774537382913e-06,
-      "loss": 0.3498,
-      "step": 11160
-    },
-    {
-      "epoch": 0.7294294490556172,
-      "grad_norm": 0.4691733121871948,
-      "learning_rate": 8.774545537505321e-06,
-      "loss": 0.4267,
-      "step": 11161
-    },
-    {
-      "epoch": 0.7294948042611594,
-      "grad_norm": 0.45611223578453064,
-      "learning_rate": 8.774316519217963e-06,
-      "loss": 0.3928,
-      "step": 11162
-    },
-    {
-      "epoch": 0.7295601594667015,
-      "grad_norm": 0.44422146677970886,
-      "learning_rate": 8.774087482521955e-06,
-      "loss": 0.3437,
-      "step": 11163
-    },
-    {
-      "epoch": 0.7296255146722437,
-      "grad_norm": 0.4369373321533203,
-      "learning_rate": 8.773858427418417e-06,
-      "loss": 0.3521,
-      "step": 11164
-    },
-    {
-      "epoch": 0.7296908698777858,
-      "grad_norm": 0.45053038001060486,
-      "learning_rate": 8.773629353908463e-06,
-      "loss": 0.3761,
-      "step": 11165
-    },
-    {
-      "epoch": 0.7297562250833279,
-      "grad_norm": 0.44090405106544495,
-      "learning_rate": 8.773400261993211e-06,
-      "loss": 0.388,
-      "step": 11166
-    },
-    {
-      "epoch": 0.72982158028887,
-      "grad_norm": 0.4256582260131836,
-      "learning_rate": 8.77317115167378e-06,
-      "loss": 0.354,
-      "step": 11167
-    },
-    {
-      "epoch": 0.7298869354944121,
-      "grad_norm": 0.4984571933746338,
-      "learning_rate": 8.772942022951285e-06,
-      "loss": 0.4391,
-      "step": 11168
-    },
-    {
-      "epoch": 0.7299522906999543,
-      "grad_norm": 0.46019721031188965,
-      "learning_rate": 8.772712875826842e-06,
-      "loss": 0.3867,
-      "step": 11169
-    },
-    {
-      "epoch": 0.7300176459054963,
-      "grad_norm": 0.5745295882225037,
-      "learning_rate": 8.772483710301577e-06,
-      "loss": 0.3707,
-      "step": 11170
-    },
-    {
-      "epoch": 0.7300830011110385,
-      "grad_norm": 0.4176986515522003,
-      "learning_rate": 8.772254526376599e-06,
-      "loss": 0.351,
-      "step": 11171
-    },
-    {
-      "epoch": 0.7301483563165806,
-      "grad_norm": 0.4469206929206848,
-      "learning_rate": 8.772025324053027e-06,
-      "loss": 0.413,
-      "step": 11172
-    },
-    {
-      "epoch": 0.7302137115221228,
-      "grad_norm": 0.48123699426651,
-      "learning_rate": 8.771796103331984e-06,
-      "loss": 0.4142,
-      "step": 11173
-    },
-    {
-      "epoch": 0.7302790667276648,
-      "grad_norm": 0.41997450590133667,
-      "learning_rate": 8.771566864214583e-06,
-      "loss": 0.3341,
-      "step": 11174
-    },
-    {
-      "epoch": 0.730344421933207,
-      "grad_norm": 0.4403407573699951,
-      "learning_rate": 8.771337606701944e-06,
-      "loss": 0.374,
-      "step": 11175
-    },
-    {
-      "epoch": 0.7304097771387491,
-      "grad_norm": 0.4437764286994934,
-      "learning_rate": 8.771108330795185e-06,
-      "loss": 0.3682,
-      "step": 11176
-    },
-    {
-      "epoch": 0.7304751323442912,
-      "grad_norm": 0.41584858298301697,
-      "learning_rate": 8.770879036495424e-06,
-      "loss": 0.3315,
-      "step": 11177
-    },
-    {
-      "epoch": 0.7305404875498334,
-      "grad_norm": 0.43823719024658203,
-      "learning_rate": 8.77064972380378e-06,
-      "loss": 0.3654,
-      "step": 11178
-    },
-    {
-      "epoch": 0.7306058427553754,
-      "grad_norm": 0.43495121598243713,
-      "learning_rate": 8.770420392721372e-06,
-      "loss": 0.3459,
-      "step": 11179
-    },
-    {
-      "epoch": 0.7306711979609176,
-      "grad_norm": 0.46856689453125,
-      "learning_rate": 8.770191043249316e-06,
-      "loss": 0.4379,
-      "step": 11180
-    },
-    {
-      "epoch": 0.7307365531664597,
-      "grad_norm": 0.4437059462070465,
-      "learning_rate": 8.769961675388731e-06,
-      "loss": 0.3853,
-      "step": 11181
-    },
-    {
-      "epoch": 0.7308019083720019,
-      "grad_norm": 0.43872666358947754,
-      "learning_rate": 8.76973228914074e-06,
-      "loss": 0.3891,
-      "step": 11182
-    },
-    {
-      "epoch": 0.730867263577544,
-      "grad_norm": 0.42102178931236267,
-      "learning_rate": 8.769502884506457e-06,
-      "loss": 0.3451,
-      "step": 11183
-    },
-    {
-      "epoch": 0.730932618783086,
-      "grad_norm": 0.4101921617984772,
-      "learning_rate": 8.769273461487003e-06,
-      "loss": 0.3137,
-      "step": 11184
-    },
-    {
-      "epoch": 0.7309979739886282,
-      "grad_norm": 0.45409631729125977,
-      "learning_rate": 8.769044020083497e-06,
-      "loss": 0.4182,
-      "step": 11185
-    },
-    {
-      "epoch": 0.7310633291941703,
-      "grad_norm": 0.4334227740764618,
-      "learning_rate": 8.768814560297056e-06,
-      "loss": 0.3833,
-      "step": 11186
-    },
-    {
-      "epoch": 0.7311286843997125,
-      "grad_norm": 0.4318891763687134,
-      "learning_rate": 8.768585082128802e-06,
-      "loss": 0.3325,
-      "step": 11187
-    },
-    {
-      "epoch": 0.7311940396052545,
-      "grad_norm": 0.4427895247936249,
-      "learning_rate": 8.768355585579852e-06,
-      "loss": 0.3587,
-      "step": 11188
-    },
-    {
-      "epoch": 0.7312593948107967,
-      "grad_norm": 0.4658011794090271,
-      "learning_rate": 8.768126070651328e-06,
-      "loss": 0.3791,
-      "step": 11189
-    },
-    {
-      "epoch": 0.7313247500163388,
-      "grad_norm": 0.4469871520996094,
-      "learning_rate": 8.767896537344346e-06,
-      "loss": 0.3529,
-      "step": 11190
-    },
-    {
-      "epoch": 0.731390105221881,
-      "grad_norm": 0.4537349343299866,
-      "learning_rate": 8.767666985660027e-06,
-      "loss": 0.3733,
-      "step": 11191
-    },
-    {
-      "epoch": 0.731455460427423,
-      "grad_norm": 0.6510783433914185,
-      "learning_rate": 8.767437415599493e-06,
-      "loss": 0.4483,
-      "step": 11192
-    },
-    {
-      "epoch": 0.7315208156329651,
-      "grad_norm": 0.42360860109329224,
-      "learning_rate": 8.76720782716386e-06,
-      "loss": 0.325,
-      "step": 11193
-    },
-    {
-      "epoch": 0.7315861708385073,
-      "grad_norm": 0.4338391423225403,
-      "learning_rate": 8.76697822035425e-06,
-      "loss": 0.3768,
-      "step": 11194
-    },
-    {
-      "epoch": 0.7316515260440494,
-      "grad_norm": 0.424140065908432,
-      "learning_rate": 8.766748595171783e-06,
-      "loss": 0.3347,
-      "step": 11195
-    },
-    {
-      "epoch": 0.7317168812495916,
-      "grad_norm": 0.42875343561172485,
-      "learning_rate": 8.766518951617576e-06,
-      "loss": 0.3437,
-      "step": 11196
-    },
-    {
-      "epoch": 0.7317822364551336,
-      "grad_norm": 0.46043136715888977,
-      "learning_rate": 8.766289289692753e-06,
-      "loss": 0.4028,
-      "step": 11197
-    },
-    {
-      "epoch": 0.7318475916606758,
-      "grad_norm": 0.4252198338508606,
-      "learning_rate": 8.766059609398432e-06,
-      "loss": 0.3685,
-      "step": 11198
-    },
-    {
-      "epoch": 0.7319129468662179,
-      "grad_norm": 0.4179631769657135,
-      "learning_rate": 8.765829910735733e-06,
-      "loss": 0.35,
-      "step": 11199
-    },
-    {
-      "epoch": 0.7319783020717601,
-      "grad_norm": 0.41065219044685364,
-      "learning_rate": 8.765600193705777e-06,
-      "loss": 0.384,
-      "step": 11200
-    },
-    {
-      "epoch": 0.7320436572773021,
-      "grad_norm": 0.41988229751586914,
-      "learning_rate": 8.765370458309684e-06,
-      "loss": 0.3401,
-      "step": 11201
-    },
-    {
-      "epoch": 0.7321090124828442,
-      "grad_norm": 0.42483600974082947,
-      "learning_rate": 8.765140704548576e-06,
-      "loss": 0.3849,
-      "step": 11202
-    },
-    {
-      "epoch": 0.7321743676883864,
-      "grad_norm": 0.4606691896915436,
-      "learning_rate": 8.76491093242357e-06,
-      "loss": 0.3649,
-      "step": 11203
-    },
-    {
-      "epoch": 0.7322397228939285,
-      "grad_norm": 0.4374525547027588,
-      "learning_rate": 8.764681141935792e-06,
-      "loss": 0.3538,
-      "step": 11204
-    },
-    {
-      "epoch": 0.7323050780994707,
-      "grad_norm": 0.4549223780632019,
-      "learning_rate": 8.764451333086358e-06,
-      "loss": 0.4093,
-      "step": 11205
-    },
-    {
-      "epoch": 0.7323704333050127,
-      "grad_norm": 0.42030951380729675,
-      "learning_rate": 8.764221505876393e-06,
-      "loss": 0.351,
-      "step": 11206
-    },
-    {
-      "epoch": 0.7324357885105549,
-      "grad_norm": 0.4788500964641571,
-      "learning_rate": 8.763991660307014e-06,
-      "loss": 0.4289,
-      "step": 11207
-    },
-    {
-      "epoch": 0.732501143716097,
-      "grad_norm": 0.46380653977394104,
-      "learning_rate": 8.763761796379343e-06,
-      "loss": 0.4169,
-      "step": 11208
-    },
-    {
-      "epoch": 0.732566498921639,
-      "grad_norm": 0.4581639766693115,
-      "learning_rate": 8.763531914094502e-06,
-      "loss": 0.4127,
-      "step": 11209
-    },
-    {
-      "epoch": 0.7326318541271812,
-      "grad_norm": 0.45701515674591064,
-      "learning_rate": 8.763302013453614e-06,
-      "loss": 0.4049,
-      "step": 11210
-    },
-    {
-      "epoch": 0.7326972093327233,
-      "grad_norm": 0.46352502703666687,
-      "learning_rate": 8.763072094457797e-06,
-      "loss": 0.3964,
-      "step": 11211
-    },
-    {
-      "epoch": 0.7327625645382655,
-      "grad_norm": 0.4479861557483673,
-      "learning_rate": 8.762842157108173e-06,
-      "loss": 0.3951,
-      "step": 11212
-    },
-    {
-      "epoch": 0.7328279197438076,
-      "grad_norm": 0.40743589401245117,
-      "learning_rate": 8.762612201405865e-06,
-      "loss": 0.3412,
-      "step": 11213
-    },
-    {
-      "epoch": 0.7328932749493497,
-      "grad_norm": 0.4625330865383148,
-      "learning_rate": 8.762382227351995e-06,
-      "loss": 0.3849,
-      "step": 11214
-    },
-    {
-      "epoch": 0.7329586301548918,
-      "grad_norm": 0.4576355814933777,
-      "learning_rate": 8.76215223494768e-06,
-      "loss": 0.4092,
-      "step": 11215
-    },
-    {
-      "epoch": 0.733023985360434,
-      "grad_norm": 0.4445648789405823,
-      "learning_rate": 8.76192222419405e-06,
-      "loss": 0.39,
-      "step": 11216
-    },
-    {
-      "epoch": 0.7330893405659761,
-      "grad_norm": 0.47072356939315796,
-      "learning_rate": 8.761692195092219e-06,
-      "loss": 0.388,
-      "step": 11217
-    },
-    {
-      "epoch": 0.7331546957715182,
-      "grad_norm": 0.4482540488243103,
-      "learning_rate": 8.761462147643311e-06,
-      "loss": 0.3848,
-      "step": 11218
-    },
-    {
-      "epoch": 0.7332200509770603,
-      "grad_norm": 0.4511313736438751,
-      "learning_rate": 8.761232081848452e-06,
-      "loss": 0.3986,
-      "step": 11219
-    },
-    {
-      "epoch": 0.7332854061826024,
-      "grad_norm": 0.45848849415779114,
-      "learning_rate": 8.761001997708759e-06,
-      "loss": 0.3967,
-      "step": 11220
-    },
-    {
-      "epoch": 0.7333507613881446,
-      "grad_norm": 0.4291442930698395,
-      "learning_rate": 8.760771895225358e-06,
-      "loss": 0.3428,
-      "step": 11221
-    },
-    {
-      "epoch": 0.7334161165936867,
-      "grad_norm": 0.4486805498600006,
-      "learning_rate": 8.760541774399368e-06,
-      "loss": 0.3835,
-      "step": 11222
-    },
-    {
-      "epoch": 0.7334814717992288,
-      "grad_norm": 0.40763527154922485,
-      "learning_rate": 8.760311635231913e-06,
-      "loss": 0.3184,
-      "step": 11223
-    },
-    {
-      "epoch": 0.7335468270047709,
-      "grad_norm": 0.44409674406051636,
-      "learning_rate": 8.760081477724116e-06,
-      "loss": 0.3848,
-      "step": 11224
-    },
-    {
-      "epoch": 0.7336121822103131,
-      "grad_norm": 0.4710194766521454,
-      "learning_rate": 8.7598513018771e-06,
-      "loss": 0.3996,
-      "step": 11225
-    },
-    {
-      "epoch": 0.7336775374158552,
-      "grad_norm": 0.4506535530090332,
-      "learning_rate": 8.759621107691985e-06,
-      "loss": 0.3897,
-      "step": 11226
-    },
-    {
-      "epoch": 0.7337428926213972,
-      "grad_norm": 0.508561909198761,
-      "learning_rate": 8.759390895169896e-06,
-      "loss": 0.4555,
-      "step": 11227
-    },
-    {
-      "epoch": 0.7338082478269394,
-      "grad_norm": 0.4735959768295288,
-      "learning_rate": 8.759160664311957e-06,
-      "loss": 0.4099,
-      "step": 11228
-    },
-    {
-      "epoch": 0.7338736030324815,
-      "grad_norm": 0.46254050731658936,
-      "learning_rate": 8.758930415119286e-06,
-      "loss": 0.4225,
-      "step": 11229
-    },
-    {
-      "epoch": 0.7339389582380237,
-      "grad_norm": 0.446707546710968,
-      "learning_rate": 8.75870014759301e-06,
-      "loss": 0.3645,
-      "step": 11230
-    },
-    {
-      "epoch": 0.7340043134435658,
-      "grad_norm": 0.4319542348384857,
-      "learning_rate": 8.758469861734252e-06,
-      "loss": 0.3557,
-      "step": 11231
-    },
-    {
-      "epoch": 0.7340696686491079,
-      "grad_norm": 0.47102925181388855,
-      "learning_rate": 8.758239557544135e-06,
-      "loss": 0.3969,
-      "step": 11232
-    },
-    {
-      "epoch": 0.73413502385465,
-      "grad_norm": 0.434438556432724,
-      "learning_rate": 8.758009235023782e-06,
-      "loss": 0.3691,
-      "step": 11233
-    },
-    {
-      "epoch": 0.7342003790601922,
-      "grad_norm": 0.5033477544784546,
-      "learning_rate": 8.757778894174314e-06,
-      "loss": 0.4246,
-      "step": 11234
-    },
-    {
-      "epoch": 0.7342657342657343,
-      "grad_norm": 0.42868471145629883,
-      "learning_rate": 8.757548534996858e-06,
-      "loss": 0.3827,
-      "step": 11235
-    },
-    {
-      "epoch": 0.7343310894712763,
-      "grad_norm": 0.47100287675857544,
-      "learning_rate": 8.757318157492535e-06,
-      "loss": 0.4058,
-      "step": 11236
-    },
-    {
-      "epoch": 0.7343964446768185,
-      "grad_norm": 0.5234546661376953,
-      "learning_rate": 8.75708776166247e-06,
-      "loss": 0.4786,
-      "step": 11237
-    },
-    {
-      "epoch": 0.7344617998823606,
-      "grad_norm": 0.4534815847873688,
-      "learning_rate": 8.756857347507787e-06,
-      "loss": 0.4056,
-      "step": 11238
-    },
-    {
-      "epoch": 0.7345271550879028,
-      "grad_norm": 0.49612924456596375,
-      "learning_rate": 8.75662691502961e-06,
-      "loss": 0.3897,
-      "step": 11239
-    },
-    {
-      "epoch": 0.7345925102934449,
-      "grad_norm": 0.4975704252719879,
-      "learning_rate": 8.75639646422906e-06,
-      "loss": 0.3943,
-      "step": 11240
-    },
-    {
-      "epoch": 0.734657865498987,
-      "grad_norm": 0.4549933671951294,
-      "learning_rate": 8.756165995107265e-06,
-      "loss": 0.4065,
-      "step": 11241
-    },
-    {
-      "epoch": 0.7347232207045291,
-      "grad_norm": 0.4527456760406494,
-      "learning_rate": 8.755935507665346e-06,
-      "loss": 0.3533,
-      "step": 11242
-    },
-    {
-      "epoch": 0.7347885759100712,
-      "grad_norm": 0.4188726544380188,
-      "learning_rate": 8.755705001904428e-06,
-      "loss": 0.3398,
-      "step": 11243
-    },
-    {
-      "epoch": 0.7348539311156134,
-      "grad_norm": 0.4060506820678711,
-      "learning_rate": 8.755474477825636e-06,
-      "loss": 0.3237,
-      "step": 11244
-    },
-    {
-      "epoch": 0.7349192863211554,
-      "grad_norm": 0.4342059791088104,
-      "learning_rate": 8.755243935430095e-06,
-      "loss": 0.3419,
-      "step": 11245
-    },
-    {
-      "epoch": 0.7349846415266976,
-      "grad_norm": 0.5011023283004761,
-      "learning_rate": 8.755013374718928e-06,
-      "loss": 0.4923,
-      "step": 11246
-    },
-    {
-      "epoch": 0.7350499967322397,
-      "grad_norm": 1.7879060506820679,
-      "learning_rate": 8.75478279569326e-06,
-      "loss": 0.425,
-      "step": 11247
-    },
-    {
-      "epoch": 0.7351153519377819,
-      "grad_norm": 0.44676414132118225,
-      "learning_rate": 8.754552198354214e-06,
-      "loss": 0.3395,
-      "step": 11248
-    },
-    {
-      "epoch": 0.735180707143324,
-      "grad_norm": 0.4825112223625183,
-      "learning_rate": 8.754321582702917e-06,
-      "loss": 0.4456,
-      "step": 11249
-    },
-    {
-      "epoch": 0.7352460623488661,
-      "grad_norm": 0.47368019819259644,
-      "learning_rate": 8.754090948740494e-06,
-      "loss": 0.421,
-      "step": 11250
-    },
-    {
-      "epoch": 0.7353114175544082,
-      "grad_norm": 0.4699123501777649,
-      "learning_rate": 8.753860296468069e-06,
-      "loss": 0.403,
-      "step": 11251
-    },
-    {
-      "epoch": 0.7353767727599503,
-      "grad_norm": 0.41633594036102295,
-      "learning_rate": 8.753629625886764e-06,
-      "loss": 0.3443,
-      "step": 11252
-    },
-    {
-      "epoch": 0.7354421279654925,
-      "grad_norm": 0.4335797429084778,
-      "learning_rate": 8.75339893699771e-06,
-      "loss": 0.3668,
-      "step": 11253
-    },
-    {
-      "epoch": 0.7355074831710345,
-      "grad_norm": 0.43298983573913574,
-      "learning_rate": 8.753168229802028e-06,
-      "loss": 0.3585,
-      "step": 11254
-    },
-    {
-      "epoch": 0.7355728383765767,
-      "grad_norm": 0.42844730615615845,
-      "learning_rate": 8.752937504300845e-06,
-      "loss": 0.3314,
-      "step": 11255
-    },
-    {
-      "epoch": 0.7356381935821188,
-      "grad_norm": 0.45574915409088135,
-      "learning_rate": 8.752706760495284e-06,
-      "loss": 0.3768,
-      "step": 11256
-    },
-    {
-      "epoch": 0.735703548787661,
-      "grad_norm": 0.42137467861175537,
-      "learning_rate": 8.752475998386474e-06,
-      "loss": 0.3425,
-      "step": 11257
-    },
-    {
-      "epoch": 0.735768903993203,
-      "grad_norm": 0.43730729818344116,
-      "learning_rate": 8.752245217975537e-06,
-      "loss": 0.3447,
-      "step": 11258
-    },
-    {
-      "epoch": 0.7358342591987452,
-      "grad_norm": 0.4258533716201782,
-      "learning_rate": 8.752014419263601e-06,
-      "loss": 0.3603,
-      "step": 11259
-    },
-    {
-      "epoch": 0.7358996144042873,
-      "grad_norm": 0.4704715311527252,
-      "learning_rate": 8.751783602251791e-06,
-      "loss": 0.4428,
-      "step": 11260
-    },
-    {
-      "epoch": 0.7359649696098294,
-      "grad_norm": 0.43059343099594116,
-      "learning_rate": 8.751552766941233e-06,
-      "loss": 0.3586,
-      "step": 11261
-    },
-    {
-      "epoch": 0.7360303248153716,
-      "grad_norm": 0.41468125581741333,
-      "learning_rate": 8.751321913333051e-06,
-      "loss": 0.3551,
-      "step": 11262
-    },
-    {
-      "epoch": 0.7360956800209136,
-      "grad_norm": 0.40264788269996643,
-      "learning_rate": 8.751091041428373e-06,
-      "loss": 0.2891,
-      "step": 11263
-    },
-    {
-      "epoch": 0.7361610352264558,
-      "grad_norm": 0.4065679609775543,
-      "learning_rate": 8.750860151228326e-06,
-      "loss": 0.3291,
-      "step": 11264
-    },
-    {
-      "epoch": 0.7362263904319979,
-      "grad_norm": 0.4681910276412964,
-      "learning_rate": 8.750629242734032e-06,
-      "loss": 0.3894,
-      "step": 11265
-    },
-    {
-      "epoch": 0.7362917456375401,
-      "grad_norm": 0.46393883228302,
-      "learning_rate": 8.750398315946623e-06,
-      "loss": 0.3906,
-      "step": 11266
-    },
-    {
-      "epoch": 0.7363571008430821,
-      "grad_norm": 0.49299412965774536,
-      "learning_rate": 8.750167370867219e-06,
-      "loss": 0.4371,
-      "step": 11267
-    },
-    {
-      "epoch": 0.7364224560486242,
-      "grad_norm": 0.4403276741504669,
-      "learning_rate": 8.74993640749695e-06,
-      "loss": 0.3671,
-      "step": 11268
-    },
-    {
-      "epoch": 0.7364878112541664,
-      "grad_norm": 0.4243725836277008,
-      "learning_rate": 8.749705425836945e-06,
-      "loss": 0.3669,
-      "step": 11269
-    },
-    {
-      "epoch": 0.7365531664597085,
-      "grad_norm": 0.4288892149925232,
-      "learning_rate": 8.749474425888324e-06,
-      "loss": 0.3821,
-      "step": 11270
-    },
-    {
-      "epoch": 0.7366185216652507,
-      "grad_norm": 0.4476723372936249,
-      "learning_rate": 8.74924340765222e-06,
-      "loss": 0.3891,
-      "step": 11271
-    },
-    {
-      "epoch": 0.7366838768707927,
-      "grad_norm": 0.4608297646045685,
-      "learning_rate": 8.749012371129756e-06,
-      "loss": 0.4004,
-      "step": 11272
-    },
-    {
-      "epoch": 0.7367492320763349,
-      "grad_norm": 0.4542662799358368,
-      "learning_rate": 8.748781316322058e-06,
-      "loss": 0.3932,
-      "step": 11273
-    },
-    {
-      "epoch": 0.736814587281877,
-      "grad_norm": 0.4622139632701874,
-      "learning_rate": 8.748550243230259e-06,
-      "loss": 0.3562,
-      "step": 11274
-    },
-    {
-      "epoch": 0.7368799424874192,
-      "grad_norm": 0.45630398392677307,
-      "learning_rate": 8.748319151855478e-06,
-      "loss": 0.3839,
-      "step": 11275
-    },
-    {
-      "epoch": 0.7369452976929612,
-      "grad_norm": 0.47129204869270325,
-      "learning_rate": 8.748088042198848e-06,
-      "loss": 0.4271,
-      "step": 11276
-    },
-    {
-      "epoch": 0.7370106528985033,
-      "grad_norm": 0.43397608399391174,
-      "learning_rate": 8.747856914261493e-06,
-      "loss": 0.3838,
-      "step": 11277
-    },
-    {
-      "epoch": 0.7370760081040455,
-      "grad_norm": 0.46451449394226074,
-      "learning_rate": 8.747625768044542e-06,
-      "loss": 0.3854,
-      "step": 11278
-    },
-    {
-      "epoch": 0.7371413633095876,
-      "grad_norm": 0.41327789425849915,
-      "learning_rate": 8.747394603549122e-06,
-      "loss": 0.3437,
-      "step": 11279
-    },
-    {
-      "epoch": 0.7372067185151298,
-      "grad_norm": 0.4603644907474518,
-      "learning_rate": 8.747163420776361e-06,
-      "loss": 0.3815,
-      "step": 11280
-    },
-    {
-      "epoch": 0.7372720737206718,
-      "grad_norm": 0.470671147108078,
-      "learning_rate": 8.746932219727384e-06,
-      "loss": 0.403,
-      "step": 11281
-    },
-    {
-      "epoch": 0.737337428926214,
-      "grad_norm": 0.43903297185897827,
-      "learning_rate": 8.746701000403321e-06,
-      "loss": 0.3519,
-      "step": 11282
-    },
-    {
-      "epoch": 0.7374027841317561,
-      "grad_norm": 0.46954628825187683,
-      "learning_rate": 8.7464697628053e-06,
-      "loss": 0.4205,
-      "step": 11283
-    },
-    {
-      "epoch": 0.7374681393372983,
-      "grad_norm": 0.4562249481678009,
-      "learning_rate": 8.746238506934448e-06,
-      "loss": 0.3838,
-      "step": 11284
-    },
-    {
-      "epoch": 0.7375334945428403,
-      "grad_norm": 0.42818036675453186,
-      "learning_rate": 8.746007232791893e-06,
-      "loss": 0.3689,
-      "step": 11285
-    },
-    {
-      "epoch": 0.7375988497483824,
-      "grad_norm": 0.47790461778640747,
-      "learning_rate": 8.745775940378762e-06,
-      "loss": 0.3995,
-      "step": 11286
-    },
-    {
-      "epoch": 0.7376642049539246,
-      "grad_norm": 0.46306198835372925,
-      "learning_rate": 8.745544629696185e-06,
-      "loss": 0.3946,
-      "step": 11287
-    },
-    {
-      "epoch": 0.7377295601594667,
-      "grad_norm": 0.43406930565834045,
-      "learning_rate": 8.74531330074529e-06,
-      "loss": 0.3459,
-      "step": 11288
-    },
-    {
-      "epoch": 0.7377949153650089,
-      "grad_norm": 0.44612717628479004,
-      "learning_rate": 8.745081953527203e-06,
-      "loss": 0.3576,
-      "step": 11289
-    },
-    {
-      "epoch": 0.7378602705705509,
-      "grad_norm": 0.48750752210617065,
-      "learning_rate": 8.744850588043055e-06,
-      "loss": 0.4311,
-      "step": 11290
-    },
-    {
-      "epoch": 0.7379256257760931,
-      "grad_norm": 0.49298563599586487,
-      "learning_rate": 8.744619204293974e-06,
-      "loss": 0.4336,
-      "step": 11291
-    },
-    {
-      "epoch": 0.7379909809816352,
-      "grad_norm": 0.453427255153656,
-      "learning_rate": 8.744387802281086e-06,
-      "loss": 0.3561,
-      "step": 11292
-    },
-    {
-      "epoch": 0.7380563361871773,
-      "grad_norm": 0.43164753913879395,
-      "learning_rate": 8.744156382005521e-06,
-      "loss": 0.4027,
-      "step": 11293
-    },
-    {
-      "epoch": 0.7381216913927194,
-      "grad_norm": 0.5250502228736877,
-      "learning_rate": 8.74392494346841e-06,
-      "loss": 0.4318,
-      "step": 11294
-    },
-    {
-      "epoch": 0.7381870465982615,
-      "grad_norm": 0.439285010099411,
-      "learning_rate": 8.74369348667088e-06,
-      "loss": 0.3363,
-      "step": 11295
-    },
-    {
-      "epoch": 0.7382524018038037,
-      "grad_norm": 0.45224079489707947,
-      "learning_rate": 8.74346201161406e-06,
-      "loss": 0.3764,
-      "step": 11296
-    },
-    {
-      "epoch": 0.7383177570093458,
-      "grad_norm": 0.4611002802848816,
-      "learning_rate": 8.743230518299078e-06,
-      "loss": 0.3675,
-      "step": 11297
-    },
-    {
-      "epoch": 0.738383112214888,
-      "grad_norm": 0.46360254287719727,
-      "learning_rate": 8.742999006727064e-06,
-      "loss": 0.4002,
-      "step": 11298
-    },
-    {
-      "epoch": 0.73844846742043,
-      "grad_norm": 0.4859495759010315,
-      "learning_rate": 8.742767476899148e-06,
-      "loss": 0.3767,
-      "step": 11299
-    },
-    {
-      "epoch": 0.7385138226259722,
-      "grad_norm": 0.4485567510128021,
-      "learning_rate": 8.742535928816457e-06,
-      "loss": 0.3742,
-      "step": 11300
-    },
-    {
-      "epoch": 0.7385791778315143,
-      "grad_norm": 0.45822906494140625,
-      "learning_rate": 8.742304362480123e-06,
-      "loss": 0.3788,
-      "step": 11301
-    },
-    {
-      "epoch": 0.7386445330370564,
-      "grad_norm": 0.4571850001811981,
-      "learning_rate": 8.742072777891275e-06,
-      "loss": 0.4086,
-      "step": 11302
-    },
-    {
-      "epoch": 0.7387098882425985,
-      "grad_norm": 0.4510149359703064,
-      "learning_rate": 8.74184117505104e-06,
-      "loss": 0.4104,
-      "step": 11303
-    },
-    {
-      "epoch": 0.7387752434481406,
-      "grad_norm": 0.45367246866226196,
-      "learning_rate": 8.741609553960548e-06,
-      "loss": 0.3683,
-      "step": 11304
-    },
-    {
-      "epoch": 0.7388405986536828,
-      "grad_norm": 0.4416310787200928,
-      "learning_rate": 8.741377914620933e-06,
-      "loss": 0.3951,
-      "step": 11305
-    },
-    {
-      "epoch": 0.7389059538592249,
-      "grad_norm": 0.46273839473724365,
-      "learning_rate": 8.741146257033321e-06,
-      "loss": 0.4341,
-      "step": 11306
-    },
-    {
-      "epoch": 0.738971309064767,
-      "grad_norm": 0.4469950199127197,
-      "learning_rate": 8.740914581198841e-06,
-      "loss": 0.3583,
-      "step": 11307
-    },
-    {
-      "epoch": 0.7390366642703091,
-      "grad_norm": 0.4812050759792328,
-      "learning_rate": 8.740682887118626e-06,
-      "loss": 0.4002,
-      "step": 11308
-    },
-    {
-      "epoch": 0.7391020194758513,
-      "grad_norm": 0.4215378761291504,
-      "learning_rate": 8.740451174793805e-06,
-      "loss": 0.3537,
-      "step": 11309
-    },
-    {
-      "epoch": 0.7391673746813934,
-      "grad_norm": 0.4300982654094696,
-      "learning_rate": 8.740219444225506e-06,
-      "loss": 0.3314,
-      "step": 11310
-    },
-    {
-      "epoch": 0.7392327298869354,
-      "grad_norm": 0.4511035680770874,
-      "learning_rate": 8.739987695414863e-06,
-      "loss": 0.37,
-      "step": 11311
-    },
-    {
-      "epoch": 0.7392980850924776,
-      "grad_norm": 0.48972102999687195,
-      "learning_rate": 8.739755928363004e-06,
-      "loss": 0.4092,
-      "step": 11312
-    },
-    {
-      "epoch": 0.7393634402980197,
-      "grad_norm": 0.5016107559204102,
-      "learning_rate": 8.739524143071058e-06,
-      "loss": 0.4012,
-      "step": 11313
-    },
-    {
-      "epoch": 0.7394287955035619,
-      "grad_norm": 0.4714929759502411,
-      "learning_rate": 8.73929233954016e-06,
-      "loss": 0.4362,
-      "step": 11314
-    },
-    {
-      "epoch": 0.739494150709104,
-      "grad_norm": 0.4383006989955902,
-      "learning_rate": 8.739060517771437e-06,
-      "loss": 0.3916,
-      "step": 11315
-    },
-    {
-      "epoch": 0.7395595059146461,
-      "grad_norm": 0.45401644706726074,
-      "learning_rate": 8.73882867776602e-06,
-      "loss": 0.3956,
-      "step": 11316
-    },
-    {
-      "epoch": 0.7396248611201882,
-      "grad_norm": 0.41924989223480225,
-      "learning_rate": 8.73859681952504e-06,
-      "loss": 0.3462,
-      "step": 11317
-    },
-    {
-      "epoch": 0.7396902163257304,
-      "grad_norm": 0.4588722288608551,
-      "learning_rate": 8.738364943049628e-06,
-      "loss": 0.4063,
-      "step": 11318
-    },
-    {
-      "epoch": 0.7397555715312725,
-      "grad_norm": 0.4037357568740845,
-      "learning_rate": 8.738133048340916e-06,
-      "loss": 0.3313,
-      "step": 11319
-    },
-    {
-      "epoch": 0.7398209267368145,
-      "grad_norm": 0.4058557450771332,
-      "learning_rate": 8.737901135400034e-06,
-      "loss": 0.31,
-      "step": 11320
-    },
-    {
-      "epoch": 0.7398862819423567,
-      "grad_norm": 0.4838840663433075,
-      "learning_rate": 8.737669204228112e-06,
-      "loss": 0.361,
-      "step": 11321
-    },
-    {
-      "epoch": 0.7399516371478988,
-      "grad_norm": 0.4566826820373535,
-      "learning_rate": 8.737437254826283e-06,
-      "loss": 0.4101,
-      "step": 11322
-    },
-    {
-      "epoch": 0.740016992353441,
-      "grad_norm": 0.4308919608592987,
-      "learning_rate": 8.737205287195678e-06,
-      "loss": 0.3454,
-      "step": 11323
-    },
-    {
-      "epoch": 0.7400823475589831,
-      "grad_norm": 0.5028971433639526,
-      "learning_rate": 8.736973301337428e-06,
-      "loss": 0.4994,
-      "step": 11324
-    },
-    {
-      "epoch": 0.7401477027645252,
-      "grad_norm": 0.4332844913005829,
-      "learning_rate": 8.736741297252665e-06,
-      "loss": 0.345,
-      "step": 11325
-    },
-    {
-      "epoch": 0.7402130579700673,
-      "grad_norm": 0.43167319893836975,
-      "learning_rate": 8.73650927494252e-06,
-      "loss": 0.3938,
-      "step": 11326
-    },
-    {
-      "epoch": 0.7402784131756094,
-      "grad_norm": 0.4252494275569916,
-      "learning_rate": 8.736277234408125e-06,
-      "loss": 0.3029,
-      "step": 11327
-    },
-    {
-      "epoch": 0.7403437683811516,
-      "grad_norm": 0.46765902638435364,
-      "learning_rate": 8.736045175650612e-06,
-      "loss": 0.402,
-      "step": 11328
-    },
-    {
-      "epoch": 0.7404091235866936,
-      "grad_norm": 0.4168873727321625,
-      "learning_rate": 8.735813098671111e-06,
-      "loss": 0.3234,
-      "step": 11329
-    },
-    {
-      "epoch": 0.7404744787922358,
-      "grad_norm": 0.45331600308418274,
-      "learning_rate": 8.735581003470754e-06,
-      "loss": 0.3539,
-      "step": 11330
-    },
-    {
-      "epoch": 0.7405398339977779,
-      "grad_norm": 0.45428404211997986,
-      "learning_rate": 8.735348890050678e-06,
-      "loss": 0.3962,
-      "step": 11331
-    },
-    {
-      "epoch": 0.7406051892033201,
-      "grad_norm": 0.47911834716796875,
-      "learning_rate": 8.73511675841201e-06,
-      "loss": 0.4235,
-      "step": 11332
-    },
-    {
-      "epoch": 0.7406705444088622,
-      "grad_norm": 0.4567180573940277,
-      "learning_rate": 8.734884608555882e-06,
-      "loss": 0.4001,
-      "step": 11333
-    },
-    {
-      "epoch": 0.7407358996144043,
-      "grad_norm": 1.2108349800109863,
-      "learning_rate": 8.73465244048343e-06,
-      "loss": 0.3833,
-      "step": 11334
-    },
-    {
-      "epoch": 0.7408012548199464,
-      "grad_norm": 0.45654308795928955,
-      "learning_rate": 8.734420254195784e-06,
-      "loss": 0.4051,
-      "step": 11335
-    },
-    {
-      "epoch": 0.7408666100254885,
-      "grad_norm": 0.4677703380584717,
-      "learning_rate": 8.734188049694075e-06,
-      "loss": 0.3768,
-      "step": 11336
-    },
-    {
-      "epoch": 0.7409319652310307,
-      "grad_norm": 0.5017822980880737,
-      "learning_rate": 8.733955826979439e-06,
-      "loss": 0.4115,
-      "step": 11337
-    },
-    {
-      "epoch": 0.7409973204365727,
-      "grad_norm": 0.4488432705402374,
-      "learning_rate": 8.733723586053006e-06,
-      "loss": 0.3549,
-      "step": 11338
-    },
-    {
-      "epoch": 0.7410626756421149,
-      "grad_norm": 0.4527190327644348,
-      "learning_rate": 8.733491326915909e-06,
-      "loss": 0.3453,
-      "step": 11339
-    },
-    {
-      "epoch": 0.741128030847657,
-      "grad_norm": 0.4319392144680023,
-      "learning_rate": 8.733259049569282e-06,
-      "loss": 0.3334,
-      "step": 11340
-    },
-    {
-      "epoch": 0.7411933860531992,
-      "grad_norm": 0.4645390212535858,
-      "learning_rate": 8.733026754014258e-06,
-      "loss": 0.3774,
-      "step": 11341
-    },
-    {
-      "epoch": 0.7412587412587412,
-      "grad_norm": 0.4230360984802246,
-      "learning_rate": 8.73279444025197e-06,
-      "loss": 0.3439,
-      "step": 11342
-    },
-    {
-      "epoch": 0.7413240964642834,
-      "grad_norm": 0.415467232465744,
-      "learning_rate": 8.73256210828355e-06,
-      "loss": 0.3112,
-      "step": 11343
-    },
-    {
-      "epoch": 0.7413894516698255,
-      "grad_norm": 0.4561569094657898,
-      "learning_rate": 8.732329758110128e-06,
-      "loss": 0.398,
-      "step": 11344
-    },
-    {
-      "epoch": 0.7414548068753676,
-      "grad_norm": 0.4417307674884796,
-      "learning_rate": 8.732097389732845e-06,
-      "loss": 0.381,
-      "step": 11345
-    },
-    {
-      "epoch": 0.7415201620809098,
-      "grad_norm": 0.4971565306186676,
-      "learning_rate": 8.731865003152829e-06,
-      "loss": 0.4116,
-      "step": 11346
-    },
-    {
-      "epoch": 0.7415855172864518,
-      "grad_norm": 0.4177454710006714,
-      "learning_rate": 8.731632598371213e-06,
-      "loss": 0.3468,
-      "step": 11347
-    },
-    {
-      "epoch": 0.741650872491994,
-      "grad_norm": 0.4392794668674469,
-      "learning_rate": 8.731400175389133e-06,
-      "loss": 0.3693,
-      "step": 11348
-    },
-    {
-      "epoch": 0.7417162276975361,
-      "grad_norm": 0.4654693901538849,
-      "learning_rate": 8.731167734207722e-06,
-      "loss": 0.4123,
-      "step": 11349
-    },
-    {
-      "epoch": 0.7417815829030783,
-      "grad_norm": 0.4322023093700409,
-      "learning_rate": 8.730935274828112e-06,
-      "loss": 0.3722,
-      "step": 11350
-    },
-    {
-      "epoch": 0.7418469381086203,
-      "grad_norm": 0.4423964321613312,
-      "learning_rate": 8.73070279725144e-06,
-      "loss": 0.3735,
-      "step": 11351
-    },
-    {
-      "epoch": 0.7419122933141624,
-      "grad_norm": 0.45737385749816895,
-      "learning_rate": 8.730470301478836e-06,
-      "loss": 0.3687,
-      "step": 11352
-    },
-    {
-      "epoch": 0.7419776485197046,
-      "grad_norm": 0.45981091260910034,
-      "learning_rate": 8.730237787511438e-06,
-      "loss": 0.4041,
-      "step": 11353
-    },
-    {
-      "epoch": 0.7420430037252467,
-      "grad_norm": 0.44975706934928894,
-      "learning_rate": 8.730005255350377e-06,
-      "loss": 0.4054,
-      "step": 11354
-    },
-    {
-      "epoch": 0.7421083589307889,
-      "grad_norm": 0.43052786588668823,
-      "learning_rate": 8.72977270499679e-06,
-      "loss": 0.3409,
-      "step": 11355
-    },
-    {
-      "epoch": 0.7421737141363309,
-      "grad_norm": 0.4633021652698517,
-      "learning_rate": 8.729540136451808e-06,
-      "loss": 0.456,
-      "step": 11356
-    },
-    {
-      "epoch": 0.7422390693418731,
-      "grad_norm": 0.44586482644081116,
-      "learning_rate": 8.729307549716565e-06,
-      "loss": 0.4209,
-      "step": 11357
-    },
-    {
-      "epoch": 0.7423044245474152,
-      "grad_norm": 0.44594958424568176,
-      "learning_rate": 8.7290749447922e-06,
-      "loss": 0.3821,
-      "step": 11358
-    },
-    {
-      "epoch": 0.7423697797529574,
-      "grad_norm": 0.43272367119789124,
-      "learning_rate": 8.728842321679842e-06,
-      "loss": 0.3524,
-      "step": 11359
-    },
-    {
-      "epoch": 0.7424351349584994,
-      "grad_norm": 0.46184101700782776,
-      "learning_rate": 8.72860968038063e-06,
-      "loss": 0.3372,
-      "step": 11360
-    },
-    {
-      "epoch": 0.7425004901640415,
-      "grad_norm": 0.4214572012424469,
-      "learning_rate": 8.728377020895696e-06,
-      "loss": 0.3307,
-      "step": 11361
-    },
-    {
-      "epoch": 0.7425658453695837,
-      "grad_norm": 0.4390409290790558,
-      "learning_rate": 8.728144343226177e-06,
-      "loss": 0.3887,
-      "step": 11362
-    },
-    {
-      "epoch": 0.7426312005751258,
-      "grad_norm": 0.44752877950668335,
-      "learning_rate": 8.727911647373206e-06,
-      "loss": 0.3604,
-      "step": 11363
-    },
-    {
-      "epoch": 0.742696555780668,
-      "grad_norm": 0.4602724611759186,
-      "learning_rate": 8.727678933337918e-06,
-      "loss": 0.3972,
-      "step": 11364
-    },
-    {
-      "epoch": 0.74276191098621,
-      "grad_norm": 0.4244244396686554,
-      "learning_rate": 8.727446201121447e-06,
-      "loss": 0.3764,
-      "step": 11365
-    },
-    {
-      "epoch": 0.7428272661917522,
-      "grad_norm": 0.4103724956512451,
-      "learning_rate": 8.727213450724931e-06,
-      "loss": 0.3205,
-      "step": 11366
-    },
-    {
-      "epoch": 0.7428926213972943,
-      "grad_norm": 0.45078158378601074,
-      "learning_rate": 8.726980682149503e-06,
-      "loss": 0.3665,
-      "step": 11367
-    },
-    {
-      "epoch": 0.7429579766028365,
-      "grad_norm": 0.4778234362602234,
-      "learning_rate": 8.726747895396302e-06,
-      "loss": 0.374,
-      "step": 11368
-    },
-    {
-      "epoch": 0.7430233318083785,
-      "grad_norm": 0.4740595519542694,
-      "learning_rate": 8.726515090466456e-06,
-      "loss": 0.421,
-      "step": 11369
-    },
-    {
-      "epoch": 0.7430886870139206,
-      "grad_norm": 0.43524232506752014,
-      "learning_rate": 8.726282267361109e-06,
-      "loss": 0.3649,
-      "step": 11370
-    },
-    {
-      "epoch": 0.7431540422194628,
-      "grad_norm": 0.4340595602989197,
-      "learning_rate": 8.72604942608139e-06,
-      "loss": 0.3752,
-      "step": 11371
-    },
-    {
-      "epoch": 0.7432193974250049,
-      "grad_norm": 0.4169885814189911,
-      "learning_rate": 8.725816566628437e-06,
-      "loss": 0.3461,
-      "step": 11372
-    },
-    {
-      "epoch": 0.743284752630547,
-      "grad_norm": 0.4595286250114441,
-      "learning_rate": 8.725583689003388e-06,
-      "loss": 0.4014,
-      "step": 11373
-    },
-    {
-      "epoch": 0.7433501078360891,
-      "grad_norm": 0.4522557556629181,
-      "learning_rate": 8.725350793207374e-06,
-      "loss": 0.3678,
-      "step": 11374
-    },
-    {
-      "epoch": 0.7434154630416313,
-      "grad_norm": 0.4524635076522827,
-      "learning_rate": 8.725117879241536e-06,
-      "loss": 0.3877,
-      "step": 11375
-    },
-    {
-      "epoch": 0.7434808182471734,
-      "grad_norm": 0.4340183436870575,
-      "learning_rate": 8.724884947107006e-06,
-      "loss": 0.3676,
-      "step": 11376
-    },
-    {
-      "epoch": 0.7435461734527156,
-      "grad_norm": 0.45854222774505615,
-      "learning_rate": 8.724651996804922e-06,
-      "loss": 0.4319,
-      "step": 11377
-    },
-    {
-      "epoch": 0.7436115286582576,
-      "grad_norm": 0.4483458697795868,
-      "learning_rate": 8.72441902833642e-06,
-      "loss": 0.3776,
-      "step": 11378
-    },
-    {
-      "epoch": 0.7436768838637997,
-      "grad_norm": 0.44911783933639526,
-      "learning_rate": 8.724186041702636e-06,
-      "loss": 0.369,
-      "step": 11379
-    },
-    {
-      "epoch": 0.7437422390693419,
-      "grad_norm": 0.4570876657962799,
-      "learning_rate": 8.723953036904707e-06,
-      "loss": 0.3888,
-      "step": 11380
-    },
-    {
-      "epoch": 0.743807594274884,
-      "grad_norm": 0.45264339447021484,
-      "learning_rate": 8.723720013943769e-06,
-      "loss": 0.3889,
-      "step": 11381
-    },
-    {
-      "epoch": 0.7438729494804261,
-      "grad_norm": 0.4470329284667969,
-      "learning_rate": 8.723486972820957e-06,
-      "loss": 0.3413,
-      "step": 11382
-    },
-    {
-      "epoch": 0.7439383046859682,
-      "grad_norm": 0.46752843260765076,
-      "learning_rate": 8.72325391353741e-06,
-      "loss": 0.4063,
-      "step": 11383
-    },
-    {
-      "epoch": 0.7440036598915104,
-      "grad_norm": 0.45190879702568054,
-      "learning_rate": 8.723020836094265e-06,
-      "loss": 0.4091,
-      "step": 11384
-    },
-    {
-      "epoch": 0.7440690150970525,
-      "grad_norm": 0.4984181225299835,
-      "learning_rate": 8.722787740492655e-06,
-      "loss": 0.4156,
-      "step": 11385
-    },
-    {
-      "epoch": 0.7441343703025945,
-      "grad_norm": 0.44895896315574646,
-      "learning_rate": 8.72255462673372e-06,
-      "loss": 0.3793,
-      "step": 11386
-    },
-    {
-      "epoch": 0.7441997255081367,
-      "grad_norm": 0.4525301158428192,
-      "learning_rate": 8.7223214948186e-06,
-      "loss": 0.3845,
-      "step": 11387
-    },
-    {
-      "epoch": 0.7442650807136788,
-      "grad_norm": 0.4790029525756836,
-      "learning_rate": 8.722088344748425e-06,
-      "loss": 0.4163,
-      "step": 11388
-    },
-    {
-      "epoch": 0.744330435919221,
-      "grad_norm": 0.46557191014289856,
-      "learning_rate": 8.721855176524337e-06,
-      "loss": 0.4317,
-      "step": 11389
-    },
-    {
-      "epoch": 0.7443957911247631,
-      "grad_norm": 0.49655681848526,
-      "learning_rate": 8.721621990147472e-06,
-      "loss": 0.4354,
-      "step": 11390
-    },
-    {
-      "epoch": 0.7444611463303052,
-      "grad_norm": 0.4580974578857422,
-      "learning_rate": 8.721388785618967e-06,
-      "loss": 0.4089,
-      "step": 11391
-    },
-    {
-      "epoch": 0.7445265015358473,
-      "grad_norm": 0.4186343252658844,
-      "learning_rate": 8.721155562939961e-06,
-      "loss": 0.3283,
-      "step": 11392
-    },
-    {
-      "epoch": 0.7445918567413895,
-      "grad_norm": 0.4594210684299469,
-      "learning_rate": 8.720922322111587e-06,
-      "loss": 0.3889,
-      "step": 11393
-    },
-    {
-      "epoch": 0.7446572119469316,
-      "grad_norm": 0.4190291464328766,
-      "learning_rate": 8.720689063134989e-06,
-      "loss": 0.3284,
-      "step": 11394
-    },
-    {
-      "epoch": 0.7447225671524736,
-      "grad_norm": 0.4274213910102844,
-      "learning_rate": 8.720455786011299e-06,
-      "loss": 0.354,
-      "step": 11395
-    },
-    {
-      "epoch": 0.7447879223580158,
-      "grad_norm": 0.421599805355072,
-      "learning_rate": 8.720222490741659e-06,
-      "loss": 0.3247,
-      "step": 11396
-    },
-    {
-      "epoch": 0.7448532775635579,
-      "grad_norm": 0.4587099552154541,
-      "learning_rate": 8.719989177327204e-06,
-      "loss": 0.401,
-      "step": 11397
-    },
-    {
-      "epoch": 0.7449186327691001,
-      "grad_norm": 0.535896360874176,
-      "learning_rate": 8.719755845769073e-06,
-      "loss": 0.4355,
-      "step": 11398
-    },
-    {
-      "epoch": 0.7449839879746422,
-      "grad_norm": 0.4585549533367157,
-      "learning_rate": 8.719522496068405e-06,
-      "loss": 0.3988,
-      "step": 11399
-    },
-    {
-      "epoch": 0.7450493431801843,
-      "grad_norm": 0.4203391969203949,
-      "learning_rate": 8.719289128226336e-06,
-      "loss": 0.3494,
-      "step": 11400
-    },
-    {
-      "epoch": 0.7451146983857264,
-      "grad_norm": 0.44131430983543396,
-      "learning_rate": 8.719055742244007e-06,
-      "loss": 0.35,
-      "step": 11401
-    },
-    {
-      "epoch": 0.7451800535912686,
-      "grad_norm": 0.40603551268577576,
-      "learning_rate": 8.718822338122554e-06,
-      "loss": 0.3217,
-      "step": 11402
-    },
-    {
-      "epoch": 0.7452454087968107,
-      "grad_norm": 0.45532020926475525,
-      "learning_rate": 8.718588915863116e-06,
-      "loss": 0.403,
-      "step": 11403
-    },
-    {
-      "epoch": 0.7453107640023527,
-      "grad_norm": 0.44275617599487305,
-      "learning_rate": 8.71835547546683e-06,
-      "loss": 0.3747,
-      "step": 11404
-    },
-    {
-      "epoch": 0.7453761192078949,
-      "grad_norm": 0.4695359766483307,
-      "learning_rate": 8.71812201693484e-06,
-      "loss": 0.4013,
-      "step": 11405
-    },
-    {
-      "epoch": 0.745441474413437,
-      "grad_norm": 0.46241793036460876,
-      "learning_rate": 8.717888540268279e-06,
-      "loss": 0.3933,
-      "step": 11406
-    },
-    {
-      "epoch": 0.7455068296189792,
-      "grad_norm": 0.4245399832725525,
-      "learning_rate": 8.717655045468286e-06,
-      "loss": 0.355,
-      "step": 11407
-    },
-    {
-      "epoch": 0.7455721848245213,
-      "grad_norm": 0.42461681365966797,
-      "learning_rate": 8.717421532536004e-06,
-      "loss": 0.3682,
-      "step": 11408
-    },
-    {
-      "epoch": 0.7456375400300634,
-      "grad_norm": 0.4408150911331177,
-      "learning_rate": 8.717188001472567e-06,
-      "loss": 0.3808,
-      "step": 11409
-    },
-    {
-      "epoch": 0.7457028952356055,
-      "grad_norm": 0.4594395160675049,
-      "learning_rate": 8.716954452279118e-06,
-      "loss": 0.4048,
-      "step": 11410
-    },
-    {
-      "epoch": 0.7457682504411476,
-      "grad_norm": 0.445923775434494,
-      "learning_rate": 8.716720884956793e-06,
-      "loss": 0.3763,
-      "step": 11411
-    },
-    {
-      "epoch": 0.7458336056466898,
-      "grad_norm": 0.4819449186325073,
-      "learning_rate": 8.716487299506733e-06,
-      "loss": 0.4028,
-      "step": 11412
-    },
-    {
-      "epoch": 0.7458989608522318,
-      "grad_norm": 0.43233078718185425,
-      "learning_rate": 8.716253695930079e-06,
-      "loss": 0.4006,
-      "step": 11413
-    },
-    {
-      "epoch": 0.745964316057774,
-      "grad_norm": 0.4277048707008362,
-      "learning_rate": 8.716020074227966e-06,
-      "loss": 0.3059,
-      "step": 11414
-    },
-    {
-      "epoch": 0.7460296712633161,
-      "grad_norm": 0.4572831094264984,
-      "learning_rate": 8.715786434401537e-06,
-      "loss": 0.4158,
-      "step": 11415
-    },
-    {
-      "epoch": 0.7460950264688583,
-      "grad_norm": 0.4788439869880676,
-      "learning_rate": 8.71555277645193e-06,
-      "loss": 0.4349,
-      "step": 11416
-    },
-    {
-      "epoch": 0.7461603816744004,
-      "grad_norm": 0.4642075002193451,
-      "learning_rate": 8.715319100380284e-06,
-      "loss": 0.3932,
-      "step": 11417
-    },
-    {
-      "epoch": 0.7462257368799425,
-      "grad_norm": 0.4323691129684448,
-      "learning_rate": 8.71508540618774e-06,
-      "loss": 0.3656,
-      "step": 11418
-    },
-    {
-      "epoch": 0.7462910920854846,
-      "grad_norm": 0.441805362701416,
-      "learning_rate": 8.71485169387544e-06,
-      "loss": 0.3654,
-      "step": 11419
-    },
-    {
-      "epoch": 0.7463564472910267,
-      "grad_norm": 0.46433573961257935,
-      "learning_rate": 8.714617963444518e-06,
-      "loss": 0.4275,
-      "step": 11420
-    },
-    {
-      "epoch": 0.7464218024965689,
-      "grad_norm": 0.4299795925617218,
-      "learning_rate": 8.714384214896118e-06,
-      "loss": 0.3541,
-      "step": 11421
-    },
-    {
-      "epoch": 0.7464871577021109,
-      "grad_norm": 0.44786059856414795,
-      "learning_rate": 8.714150448231379e-06,
-      "loss": 0.3939,
-      "step": 11422
-    },
-    {
-      "epoch": 0.7465525129076531,
-      "grad_norm": 0.41369113326072693,
-      "learning_rate": 8.713916663451443e-06,
-      "loss": 0.3466,
-      "step": 11423
-    },
-    {
-      "epoch": 0.7466178681131952,
-      "grad_norm": 0.4648091197013855,
-      "learning_rate": 8.713682860557448e-06,
-      "loss": 0.3972,
-      "step": 11424
-    },
-    {
-      "epoch": 0.7466832233187374,
-      "grad_norm": 0.45776987075805664,
-      "learning_rate": 8.713449039550536e-06,
-      "loss": 0.3989,
-      "step": 11425
-    },
-    {
-      "epoch": 0.7467485785242794,
-      "grad_norm": 0.4523124098777771,
-      "learning_rate": 8.713215200431845e-06,
-      "loss": 0.3766,
-      "step": 11426
-    },
-    {
-      "epoch": 0.7468139337298216,
-      "grad_norm": 0.4293292462825775,
-      "learning_rate": 8.71298134320252e-06,
-      "loss": 0.3733,
-      "step": 11427
-    },
-    {
-      "epoch": 0.7468792889353637,
-      "grad_norm": 0.42689448595046997,
-      "learning_rate": 8.712747467863696e-06,
-      "loss": 0.3553,
-      "step": 11428
-    },
-    {
-      "epoch": 0.7469446441409058,
-      "grad_norm": 0.4537889361381531,
-      "learning_rate": 8.712513574416517e-06,
-      "loss": 0.4033,
-      "step": 11429
-    },
-    {
-      "epoch": 0.747009999346448,
-      "grad_norm": 0.4374942183494568,
-      "learning_rate": 8.712279662862123e-06,
-      "loss": 0.3793,
-      "step": 11430
-    },
-    {
-      "epoch": 0.74707535455199,
-      "grad_norm": 0.42012590169906616,
-      "learning_rate": 8.712045733201655e-06,
-      "loss": 0.3506,
-      "step": 11431
-    },
-    {
-      "epoch": 0.7471407097575322,
-      "grad_norm": 0.44662463665008545,
-      "learning_rate": 8.711811785436254e-06,
-      "loss": 0.3759,
-      "step": 11432
-    },
-    {
-      "epoch": 0.7472060649630743,
-      "grad_norm": 0.48212146759033203,
-      "learning_rate": 8.711577819567062e-06,
-      "loss": 0.4619,
-      "step": 11433
-    },
-    {
-      "epoch": 0.7472714201686165,
-      "grad_norm": 0.4506356120109558,
-      "learning_rate": 8.71134383559522e-06,
-      "loss": 0.3988,
-      "step": 11434
-    },
-    {
-      "epoch": 0.7473367753741585,
-      "grad_norm": 0.43944063782691956,
-      "learning_rate": 8.711109833521865e-06,
-      "loss": 0.3839,
-      "step": 11435
-    },
-    {
-      "epoch": 0.7474021305797006,
-      "grad_norm": 0.4397209584712982,
-      "learning_rate": 8.710875813348144e-06,
-      "loss": 0.4005,
-      "step": 11436
-    },
-    {
-      "epoch": 0.7474674857852428,
-      "grad_norm": 0.45714902877807617,
-      "learning_rate": 8.710641775075194e-06,
-      "loss": 0.3947,
-      "step": 11437
-    },
-    {
-      "epoch": 0.7475328409907849,
-      "grad_norm": 0.4064410328865051,
-      "learning_rate": 8.710407718704162e-06,
-      "loss": 0.3565,
-      "step": 11438
-    },
-    {
-      "epoch": 0.7475981961963271,
-      "grad_norm": 0.4440568685531616,
-      "learning_rate": 8.710173644236185e-06,
-      "loss": 0.4005,
-      "step": 11439
-    },
-    {
-      "epoch": 0.7476635514018691,
-      "grad_norm": 0.44599300622940063,
-      "learning_rate": 8.709939551672404e-06,
-      "loss": 0.3689,
-      "step": 11440
-    },
-    {
-      "epoch": 0.7477289066074113,
-      "grad_norm": 0.45519688725471497,
-      "learning_rate": 8.709705441013965e-06,
-      "loss": 0.3883,
-      "step": 11441
-    },
-    {
-      "epoch": 0.7477942618129534,
-      "grad_norm": 0.4230014681816101,
-      "learning_rate": 8.709471312262005e-06,
-      "loss": 0.3695,
-      "step": 11442
-    },
-    {
-      "epoch": 0.7478596170184956,
-      "grad_norm": 0.45030948519706726,
-      "learning_rate": 8.70923716541767e-06,
-      "loss": 0.3881,
-      "step": 11443
-    },
-    {
-      "epoch": 0.7479249722240376,
-      "grad_norm": 0.4607604146003723,
-      "learning_rate": 8.7090030004821e-06,
-      "loss": 0.4196,
-      "step": 11444
-    },
-    {
-      "epoch": 0.7479903274295797,
-      "grad_norm": 0.4362353980541229,
-      "learning_rate": 8.708768817456437e-06,
-      "loss": 0.3863,
-      "step": 11445
-    },
-    {
-      "epoch": 0.7480556826351219,
-      "grad_norm": 0.406549870967865,
-      "learning_rate": 8.708534616341824e-06,
-      "loss": 0.3502,
-      "step": 11446
-    },
-    {
-      "epoch": 0.748121037840664,
-      "grad_norm": 0.46241170167922974,
-      "learning_rate": 8.708300397139403e-06,
-      "loss": 0.4045,
-      "step": 11447
-    },
-    {
-      "epoch": 0.7481863930462062,
-      "grad_norm": 0.44195207953453064,
-      "learning_rate": 8.708066159850315e-06,
-      "loss": 0.3711,
-      "step": 11448
-    },
-    {
-      "epoch": 0.7482517482517482,
-      "grad_norm": 0.4226051867008209,
-      "learning_rate": 8.707831904475706e-06,
-      "loss": 0.38,
-      "step": 11449
-    },
-    {
-      "epoch": 0.7483171034572904,
-      "grad_norm": 0.44911104440689087,
-      "learning_rate": 8.707597631016714e-06,
-      "loss": 0.4057,
-      "step": 11450
-    },
-    {
-      "epoch": 0.7483824586628325,
-      "grad_norm": 0.42989981174468994,
-      "learning_rate": 8.707363339474486e-06,
-      "loss": 0.376,
-      "step": 11451
-    },
-    {
-      "epoch": 0.7484478138683747,
-      "grad_norm": 0.4480683207511902,
-      "learning_rate": 8.70712902985016e-06,
-      "loss": 0.3671,
-      "step": 11452
-    },
-    {
-      "epoch": 0.7485131690739167,
-      "grad_norm": 0.44000551104545593,
-      "learning_rate": 8.706894702144883e-06,
-      "loss": 0.3978,
-      "step": 11453
-    },
-    {
-      "epoch": 0.7485785242794588,
-      "grad_norm": 0.4732186794281006,
-      "learning_rate": 8.706660356359796e-06,
-      "loss": 0.3743,
-      "step": 11454
-    },
-    {
-      "epoch": 0.748643879485001,
-      "grad_norm": 0.4579940736293793,
-      "learning_rate": 8.706425992496043e-06,
-      "loss": 0.413,
-      "step": 11455
-    },
-    {
-      "epoch": 0.7487092346905431,
-      "grad_norm": 0.4685218334197998,
-      "learning_rate": 8.706191610554767e-06,
-      "loss": 0.4361,
-      "step": 11456
-    },
-    {
-      "epoch": 0.7487745898960853,
-      "grad_norm": 0.413667768239975,
-      "learning_rate": 8.70595721053711e-06,
-      "loss": 0.311,
-      "step": 11457
-    },
-    {
-      "epoch": 0.7488399451016273,
-      "grad_norm": 0.43470972776412964,
-      "learning_rate": 8.705722792444215e-06,
-      "loss": 0.3814,
-      "step": 11458
-    },
-    {
-      "epoch": 0.7489053003071695,
-      "grad_norm": 0.44430139660835266,
-      "learning_rate": 8.705488356277227e-06,
-      "loss": 0.3979,
-      "step": 11459
-    },
-    {
-      "epoch": 0.7489706555127116,
-      "grad_norm": 0.44423583149909973,
-      "learning_rate": 8.705253902037289e-06,
-      "loss": 0.4073,
-      "step": 11460
-    },
-    {
-      "epoch": 0.7490360107182538,
-      "grad_norm": 0.4940393567085266,
-      "learning_rate": 8.705019429725543e-06,
-      "loss": 0.4367,
-      "step": 11461
-    },
-    {
-      "epoch": 0.7491013659237958,
-      "grad_norm": 0.5035269260406494,
-      "learning_rate": 8.704784939343134e-06,
-      "loss": 0.4584,
-      "step": 11462
-    },
-    {
-      "epoch": 0.7491667211293379,
-      "grad_norm": 0.47204816341400146,
-      "learning_rate": 8.704550430891206e-06,
-      "loss": 0.3979,
-      "step": 11463
-    },
-    {
-      "epoch": 0.7492320763348801,
-      "grad_norm": 0.43368858098983765,
-      "learning_rate": 8.704315904370901e-06,
-      "loss": 0.3773,
-      "step": 11464
-    },
-    {
-      "epoch": 0.7492974315404222,
-      "grad_norm": 0.4607824385166168,
-      "learning_rate": 8.704081359783365e-06,
-      "loss": 0.3685,
-      "step": 11465
-    },
-    {
-      "epoch": 0.7493627867459643,
-      "grad_norm": 0.462039053440094,
-      "learning_rate": 8.70384679712974e-06,
-      "loss": 0.436,
-      "step": 11466
-    },
-    {
-      "epoch": 0.7494281419515064,
-      "grad_norm": 0.4594377279281616,
-      "learning_rate": 8.703612216411172e-06,
-      "loss": 0.4097,
-      "step": 11467
-    },
-    {
-      "epoch": 0.7494934971570486,
-      "grad_norm": 0.47541284561157227,
-      "learning_rate": 8.703377617628804e-06,
-      "loss": 0.3993,
-      "step": 11468
-    },
-    {
-      "epoch": 0.7495588523625907,
-      "grad_norm": 0.4668521583080292,
-      "learning_rate": 8.70314300078378e-06,
-      "loss": 0.4087,
-      "step": 11469
-    },
-    {
-      "epoch": 0.7496242075681327,
-      "grad_norm": 0.44209030270576477,
-      "learning_rate": 8.702908365877245e-06,
-      "loss": 0.3757,
-      "step": 11470
-    },
-    {
-      "epoch": 0.7496895627736749,
-      "grad_norm": 0.44060683250427246,
-      "learning_rate": 8.702673712910344e-06,
-      "loss": 0.3288,
-      "step": 11471
-    },
-    {
-      "epoch": 0.749754917979217,
-      "grad_norm": 0.4443775415420532,
-      "learning_rate": 8.702439041884219e-06,
-      "loss": 0.3851,
-      "step": 11472
-    },
-    {
-      "epoch": 0.7498202731847592,
-      "grad_norm": 0.4726814031600952,
-      "learning_rate": 8.702204352800016e-06,
-      "loss": 0.4268,
-      "step": 11473
-    },
-    {
-      "epoch": 0.7498856283903013,
-      "grad_norm": 0.48055610060691833,
-      "learning_rate": 8.701969645658881e-06,
-      "loss": 0.4545,
-      "step": 11474
-    },
-    {
-      "epoch": 0.7499509835958434,
-      "grad_norm": 0.3886030614376068,
-      "learning_rate": 8.701734920461957e-06,
-      "loss": 0.2669,
-      "step": 11475
-    },
-    {
-      "epoch": 0.7500163388013855,
-      "grad_norm": 0.4247395396232605,
-      "learning_rate": 8.70150017721039e-06,
-      "loss": 0.3886,
-      "step": 11476
-    },
-    {
-      "epoch": 0.7500816940069277,
-      "grad_norm": 0.4572971761226654,
-      "learning_rate": 8.701265415905324e-06,
-      "loss": 0.4013,
-      "step": 11477
-    },
-    {
-      "epoch": 0.7501470492124698,
-      "grad_norm": 0.4315987229347229,
-      "learning_rate": 8.701030636547905e-06,
-      "loss": 0.3727,
-      "step": 11478
-    },
-    {
-      "epoch": 0.7502124044180118,
-      "grad_norm": 0.47833821177482605,
-      "learning_rate": 8.700795839139276e-06,
-      "loss": 0.4174,
-      "step": 11479
-    },
-    {
-      "epoch": 0.750277759623554,
-      "grad_norm": 0.42190808057785034,
-      "learning_rate": 8.700561023680584e-06,
-      "loss": 0.3662,
-      "step": 11480
-    },
-    {
-      "epoch": 0.7503431148290961,
-      "grad_norm": 0.4663477838039398,
-      "learning_rate": 8.700326190172974e-06,
-      "loss": 0.4081,
-      "step": 11481
-    },
-    {
-      "epoch": 0.7504084700346383,
-      "grad_norm": 0.40893182158470154,
-      "learning_rate": 8.70009133861759e-06,
-      "loss": 0.3448,
-      "step": 11482
-    },
-    {
-      "epoch": 0.7504738252401804,
-      "grad_norm": 0.4747457802295685,
-      "learning_rate": 8.699856469015581e-06,
-      "loss": 0.3802,
-      "step": 11483
-    },
-    {
-      "epoch": 0.7505391804457225,
-      "grad_norm": 0.4700453281402588,
-      "learning_rate": 8.69962158136809e-06,
-      "loss": 0.4044,
-      "step": 11484
-    },
-    {
-      "epoch": 0.7506045356512646,
-      "grad_norm": 0.4343928396701813,
-      "learning_rate": 8.699386675676263e-06,
-      "loss": 0.3571,
-      "step": 11485
-    },
-    {
-      "epoch": 0.7506698908568068,
-      "grad_norm": 0.45376333594322205,
-      "learning_rate": 8.699151751941245e-06,
-      "loss": 0.4145,
-      "step": 11486
-    },
-    {
-      "epoch": 0.7507352460623489,
-      "grad_norm": 0.4375240206718445,
-      "learning_rate": 8.69891681016418e-06,
-      "loss": 0.4194,
-      "step": 11487
-    },
-    {
-      "epoch": 0.7508006012678909,
-      "grad_norm": 0.4365542531013489,
-      "learning_rate": 8.698681850346218e-06,
-      "loss": 0.3651,
-      "step": 11488
-    },
-    {
-      "epoch": 0.7508659564734331,
-      "grad_norm": 0.448963463306427,
-      "learning_rate": 8.698446872488504e-06,
-      "loss": 0.3901,
-      "step": 11489
-    },
-    {
-      "epoch": 0.7509313116789752,
-      "grad_norm": 0.4558291733264923,
-      "learning_rate": 8.698211876592184e-06,
-      "loss": 0.4189,
-      "step": 11490
-    },
-    {
-      "epoch": 0.7509966668845174,
-      "grad_norm": 0.4616054892539978,
-      "learning_rate": 8.697976862658401e-06,
-      "loss": 0.3803,
-      "step": 11491
-    },
-    {
-      "epoch": 0.7510620220900595,
-      "grad_norm": 0.44523996114730835,
-      "learning_rate": 8.697741830688307e-06,
-      "loss": 0.389,
-      "step": 11492
-    },
-    {
-      "epoch": 0.7511273772956016,
-      "grad_norm": 0.4284723997116089,
-      "learning_rate": 8.697506780683043e-06,
-      "loss": 0.3622,
-      "step": 11493
-    },
-    {
-      "epoch": 0.7511927325011437,
-      "grad_norm": 0.423247754573822,
-      "learning_rate": 8.697271712643758e-06,
-      "loss": 0.2948,
-      "step": 11494
-    },
-    {
-      "epoch": 0.7512580877066858,
-      "grad_norm": 0.4341980814933777,
-      "learning_rate": 8.697036626571598e-06,
-      "loss": 0.3732,
-      "step": 11495
-    },
-    {
-      "epoch": 0.751323442912228,
-      "grad_norm": 0.4351692199707031,
-      "learning_rate": 8.696801522467708e-06,
-      "loss": 0.3614,
-      "step": 11496
-    },
-    {
-      "epoch": 0.75138879811777,
-      "grad_norm": 0.47556573152542114,
-      "learning_rate": 8.696566400333239e-06,
-      "loss": 0.4038,
-      "step": 11497
-    },
-    {
-      "epoch": 0.7514541533233122,
-      "grad_norm": 0.4805038571357727,
-      "learning_rate": 8.696331260169332e-06,
-      "loss": 0.4759,
-      "step": 11498
-    },
-    {
-      "epoch": 0.7515195085288543,
-      "grad_norm": 0.4455169141292572,
-      "learning_rate": 8.696096101977141e-06,
-      "loss": 0.3553,
-      "step": 11499
-    },
-    {
-      "epoch": 0.7515848637343965,
-      "grad_norm": 0.41066446900367737,
-      "learning_rate": 8.695860925757807e-06,
-      "loss": 0.3282,
-      "step": 11500
-    },
-    {
-      "epoch": 0.7516502189399386,
-      "grad_norm": 0.47480136156082153,
-      "learning_rate": 8.695625731512477e-06,
-      "loss": 0.4168,
-      "step": 11501
-    },
-    {
-      "epoch": 0.7517155741454807,
-      "grad_norm": 0.4357443153858185,
-      "learning_rate": 8.695390519242302e-06,
-      "loss": 0.3992,
-      "step": 11502
-    },
-    {
-      "epoch": 0.7517809293510228,
-      "grad_norm": 0.4462246894836426,
-      "learning_rate": 8.695155288948425e-06,
-      "loss": 0.3938,
-      "step": 11503
-    },
-    {
-      "epoch": 0.7518462845565649,
-      "grad_norm": 0.4426057040691376,
-      "learning_rate": 8.694920040632e-06,
-      "loss": 0.3831,
-      "step": 11504
-    },
-    {
-      "epoch": 0.7519116397621071,
-      "grad_norm": 0.44983747601509094,
-      "learning_rate": 8.694684774294167e-06,
-      "loss": 0.3367,
-      "step": 11505
-    },
-    {
-      "epoch": 0.7519769949676491,
-      "grad_norm": 0.4833635985851288,
-      "learning_rate": 8.694449489936076e-06,
-      "loss": 0.4439,
-      "step": 11506
-    },
-    {
-      "epoch": 0.7520423501731913,
-      "grad_norm": 0.453665167093277,
-      "learning_rate": 8.694214187558875e-06,
-      "loss": 0.3747,
-      "step": 11507
-    },
-    {
-      "epoch": 0.7521077053787334,
-      "grad_norm": 0.4427410662174225,
-      "learning_rate": 8.693978867163712e-06,
-      "loss": 0.3837,
-      "step": 11508
-    },
-    {
-      "epoch": 0.7521730605842756,
-      "grad_norm": 0.44865310192108154,
-      "learning_rate": 8.693743528751734e-06,
-      "loss": 0.4089,
-      "step": 11509
-    },
-    {
-      "epoch": 0.7522384157898176,
-      "grad_norm": 0.4555530846118927,
-      "learning_rate": 8.69350817232409e-06,
-      "loss": 0.3709,
-      "step": 11510
-    },
-    {
-      "epoch": 0.7523037709953598,
-      "grad_norm": 0.45255982875823975,
-      "learning_rate": 8.693272797881926e-06,
-      "loss": 0.3983,
-      "step": 11511
-    },
-    {
-      "epoch": 0.7523691262009019,
-      "grad_norm": 0.4345415532588959,
-      "learning_rate": 8.693037405426392e-06,
-      "loss": 0.3691,
-      "step": 11512
-    },
-    {
-      "epoch": 0.752434481406444,
-      "grad_norm": 0.4903244078159332,
-      "learning_rate": 8.692801994958636e-06,
-      "loss": 0.3971,
-      "step": 11513
-    },
-    {
-      "epoch": 0.7524998366119862,
-      "grad_norm": 0.4598952531814575,
-      "learning_rate": 8.692566566479803e-06,
-      "loss": 0.3886,
-      "step": 11514
-    },
-    {
-      "epoch": 0.7525651918175282,
-      "grad_norm": 0.4512850046157837,
-      "learning_rate": 8.692331119991046e-06,
-      "loss": 0.3896,
-      "step": 11515
-    },
-    {
-      "epoch": 0.7526305470230704,
-      "grad_norm": 0.4707580804824829,
-      "learning_rate": 8.69209565549351e-06,
-      "loss": 0.3967,
-      "step": 11516
-    },
-    {
-      "epoch": 0.7526959022286125,
-      "grad_norm": 0.41047024726867676,
-      "learning_rate": 8.691860172988344e-06,
-      "loss": 0.3369,
-      "step": 11517
-    },
-    {
-      "epoch": 0.7527612574341547,
-      "grad_norm": 0.4959609806537628,
-      "learning_rate": 8.691624672476698e-06,
-      "loss": 0.4684,
-      "step": 11518
-    },
-    {
-      "epoch": 0.7528266126396967,
-      "grad_norm": 0.5354952216148376,
-      "learning_rate": 8.691389153959717e-06,
-      "loss": 0.4638,
-      "step": 11519
-    },
-    {
-      "epoch": 0.7528919678452388,
-      "grad_norm": 0.42644286155700684,
-      "learning_rate": 8.691153617438555e-06,
-      "loss": 0.332,
-      "step": 11520
-    },
-    {
-      "epoch": 0.752957323050781,
-      "grad_norm": 0.5352382659912109,
-      "learning_rate": 8.690918062914357e-06,
-      "loss": 0.3896,
-      "step": 11521
-    },
-    {
-      "epoch": 0.7530226782563231,
-      "grad_norm": 0.49492889642715454,
-      "learning_rate": 8.690682490388273e-06,
-      "loss": 0.4288,
-      "step": 11522
-    },
-    {
-      "epoch": 0.7530880334618653,
-      "grad_norm": 0.4558436870574951,
-      "learning_rate": 8.690446899861453e-06,
-      "loss": 0.3642,
-      "step": 11523
-    },
-    {
-      "epoch": 0.7531533886674073,
-      "grad_norm": 0.44027870893478394,
-      "learning_rate": 8.690211291335045e-06,
-      "loss": 0.3486,
-      "step": 11524
-    },
-    {
-      "epoch": 0.7532187438729495,
-      "grad_norm": 0.4419024586677551,
-      "learning_rate": 8.689975664810197e-06,
-      "loss": 0.3875,
-      "step": 11525
-    },
-    {
-      "epoch": 0.7532840990784916,
-      "grad_norm": 0.45365065336227417,
-      "learning_rate": 8.689740020288059e-06,
-      "loss": 0.3595,
-      "step": 11526
-    },
-    {
-      "epoch": 0.7533494542840338,
-      "grad_norm": 0.44982078671455383,
-      "learning_rate": 8.689504357769781e-06,
-      "loss": 0.325,
-      "step": 11527
-    },
-    {
-      "epoch": 0.7534148094895758,
-      "grad_norm": 0.45227643847465515,
-      "learning_rate": 8.689268677256514e-06,
-      "loss": 0.3648,
-      "step": 11528
-    },
-    {
-      "epoch": 0.7534801646951179,
-      "grad_norm": 0.4553734362125397,
-      "learning_rate": 8.689032978749402e-06,
-      "loss": 0.3971,
-      "step": 11529
-    },
-    {
-      "epoch": 0.7535455199006601,
-      "grad_norm": 0.4540592432022095,
-      "learning_rate": 8.6887972622496e-06,
-      "loss": 0.4231,
-      "step": 11530
-    },
-    {
-      "epoch": 0.7536108751062022,
-      "grad_norm": 0.3962876498699188,
-      "learning_rate": 8.688561527758257e-06,
-      "loss": 0.3009,
-      "step": 11531
-    },
-    {
-      "epoch": 0.7536762303117444,
-      "grad_norm": 0.4543502926826477,
-      "learning_rate": 8.68832577527652e-06,
-      "loss": 0.351,
-      "step": 11532
-    },
-    {
-      "epoch": 0.7537415855172864,
-      "grad_norm": 0.4949527978897095,
-      "learning_rate": 8.68809000480554e-06,
-      "loss": 0.4191,
-      "step": 11533
-    },
-    {
-      "epoch": 0.7538069407228286,
-      "grad_norm": 0.42827823758125305,
-      "learning_rate": 8.68785421634647e-06,
-      "loss": 0.3214,
-      "step": 11534
-    },
-    {
-      "epoch": 0.7538722959283707,
-      "grad_norm": 0.4728573262691498,
-      "learning_rate": 8.687618409900455e-06,
-      "loss": 0.4247,
-      "step": 11535
-    },
-    {
-      "epoch": 0.7539376511339129,
-      "grad_norm": 0.4322061836719513,
-      "learning_rate": 8.687382585468648e-06,
-      "loss": 0.3699,
-      "step": 11536
-    },
-    {
-      "epoch": 0.7540030063394549,
-      "grad_norm": 0.45697495341300964,
-      "learning_rate": 8.6871467430522e-06,
-      "loss": 0.4033,
-      "step": 11537
-    },
-    {
-      "epoch": 0.754068361544997,
-      "grad_norm": 0.446732759475708,
-      "learning_rate": 8.686910882652257e-06,
-      "loss": 0.3816,
-      "step": 11538
-    },
-    {
-      "epoch": 0.7541337167505392,
-      "grad_norm": 0.4661967158317566,
-      "learning_rate": 8.686675004269974e-06,
-      "loss": 0.401,
-      "step": 11539
-    },
-    {
-      "epoch": 0.7541990719560813,
-      "grad_norm": 0.46498775482177734,
-      "learning_rate": 8.6864391079065e-06,
-      "loss": 0.4261,
-      "step": 11540
-    },
-    {
-      "epoch": 0.7542644271616235,
-      "grad_norm": 0.4342498183250427,
-      "learning_rate": 8.686203193562985e-06,
-      "loss": 0.3716,
-      "step": 11541
-    },
-    {
-      "epoch": 0.7543297823671655,
-      "grad_norm": 0.42994990944862366,
-      "learning_rate": 8.685967261240583e-06,
-      "loss": 0.3809,
-      "step": 11542
-    },
-    {
-      "epoch": 0.7543951375727077,
-      "grad_norm": 0.4623600244522095,
-      "learning_rate": 8.685731310940437e-06,
-      "loss": 0.4416,
-      "step": 11543
-    },
-    {
-      "epoch": 0.7544604927782498,
-      "grad_norm": 0.40724843740463257,
-      "learning_rate": 8.685495342663706e-06,
-      "loss": 0.3419,
-      "step": 11544
-    },
-    {
-      "epoch": 0.754525847983792,
-      "grad_norm": 0.43778303265571594,
-      "learning_rate": 8.685259356411534e-06,
-      "loss": 0.3604,
-      "step": 11545
-    },
-    {
-      "epoch": 0.754591203189334,
-      "grad_norm": 0.4433067739009857,
-      "learning_rate": 8.685023352185078e-06,
-      "loss": 0.3869,
-      "step": 11546
-    },
-    {
-      "epoch": 0.7546565583948761,
-      "grad_norm": 0.4431113302707672,
-      "learning_rate": 8.684787329985488e-06,
-      "loss": 0.3949,
-      "step": 11547
-    },
-    {
-      "epoch": 0.7547219136004183,
-      "grad_norm": 0.44400447607040405,
-      "learning_rate": 8.684551289813911e-06,
-      "loss": 0.3815,
-      "step": 11548
-    },
-    {
-      "epoch": 0.7547872688059604,
-      "grad_norm": 0.474870890378952,
-      "learning_rate": 8.6843152316715e-06,
-      "loss": 0.4021,
-      "step": 11549
-    },
-    {
-      "epoch": 0.7548526240115025,
-      "grad_norm": 0.4529237449169159,
-      "learning_rate": 8.684079155559411e-06,
-      "loss": 0.3408,
-      "step": 11550
-    },
-    {
-      "epoch": 0.7549179792170446,
-      "grad_norm": 0.4417824149131775,
-      "learning_rate": 8.683843061478789e-06,
-      "loss": 0.3718,
-      "step": 11551
-    },
-    {
-      "epoch": 0.7549833344225868,
-      "grad_norm": 0.44871315360069275,
-      "learning_rate": 8.683606949430788e-06,
-      "loss": 0.3567,
-      "step": 11552
-    },
-    {
-      "epoch": 0.7550486896281289,
-      "grad_norm": 0.43852290511131287,
-      "learning_rate": 8.683370819416561e-06,
-      "loss": 0.3519,
-      "step": 11553
-    },
-    {
-      "epoch": 0.755114044833671,
-      "grad_norm": 0.43814173340797424,
-      "learning_rate": 8.683134671437257e-06,
-      "loss": 0.3662,
-      "step": 11554
-    },
-    {
-      "epoch": 0.7551794000392131,
-      "grad_norm": 0.4667989909648895,
-      "learning_rate": 8.68289850549403e-06,
-      "loss": 0.4218,
-      "step": 11555
-    },
-    {
-      "epoch": 0.7552447552447552,
-      "grad_norm": 0.470702588558197,
-      "learning_rate": 8.68266232158803e-06,
-      "loss": 0.4268,
-      "step": 11556
-    },
-    {
-      "epoch": 0.7553101104502974,
-      "grad_norm": 0.42009657621383667,
-      "learning_rate": 8.682426119720412e-06,
-      "loss": 0.3181,
-      "step": 11557
-    },
-    {
-      "epoch": 0.7553754656558395,
-      "grad_norm": 0.43755048513412476,
-      "learning_rate": 8.682189899892326e-06,
-      "loss": 0.3637,
-      "step": 11558
-    },
-    {
-      "epoch": 0.7554408208613816,
-      "grad_norm": 0.4409218430519104,
-      "learning_rate": 8.681953662104925e-06,
-      "loss": 0.3784,
-      "step": 11559
-    },
-    {
-      "epoch": 0.7555061760669237,
-      "grad_norm": 0.40497660636901855,
-      "learning_rate": 8.681717406359359e-06,
-      "loss": 0.3233,
-      "step": 11560
-    },
-    {
-      "epoch": 0.7555715312724659,
-      "grad_norm": 0.42020371556282043,
-      "learning_rate": 8.681481132656782e-06,
-      "loss": 0.3554,
-      "step": 11561
-    },
-    {
-      "epoch": 0.755636886478008,
-      "grad_norm": 0.42821750044822693,
-      "learning_rate": 8.681244840998347e-06,
-      "loss": 0.3715,
-      "step": 11562
-    },
-    {
-      "epoch": 0.75570224168355,
-      "grad_norm": 0.4186302423477173,
-      "learning_rate": 8.681008531385204e-06,
-      "loss": 0.3454,
-      "step": 11563
-    },
-    {
-      "epoch": 0.7557675968890922,
-      "grad_norm": 0.48060446977615356,
-      "learning_rate": 8.680772203818507e-06,
-      "loss": 0.3975,
-      "step": 11564
-    },
-    {
-      "epoch": 0.7558329520946343,
-      "grad_norm": 0.44453316926956177,
-      "learning_rate": 8.680535858299409e-06,
-      "loss": 0.3829,
-      "step": 11565
-    },
-    {
-      "epoch": 0.7558983073001765,
-      "grad_norm": 0.4027644693851471,
-      "learning_rate": 8.680299494829063e-06,
-      "loss": 0.343,
-      "step": 11566
-    },
-    {
-      "epoch": 0.7559636625057186,
-      "grad_norm": 0.4423559010028839,
-      "learning_rate": 8.680063113408622e-06,
-      "loss": 0.3849,
-      "step": 11567
-    },
-    {
-      "epoch": 0.7560290177112607,
-      "grad_norm": 0.4409993886947632,
-      "learning_rate": 8.679826714039238e-06,
-      "loss": 0.4017,
-      "step": 11568
-    },
-    {
-      "epoch": 0.7560943729168028,
-      "grad_norm": 0.4309713840484619,
-      "learning_rate": 8.679590296722065e-06,
-      "loss": 0.3557,
-      "step": 11569
-    },
-    {
-      "epoch": 0.756159728122345,
-      "grad_norm": 0.4162459373474121,
-      "learning_rate": 8.679353861458252e-06,
-      "loss": 0.3479,
-      "step": 11570
-    },
-    {
-      "epoch": 0.7562250833278871,
-      "grad_norm": 0.4345146119594574,
-      "learning_rate": 8.67911740824896e-06,
-      "loss": 0.3962,
-      "step": 11571
-    },
-    {
-      "epoch": 0.7562904385334291,
-      "grad_norm": 0.4284350275993347,
-      "learning_rate": 8.678880937095336e-06,
-      "loss": 0.3865,
-      "step": 11572
-    },
-    {
-      "epoch": 0.7563557937389713,
-      "grad_norm": 0.427379310131073,
-      "learning_rate": 8.678644447998535e-06,
-      "loss": 0.3639,
-      "step": 11573
-    },
-    {
-      "epoch": 0.7564211489445134,
-      "grad_norm": 0.4237443804740906,
-      "learning_rate": 8.67840794095971e-06,
-      "loss": 0.3371,
-      "step": 11574
-    },
-    {
-      "epoch": 0.7564865041500556,
-      "grad_norm": 0.4654266834259033,
-      "learning_rate": 8.678171415980017e-06,
-      "loss": 0.3696,
-      "step": 11575
-    },
-    {
-      "epoch": 0.7565518593555977,
-      "grad_norm": 0.4269372522830963,
-      "learning_rate": 8.677934873060606e-06,
-      "loss": 0.352,
-      "step": 11576
-    },
-    {
-      "epoch": 0.7566172145611398,
-      "grad_norm": 0.44193196296691895,
-      "learning_rate": 8.677698312202634e-06,
-      "loss": 0.383,
-      "step": 11577
-    },
-    {
-      "epoch": 0.7566825697666819,
-      "grad_norm": 0.44568783044815063,
-      "learning_rate": 8.677461733407251e-06,
-      "loss": 0.3944,
-      "step": 11578
-    },
-    {
-      "epoch": 0.756747924972224,
-      "grad_norm": 0.4438956677913666,
-      "learning_rate": 8.677225136675616e-06,
-      "loss": 0.3602,
-      "step": 11579
-    },
-    {
-      "epoch": 0.7568132801777662,
-      "grad_norm": 0.4461032748222351,
-      "learning_rate": 8.676988522008878e-06,
-      "loss": 0.3527,
-      "step": 11580
-    },
-    {
-      "epoch": 0.7568786353833082,
-      "grad_norm": 0.45144277811050415,
-      "learning_rate": 8.676751889408192e-06,
-      "loss": 0.3758,
-      "step": 11581
-    },
-    {
-      "epoch": 0.7569439905888504,
-      "grad_norm": 0.3841136693954468,
-      "learning_rate": 8.676515238874716e-06,
-      "loss": 0.3076,
-      "step": 11582
-    },
-    {
-      "epoch": 0.7570093457943925,
-      "grad_norm": 0.4647257328033447,
-      "learning_rate": 8.676278570409602e-06,
-      "loss": 0.4046,
-      "step": 11583
-    },
-    {
-      "epoch": 0.7570747009999347,
-      "grad_norm": 0.4637050926685333,
-      "learning_rate": 8.676041884014001e-06,
-      "loss": 0.385,
-      "step": 11584
-    },
-    {
-      "epoch": 0.7571400562054768,
-      "grad_norm": 0.4185953736305237,
-      "learning_rate": 8.675805179689073e-06,
-      "loss": 0.3431,
-      "step": 11585
-    },
-    {
-      "epoch": 0.7572054114110189,
-      "grad_norm": 0.4225028157234192,
-      "learning_rate": 8.675568457435967e-06,
-      "loss": 0.3503,
-      "step": 11586
-    },
-    {
-      "epoch": 0.757270766616561,
-      "grad_norm": 0.41787388920783997,
-      "learning_rate": 8.67533171725584e-06,
-      "loss": 0.3315,
-      "step": 11587
-    },
-    {
-      "epoch": 0.7573361218221031,
-      "grad_norm": 0.4110637307167053,
-      "learning_rate": 8.67509495914985e-06,
-      "loss": 0.327,
-      "step": 11588
-    },
-    {
-      "epoch": 0.7574014770276453,
-      "grad_norm": 0.4453466534614563,
-      "learning_rate": 8.674858183119147e-06,
-      "loss": 0.3696,
-      "step": 11589
-    },
-    {
-      "epoch": 0.7574668322331873,
-      "grad_norm": 0.41781923174858093,
-      "learning_rate": 8.674621389164887e-06,
-      "loss": 0.3557,
-      "step": 11590
-    },
-    {
-      "epoch": 0.7575321874387295,
-      "grad_norm": 0.45862486958503723,
-      "learning_rate": 8.674384577288228e-06,
-      "loss": 0.3765,
-      "step": 11591
-    },
-    {
-      "epoch": 0.7575975426442716,
-      "grad_norm": 0.4591810703277588,
-      "learning_rate": 8.67414774749032e-06,
-      "loss": 0.4035,
-      "step": 11592
-    },
-    {
-      "epoch": 0.7576628978498138,
-      "grad_norm": 0.45626866817474365,
-      "learning_rate": 8.673910899772323e-06,
-      "loss": 0.3924,
-      "step": 11593
-    },
-    {
-      "epoch": 0.7577282530553558,
-      "grad_norm": 0.43373769521713257,
-      "learning_rate": 8.673674034135386e-06,
-      "loss": 0.382,
-      "step": 11594
-    },
-    {
-      "epoch": 0.757793608260898,
-      "grad_norm": 0.4272853136062622,
-      "learning_rate": 8.673437150580671e-06,
-      "loss": 0.3436,
-      "step": 11595
-    },
-    {
-      "epoch": 0.7578589634664401,
-      "grad_norm": 0.4375975430011749,
-      "learning_rate": 8.673200249109329e-06,
-      "loss": 0.3753,
-      "step": 11596
-    },
-    {
-      "epoch": 0.7579243186719822,
-      "grad_norm": 0.43419358134269714,
-      "learning_rate": 8.672963329722518e-06,
-      "loss": 0.3626,
-      "step": 11597
-    },
-    {
-      "epoch": 0.7579896738775244,
-      "grad_norm": 0.4664026200771332,
-      "learning_rate": 8.672726392421391e-06,
-      "loss": 0.3891,
-      "step": 11598
-    },
-    {
-      "epoch": 0.7580550290830664,
-      "grad_norm": 0.4879235327243805,
-      "learning_rate": 8.672489437207106e-06,
-      "loss": 0.4508,
-      "step": 11599
-    },
-    {
-      "epoch": 0.7581203842886086,
-      "grad_norm": 0.4599400758743286,
-      "learning_rate": 8.672252464080817e-06,
-      "loss": 0.3914,
-      "step": 11600
-    },
-    {
-      "epoch": 0.7581857394941507,
-      "grad_norm": 0.44503307342529297,
-      "learning_rate": 8.672015473043683e-06,
-      "loss": 0.3927,
-      "step": 11601
-    },
-    {
-      "epoch": 0.7582510946996929,
-      "grad_norm": 0.43716150522232056,
-      "learning_rate": 8.671778464096855e-06,
-      "loss": 0.3692,
-      "step": 11602
-    },
-    {
-      "epoch": 0.758316449905235,
-      "grad_norm": 0.4063815176486969,
-      "learning_rate": 8.671541437241493e-06,
-      "loss": 0.3268,
-      "step": 11603
-    },
-    {
-      "epoch": 0.758381805110777,
-      "grad_norm": 0.4981492757797241,
-      "learning_rate": 8.671304392478749e-06,
-      "loss": 0.4306,
-      "step": 11604
-    },
-    {
-      "epoch": 0.7584471603163192,
-      "grad_norm": 0.43988507986068726,
-      "learning_rate": 8.671067329809783e-06,
-      "loss": 0.3765,
-      "step": 11605
-    },
-    {
-      "epoch": 0.7585125155218613,
-      "grad_norm": 0.4441068172454834,
-      "learning_rate": 8.67083024923575e-06,
-      "loss": 0.3436,
-      "step": 11606
-    },
-    {
-      "epoch": 0.7585778707274035,
-      "grad_norm": 0.47273120284080505,
-      "learning_rate": 8.670593150757806e-06,
-      "loss": 0.4095,
-      "step": 11607
-    },
-    {
-      "epoch": 0.7586432259329455,
-      "grad_norm": 0.4163842499256134,
-      "learning_rate": 8.670356034377109e-06,
-      "loss": 0.357,
-      "step": 11608
-    },
-    {
-      "epoch": 0.7587085811384877,
-      "grad_norm": 0.43506866693496704,
-      "learning_rate": 8.670118900094812e-06,
-      "loss": 0.3778,
-      "step": 11609
-    },
-    {
-      "epoch": 0.7587739363440298,
-      "grad_norm": 0.46837303042411804,
-      "learning_rate": 8.669881747912074e-06,
-      "loss": 0.4012,
-      "step": 11610
-    },
-    {
-      "epoch": 0.758839291549572,
-      "grad_norm": 0.4551672041416168,
-      "learning_rate": 8.669644577830052e-06,
-      "loss": 0.3732,
-      "step": 11611
-    },
-    {
-      "epoch": 0.758904646755114,
-      "grad_norm": 0.4556867182254791,
-      "learning_rate": 8.669407389849902e-06,
-      "loss": 0.3865,
-      "step": 11612
-    },
-    {
-      "epoch": 0.7589700019606561,
-      "grad_norm": 0.6006876826286316,
-      "learning_rate": 8.66917018397278e-06,
-      "loss": 0.3906,
-      "step": 11613
-    },
-    {
-      "epoch": 0.7590353571661983,
-      "grad_norm": 0.4503045976161957,
-      "learning_rate": 8.668932960199846e-06,
-      "loss": 0.3708,
-      "step": 11614
-    },
-    {
-      "epoch": 0.7591007123717404,
-      "grad_norm": 0.42857062816619873,
-      "learning_rate": 8.668695718532254e-06,
-      "loss": 0.3392,
-      "step": 11615
-    },
-    {
-      "epoch": 0.7591660675772826,
-      "grad_norm": 0.4834391176700592,
-      "learning_rate": 8.668458458971162e-06,
-      "loss": 0.4009,
-      "step": 11616
-    },
-    {
-      "epoch": 0.7592314227828246,
-      "grad_norm": 0.4342235326766968,
-      "learning_rate": 8.668221181517726e-06,
-      "loss": 0.3548,
-      "step": 11617
-    },
-    {
-      "epoch": 0.7592967779883668,
-      "grad_norm": 0.49423882365226746,
-      "learning_rate": 8.667983886173106e-06,
-      "loss": 0.4,
-      "step": 11618
-    },
-    {
-      "epoch": 0.7593621331939089,
-      "grad_norm": 0.4574379324913025,
-      "learning_rate": 8.667746572938458e-06,
-      "loss": 0.4107,
-      "step": 11619
-    },
-    {
-      "epoch": 0.7594274883994511,
-      "grad_norm": 0.46529674530029297,
-      "learning_rate": 8.667509241814938e-06,
-      "loss": 0.3733,
-      "step": 11620
-    },
-    {
-      "epoch": 0.7594928436049931,
-      "grad_norm": 0.41194331645965576,
-      "learning_rate": 8.667271892803706e-06,
-      "loss": 0.3436,
-      "step": 11621
-    },
-    {
-      "epoch": 0.7595581988105352,
-      "grad_norm": 0.412412166595459,
-      "learning_rate": 8.667034525905918e-06,
-      "loss": 0.336,
-      "step": 11622
-    },
-    {
-      "epoch": 0.7596235540160774,
-      "grad_norm": 0.43101462721824646,
-      "learning_rate": 8.666797141122731e-06,
-      "loss": 0.3668,
-      "step": 11623
-    },
-    {
-      "epoch": 0.7596889092216195,
-      "grad_norm": 0.4349445700645447,
-      "learning_rate": 8.666559738455306e-06,
-      "loss": 0.3421,
-      "step": 11624
-    },
-    {
-      "epoch": 0.7597542644271617,
-      "grad_norm": 0.4164186716079712,
-      "learning_rate": 8.666322317904798e-06,
-      "loss": 0.3516,
-      "step": 11625
-    },
-    {
-      "epoch": 0.7598196196327037,
-      "grad_norm": 0.4714847505092621,
-      "learning_rate": 8.666084879472367e-06,
-      "loss": 0.3593,
-      "step": 11626
-    },
-    {
-      "epoch": 0.7598849748382459,
-      "grad_norm": 0.4621010720729828,
-      "learning_rate": 8.665847423159168e-06,
-      "loss": 0.3688,
-      "step": 11627
-    },
-    {
-      "epoch": 0.759950330043788,
-      "grad_norm": 0.41812488436698914,
-      "learning_rate": 8.665609948966363e-06,
-      "loss": 0.3475,
-      "step": 11628
-    },
-    {
-      "epoch": 0.7600156852493302,
-      "grad_norm": 0.4732639789581299,
-      "learning_rate": 8.665372456895108e-06,
-      "loss": 0.3967,
-      "step": 11629
-    },
-    {
-      "epoch": 0.7600810404548722,
-      "grad_norm": 0.5000624060630798,
-      "learning_rate": 8.66513494694656e-06,
-      "loss": 0.4287,
-      "step": 11630
-    },
-    {
-      "epoch": 0.7601463956604143,
-      "grad_norm": 0.44705361127853394,
-      "learning_rate": 8.664897419121881e-06,
-      "loss": 0.366,
-      "step": 11631
-    },
-    {
-      "epoch": 0.7602117508659565,
-      "grad_norm": 0.4290550649166107,
-      "learning_rate": 8.664659873422228e-06,
-      "loss": 0.3688,
-      "step": 11632
-    },
-    {
-      "epoch": 0.7602771060714986,
-      "grad_norm": 0.5169782638549805,
-      "learning_rate": 8.664422309848758e-06,
-      "loss": 0.4368,
-      "step": 11633
-    },
-    {
-      "epoch": 0.7603424612770407,
-      "grad_norm": 0.42105633020401,
-      "learning_rate": 8.66418472840263e-06,
-      "loss": 0.353,
-      "step": 11634
-    },
-    {
-      "epoch": 0.7604078164825828,
-      "grad_norm": 0.4356044828891754,
-      "learning_rate": 8.663947129085006e-06,
-      "loss": 0.3447,
-      "step": 11635
-    },
-    {
-      "epoch": 0.760473171688125,
-      "grad_norm": 0.44438812136650085,
-      "learning_rate": 8.663709511897043e-06,
-      "loss": 0.3641,
-      "step": 11636
-    },
-    {
-      "epoch": 0.7605385268936671,
-      "grad_norm": 0.5030198097229004,
-      "learning_rate": 8.663471876839898e-06,
-      "loss": 0.4122,
-      "step": 11637
-    },
-    {
-      "epoch": 0.7606038820992091,
-      "grad_norm": 0.4163077175617218,
-      "learning_rate": 8.663234223914732e-06,
-      "loss": 0.2977,
-      "step": 11638
-    },
-    {
-      "epoch": 0.7606692373047513,
-      "grad_norm": 0.48841458559036255,
-      "learning_rate": 8.662996553122702e-06,
-      "loss": 0.3781,
-      "step": 11639
-    },
-    {
-      "epoch": 0.7607345925102934,
-      "grad_norm": 0.42547884583473206,
-      "learning_rate": 8.662758864464971e-06,
-      "loss": 0.3418,
-      "step": 11640
-    },
-    {
-      "epoch": 0.7607999477158356,
-      "grad_norm": 0.39035564661026,
-      "learning_rate": 8.662521157942694e-06,
-      "loss": 0.2919,
-      "step": 11641
-    },
-    {
-      "epoch": 0.7608653029213777,
-      "grad_norm": 0.47192656993865967,
-      "learning_rate": 8.662283433557033e-06,
-      "loss": 0.413,
-      "step": 11642
-    },
-    {
-      "epoch": 0.7609306581269198,
-      "grad_norm": 0.44570621848106384,
-      "learning_rate": 8.66204569130915e-06,
-      "loss": 0.4303,
-      "step": 11643
-    },
-    {
-      "epoch": 0.7609960133324619,
-      "grad_norm": 0.45997560024261475,
-      "learning_rate": 8.661807931200199e-06,
-      "loss": 0.3719,
-      "step": 11644
-    },
-    {
-      "epoch": 0.7610613685380041,
-      "grad_norm": 0.47064775228500366,
-      "learning_rate": 8.66157015323134e-06,
-      "loss": 0.4037,
-      "step": 11645
-    },
-    {
-      "epoch": 0.7611267237435462,
-      "grad_norm": 0.42042768001556396,
-      "learning_rate": 8.661332357403738e-06,
-      "loss": 0.3108,
-      "step": 11646
-    },
-    {
-      "epoch": 0.7611920789490882,
-      "grad_norm": 0.45020681619644165,
-      "learning_rate": 8.66109454371855e-06,
-      "loss": 0.3675,
-      "step": 11647
-    },
-    {
-      "epoch": 0.7612574341546304,
-      "grad_norm": 0.4344450831413269,
-      "learning_rate": 8.660856712176933e-06,
-      "loss": 0.3771,
-      "step": 11648
-    },
-    {
-      "epoch": 0.7613227893601725,
-      "grad_norm": 0.43782344460487366,
-      "learning_rate": 8.660618862780051e-06,
-      "loss": 0.3482,
-      "step": 11649
-    },
-    {
-      "epoch": 0.7613881445657147,
-      "grad_norm": 0.45356184244155884,
-      "learning_rate": 8.660380995529063e-06,
-      "loss": 0.3907,
-      "step": 11650
-    },
-    {
-      "epoch": 0.7614534997712568,
-      "grad_norm": 0.4710615277290344,
-      "learning_rate": 8.660143110425127e-06,
-      "loss": 0.3906,
-      "step": 11651
-    },
-    {
-      "epoch": 0.7615188549767989,
-      "grad_norm": 0.431894987821579,
-      "learning_rate": 8.659905207469408e-06,
-      "loss": 0.3656,
-      "step": 11652
-    },
-    {
-      "epoch": 0.761584210182341,
-      "grad_norm": 0.4785623550415039,
-      "learning_rate": 8.659667286663062e-06,
-      "loss": 0.4397,
-      "step": 11653
-    },
-    {
-      "epoch": 0.7616495653878832,
-      "grad_norm": 0.4310261309146881,
-      "learning_rate": 8.65942934800725e-06,
-      "loss": 0.3557,
-      "step": 11654
-    },
-    {
-      "epoch": 0.7617149205934253,
-      "grad_norm": 0.436513751745224,
-      "learning_rate": 8.659191391503135e-06,
-      "loss": 0.3647,
-      "step": 11655
-    },
-    {
-      "epoch": 0.7617802757989673,
-      "grad_norm": 0.44582149386405945,
-      "learning_rate": 8.658953417151874e-06,
-      "loss": 0.4017,
-      "step": 11656
-    },
-    {
-      "epoch": 0.7618456310045095,
-      "grad_norm": 0.5000804662704468,
-      "learning_rate": 8.65871542495463e-06,
-      "loss": 0.4561,
-      "step": 11657
-    },
-    {
-      "epoch": 0.7619109862100516,
-      "grad_norm": 0.46625787019729614,
-      "learning_rate": 8.658477414912564e-06,
-      "loss": 0.4236,
-      "step": 11658
-    },
-    {
-      "epoch": 0.7619763414155938,
-      "grad_norm": 0.4202720522880554,
-      "learning_rate": 8.658239387026836e-06,
-      "loss": 0.3506,
-      "step": 11659
-    },
-    {
-      "epoch": 0.7620416966211359,
-      "grad_norm": 0.44120508432388306,
-      "learning_rate": 8.658001341298608e-06,
-      "loss": 0.3746,
-      "step": 11660
-    },
-    {
-      "epoch": 0.762107051826678,
-      "grad_norm": 0.49611538648605347,
-      "learning_rate": 8.65776327772904e-06,
-      "loss": 0.4009,
-      "step": 11661
-    },
-    {
-      "epoch": 0.7621724070322201,
-      "grad_norm": 0.46743810176849365,
-      "learning_rate": 8.657525196319292e-06,
-      "loss": 0.4368,
-      "step": 11662
-    },
-    {
-      "epoch": 0.7622377622377622,
-      "grad_norm": 0.40163370966911316,
-      "learning_rate": 8.657287097070528e-06,
-      "loss": 0.336,
-      "step": 11663
-    },
-    {
-      "epoch": 0.7623031174433044,
-      "grad_norm": 0.44429242610931396,
-      "learning_rate": 8.657048979983906e-06,
-      "loss": 0.418,
-      "step": 11664
-    },
-    {
-      "epoch": 0.7623684726488464,
-      "grad_norm": 0.5027108192443848,
-      "learning_rate": 8.656810845060591e-06,
-      "loss": 0.4699,
-      "step": 11665
-    },
-    {
-      "epoch": 0.7624338278543886,
-      "grad_norm": 0.4208289384841919,
-      "learning_rate": 8.656572692301742e-06,
-      "loss": 0.3212,
-      "step": 11666
-    },
-    {
-      "epoch": 0.7624991830599307,
-      "grad_norm": 0.41135329008102417,
-      "learning_rate": 8.65633452170852e-06,
-      "loss": 0.345,
-      "step": 11667
-    },
-    {
-      "epoch": 0.7625645382654729,
-      "grad_norm": 0.45532602071762085,
-      "learning_rate": 8.656096333282092e-06,
-      "loss": 0.3948,
-      "step": 11668
-    },
-    {
-      "epoch": 0.762629893471015,
-      "grad_norm": 0.477816104888916,
-      "learning_rate": 8.65585812702361e-06,
-      "loss": 0.4097,
-      "step": 11669
-    },
-    {
-      "epoch": 0.7626952486765571,
-      "grad_norm": 0.4528370201587677,
-      "learning_rate": 8.655619902934244e-06,
-      "loss": 0.3611,
-      "step": 11670
-    },
-    {
-      "epoch": 0.7627606038820992,
-      "grad_norm": 0.4283353090286255,
-      "learning_rate": 8.655381661015154e-06,
-      "loss": 0.333,
-      "step": 11671
-    },
-    {
-      "epoch": 0.7628259590876413,
-      "grad_norm": 0.46747609972953796,
-      "learning_rate": 8.6551434012675e-06,
-      "loss": 0.3931,
-      "step": 11672
-    },
-    {
-      "epoch": 0.7628913142931835,
-      "grad_norm": 0.40733906626701355,
-      "learning_rate": 8.654905123692448e-06,
-      "loss": 0.3223,
-      "step": 11673
-    },
-    {
-      "epoch": 0.7629566694987255,
-      "grad_norm": 0.4334840774536133,
-      "learning_rate": 8.654666828291155e-06,
-      "loss": 0.3357,
-      "step": 11674
-    },
-    {
-      "epoch": 0.7630220247042677,
-      "grad_norm": 0.4467930197715759,
-      "learning_rate": 8.654428515064787e-06,
-      "loss": 0.356,
-      "step": 11675
-    },
-    {
-      "epoch": 0.7630873799098098,
-      "grad_norm": 0.4429667592048645,
-      "learning_rate": 8.654190184014503e-06,
-      "loss": 0.3475,
-      "step": 11676
-    },
-    {
-      "epoch": 0.763152735115352,
-      "grad_norm": 0.4922010898590088,
-      "learning_rate": 8.65395183514147e-06,
-      "loss": 0.4329,
-      "step": 11677
-    },
-    {
-      "epoch": 0.763218090320894,
-      "grad_norm": 0.4463573694229126,
-      "learning_rate": 8.653713468446847e-06,
-      "loss": 0.3797,
-      "step": 11678
-    },
-    {
-      "epoch": 0.7632834455264362,
-      "grad_norm": 0.4786367118358612,
-      "learning_rate": 8.653475083931798e-06,
-      "loss": 0.4105,
-      "step": 11679
-    },
-    {
-      "epoch": 0.7633488007319783,
-      "grad_norm": 0.5591999888420105,
-      "learning_rate": 8.653236681597486e-06,
-      "loss": 0.3655,
-      "step": 11680
-    },
-    {
-      "epoch": 0.7634141559375204,
-      "grad_norm": 0.4530841112136841,
-      "learning_rate": 8.652998261445072e-06,
-      "loss": 0.3665,
-      "step": 11681
-    },
-    {
-      "epoch": 0.7634795111430626,
-      "grad_norm": 0.46425333619117737,
-      "learning_rate": 8.65275982347572e-06,
-      "loss": 0.415,
-      "step": 11682
-    },
-    {
-      "epoch": 0.7635448663486046,
-      "grad_norm": 0.4381583333015442,
-      "learning_rate": 8.652521367690592e-06,
-      "loss": 0.3453,
-      "step": 11683
-    },
-    {
-      "epoch": 0.7636102215541468,
-      "grad_norm": 0.4222116470336914,
-      "learning_rate": 8.652282894090853e-06,
-      "loss": 0.3456,
-      "step": 11684
-    },
-    {
-      "epoch": 0.7636755767596889,
-      "grad_norm": 0.4758513271808624,
-      "learning_rate": 8.652044402677666e-06,
-      "loss": 0.4071,
-      "step": 11685
-    },
-    {
-      "epoch": 0.7637409319652311,
-      "grad_norm": 0.43865343928337097,
-      "learning_rate": 8.651805893452192e-06,
-      "loss": 0.3294,
-      "step": 11686
-    },
-    {
-      "epoch": 0.7638062871707731,
-      "grad_norm": 0.47241854667663574,
-      "learning_rate": 8.651567366415596e-06,
-      "loss": 0.3922,
-      "step": 11687
-    },
-    {
-      "epoch": 0.7638716423763152,
-      "grad_norm": 0.5159560441970825,
-      "learning_rate": 8.651328821569041e-06,
-      "loss": 0.3647,
-      "step": 11688
-    },
-    {
-      "epoch": 0.7639369975818574,
-      "grad_norm": 0.45398226380348206,
-      "learning_rate": 8.65109025891369e-06,
-      "loss": 0.3486,
-      "step": 11689
-    },
-    {
-      "epoch": 0.7640023527873995,
-      "grad_norm": 0.45617467164993286,
-      "learning_rate": 8.650851678450707e-06,
-      "loss": 0.3656,
-      "step": 11690
-    },
-    {
-      "epoch": 0.7640677079929417,
-      "grad_norm": 0.4122573733329773,
-      "learning_rate": 8.650613080181256e-06,
-      "loss": 0.3143,
-      "step": 11691
-    },
-    {
-      "epoch": 0.7641330631984837,
-      "grad_norm": 0.45565274357795715,
-      "learning_rate": 8.650374464106499e-06,
-      "loss": 0.3446,
-      "step": 11692
-    },
-    {
-      "epoch": 0.7641984184040259,
-      "grad_norm": 0.43330758810043335,
-      "learning_rate": 8.650135830227601e-06,
-      "loss": 0.3757,
-      "step": 11693
-    },
-    {
-      "epoch": 0.764263773609568,
-      "grad_norm": 0.4667772650718689,
-      "learning_rate": 8.64989717854573e-06,
-      "loss": 0.3864,
-      "step": 11694
-    },
-    {
-      "epoch": 0.7643291288151102,
-      "grad_norm": 0.5273546576499939,
-      "learning_rate": 8.649658509062042e-06,
-      "loss": 0.4488,
-      "step": 11695
-    },
-    {
-      "epoch": 0.7643944840206522,
-      "grad_norm": 0.515292763710022,
-      "learning_rate": 8.649419821777705e-06,
-      "loss": 0.4618,
-      "step": 11696
-    },
-    {
-      "epoch": 0.7644598392261943,
-      "grad_norm": 0.4481314420700073,
-      "learning_rate": 8.649181116693886e-06,
-      "loss": 0.3739,
-      "step": 11697
-    },
-    {
-      "epoch": 0.7645251944317365,
-      "grad_norm": 0.45073914527893066,
-      "learning_rate": 8.648942393811744e-06,
-      "loss": 0.3912,
-      "step": 11698
-    },
-    {
-      "epoch": 0.7645905496372786,
-      "grad_norm": 0.5078251957893372,
-      "learning_rate": 8.648703653132447e-06,
-      "loss": 0.3904,
-      "step": 11699
-    },
-    {
-      "epoch": 0.7646559048428208,
-      "grad_norm": 0.39665457606315613,
-      "learning_rate": 8.648464894657158e-06,
-      "loss": 0.3211,
-      "step": 11700
-    },
-    {
-      "epoch": 0.7647212600483628,
-      "grad_norm": 0.45499134063720703,
-      "learning_rate": 8.648226118387041e-06,
-      "loss": 0.3692,
-      "step": 11701
-    },
-    {
-      "epoch": 0.764786615253905,
-      "grad_norm": 0.4568077623844147,
-      "learning_rate": 8.647987324323264e-06,
-      "loss": 0.3588,
-      "step": 11702
-    },
-    {
-      "epoch": 0.7648519704594471,
-      "grad_norm": 0.46008485555648804,
-      "learning_rate": 8.647748512466986e-06,
-      "loss": 0.4029,
-      "step": 11703
-    },
-    {
-      "epoch": 0.7649173256649893,
-      "grad_norm": 0.45498141646385193,
-      "learning_rate": 8.647509682819377e-06,
-      "loss": 0.3804,
-      "step": 11704
-    },
-    {
-      "epoch": 0.7649826808705313,
-      "grad_norm": 0.4846314787864685,
-      "learning_rate": 8.647270835381598e-06,
-      "loss": 0.4366,
-      "step": 11705
-    },
-    {
-      "epoch": 0.7650480360760734,
-      "grad_norm": 0.44207775592803955,
-      "learning_rate": 8.647031970154817e-06,
-      "loss": 0.3541,
-      "step": 11706
-    },
-    {
-      "epoch": 0.7651133912816156,
-      "grad_norm": 0.44153493642807007,
-      "learning_rate": 8.646793087140197e-06,
-      "loss": 0.3694,
-      "step": 11707
-    },
-    {
-      "epoch": 0.7651787464871577,
-      "grad_norm": 0.4551956057548523,
-      "learning_rate": 8.646554186338902e-06,
-      "loss": 0.3745,
-      "step": 11708
-    },
-    {
-      "epoch": 0.7652441016926999,
-      "grad_norm": 0.499629408121109,
-      "learning_rate": 8.646315267752102e-06,
-      "loss": 0.4758,
-      "step": 11709
-    },
-    {
-      "epoch": 0.7653094568982419,
-      "grad_norm": 0.44776612520217896,
-      "learning_rate": 8.646076331380957e-06,
-      "loss": 0.3833,
-      "step": 11710
-    },
-    {
-      "epoch": 0.7653748121037841,
-      "grad_norm": 0.48136380314826965,
-      "learning_rate": 8.645837377226635e-06,
-      "loss": 0.4217,
-      "step": 11711
-    },
-    {
-      "epoch": 0.7654401673093262,
-      "grad_norm": 0.4859139919281006,
-      "learning_rate": 8.645598405290303e-06,
-      "loss": 0.4277,
-      "step": 11712
-    },
-    {
-      "epoch": 0.7655055225148684,
-      "grad_norm": 0.4509076774120331,
-      "learning_rate": 8.645359415573122e-06,
-      "loss": 0.3796,
-      "step": 11713
-    },
-    {
-      "epoch": 0.7655708777204104,
-      "grad_norm": 0.4640675187110901,
-      "learning_rate": 8.645120408076262e-06,
-      "loss": 0.4104,
-      "step": 11714
-    },
-    {
-      "epoch": 0.7656362329259525,
-      "grad_norm": 0.4932800233364105,
-      "learning_rate": 8.644881382800888e-06,
-      "loss": 0.4048,
-      "step": 11715
-    },
-    {
-      "epoch": 0.7657015881314947,
-      "grad_norm": 0.431494802236557,
-      "learning_rate": 8.644642339748161e-06,
-      "loss": 0.3651,
-      "step": 11716
-    },
-    {
-      "epoch": 0.7657669433370368,
-      "grad_norm": 0.4577261209487915,
-      "learning_rate": 8.644403278919254e-06,
-      "loss": 0.3665,
-      "step": 11717
-    },
-    {
-      "epoch": 0.765832298542579,
-      "grad_norm": 0.4488811194896698,
-      "learning_rate": 8.644164200315327e-06,
-      "loss": 0.3928,
-      "step": 11718
-    },
-    {
-      "epoch": 0.765897653748121,
-      "grad_norm": 0.45216381549835205,
-      "learning_rate": 8.643925103937552e-06,
-      "loss": 0.3864,
-      "step": 11719
-    },
-    {
-      "epoch": 0.7659630089536632,
-      "grad_norm": 0.4638056755065918,
-      "learning_rate": 8.64368598978709e-06,
-      "loss": 0.4073,
-      "step": 11720
-    },
-    {
-      "epoch": 0.7660283641592053,
-      "grad_norm": 0.4283357262611389,
-      "learning_rate": 8.64344685786511e-06,
-      "loss": 0.3315,
-      "step": 11721
-    },
-    {
-      "epoch": 0.7660937193647473,
-      "grad_norm": 0.4663817286491394,
-      "learning_rate": 8.643207708172776e-06,
-      "loss": 0.4222,
-      "step": 11722
-    },
-    {
-      "epoch": 0.7661590745702895,
-      "grad_norm": 0.4846576452255249,
-      "learning_rate": 8.642968540711257e-06,
-      "loss": 0.4261,
-      "step": 11723
-    },
-    {
-      "epoch": 0.7662244297758316,
-      "grad_norm": 0.4734724164009094,
-      "learning_rate": 8.642729355481719e-06,
-      "loss": 0.3842,
-      "step": 11724
-    },
-    {
-      "epoch": 0.7662897849813738,
-      "grad_norm": 0.4373061954975128,
-      "learning_rate": 8.642490152485326e-06,
-      "loss": 0.3705,
-      "step": 11725
-    },
-    {
-      "epoch": 0.7663551401869159,
-      "grad_norm": 0.4628601670265198,
-      "learning_rate": 8.642250931723247e-06,
-      "loss": 0.395,
-      "step": 11726
-    },
-    {
-      "epoch": 0.766420495392458,
-      "grad_norm": 0.4402123987674713,
-      "learning_rate": 8.64201169319665e-06,
-      "loss": 0.3592,
-      "step": 11727
-    },
-    {
-      "epoch": 0.7664858505980001,
-      "grad_norm": 0.4120115637779236,
-      "learning_rate": 8.641772436906698e-06,
-      "loss": 0.3524,
-      "step": 11728
-    },
-    {
-      "epoch": 0.7665512058035423,
-      "grad_norm": 0.48394501209259033,
-      "learning_rate": 8.641533162854561e-06,
-      "loss": 0.4382,
-      "step": 11729
-    },
-    {
-      "epoch": 0.7666165610090844,
-      "grad_norm": 0.47838151454925537,
-      "learning_rate": 8.641293871041407e-06,
-      "loss": 0.3989,
-      "step": 11730
-    },
-    {
-      "epoch": 0.7666819162146264,
-      "grad_norm": 0.42966246604919434,
-      "learning_rate": 8.6410545614684e-06,
-      "loss": 0.3814,
-      "step": 11731
-    },
-    {
-      "epoch": 0.7667472714201686,
-      "grad_norm": 0.4653118848800659,
-      "learning_rate": 8.640815234136708e-06,
-      "loss": 0.4118,
-      "step": 11732
-    },
-    {
-      "epoch": 0.7668126266257107,
-      "grad_norm": 0.44381260871887207,
-      "learning_rate": 8.6405758890475e-06,
-      "loss": 0.3608,
-      "step": 11733
-    },
-    {
-      "epoch": 0.7668779818312529,
-      "grad_norm": 0.45664727687835693,
-      "learning_rate": 8.640336526201942e-06,
-      "loss": 0.4156,
-      "step": 11734
-    },
-    {
-      "epoch": 0.766943337036795,
-      "grad_norm": 0.45807671546936035,
-      "learning_rate": 8.6400971456012e-06,
-      "loss": 0.3962,
-      "step": 11735
-    },
-    {
-      "epoch": 0.7670086922423371,
-      "grad_norm": 0.43573200702667236,
-      "learning_rate": 8.639857747246444e-06,
-      "loss": 0.3771,
-      "step": 11736
-    },
-    {
-      "epoch": 0.7670740474478792,
-      "grad_norm": 0.4622085690498352,
-      "learning_rate": 8.639618331138842e-06,
-      "loss": 0.405,
-      "step": 11737
-    },
-    {
-      "epoch": 0.7671394026534214,
-      "grad_norm": 0.4761989712715149,
-      "learning_rate": 8.63937889727956e-06,
-      "loss": 0.4079,
-      "step": 11738
-    },
-    {
-      "epoch": 0.7672047578589635,
-      "grad_norm": 0.4308473765850067,
-      "learning_rate": 8.639139445669765e-06,
-      "loss": 0.3626,
-      "step": 11739
-    },
-    {
-      "epoch": 0.7672701130645055,
-      "grad_norm": 0.46983495354652405,
-      "learning_rate": 8.638899976310628e-06,
-      "loss": 0.4076,
-      "step": 11740
-    },
-    {
-      "epoch": 0.7673354682700477,
-      "grad_norm": 0.43866461515426636,
-      "learning_rate": 8.638660489203314e-06,
-      "loss": 0.3467,
-      "step": 11741
-    },
-    {
-      "epoch": 0.7674008234755898,
-      "grad_norm": 0.46919476985931396,
-      "learning_rate": 8.638420984348992e-06,
-      "loss": 0.4043,
-      "step": 11742
-    },
-    {
-      "epoch": 0.767466178681132,
-      "grad_norm": 0.508739173412323,
-      "learning_rate": 8.638181461748831e-06,
-      "loss": 0.4371,
-      "step": 11743
-    },
-    {
-      "epoch": 0.7675315338866741,
-      "grad_norm": 0.4344213008880615,
-      "learning_rate": 8.637941921403998e-06,
-      "loss": 0.3567,
-      "step": 11744
-    },
-    {
-      "epoch": 0.7675968890922162,
-      "grad_norm": 0.445400208234787,
-      "learning_rate": 8.637702363315663e-06,
-      "loss": 0.399,
-      "step": 11745
-    },
-    {
-      "epoch": 0.7676622442977583,
-      "grad_norm": 0.43348976969718933,
-      "learning_rate": 8.637462787484994e-06,
-      "loss": 0.3566,
-      "step": 11746
-    },
-    {
-      "epoch": 0.7677275995033004,
-      "grad_norm": 0.47843995690345764,
-      "learning_rate": 8.637223193913157e-06,
-      "loss": 0.412,
-      "step": 11747
-    },
-    {
-      "epoch": 0.7677929547088426,
-      "grad_norm": 0.5175793170928955,
-      "learning_rate": 8.636983582601324e-06,
-      "loss": 0.4553,
-      "step": 11748
-    },
-    {
-      "epoch": 0.7678583099143846,
-      "grad_norm": 0.4482249617576599,
-      "learning_rate": 8.636743953550662e-06,
-      "loss": 0.3796,
-      "step": 11749
-    },
-    {
-      "epoch": 0.7679236651199268,
-      "grad_norm": 0.43500909209251404,
-      "learning_rate": 8.636504306762339e-06,
-      "loss": 0.3559,
-      "step": 11750
-    },
-    {
-      "epoch": 0.7679890203254689,
-      "grad_norm": 0.4549643397331238,
-      "learning_rate": 8.636264642237523e-06,
-      "loss": 0.3493,
-      "step": 11751
-    },
-    {
-      "epoch": 0.7680543755310111,
-      "grad_norm": 0.4634072184562683,
-      "learning_rate": 8.636024959977387e-06,
-      "loss": 0.3774,
-      "step": 11752
-    },
-    {
-      "epoch": 0.7681197307365532,
-      "grad_norm": 0.4879359304904938,
-      "learning_rate": 8.635785259983097e-06,
-      "loss": 0.4071,
-      "step": 11753
-    },
-    {
-      "epoch": 0.7681850859420953,
-      "grad_norm": 0.4796138107776642,
-      "learning_rate": 8.635545542255823e-06,
-      "loss": 0.3805,
-      "step": 11754
-    },
-    {
-      "epoch": 0.7682504411476374,
-      "grad_norm": 0.4613151550292969,
-      "learning_rate": 8.635305806796733e-06,
-      "loss": 0.4339,
-      "step": 11755
-    },
-    {
-      "epoch": 0.7683157963531795,
-      "grad_norm": 0.5890063047409058,
-      "learning_rate": 8.635066053607e-06,
-      "loss": 0.4169,
-      "step": 11756
-    },
-    {
-      "epoch": 0.7683811515587217,
-      "grad_norm": 0.45519086718559265,
-      "learning_rate": 8.634826282687787e-06,
-      "loss": 0.3775,
-      "step": 11757
-    },
-    {
-      "epoch": 0.7684465067642637,
-      "grad_norm": 0.45684412121772766,
-      "learning_rate": 8.63458649404027e-06,
-      "loss": 0.3844,
-      "step": 11758
-    },
-    {
-      "epoch": 0.7685118619698059,
-      "grad_norm": 0.42663997411727905,
-      "learning_rate": 8.634346687665613e-06,
-      "loss": 0.3461,
-      "step": 11759
-    },
-    {
-      "epoch": 0.768577217175348,
-      "grad_norm": 0.45491963624954224,
-      "learning_rate": 8.634106863564988e-06,
-      "loss": 0.4126,
-      "step": 11760
-    },
-    {
-      "epoch": 0.7686425723808902,
-      "grad_norm": 0.4103669226169586,
-      "learning_rate": 8.633867021739567e-06,
-      "loss": 0.3376,
-      "step": 11761
-    },
-    {
-      "epoch": 0.7687079275864322,
-      "grad_norm": 0.46748030185699463,
-      "learning_rate": 8.633627162190516e-06,
-      "loss": 0.4257,
-      "step": 11762
-    },
-    {
-      "epoch": 0.7687732827919744,
-      "grad_norm": 0.4791647493839264,
-      "learning_rate": 8.633387284919007e-06,
-      "loss": 0.416,
-      "step": 11763
-    },
-    {
-      "epoch": 0.7688386379975165,
-      "grad_norm": 0.4199827015399933,
-      "learning_rate": 8.63314738992621e-06,
-      "loss": 0.3352,
-      "step": 11764
-    },
-    {
-      "epoch": 0.7689039932030586,
-      "grad_norm": 0.4442897439002991,
-      "learning_rate": 8.632907477213293e-06,
-      "loss": 0.3389,
-      "step": 11765
-    },
-    {
-      "epoch": 0.7689693484086008,
-      "grad_norm": 0.45387426018714905,
-      "learning_rate": 8.63266754678143e-06,
-      "loss": 0.3596,
-      "step": 11766
-    },
-    {
-      "epoch": 0.7690347036141428,
-      "grad_norm": 0.4821358323097229,
-      "learning_rate": 8.632427598631787e-06,
-      "loss": 0.4174,
-      "step": 11767
-    },
-    {
-      "epoch": 0.769100058819685,
-      "grad_norm": 0.4582551419734955,
-      "learning_rate": 8.632187632765538e-06,
-      "loss": 0.3776,
-      "step": 11768
-    },
-    {
-      "epoch": 0.7691654140252271,
-      "grad_norm": 0.4344165027141571,
-      "learning_rate": 8.63194764918385e-06,
-      "loss": 0.3635,
-      "step": 11769
-    },
-    {
-      "epoch": 0.7692307692307693,
-      "grad_norm": 0.5245004892349243,
-      "learning_rate": 8.631707647887895e-06,
-      "loss": 0.3813,
-      "step": 11770
-    },
-    {
-      "epoch": 0.7692961244363113,
-      "grad_norm": 0.4463549256324768,
-      "learning_rate": 8.631467628878844e-06,
-      "loss": 0.368,
-      "step": 11771
-    },
-    {
-      "epoch": 0.7693614796418534,
-      "grad_norm": 0.43788066506385803,
-      "learning_rate": 8.631227592157869e-06,
-      "loss": 0.349,
-      "step": 11772
-    },
-    {
-      "epoch": 0.7694268348473956,
-      "grad_norm": 0.5027048587799072,
-      "learning_rate": 8.630987537726136e-06,
-      "loss": 0.4162,
-      "step": 11773
-    },
-    {
-      "epoch": 0.7694921900529377,
-      "grad_norm": 0.4503048062324524,
-      "learning_rate": 8.630747465584821e-06,
-      "loss": 0.3879,
-      "step": 11774
-    },
-    {
-      "epoch": 0.7695575452584799,
-      "grad_norm": 0.45998138189315796,
-      "learning_rate": 8.630507375735093e-06,
-      "loss": 0.3644,
-      "step": 11775
-    },
-    {
-      "epoch": 0.7696229004640219,
-      "grad_norm": 0.4338878393173218,
-      "learning_rate": 8.630267268178121e-06,
-      "loss": 0.3608,
-      "step": 11776
-    },
-    {
-      "epoch": 0.7696882556695641,
-      "grad_norm": 0.42915183305740356,
-      "learning_rate": 8.630027142915081e-06,
-      "loss": 0.353,
-      "step": 11777
-    },
-    {
-      "epoch": 0.7697536108751062,
-      "grad_norm": 0.49270331859588623,
-      "learning_rate": 8.629786999947138e-06,
-      "loss": 0.4232,
-      "step": 11778
-    },
-    {
-      "epoch": 0.7698189660806484,
-      "grad_norm": 0.3997019827365875,
-      "learning_rate": 8.629546839275467e-06,
-      "loss": 0.3377,
-      "step": 11779
-    },
-    {
-      "epoch": 0.7698843212861904,
-      "grad_norm": 0.44041168689727783,
-      "learning_rate": 8.62930666090124e-06,
-      "loss": 0.4241,
-      "step": 11780
-    },
-    {
-      "epoch": 0.7699496764917325,
-      "grad_norm": 0.43760037422180176,
-      "learning_rate": 8.629066464825625e-06,
-      "loss": 0.38,
-      "step": 11781
-    },
-    {
-      "epoch": 0.7700150316972747,
-      "grad_norm": 0.38996458053588867,
-      "learning_rate": 8.628826251049797e-06,
-      "loss": 0.3225,
-      "step": 11782
-    },
-    {
-      "epoch": 0.7700803869028168,
-      "grad_norm": 0.465849369764328,
-      "learning_rate": 8.628586019574927e-06,
-      "loss": 0.3934,
-      "step": 11783
-    },
-    {
-      "epoch": 0.770145742108359,
-      "grad_norm": 0.41999539732933044,
-      "learning_rate": 8.628345770402185e-06,
-      "loss": 0.3513,
-      "step": 11784
-    },
-    {
-      "epoch": 0.770211097313901,
-      "grad_norm": 0.40005627274513245,
-      "learning_rate": 8.628105503532742e-06,
-      "loss": 0.3082,
-      "step": 11785
-    },
-    {
-      "epoch": 0.7702764525194432,
-      "grad_norm": 0.4402957260608673,
-      "learning_rate": 8.627865218967775e-06,
-      "loss": 0.3717,
-      "step": 11786
-    },
-    {
-      "epoch": 0.7703418077249853,
-      "grad_norm": 0.45970413088798523,
-      "learning_rate": 8.62762491670845e-06,
-      "loss": 0.3922,
-      "step": 11787
-    },
-    {
-      "epoch": 0.7704071629305275,
-      "grad_norm": 0.4797110855579376,
-      "learning_rate": 8.627384596755942e-06,
-      "loss": 0.4188,
-      "step": 11788
-    },
-    {
-      "epoch": 0.7704725181360695,
-      "grad_norm": 0.45043545961380005,
-      "learning_rate": 8.627144259111423e-06,
-      "loss": 0.3721,
-      "step": 11789
-    },
-    {
-      "epoch": 0.7705378733416116,
-      "grad_norm": 0.45030996203422546,
-      "learning_rate": 8.626903903776064e-06,
-      "loss": 0.3714,
-      "step": 11790
-    },
-    {
-      "epoch": 0.7706032285471538,
-      "grad_norm": 0.4351586103439331,
-      "learning_rate": 8.62666353075104e-06,
-      "loss": 0.3561,
-      "step": 11791
-    },
-    {
-      "epoch": 0.7706685837526959,
-      "grad_norm": 0.4533548653125763,
-      "learning_rate": 8.626423140037522e-06,
-      "loss": 0.3344,
-      "step": 11792
-    },
-    {
-      "epoch": 0.770733938958238,
-      "grad_norm": 0.46541261672973633,
-      "learning_rate": 8.626182731636678e-06,
-      "loss": 0.3848,
-      "step": 11793
-    },
-    {
-      "epoch": 0.7707992941637801,
-      "grad_norm": 0.45900705456733704,
-      "learning_rate": 8.625942305549688e-06,
-      "loss": 0.3664,
-      "step": 11794
-    },
-    {
-      "epoch": 0.7708646493693223,
-      "grad_norm": 0.4331720769405365,
-      "learning_rate": 8.625701861777721e-06,
-      "loss": 0.3679,
-      "step": 11795
-    },
-    {
-      "epoch": 0.7709300045748644,
-      "grad_norm": 0.4515995979309082,
-      "learning_rate": 8.62546140032195e-06,
-      "loss": 0.3557,
-      "step": 11796
-    },
-    {
-      "epoch": 0.7709953597804066,
-      "grad_norm": 0.44696739315986633,
-      "learning_rate": 8.625220921183546e-06,
-      "loss": 0.3791,
-      "step": 11797
-    },
-    {
-      "epoch": 0.7710607149859486,
-      "grad_norm": 0.4370102882385254,
-      "learning_rate": 8.624980424363684e-06,
-      "loss": 0.3438,
-      "step": 11798
-    },
-    {
-      "epoch": 0.7711260701914907,
-      "grad_norm": 0.4815715551376343,
-      "learning_rate": 8.62473990986354e-06,
-      "loss": 0.4063,
-      "step": 11799
-    },
-    {
-      "epoch": 0.7711914253970329,
-      "grad_norm": 0.4522111117839813,
-      "learning_rate": 8.624499377684279e-06,
-      "loss": 0.4132,
-      "step": 11800
-    },
-    {
-      "epoch": 0.771256780602575,
-      "grad_norm": 0.43564996123313904,
-      "learning_rate": 8.624258827827082e-06,
-      "loss": 0.3764,
-      "step": 11801
-    },
-    {
-      "epoch": 0.7713221358081171,
-      "grad_norm": 0.42831745743751526,
-      "learning_rate": 8.624018260293118e-06,
-      "loss": 0.367,
-      "step": 11802
-    },
-    {
-      "epoch": 0.7713874910136592,
-      "grad_norm": 0.44943052530288696,
-      "learning_rate": 8.623777675083562e-06,
-      "loss": 0.3461,
-      "step": 11803
-    },
-    {
-      "epoch": 0.7714528462192014,
-      "grad_norm": 0.5182769894599915,
-      "learning_rate": 8.623537072199587e-06,
-      "loss": 0.4253,
-      "step": 11804
-    },
-    {
-      "epoch": 0.7715182014247435,
-      "grad_norm": 0.4539603888988495,
-      "learning_rate": 8.623296451642365e-06,
-      "loss": 0.4197,
-      "step": 11805
-    },
-    {
-      "epoch": 0.7715835566302855,
-      "grad_norm": 0.4099823832511902,
-      "learning_rate": 8.623055813413072e-06,
-      "loss": 0.3466,
-      "step": 11806
-    },
-    {
-      "epoch": 0.7716489118358277,
-      "grad_norm": 0.44863128662109375,
-      "learning_rate": 8.62281515751288e-06,
-      "loss": 0.3664,
-      "step": 11807
-    },
-    {
-      "epoch": 0.7717142670413698,
-      "grad_norm": 0.7694123387336731,
-      "learning_rate": 8.622574483942965e-06,
-      "loss": 0.3734,
-      "step": 11808
-    },
-    {
-      "epoch": 0.771779622246912,
-      "grad_norm": 0.43434423208236694,
-      "learning_rate": 8.622333792704499e-06,
-      "loss": 0.3694,
-      "step": 11809
-    },
-    {
-      "epoch": 0.7718449774524541,
-      "grad_norm": 0.42186304926872253,
-      "learning_rate": 8.622093083798654e-06,
-      "loss": 0.3346,
-      "step": 11810
-    },
-    {
-      "epoch": 0.7719103326579962,
-      "grad_norm": 0.4894380569458008,
-      "learning_rate": 8.621852357226608e-06,
-      "loss": 0.4371,
-      "step": 11811
-    },
-    {
-      "epoch": 0.7719756878635383,
-      "grad_norm": 0.44940561056137085,
-      "learning_rate": 8.621611612989533e-06,
-      "loss": 0.4001,
-      "step": 11812
-    },
-    {
-      "epoch": 0.7720410430690805,
-      "grad_norm": 0.4715287387371063,
-      "learning_rate": 8.621370851088603e-06,
-      "loss": 0.3795,
-      "step": 11813
-    },
-    {
-      "epoch": 0.7721063982746226,
-      "grad_norm": 0.43742066621780396,
-      "learning_rate": 8.621130071524995e-06,
-      "loss": 0.356,
-      "step": 11814
-    },
-    {
-      "epoch": 0.7721717534801646,
-      "grad_norm": 0.44761908054351807,
-      "learning_rate": 8.620889274299879e-06,
-      "loss": 0.3809,
-      "step": 11815
-    },
-    {
-      "epoch": 0.7722371086857068,
-      "grad_norm": 0.41823145747184753,
-      "learning_rate": 8.620648459414431e-06,
-      "loss": 0.3349,
-      "step": 11816
-    },
-    {
-      "epoch": 0.7723024638912489,
-      "grad_norm": 0.4446566700935364,
-      "learning_rate": 8.620407626869828e-06,
-      "loss": 0.3676,
-      "step": 11817
-    },
-    {
-      "epoch": 0.7723678190967911,
-      "grad_norm": 0.4725385904312134,
-      "learning_rate": 8.620166776667242e-06,
-      "loss": 0.3795,
-      "step": 11818
-    },
-    {
-      "epoch": 0.7724331743023332,
-      "grad_norm": 0.4345307946205139,
-      "learning_rate": 8.619925908807848e-06,
-      "loss": 0.3829,
-      "step": 11819
-    },
-    {
-      "epoch": 0.7724985295078753,
-      "grad_norm": 0.4570246934890747,
-      "learning_rate": 8.61968502329282e-06,
-      "loss": 0.3506,
-      "step": 11820
-    },
-    {
-      "epoch": 0.7725638847134174,
-      "grad_norm": 0.47539472579956055,
-      "learning_rate": 8.619444120123337e-06,
-      "loss": 0.4179,
-      "step": 11821
-    },
-    {
-      "epoch": 0.7726292399189596,
-      "grad_norm": 0.44697678089141846,
-      "learning_rate": 8.619203199300571e-06,
-      "loss": 0.3929,
-      "step": 11822
-    },
-    {
-      "epoch": 0.7726945951245017,
-      "grad_norm": 0.5241457223892212,
-      "learning_rate": 8.618962260825696e-06,
-      "loss": 0.3915,
-      "step": 11823
-    },
-    {
-      "epoch": 0.7727599503300437,
-      "grad_norm": 0.4437718689441681,
-      "learning_rate": 8.61872130469989e-06,
-      "loss": 0.3843,
-      "step": 11824
-    },
-    {
-      "epoch": 0.7728253055355859,
-      "grad_norm": 0.41975370049476624,
-      "learning_rate": 8.618480330924323e-06,
-      "loss": 0.3247,
-      "step": 11825
-    },
-    {
-      "epoch": 0.772890660741128,
-      "grad_norm": 0.4372032582759857,
-      "learning_rate": 8.618239339500177e-06,
-      "loss": 0.3659,
-      "step": 11826
-    },
-    {
-      "epoch": 0.7729560159466702,
-      "grad_norm": 0.4222608208656311,
-      "learning_rate": 8.617998330428622e-06,
-      "loss": 0.3426,
-      "step": 11827
-    },
-    {
-      "epoch": 0.7730213711522123,
-      "grad_norm": 0.44943708181381226,
-      "learning_rate": 8.617757303710839e-06,
-      "loss": 0.3547,
-      "step": 11828
-    },
-    {
-      "epoch": 0.7730867263577544,
-      "grad_norm": 0.44556036591529846,
-      "learning_rate": 8.617516259347997e-06,
-      "loss": 0.3786,
-      "step": 11829
-    },
-    {
-      "epoch": 0.7731520815632965,
-      "grad_norm": 0.4915030896663666,
-      "learning_rate": 8.617275197341277e-06,
-      "loss": 0.3299,
-      "step": 11830
-    },
-    {
-      "epoch": 0.7732174367688386,
-      "grad_norm": 0.4466126263141632,
-      "learning_rate": 8.617034117691852e-06,
-      "loss": 0.3947,
-      "step": 11831
-    },
-    {
-      "epoch": 0.7732827919743808,
-      "grad_norm": 0.4619917571544647,
-      "learning_rate": 8.616793020400898e-06,
-      "loss": 0.38,
-      "step": 11832
-    },
-    {
-      "epoch": 0.7733481471799228,
-      "grad_norm": 0.43117502331733704,
-      "learning_rate": 8.616551905469592e-06,
-      "loss": 0.3548,
-      "step": 11833
-    },
-    {
-      "epoch": 0.773413502385465,
-      "grad_norm": 0.46063923835754395,
-      "learning_rate": 8.61631077289911e-06,
-      "loss": 0.3802,
-      "step": 11834
-    },
-    {
-      "epoch": 0.7734788575910071,
-      "grad_norm": 0.43890947103500366,
-      "learning_rate": 8.616069622690627e-06,
-      "loss": 0.3819,
-      "step": 11835
-    },
-    {
-      "epoch": 0.7735442127965493,
-      "grad_norm": 0.43641477823257446,
-      "learning_rate": 8.61582845484532e-06,
-      "loss": 0.38,
-      "step": 11836
-    },
-    {
-      "epoch": 0.7736095680020914,
-      "grad_norm": 0.42910319566726685,
-      "learning_rate": 8.615587269364365e-06,
-      "loss": 0.3313,
-      "step": 11837
-    },
-    {
-      "epoch": 0.7736749232076335,
-      "grad_norm": 0.43791988492012024,
-      "learning_rate": 8.615346066248938e-06,
-      "loss": 0.3416,
-      "step": 11838
-    },
-    {
-      "epoch": 0.7737402784131756,
-      "grad_norm": 0.4538554549217224,
-      "learning_rate": 8.615104845500215e-06,
-      "loss": 0.3836,
-      "step": 11839
-    },
-    {
-      "epoch": 0.7738056336187177,
-      "grad_norm": 0.46823617815971375,
-      "learning_rate": 8.614863607119374e-06,
-      "loss": 0.3878,
-      "step": 11840
-    },
-    {
-      "epoch": 0.7738709888242599,
-      "grad_norm": 0.47379255294799805,
-      "learning_rate": 8.614622351107592e-06,
-      "loss": 0.4466,
-      "step": 11841
-    },
-    {
-      "epoch": 0.7739363440298019,
-      "grad_norm": 0.4573115408420563,
-      "learning_rate": 8.614381077466043e-06,
-      "loss": 0.4009,
-      "step": 11842
-    },
-    {
-      "epoch": 0.7740016992353441,
-      "grad_norm": 0.4688892960548401,
-      "learning_rate": 8.614139786195905e-06,
-      "loss": 0.3851,
-      "step": 11843
-    },
-    {
-      "epoch": 0.7740670544408862,
-      "grad_norm": 0.4632177948951721,
-      "learning_rate": 8.613898477298356e-06,
-      "loss": 0.4063,
-      "step": 11844
-    },
-    {
-      "epoch": 0.7741324096464284,
-      "grad_norm": 0.4439104199409485,
-      "learning_rate": 8.613657150774573e-06,
-      "loss": 0.3591,
-      "step": 11845
-    },
-    {
-      "epoch": 0.7741977648519704,
-      "grad_norm": 0.45160600543022156,
-      "learning_rate": 8.61341580662573e-06,
-      "loss": 0.402,
-      "step": 11846
-    },
-    {
-      "epoch": 0.7742631200575126,
-      "grad_norm": 0.4168391525745392,
-      "learning_rate": 8.61317444485301e-06,
-      "loss": 0.3711,
-      "step": 11847
-    },
-    {
-      "epoch": 0.7743284752630547,
-      "grad_norm": 0.40535280108451843,
-      "learning_rate": 8.612933065457583e-06,
-      "loss": 0.3128,
-      "step": 11848
-    },
-    {
-      "epoch": 0.7743938304685968,
-      "grad_norm": 0.43484732508659363,
-      "learning_rate": 8.612691668440631e-06,
-      "loss": 0.3893,
-      "step": 11849
-    },
-    {
-      "epoch": 0.774459185674139,
-      "grad_norm": 0.4906025230884552,
-      "learning_rate": 8.61245025380333e-06,
-      "loss": 0.4006,
-      "step": 11850
-    },
-    {
-      "epoch": 0.774524540879681,
-      "grad_norm": 0.42672377824783325,
-      "learning_rate": 8.61220882154686e-06,
-      "loss": 0.3681,
-      "step": 11851
-    },
-    {
-      "epoch": 0.7745898960852232,
-      "grad_norm": 0.44538670778274536,
-      "learning_rate": 8.611967371672392e-06,
-      "loss": 0.3864,
-      "step": 11852
-    },
-    {
-      "epoch": 0.7746552512907653,
-      "grad_norm": 0.422595351934433,
-      "learning_rate": 8.61172590418111e-06,
-      "loss": 0.3771,
-      "step": 11853
-    },
-    {
-      "epoch": 0.7747206064963075,
-      "grad_norm": 0.4126081168651581,
-      "learning_rate": 8.611484419074189e-06,
-      "loss": 0.3522,
-      "step": 11854
-    },
-    {
-      "epoch": 0.7747859617018495,
-      "grad_norm": 0.44276732206344604,
-      "learning_rate": 8.611242916352809e-06,
-      "loss": 0.3568,
-      "step": 11855
-    },
-    {
-      "epoch": 0.7748513169073916,
-      "grad_norm": 0.44415268301963806,
-      "learning_rate": 8.611001396018144e-06,
-      "loss": 0.347,
-      "step": 11856
-    },
-    {
-      "epoch": 0.7749166721129338,
-      "grad_norm": 0.45888668298721313,
-      "learning_rate": 8.610759858071376e-06,
-      "loss": 0.3663,
-      "step": 11857
-    },
-    {
-      "epoch": 0.7749820273184759,
-      "grad_norm": 0.4071204364299774,
-      "learning_rate": 8.61051830251368e-06,
-      "loss": 0.3298,
-      "step": 11858
-    },
-    {
-      "epoch": 0.7750473825240181,
-      "grad_norm": 0.4355985224246979,
-      "learning_rate": 8.610276729346236e-06,
-      "loss": 0.3539,
-      "step": 11859
-    },
-    {
-      "epoch": 0.7751127377295601,
-      "grad_norm": 0.41884443163871765,
-      "learning_rate": 8.61003513857022e-06,
-      "loss": 0.3401,
-      "step": 11860
-    },
-    {
-      "epoch": 0.7751780929351023,
-      "grad_norm": 0.47849681973457336,
-      "learning_rate": 8.609793530186815e-06,
-      "loss": 0.4239,
-      "step": 11861
-    },
-    {
-      "epoch": 0.7752434481406444,
-      "grad_norm": 0.4448944628238678,
-      "learning_rate": 8.609551904197196e-06,
-      "loss": 0.3872,
-      "step": 11862
-    },
-    {
-      "epoch": 0.7753088033461866,
-      "grad_norm": 0.4501670300960541,
-      "learning_rate": 8.60931026060254e-06,
-      "loss": 0.3578,
-      "step": 11863
-    },
-    {
-      "epoch": 0.7753741585517286,
-      "grad_norm": 0.4469984471797943,
-      "learning_rate": 8.609068599404028e-06,
-      "loss": 0.3946,
-      "step": 11864
-    },
-    {
-      "epoch": 0.7754395137572707,
-      "grad_norm": 0.44684213399887085,
-      "learning_rate": 8.608826920602838e-06,
-      "loss": 0.381,
-      "step": 11865
-    },
-    {
-      "epoch": 0.7755048689628129,
-      "grad_norm": 0.4682392179965973,
-      "learning_rate": 8.60858522420015e-06,
-      "loss": 0.4327,
-      "step": 11866
-    },
-    {
-      "epoch": 0.775570224168355,
-      "grad_norm": 0.4276103079319,
-      "learning_rate": 8.608343510197141e-06,
-      "loss": 0.3352,
-      "step": 11867
-    },
-    {
-      "epoch": 0.7756355793738972,
-      "grad_norm": 0.4345168471336365,
-      "learning_rate": 8.60810177859499e-06,
-      "loss": 0.3869,
-      "step": 11868
-    },
-    {
-      "epoch": 0.7757009345794392,
-      "grad_norm": 0.42673447728157043,
-      "learning_rate": 8.607860029394879e-06,
-      "loss": 0.3634,
-      "step": 11869
-    },
-    {
-      "epoch": 0.7757662897849814,
-      "grad_norm": 0.4504968822002411,
-      "learning_rate": 8.607618262597982e-06,
-      "loss": 0.3999,
-      "step": 11870
-    },
-    {
-      "epoch": 0.7758316449905235,
-      "grad_norm": 0.443198561668396,
-      "learning_rate": 8.607376478205482e-06,
-      "loss": 0.3923,
-      "step": 11871
-    },
-    {
-      "epoch": 0.7758970001960657,
-      "grad_norm": 0.45323601365089417,
-      "learning_rate": 8.607134676218558e-06,
-      "loss": 0.4069,
-      "step": 11872
-    },
-    {
-      "epoch": 0.7759623554016077,
-      "grad_norm": 0.46169528365135193,
-      "learning_rate": 8.606892856638388e-06,
-      "loss": 0.3533,
-      "step": 11873
-    },
-    {
-      "epoch": 0.7760277106071498,
-      "grad_norm": 0.4142061471939087,
-      "learning_rate": 8.606651019466153e-06,
-      "loss": 0.3583,
-      "step": 11874
-    },
-    {
-      "epoch": 0.776093065812692,
-      "grad_norm": 0.4259793162345886,
-      "learning_rate": 8.60640916470303e-06,
-      "loss": 0.3554,
-      "step": 11875
-    },
-    {
-      "epoch": 0.7761584210182341,
-      "grad_norm": 0.4354163408279419,
-      "learning_rate": 8.6061672923502e-06,
-      "loss": 0.3571,
-      "step": 11876
-    },
-    {
-      "epoch": 0.7762237762237763,
-      "grad_norm": 0.4246130883693695,
-      "learning_rate": 8.605925402408843e-06,
-      "loss": 0.3546,
-      "step": 11877
-    },
-    {
-      "epoch": 0.7762891314293183,
-      "grad_norm": 0.4725601375102997,
-      "learning_rate": 8.60568349488014e-06,
-      "loss": 0.3952,
-      "step": 11878
-    },
-    {
-      "epoch": 0.7763544866348605,
-      "grad_norm": 0.44709452986717224,
-      "learning_rate": 8.605441569765266e-06,
-      "loss": 0.3688,
-      "step": 11879
-    },
-    {
-      "epoch": 0.7764198418404026,
-      "grad_norm": 0.4556303918361664,
-      "learning_rate": 8.605199627065409e-06,
-      "loss": 0.3664,
-      "step": 11880
-    },
-    {
-      "epoch": 0.7764851970459448,
-      "grad_norm": 0.4555344581604004,
-      "learning_rate": 8.604957666781741e-06,
-      "loss": 0.398,
-      "step": 11881
-    },
-    {
-      "epoch": 0.7765505522514868,
-      "grad_norm": 0.44141873717308044,
-      "learning_rate": 8.604715688915449e-06,
-      "loss": 0.3664,
-      "step": 11882
-    },
-    {
-      "epoch": 0.7766159074570289,
-      "grad_norm": 0.44355231523513794,
-      "learning_rate": 8.604473693467707e-06,
-      "loss": 0.3717,
-      "step": 11883
-    },
-    {
-      "epoch": 0.7766812626625711,
-      "grad_norm": 0.4414403736591339,
-      "learning_rate": 8.6042316804397e-06,
-      "loss": 0.3954,
-      "step": 11884
-    },
-    {
-      "epoch": 0.7767466178681132,
-      "grad_norm": 0.39469870924949646,
-      "learning_rate": 8.603989649832602e-06,
-      "loss": 0.3054,
-      "step": 11885
-    },
-    {
-      "epoch": 0.7768119730736553,
-      "grad_norm": 0.4306773543357849,
-      "learning_rate": 8.603747601647601e-06,
-      "loss": 0.352,
-      "step": 11886
-    },
-    {
-      "epoch": 0.7768773282791974,
-      "grad_norm": 0.4756735861301422,
-      "learning_rate": 8.603505535885877e-06,
-      "loss": 0.4161,
-      "step": 11887
-    },
-    {
-      "epoch": 0.7769426834847396,
-      "grad_norm": 0.504156231880188,
-      "learning_rate": 8.603263452548604e-06,
-      "loss": 0.4627,
-      "step": 11888
-    },
-    {
-      "epoch": 0.7770080386902817,
-      "grad_norm": 0.44968071579933167,
-      "learning_rate": 8.60302135163697e-06,
-      "loss": 0.3838,
-      "step": 11889
-    },
-    {
-      "epoch": 0.7770733938958237,
-      "grad_norm": 0.4513949155807495,
-      "learning_rate": 8.60277923315215e-06,
-      "loss": 0.4074,
-      "step": 11890
-    },
-    {
-      "epoch": 0.7771387491013659,
-      "grad_norm": 0.43221279978752136,
-      "learning_rate": 8.60253709709533e-06,
-      "loss": 0.3711,
-      "step": 11891
-    },
-    {
-      "epoch": 0.777204104306908,
-      "grad_norm": 0.4418954849243164,
-      "learning_rate": 8.602294943467686e-06,
-      "loss": 0.3901,
-      "step": 11892
-    },
-    {
-      "epoch": 0.7772694595124502,
-      "grad_norm": 0.4856281280517578,
-      "learning_rate": 8.602052772270405e-06,
-      "loss": 0.403,
-      "step": 11893
-    },
-    {
-      "epoch": 0.7773348147179923,
-      "grad_norm": 0.45415908098220825,
-      "learning_rate": 8.601810583504662e-06,
-      "loss": 0.3739,
-      "step": 11894
-    },
-    {
-      "epoch": 0.7774001699235344,
-      "grad_norm": 0.473394513130188,
-      "learning_rate": 8.60156837717164e-06,
-      "loss": 0.4296,
-      "step": 11895
-    },
-    {
-      "epoch": 0.7774655251290765,
-      "grad_norm": 0.4633063077926636,
-      "learning_rate": 8.601326153272524e-06,
-      "loss": 0.3543,
-      "step": 11896
-    },
-    {
-      "epoch": 0.7775308803346187,
-      "grad_norm": 0.5050225853919983,
-      "learning_rate": 8.601083911808492e-06,
-      "loss": 0.4812,
-      "step": 11897
-    },
-    {
-      "epoch": 0.7775962355401608,
-      "grad_norm": 0.4297986924648285,
-      "learning_rate": 8.600841652780726e-06,
-      "loss": 0.3362,
-      "step": 11898
-    },
-    {
-      "epoch": 0.7776615907457028,
-      "grad_norm": 0.4877097010612488,
-      "learning_rate": 8.600599376190408e-06,
-      "loss": 0.4079,
-      "step": 11899
-    },
-    {
-      "epoch": 0.777726945951245,
-      "grad_norm": 0.4294080436229706,
-      "learning_rate": 8.60035708203872e-06,
-      "loss": 0.3678,
-      "step": 11900
-    },
-    {
-      "epoch": 0.7777923011567871,
-      "grad_norm": 0.45811372995376587,
-      "learning_rate": 8.600114770326842e-06,
-      "loss": 0.3645,
-      "step": 11901
-    },
-    {
-      "epoch": 0.7778576563623293,
-      "grad_norm": 0.44701892137527466,
-      "learning_rate": 8.599872441055957e-06,
-      "loss": 0.3905,
-      "step": 11902
-    },
-    {
-      "epoch": 0.7779230115678714,
-      "grad_norm": 0.4324595034122467,
-      "learning_rate": 8.599630094227247e-06,
-      "loss": 0.3514,
-      "step": 11903
-    },
-    {
-      "epoch": 0.7779883667734135,
-      "grad_norm": 0.4528386890888214,
-      "learning_rate": 8.599387729841894e-06,
-      "loss": 0.4042,
-      "step": 11904
-    },
-    {
-      "epoch": 0.7780537219789556,
-      "grad_norm": 0.4724828004837036,
-      "learning_rate": 8.599145347901082e-06,
-      "loss": 0.3706,
-      "step": 11905
-    },
-    {
-      "epoch": 0.7781190771844978,
-      "grad_norm": 0.44598785042762756,
-      "learning_rate": 8.59890294840599e-06,
-      "loss": 0.3439,
-      "step": 11906
-    },
-    {
-      "epoch": 0.7781844323900399,
-      "grad_norm": 0.4361238181591034,
-      "learning_rate": 8.598660531357802e-06,
-      "loss": 0.3724,
-      "step": 11907
-    },
-    {
-      "epoch": 0.7782497875955819,
-      "grad_norm": 0.40344589948654175,
-      "learning_rate": 8.5984180967577e-06,
-      "loss": 0.3263,
-      "step": 11908
-    },
-    {
-      "epoch": 0.7783151428011241,
-      "grad_norm": 0.4686245322227478,
-      "learning_rate": 8.598175644606865e-06,
-      "loss": 0.4181,
-      "step": 11909
-    },
-    {
-      "epoch": 0.7783804980066662,
-      "grad_norm": 0.408418744802475,
-      "learning_rate": 8.59793317490648e-06,
-      "loss": 0.3287,
-      "step": 11910
-    },
-    {
-      "epoch": 0.7784458532122084,
-      "grad_norm": 0.44714394211769104,
-      "learning_rate": 8.597690687657732e-06,
-      "loss": 0.3932,
-      "step": 11911
-    },
-    {
-      "epoch": 0.7785112084177505,
-      "grad_norm": 0.4426407814025879,
-      "learning_rate": 8.597448182861797e-06,
-      "loss": 0.3653,
-      "step": 11912
-    },
-    {
-      "epoch": 0.7785765636232926,
-      "grad_norm": 0.4467528760433197,
-      "learning_rate": 8.597205660519863e-06,
-      "loss": 0.3644,
-      "step": 11913
-    },
-    {
-      "epoch": 0.7786419188288347,
-      "grad_norm": 0.3986530005931854,
-      "learning_rate": 8.596963120633109e-06,
-      "loss": 0.3087,
-      "step": 11914
-    },
-    {
-      "epoch": 0.7787072740343768,
-      "grad_norm": 0.4246610999107361,
-      "learning_rate": 8.59672056320272e-06,
-      "loss": 0.361,
-      "step": 11915
-    },
-    {
-      "epoch": 0.778772629239919,
-      "grad_norm": 0.42618927359580994,
-      "learning_rate": 8.59647798822988e-06,
-      "loss": 0.3509,
-      "step": 11916
-    },
-    {
-      "epoch": 0.778837984445461,
-      "grad_norm": 0.45555341243743896,
-      "learning_rate": 8.596235395715771e-06,
-      "loss": 0.4139,
-      "step": 11917
-    },
-    {
-      "epoch": 0.7789033396510032,
-      "grad_norm": 0.43304362893104553,
-      "learning_rate": 8.595992785661575e-06,
-      "loss": 0.3584,
-      "step": 11918
-    },
-    {
-      "epoch": 0.7789686948565453,
-      "grad_norm": 0.4583284258842468,
-      "learning_rate": 8.595750158068477e-06,
-      "loss": 0.406,
-      "step": 11919
-    },
-    {
-      "epoch": 0.7790340500620875,
-      "grad_norm": 0.4887494444847107,
-      "learning_rate": 8.595507512937659e-06,
-      "loss": 0.3575,
-      "step": 11920
-    },
-    {
-      "epoch": 0.7790994052676296,
-      "grad_norm": 0.4497702717781067,
-      "learning_rate": 8.595264850270306e-06,
-      "loss": 0.3533,
-      "step": 11921
-    },
-    {
-      "epoch": 0.7791647604731717,
-      "grad_norm": 0.42957308888435364,
-      "learning_rate": 8.595022170067602e-06,
-      "loss": 0.4051,
-      "step": 11922
-    },
-    {
-      "epoch": 0.7792301156787138,
-      "grad_norm": 0.478105753660202,
-      "learning_rate": 8.594779472330727e-06,
-      "loss": 0.3959,
-      "step": 11923
-    },
-    {
-      "epoch": 0.7792954708842559,
-      "grad_norm": 0.46036097407341003,
-      "learning_rate": 8.59453675706087e-06,
-      "loss": 0.3813,
-      "step": 11924
-    },
-    {
-      "epoch": 0.7793608260897981,
-      "grad_norm": 0.4499993920326233,
-      "learning_rate": 8.594294024259209e-06,
-      "loss": 0.3961,
-      "step": 11925
-    },
-    {
-      "epoch": 0.7794261812953401,
-      "grad_norm": 0.48920682072639465,
-      "learning_rate": 8.594051273926934e-06,
-      "loss": 0.4296,
-      "step": 11926
-    },
-    {
-      "epoch": 0.7794915365008823,
-      "grad_norm": 0.4173191487789154,
-      "learning_rate": 8.593808506065222e-06,
-      "loss": 0.3475,
-      "step": 11927
-    },
-    {
-      "epoch": 0.7795568917064244,
-      "grad_norm": 0.3997310400009155,
-      "learning_rate": 8.593565720675263e-06,
-      "loss": 0.3031,
-      "step": 11928
-    },
-    {
-      "epoch": 0.7796222469119666,
-      "grad_norm": 0.45031315088272095,
-      "learning_rate": 8.59332291775824e-06,
-      "loss": 0.3866,
-      "step": 11929
-    },
-    {
-      "epoch": 0.7796876021175086,
-      "grad_norm": 0.45433446764945984,
-      "learning_rate": 8.593080097315335e-06,
-      "loss": 0.3941,
-      "step": 11930
-    },
-    {
-      "epoch": 0.7797529573230508,
-      "grad_norm": 0.4518393576145172,
-      "learning_rate": 8.592837259347734e-06,
-      "loss": 0.3848,
-      "step": 11931
-    },
-    {
-      "epoch": 0.7798183125285929,
-      "grad_norm": 0.43654945492744446,
-      "learning_rate": 8.59259440385662e-06,
-      "loss": 0.3693,
-      "step": 11932
-    },
-    {
-      "epoch": 0.779883667734135,
-      "grad_norm": 0.456781804561615,
-      "learning_rate": 8.59235153084318e-06,
-      "loss": 0.4074,
-      "step": 11933
-    },
-    {
-      "epoch": 0.7799490229396772,
-      "grad_norm": 0.4224834740161896,
-      "learning_rate": 8.592108640308596e-06,
-      "loss": 0.3505,
-      "step": 11934
-    },
-    {
-      "epoch": 0.7800143781452192,
-      "grad_norm": 0.45787283778190613,
-      "learning_rate": 8.591865732254054e-06,
-      "loss": 0.3908,
-      "step": 11935
-    },
-    {
-      "epoch": 0.7800797333507614,
-      "grad_norm": 0.4545110762119293,
-      "learning_rate": 8.59162280668074e-06,
-      "loss": 0.3792,
-      "step": 11936
-    },
-    {
-      "epoch": 0.7801450885563035,
-      "grad_norm": 0.4336988031864166,
-      "learning_rate": 8.591379863589836e-06,
-      "loss": 0.3984,
-      "step": 11937
-    },
-    {
-      "epoch": 0.7802104437618457,
-      "grad_norm": 0.42259207367897034,
-      "learning_rate": 8.591136902982526e-06,
-      "loss": 0.3212,
-      "step": 11938
-    },
-    {
-      "epoch": 0.7802757989673877,
-      "grad_norm": 0.46574777364730835,
-      "learning_rate": 8.59089392486e-06,
-      "loss": 0.3856,
-      "step": 11939
-    },
-    {
-      "epoch": 0.7803411541729298,
-      "grad_norm": 0.4679214060306549,
-      "learning_rate": 8.590650929223441e-06,
-      "loss": 0.4304,
-      "step": 11940
-    },
-    {
-      "epoch": 0.780406509378472,
-      "grad_norm": 0.45430073142051697,
-      "learning_rate": 8.590407916074031e-06,
-      "loss": 0.3464,
-      "step": 11941
-    },
-    {
-      "epoch": 0.7804718645840141,
-      "grad_norm": 0.45531001687049866,
-      "learning_rate": 8.59016488541296e-06,
-      "loss": 0.4346,
-      "step": 11942
-    },
-    {
-      "epoch": 0.7805372197895563,
-      "grad_norm": 0.5119999647140503,
-      "learning_rate": 8.58992183724141e-06,
-      "loss": 0.4221,
-      "step": 11943
-    },
-    {
-      "epoch": 0.7806025749950983,
-      "grad_norm": 0.4710194766521454,
-      "learning_rate": 8.589678771560568e-06,
-      "loss": 0.409,
-      "step": 11944
-    },
-    {
-      "epoch": 0.7806679302006405,
-      "grad_norm": 0.42829418182373047,
-      "learning_rate": 8.58943568837162e-06,
-      "loss": 0.3526,
-      "step": 11945
-    },
-    {
-      "epoch": 0.7807332854061826,
-      "grad_norm": 0.45222848653793335,
-      "learning_rate": 8.589192587675747e-06,
-      "loss": 0.4011,
-      "step": 11946
-    },
-    {
-      "epoch": 0.7807986406117248,
-      "grad_norm": 0.40631863474845886,
-      "learning_rate": 8.588949469474141e-06,
-      "loss": 0.3294,
-      "step": 11947
-    },
-    {
-      "epoch": 0.7808639958172668,
-      "grad_norm": 0.4742248058319092,
-      "learning_rate": 8.588706333767984e-06,
-      "loss": 0.4256,
-      "step": 11948
-    },
-    {
-      "epoch": 0.7809293510228089,
-      "grad_norm": 0.442628413438797,
-      "learning_rate": 8.588463180558464e-06,
-      "loss": 0.3713,
-      "step": 11949
-    },
-    {
-      "epoch": 0.7809947062283511,
-      "grad_norm": 0.4321967363357544,
-      "learning_rate": 8.588220009846766e-06,
-      "loss": 0.3699,
-      "step": 11950
-    },
-    {
-      "epoch": 0.7810600614338932,
-      "grad_norm": 0.4628668427467346,
-      "learning_rate": 8.587976821634076e-06,
-      "loss": 0.3675,
-      "step": 11951
-    },
-    {
-      "epoch": 0.7811254166394354,
-      "grad_norm": 0.4346633851528168,
-      "learning_rate": 8.587733615921579e-06,
-      "loss": 0.3761,
-      "step": 11952
-    },
-    {
-      "epoch": 0.7811907718449774,
-      "grad_norm": 0.4312704801559448,
-      "learning_rate": 8.587490392710464e-06,
-      "loss": 0.3458,
-      "step": 11953
-    },
-    {
-      "epoch": 0.7812561270505196,
-      "grad_norm": 0.46004918217658997,
-      "learning_rate": 8.587247152001914e-06,
-      "loss": 0.3894,
-      "step": 11954
-    },
-    {
-      "epoch": 0.7813214822560617,
-      "grad_norm": 0.45357799530029297,
-      "learning_rate": 8.587003893797117e-06,
-      "loss": 0.4008,
-      "step": 11955
-    },
-    {
-      "epoch": 0.7813868374616039,
-      "grad_norm": 0.41232830286026,
-      "learning_rate": 8.586760618097261e-06,
-      "loss": 0.3083,
-      "step": 11956
-    },
-    {
-      "epoch": 0.7814521926671459,
-      "grad_norm": 0.44780534505844116,
-      "learning_rate": 8.586517324903529e-06,
-      "loss": 0.3812,
-      "step": 11957
-    },
-    {
-      "epoch": 0.781517547872688,
-      "grad_norm": 0.4769149124622345,
-      "learning_rate": 8.58627401421711e-06,
-      "loss": 0.3674,
-      "step": 11958
-    },
-    {
-      "epoch": 0.7815829030782302,
-      "grad_norm": 0.43085014820098877,
-      "learning_rate": 8.58603068603919e-06,
-      "loss": 0.3468,
-      "step": 11959
-    },
-    {
-      "epoch": 0.7816482582837723,
-      "grad_norm": 0.4569220244884491,
-      "learning_rate": 8.585787340370955e-06,
-      "loss": 0.3927,
-      "step": 11960
-    },
-    {
-      "epoch": 0.7817136134893145,
-      "grad_norm": 0.7949502468109131,
-      "learning_rate": 8.585543977213595e-06,
-      "loss": 0.3838,
-      "step": 11961
-    },
-    {
-      "epoch": 0.7817789686948565,
-      "grad_norm": 0.4006519317626953,
-      "learning_rate": 8.585300596568294e-06,
-      "loss": 0.3475,
-      "step": 11962
-    },
-    {
-      "epoch": 0.7818443239003987,
-      "grad_norm": 0.4714229106903076,
-      "learning_rate": 8.585057198436239e-06,
-      "loss": 0.408,
-      "step": 11963
-    },
-    {
-      "epoch": 0.7819096791059408,
-      "grad_norm": 0.42171719670295715,
-      "learning_rate": 8.58481378281862e-06,
-      "loss": 0.3405,
-      "step": 11964
-    },
-    {
-      "epoch": 0.781975034311483,
-      "grad_norm": 0.4483102858066559,
-      "learning_rate": 8.584570349716623e-06,
-      "loss": 0.3699,
-      "step": 11965
-    },
-    {
-      "epoch": 0.782040389517025,
-      "grad_norm": 0.49136093258857727,
-      "learning_rate": 8.584326899131433e-06,
-      "loss": 0.3813,
-      "step": 11966
-    },
-    {
-      "epoch": 0.7821057447225671,
-      "grad_norm": 0.44293758273124695,
-      "learning_rate": 8.584083431064238e-06,
-      "loss": 0.3483,
-      "step": 11967
-    },
-    {
-      "epoch": 0.7821710999281093,
-      "grad_norm": 0.4744596779346466,
-      "learning_rate": 8.583839945516229e-06,
-      "loss": 0.3792,
-      "step": 11968
-    },
-    {
-      "epoch": 0.7822364551336514,
-      "grad_norm": 0.6347219347953796,
-      "learning_rate": 8.583596442488588e-06,
-      "loss": 0.3338,
-      "step": 11969
-    },
-    {
-      "epoch": 0.7823018103391935,
-      "grad_norm": 0.44006791710853577,
-      "learning_rate": 8.583352921982507e-06,
-      "loss": 0.3547,
-      "step": 11970
-    },
-    {
-      "epoch": 0.7823671655447356,
-      "grad_norm": 0.4571622908115387,
-      "learning_rate": 8.583109383999173e-06,
-      "loss": 0.4068,
-      "step": 11971
-    },
-    {
-      "epoch": 0.7824325207502778,
-      "grad_norm": 0.4679630398750305,
-      "learning_rate": 8.582865828539773e-06,
-      "loss": 0.3377,
-      "step": 11972
-    },
-    {
-      "epoch": 0.7824978759558199,
-      "grad_norm": 0.470459908246994,
-      "learning_rate": 8.582622255605494e-06,
-      "loss": 0.4005,
-      "step": 11973
-    },
-    {
-      "epoch": 0.782563231161362,
-      "grad_norm": 0.44971734285354614,
-      "learning_rate": 8.582378665197526e-06,
-      "loss": 0.3851,
-      "step": 11974
-    },
-    {
-      "epoch": 0.7826285863669041,
-      "grad_norm": 0.4586564898490906,
-      "learning_rate": 8.582135057317055e-06,
-      "loss": 0.4114,
-      "step": 11975
-    },
-    {
-      "epoch": 0.7826939415724462,
-      "grad_norm": 0.5088337063789368,
-      "learning_rate": 8.581891431965272e-06,
-      "loss": 0.4243,
-      "step": 11976
-    },
-    {
-      "epoch": 0.7827592967779884,
-      "grad_norm": 0.45787137746810913,
-      "learning_rate": 8.581647789143364e-06,
-      "loss": 0.391,
-      "step": 11977
-    },
-    {
-      "epoch": 0.7828246519835305,
-      "grad_norm": 0.45902103185653687,
-      "learning_rate": 8.581404128852517e-06,
-      "loss": 0.3869,
-      "step": 11978
-    },
-    {
-      "epoch": 0.7828900071890726,
-      "grad_norm": 0.45155927538871765,
-      "learning_rate": 8.581160451093922e-06,
-      "loss": 0.349,
-      "step": 11979
-    },
-    {
-      "epoch": 0.7829553623946147,
-      "grad_norm": 0.4716495871543884,
-      "learning_rate": 8.58091675586877e-06,
-      "loss": 0.4086,
-      "step": 11980
-    },
-    {
-      "epoch": 0.7830207176001569,
-      "grad_norm": 0.4456287622451782,
-      "learning_rate": 8.580673043178242e-06,
-      "loss": 0.4151,
-      "step": 11981
-    },
-    {
-      "epoch": 0.783086072805699,
-      "grad_norm": 0.45639076828956604,
-      "learning_rate": 8.580429313023532e-06,
-      "loss": 0.3948,
-      "step": 11982
-    },
-    {
-      "epoch": 0.783151428011241,
-      "grad_norm": 0.46890559792518616,
-      "learning_rate": 8.58018556540583e-06,
-      "loss": 0.4031,
-      "step": 11983
-    },
-    {
-      "epoch": 0.7832167832167832,
-      "grad_norm": 0.48633939027786255,
-      "learning_rate": 8.579941800326322e-06,
-      "loss": 0.4006,
-      "step": 11984
-    },
-    {
-      "epoch": 0.7832821384223253,
-      "grad_norm": 0.45148807764053345,
-      "learning_rate": 8.579698017786196e-06,
-      "loss": 0.4164,
-      "step": 11985
-    },
-    {
-      "epoch": 0.7833474936278675,
-      "grad_norm": 0.4746284484863281,
-      "learning_rate": 8.579454217786644e-06,
-      "loss": 0.4326,
-      "step": 11986
-    },
-    {
-      "epoch": 0.7834128488334096,
-      "grad_norm": 0.46255096793174744,
-      "learning_rate": 8.579210400328852e-06,
-      "loss": 0.3971,
-      "step": 11987
-    },
-    {
-      "epoch": 0.7834782040389517,
-      "grad_norm": 0.45016103982925415,
-      "learning_rate": 8.578966565414014e-06,
-      "loss": 0.3565,
-      "step": 11988
-    },
-    {
-      "epoch": 0.7835435592444938,
-      "grad_norm": 0.443569540977478,
-      "learning_rate": 8.578722713043312e-06,
-      "loss": 0.3509,
-      "step": 11989
-    },
-    {
-      "epoch": 0.783608914450036,
-      "grad_norm": 0.4615786373615265,
-      "learning_rate": 8.578478843217944e-06,
-      "loss": 0.3812,
-      "step": 11990
-    },
-    {
-      "epoch": 0.7836742696555781,
-      "grad_norm": 0.43236657977104187,
-      "learning_rate": 8.578234955939094e-06,
-      "loss": 0.3713,
-      "step": 11991
-    },
-    {
-      "epoch": 0.7837396248611201,
-      "grad_norm": 0.4728842079639435,
-      "learning_rate": 8.57799105120795e-06,
-      "loss": 0.4118,
-      "step": 11992
-    },
-    {
-      "epoch": 0.7838049800666623,
-      "grad_norm": 0.44954821467399597,
-      "learning_rate": 8.577747129025706e-06,
-      "loss": 0.3861,
-      "step": 11993
-    },
-    {
-      "epoch": 0.7838703352722044,
-      "grad_norm": 0.4186398684978485,
-      "learning_rate": 8.577503189393549e-06,
-      "loss": 0.3541,
-      "step": 11994
-    },
-    {
-      "epoch": 0.7839356904777466,
-      "grad_norm": 0.45853495597839355,
-      "learning_rate": 8.57725923231267e-06,
-      "loss": 0.3808,
-      "step": 11995
-    },
-    {
-      "epoch": 0.7840010456832887,
-      "grad_norm": 0.43140414357185364,
-      "learning_rate": 8.577015257784258e-06,
-      "loss": 0.3362,
-      "step": 11996
-    },
-    {
-      "epoch": 0.7840664008888308,
-      "grad_norm": 0.439466267824173,
-      "learning_rate": 8.576771265809504e-06,
-      "loss": 0.3711,
-      "step": 11997
-    },
-    {
-      "epoch": 0.7841317560943729,
-      "grad_norm": 0.45974549651145935,
-      "learning_rate": 8.576527256389598e-06,
-      "loss": 0.377,
-      "step": 11998
-    },
-    {
-      "epoch": 0.784197111299915,
-      "grad_norm": 0.4531913101673126,
-      "learning_rate": 8.576283229525728e-06,
-      "loss": 0.3755,
-      "step": 11999
-    },
-    {
-      "epoch": 0.7842624665054572,
-      "grad_norm": 0.40491098165512085,
-      "learning_rate": 8.576039185219087e-06,
-      "loss": 0.3634,
-      "step": 12000
-    },
-    {
-      "epoch": 0.7843278217109992,
-      "grad_norm": 0.3893328011035919,
-      "learning_rate": 8.575795123470863e-06,
-      "loss": 0.3201,
-      "step": 12001
-    },
-    {
-      "epoch": 0.7843931769165414,
-      "grad_norm": 0.47420990467071533,
-      "learning_rate": 8.57555104428225e-06,
-      "loss": 0.4297,
-      "step": 12002
-    },
-    {
-      "epoch": 0.7844585321220835,
-      "grad_norm": 0.43641334772109985,
-      "learning_rate": 8.575306947654431e-06,
-      "loss": 0.3679,
-      "step": 12003
-    },
-    {
-      "epoch": 0.7845238873276257,
-      "grad_norm": 0.4373196065425873,
-      "learning_rate": 8.575062833588606e-06,
-      "loss": 0.3955,
-      "step": 12004
-    },
-    {
-      "epoch": 0.7845892425331678,
-      "grad_norm": 0.46208757162094116,
-      "learning_rate": 8.57481870208596e-06,
-      "loss": 0.4197,
-      "step": 12005
-    },
-    {
-      "epoch": 0.7846545977387099,
-      "grad_norm": 0.4125015437602997,
-      "learning_rate": 8.574574553147683e-06,
-      "loss": 0.3479,
-      "step": 12006
-    },
-    {
-      "epoch": 0.784719952944252,
-      "grad_norm": 0.45740821957588196,
-      "learning_rate": 8.574330386774968e-06,
-      "loss": 0.3843,
-      "step": 12007
-    },
-    {
-      "epoch": 0.7847853081497941,
-      "grad_norm": 0.42983338236808777,
-      "learning_rate": 8.574086202969006e-06,
-      "loss": 0.3678,
-      "step": 12008
-    },
-    {
-      "epoch": 0.7848506633553363,
-      "grad_norm": 0.4891479015350342,
-      "learning_rate": 8.573842001730987e-06,
-      "loss": 0.4414,
-      "step": 12009
-    },
-    {
-      "epoch": 0.7849160185608783,
-      "grad_norm": 0.4560488760471344,
-      "learning_rate": 8.573597783062104e-06,
-      "loss": 0.3986,
-      "step": 12010
-    },
-    {
-      "epoch": 0.7849813737664205,
-      "grad_norm": 0.427053302526474,
-      "learning_rate": 8.573353546963545e-06,
-      "loss": 0.3535,
-      "step": 12011
-    },
-    {
-      "epoch": 0.7850467289719626,
-      "grad_norm": 0.4430866241455078,
-      "learning_rate": 8.573109293436503e-06,
-      "loss": 0.3594,
-      "step": 12012
-    },
-    {
-      "epoch": 0.7851120841775048,
-      "grad_norm": 0.4583944082260132,
-      "learning_rate": 8.57286502248217e-06,
-      "loss": 0.393,
-      "step": 12013
-    },
-    {
-      "epoch": 0.7851774393830468,
-      "grad_norm": 0.4045864939689636,
-      "learning_rate": 8.572620734101735e-06,
-      "loss": 0.3289,
-      "step": 12014
-    },
-    {
-      "epoch": 0.785242794588589,
-      "grad_norm": 0.4391320049762726,
-      "learning_rate": 8.572376428296393e-06,
-      "loss": 0.3879,
-      "step": 12015
-    },
-    {
-      "epoch": 0.7853081497941311,
-      "grad_norm": 0.4494188725948334,
-      "learning_rate": 8.572132105067332e-06,
-      "loss": 0.3913,
-      "step": 12016
-    },
-    {
-      "epoch": 0.7853735049996732,
-      "grad_norm": 0.4446926414966583,
-      "learning_rate": 8.571887764415747e-06,
-      "loss": 0.357,
-      "step": 12017
-    },
-    {
-      "epoch": 0.7854388602052154,
-      "grad_norm": 0.42256808280944824,
-      "learning_rate": 8.571643406342825e-06,
-      "loss": 0.3386,
-      "step": 12018
-    },
-    {
-      "epoch": 0.7855042154107574,
-      "grad_norm": 0.4125842750072479,
-      "learning_rate": 8.571399030849763e-06,
-      "loss": 0.3023,
-      "step": 12019
-    },
-    {
-      "epoch": 0.7855695706162996,
-      "grad_norm": 0.446566641330719,
-      "learning_rate": 8.571154637937751e-06,
-      "loss": 0.3649,
-      "step": 12020
-    },
-    {
-      "epoch": 0.7856349258218417,
-      "grad_norm": 0.45315021276474,
-      "learning_rate": 8.570910227607979e-06,
-      "loss": 0.344,
-      "step": 12021
-    },
-    {
-      "epoch": 0.7857002810273839,
-      "grad_norm": 0.4215909242630005,
-      "learning_rate": 8.570665799861642e-06,
-      "loss": 0.3381,
-      "step": 12022
-    },
-    {
-      "epoch": 0.785765636232926,
-      "grad_norm": 0.4065757989883423,
-      "learning_rate": 8.57042135469993e-06,
-      "loss": 0.3218,
-      "step": 12023
-    },
-    {
-      "epoch": 0.7858309914384681,
-      "grad_norm": 0.4728504717350006,
-      "learning_rate": 8.570176892124037e-06,
-      "loss": 0.4315,
-      "step": 12024
-    },
-    {
-      "epoch": 0.7858963466440102,
-      "grad_norm": 0.48537716269493103,
-      "learning_rate": 8.569932412135153e-06,
-      "loss": 0.4065,
-      "step": 12025
-    },
-    {
-      "epoch": 0.7859617018495523,
-      "grad_norm": 0.4314275085926056,
-      "learning_rate": 8.569687914734474e-06,
-      "loss": 0.3648,
-      "step": 12026
-    },
-    {
-      "epoch": 0.7860270570550945,
-      "grad_norm": 0.4772936999797821,
-      "learning_rate": 8.56944339992319e-06,
-      "loss": 0.4191,
-      "step": 12027
-    },
-    {
-      "epoch": 0.7860924122606365,
-      "grad_norm": 0.4536139667034149,
-      "learning_rate": 8.569198867702493e-06,
-      "loss": 0.4019,
-      "step": 12028
-    },
-    {
-      "epoch": 0.7861577674661787,
-      "grad_norm": 0.45113837718963623,
-      "learning_rate": 8.568954318073577e-06,
-      "loss": 0.3901,
-      "step": 12029
-    },
-    {
-      "epoch": 0.7862231226717208,
-      "grad_norm": 0.4572070837020874,
-      "learning_rate": 8.568709751037634e-06,
-      "loss": 0.4117,
-      "step": 12030
-    },
-    {
-      "epoch": 0.786288477877263,
-      "grad_norm": 0.47010403871536255,
-      "learning_rate": 8.568465166595857e-06,
-      "loss": 0.387,
-      "step": 12031
-    },
-    {
-      "epoch": 0.786353833082805,
-      "grad_norm": 0.39718037843704224,
-      "learning_rate": 8.56822056474944e-06,
-      "loss": 0.3084,
-      "step": 12032
-    },
-    {
-      "epoch": 0.7864191882883471,
-      "grad_norm": 0.4905865788459778,
-      "learning_rate": 8.567975945499575e-06,
-      "loss": 0.3872,
-      "step": 12033
-    },
-    {
-      "epoch": 0.7864845434938893,
-      "grad_norm": 0.4178222417831421,
-      "learning_rate": 8.567731308847456e-06,
-      "loss": 0.3547,
-      "step": 12034
-    },
-    {
-      "epoch": 0.7865498986994314,
-      "grad_norm": 0.43827691674232483,
-      "learning_rate": 8.567486654794274e-06,
-      "loss": 0.3531,
-      "step": 12035
-    },
-    {
-      "epoch": 0.7866152539049736,
-      "grad_norm": 0.49150943756103516,
-      "learning_rate": 8.567241983341223e-06,
-      "loss": 0.4313,
-      "step": 12036
-    },
-    {
-      "epoch": 0.7866806091105156,
-      "grad_norm": 0.4697049856185913,
-      "learning_rate": 8.566997294489498e-06,
-      "loss": 0.4167,
-      "step": 12037
-    },
-    {
-      "epoch": 0.7867459643160578,
-      "grad_norm": 0.4594012498855591,
-      "learning_rate": 8.566752588240292e-06,
-      "loss": 0.3817,
-      "step": 12038
-    },
-    {
-      "epoch": 0.7868113195215999,
-      "grad_norm": 0.4005836546421051,
-      "learning_rate": 8.566507864594799e-06,
-      "loss": 0.3363,
-      "step": 12039
-    },
-    {
-      "epoch": 0.7868766747271421,
-      "grad_norm": 0.44431227445602417,
-      "learning_rate": 8.566263123554211e-06,
-      "loss": 0.3806,
-      "step": 12040
-    },
-    {
-      "epoch": 0.7869420299326841,
-      "grad_norm": 0.4370727837085724,
-      "learning_rate": 8.56601836511972e-06,
-      "loss": 0.3657,
-      "step": 12041
-    },
-    {
-      "epoch": 0.7870073851382262,
-      "grad_norm": 0.41017067432403564,
-      "learning_rate": 8.565773589292524e-06,
-      "loss": 0.278,
-      "step": 12042
-    },
-    {
-      "epoch": 0.7870727403437684,
-      "grad_norm": 0.47175443172454834,
-      "learning_rate": 8.565528796073815e-06,
-      "loss": 0.4219,
-      "step": 12043
-    },
-    {
-      "epoch": 0.7871380955493105,
-      "grad_norm": 0.42466551065444946,
-      "learning_rate": 8.565283985464785e-06,
-      "loss": 0.3525,
-      "step": 12044
-    },
-    {
-      "epoch": 0.7872034507548527,
-      "grad_norm": 0.5120432376861572,
-      "learning_rate": 8.565039157466632e-06,
-      "loss": 0.4914,
-      "step": 12045
-    },
-    {
-      "epoch": 0.7872688059603947,
-      "grad_norm": 0.45365110039711,
-      "learning_rate": 8.564794312080546e-06,
-      "loss": 0.408,
-      "step": 12046
-    },
-    {
-      "epoch": 0.7873341611659369,
-      "grad_norm": 0.43041422963142395,
-      "learning_rate": 8.564549449307725e-06,
-      "loss": 0.3815,
-      "step": 12047
-    },
-    {
-      "epoch": 0.787399516371479,
-      "grad_norm": 0.4978421926498413,
-      "learning_rate": 8.564304569149362e-06,
-      "loss": 0.3623,
-      "step": 12048
-    },
-    {
-      "epoch": 0.7874648715770212,
-      "grad_norm": 0.46047160029411316,
-      "learning_rate": 8.564059671606648e-06,
-      "loss": 0.3712,
-      "step": 12049
-    },
-    {
-      "epoch": 0.7875302267825632,
-      "grad_norm": 0.41224947571754456,
-      "learning_rate": 8.563814756680782e-06,
-      "loss": 0.3372,
-      "step": 12050
-    },
-    {
-      "epoch": 0.7875955819881053,
-      "grad_norm": 0.4488600194454193,
-      "learning_rate": 8.563569824372957e-06,
-      "loss": 0.3705,
-      "step": 12051
-    },
-    {
-      "epoch": 0.7876609371936475,
-      "grad_norm": 0.49246707558631897,
-      "learning_rate": 8.563324874684367e-06,
-      "loss": 0.4294,
-      "step": 12052
-    },
-    {
-      "epoch": 0.7877262923991896,
-      "grad_norm": 0.45360997319221497,
-      "learning_rate": 8.563079907616208e-06,
-      "loss": 0.395,
-      "step": 12053
-    },
-    {
-      "epoch": 0.7877916476047317,
-      "grad_norm": 0.45833253860473633,
-      "learning_rate": 8.562834923169673e-06,
-      "loss": 0.3672,
-      "step": 12054
-    },
-    {
-      "epoch": 0.7878570028102738,
-      "grad_norm": 0.47918999195098877,
-      "learning_rate": 8.562589921345958e-06,
-      "loss": 0.3636,
-      "step": 12055
-    },
-    {
-      "epoch": 0.787922358015816,
-      "grad_norm": 0.4814555048942566,
-      "learning_rate": 8.562344902146259e-06,
-      "loss": 0.3856,
-      "step": 12056
-    },
-    {
-      "epoch": 0.7879877132213581,
-      "grad_norm": 0.4582223892211914,
-      "learning_rate": 8.562099865571767e-06,
-      "loss": 0.3809,
-      "step": 12057
-    },
-    {
-      "epoch": 0.7880530684269001,
-      "grad_norm": 0.4686221480369568,
-      "learning_rate": 8.561854811623683e-06,
-      "loss": 0.4026,
-      "step": 12058
-    },
-    {
-      "epoch": 0.7881184236324423,
-      "grad_norm": 0.43056002259254456,
-      "learning_rate": 8.561609740303197e-06,
-      "loss": 0.336,
-      "step": 12059
-    },
-    {
-      "epoch": 0.7881837788379844,
-      "grad_norm": 0.4640575051307678,
-      "learning_rate": 8.561364651611507e-06,
-      "loss": 0.4138,
-      "step": 12060
-    },
-    {
-      "epoch": 0.7882491340435266,
-      "grad_norm": 0.4174641966819763,
-      "learning_rate": 8.561119545549807e-06,
-      "loss": 0.343,
-      "step": 12061
-    },
-    {
-      "epoch": 0.7883144892490687,
-      "grad_norm": 0.42930740118026733,
-      "learning_rate": 8.560874422119295e-06,
-      "loss": 0.3573,
-      "step": 12062
-    },
-    {
-      "epoch": 0.7883798444546108,
-      "grad_norm": 0.5078632235527039,
-      "learning_rate": 8.560629281321163e-06,
-      "loss": 0.4038,
-      "step": 12063
-    },
-    {
-      "epoch": 0.7884451996601529,
-      "grad_norm": 0.444355845451355,
-      "learning_rate": 8.56038412315661e-06,
-      "loss": 0.393,
-      "step": 12064
-    },
-    {
-      "epoch": 0.7885105548656951,
-      "grad_norm": 0.41402336955070496,
-      "learning_rate": 8.560138947626831e-06,
-      "loss": 0.354,
-      "step": 12065
-    },
-    {
-      "epoch": 0.7885759100712372,
-      "grad_norm": 0.45499247312545776,
-      "learning_rate": 8.559893754733019e-06,
-      "loss": 0.3967,
-      "step": 12066
-    },
-    {
-      "epoch": 0.7886412652767792,
-      "grad_norm": 0.4375172555446625,
-      "learning_rate": 8.559648544476371e-06,
-      "loss": 0.3498,
-      "step": 12067
-    },
-    {
-      "epoch": 0.7887066204823214,
-      "grad_norm": 0.4388751983642578,
-      "learning_rate": 8.559403316858086e-06,
-      "loss": 0.3374,
-      "step": 12068
-    },
-    {
-      "epoch": 0.7887719756878635,
-      "grad_norm": 0.41842880845069885,
-      "learning_rate": 8.55915807187936e-06,
-      "loss": 0.3576,
-      "step": 12069
-    },
-    {
-      "epoch": 0.7888373308934057,
-      "grad_norm": 0.4384239614009857,
-      "learning_rate": 8.558912809541382e-06,
-      "loss": 0.3983,
-      "step": 12070
-    },
-    {
-      "epoch": 0.7889026860989478,
-      "grad_norm": 0.4617427587509155,
-      "learning_rate": 8.558667529845358e-06,
-      "loss": 0.4209,
-      "step": 12071
-    },
-    {
-      "epoch": 0.7889680413044899,
-      "grad_norm": 0.44441232085227966,
-      "learning_rate": 8.558422232792477e-06,
-      "loss": 0.3714,
-      "step": 12072
-    },
-    {
-      "epoch": 0.789033396510032,
-      "grad_norm": 0.4267987608909607,
-      "learning_rate": 8.558176918383936e-06,
-      "loss": 0.3215,
-      "step": 12073
-    },
-    {
-      "epoch": 0.7890987517155742,
-      "grad_norm": 0.46230125427246094,
-      "learning_rate": 8.557931586620936e-06,
-      "loss": 0.3761,
-      "step": 12074
-    },
-    {
-      "epoch": 0.7891641069211163,
-      "grad_norm": 0.4586085379123688,
-      "learning_rate": 8.557686237504672e-06,
-      "loss": 0.3891,
-      "step": 12075
-    },
-    {
-      "epoch": 0.7892294621266583,
-      "grad_norm": 0.4671705961227417,
-      "learning_rate": 8.557440871036338e-06,
-      "loss": 0.3737,
-      "step": 12076
-    },
-    {
-      "epoch": 0.7892948173322005,
-      "grad_norm": 0.4593253433704376,
-      "learning_rate": 8.557195487217134e-06,
-      "loss": 0.4024,
-      "step": 12077
-    },
-    {
-      "epoch": 0.7893601725377426,
-      "grad_norm": 0.45820721983909607,
-      "learning_rate": 8.556950086048255e-06,
-      "loss": 0.3694,
-      "step": 12078
-    },
-    {
-      "epoch": 0.7894255277432848,
-      "grad_norm": 0.4556812047958374,
-      "learning_rate": 8.556704667530897e-06,
-      "loss": 0.3528,
-      "step": 12079
-    },
-    {
-      "epoch": 0.7894908829488269,
-      "grad_norm": 0.46928471326828003,
-      "learning_rate": 8.556459231666259e-06,
-      "loss": 0.3941,
-      "step": 12080
-    },
-    {
-      "epoch": 0.789556238154369,
-      "grad_norm": 0.4141043424606323,
-      "learning_rate": 8.556213778455535e-06,
-      "loss": 0.3726,
-      "step": 12081
-    },
-    {
-      "epoch": 0.7896215933599111,
-      "grad_norm": 0.45621001720428467,
-      "learning_rate": 8.555968307899927e-06,
-      "loss": 0.38,
-      "step": 12082
-    },
-    {
-      "epoch": 0.7896869485654532,
-      "grad_norm": 0.5583800077438354,
-      "learning_rate": 8.55572282000063e-06,
-      "loss": 0.4037,
-      "step": 12083
-    },
-    {
-      "epoch": 0.7897523037709954,
-      "grad_norm": 0.4414021968841553,
-      "learning_rate": 8.55547731475884e-06,
-      "loss": 0.3924,
-      "step": 12084
-    },
-    {
-      "epoch": 0.7898176589765374,
-      "grad_norm": 0.44545963406562805,
-      "learning_rate": 8.555231792175755e-06,
-      "loss": 0.3902,
-      "step": 12085
-    },
-    {
-      "epoch": 0.7898830141820796,
-      "grad_norm": 0.4575885832309723,
-      "learning_rate": 8.554986252252573e-06,
-      "loss": 0.374,
-      "step": 12086
-    },
-    {
-      "epoch": 0.7899483693876217,
-      "grad_norm": 0.43306130170822144,
-      "learning_rate": 8.554740694990493e-06,
-      "loss": 0.3648,
-      "step": 12087
-    },
-    {
-      "epoch": 0.7900137245931639,
-      "grad_norm": 0.43865975737571716,
-      "learning_rate": 8.55449512039071e-06,
-      "loss": 0.3752,
-      "step": 12088
-    },
-    {
-      "epoch": 0.790079079798706,
-      "grad_norm": 0.40036046504974365,
-      "learning_rate": 8.554249528454422e-06,
-      "loss": 0.3302,
-      "step": 12089
-    },
-    {
-      "epoch": 0.7901444350042481,
-      "grad_norm": 0.4885970950126648,
-      "learning_rate": 8.554003919182829e-06,
-      "loss": 0.4241,
-      "step": 12090
-    },
-    {
-      "epoch": 0.7902097902097902,
-      "grad_norm": 0.43497803807258606,
-      "learning_rate": 8.553758292577128e-06,
-      "loss": 0.3737,
-      "step": 12091
-    },
-    {
-      "epoch": 0.7902751454153323,
-      "grad_norm": 0.4482162892818451,
-      "learning_rate": 8.553512648638515e-06,
-      "loss": 0.3932,
-      "step": 12092
-    },
-    {
-      "epoch": 0.7903405006208745,
-      "grad_norm": 0.42345380783081055,
-      "learning_rate": 8.55326698736819e-06,
-      "loss": 0.3651,
-      "step": 12093
-    },
-    {
-      "epoch": 0.7904058558264165,
-      "grad_norm": 0.4518062174320221,
-      "learning_rate": 8.553021308767353e-06,
-      "loss": 0.4081,
-      "step": 12094
-    },
-    {
-      "epoch": 0.7904712110319587,
-      "grad_norm": 0.41406533122062683,
-      "learning_rate": 8.552775612837198e-06,
-      "loss": 0.319,
-      "step": 12095
-    },
-    {
-      "epoch": 0.7905365662375008,
-      "grad_norm": 0.43780308961868286,
-      "learning_rate": 8.552529899578927e-06,
-      "loss": 0.3725,
-      "step": 12096
-    },
-    {
-      "epoch": 0.790601921443043,
-      "grad_norm": 0.4579823315143585,
-      "learning_rate": 8.552284168993736e-06,
-      "loss": 0.3625,
-      "step": 12097
-    },
-    {
-      "epoch": 0.790667276648585,
-      "grad_norm": 0.42775416374206543,
-      "learning_rate": 8.552038421082824e-06,
-      "loss": 0.357,
-      "step": 12098
-    },
-    {
-      "epoch": 0.7907326318541272,
-      "grad_norm": 0.4572090804576874,
-      "learning_rate": 8.551792655847392e-06,
-      "loss": 0.3869,
-      "step": 12099
-    },
-    {
-      "epoch": 0.7907979870596693,
-      "grad_norm": 0.442373126745224,
-      "learning_rate": 8.551546873288636e-06,
-      "loss": 0.3707,
-      "step": 12100
-    },
-    {
-      "epoch": 0.7908633422652114,
-      "grad_norm": 0.42555850744247437,
-      "learning_rate": 8.551301073407755e-06,
-      "loss": 0.3863,
-      "step": 12101
-    },
-    {
-      "epoch": 0.7909286974707536,
-      "grad_norm": 0.43132033944129944,
-      "learning_rate": 8.551055256205947e-06,
-      "loss": 0.3972,
-      "step": 12102
-    },
-    {
-      "epoch": 0.7909940526762956,
-      "grad_norm": 0.44253009557724,
-      "learning_rate": 8.550809421684415e-06,
-      "loss": 0.4081,
-      "step": 12103
-    },
-    {
-      "epoch": 0.7910594078818378,
-      "grad_norm": 0.47049644589424133,
-      "learning_rate": 8.550563569844354e-06,
-      "loss": 0.4063,
-      "step": 12104
-    },
-    {
-      "epoch": 0.7911247630873799,
-      "grad_norm": 0.469762921333313,
-      "learning_rate": 8.550317700686965e-06,
-      "loss": 0.4042,
-      "step": 12105
-    },
-    {
-      "epoch": 0.7911901182929221,
-      "grad_norm": 0.4723997116088867,
-      "learning_rate": 8.550071814213447e-06,
-      "loss": 0.4079,
-      "step": 12106
-    },
-    {
-      "epoch": 0.7912554734984641,
-      "grad_norm": 0.43045368790626526,
-      "learning_rate": 8.549825910424999e-06,
-      "loss": 0.3287,
-      "step": 12107
-    },
-    {
-      "epoch": 0.7913208287040063,
-      "grad_norm": 0.43018412590026855,
-      "learning_rate": 8.549579989322818e-06,
-      "loss": 0.3721,
-      "step": 12108
-    },
-    {
-      "epoch": 0.7913861839095484,
-      "grad_norm": 0.46723949909210205,
-      "learning_rate": 8.549334050908108e-06,
-      "loss": 0.4454,
-      "step": 12109
-    },
-    {
-      "epoch": 0.7914515391150905,
-      "grad_norm": 0.44700801372528076,
-      "learning_rate": 8.549088095182064e-06,
-      "loss": 0.3945,
-      "step": 12110
-    },
-    {
-      "epoch": 0.7915168943206327,
-      "grad_norm": 0.45005619525909424,
-      "learning_rate": 8.54884212214589e-06,
-      "loss": 0.3962,
-      "step": 12111
-    },
-    {
-      "epoch": 0.7915822495261747,
-      "grad_norm": 0.47676965594291687,
-      "learning_rate": 8.548596131800782e-06,
-      "loss": 0.4342,
-      "step": 12112
-    },
-    {
-      "epoch": 0.7916476047317169,
-      "grad_norm": 0.41420450806617737,
-      "learning_rate": 8.548350124147941e-06,
-      "loss": 0.3374,
-      "step": 12113
-    },
-    {
-      "epoch": 0.791712959937259,
-      "grad_norm": 0.42397040128707886,
-      "learning_rate": 8.54810409918857e-06,
-      "loss": 0.3492,
-      "step": 12114
-    },
-    {
-      "epoch": 0.7917783151428012,
-      "grad_norm": 0.4471323788166046,
-      "learning_rate": 8.547858056923863e-06,
-      "loss": 0.3641,
-      "step": 12115
-    },
-    {
-      "epoch": 0.7918436703483432,
-      "grad_norm": 0.44073548913002014,
-      "learning_rate": 8.547611997355025e-06,
-      "loss": 0.4107,
-      "step": 12116
-    },
-    {
-      "epoch": 0.7919090255538853,
-      "grad_norm": 0.4198542535305023,
-      "learning_rate": 8.547365920483253e-06,
-      "loss": 0.3616,
-      "step": 12117
-    },
-    {
-      "epoch": 0.7919743807594275,
-      "grad_norm": 0.4398666322231293,
-      "learning_rate": 8.547119826309751e-06,
-      "loss": 0.3616,
-      "step": 12118
-    },
-    {
-      "epoch": 0.7920397359649696,
-      "grad_norm": 0.47608432173728943,
-      "learning_rate": 8.546873714835714e-06,
-      "loss": 0.4092,
-      "step": 12119
-    },
-    {
-      "epoch": 0.7921050911705118,
-      "grad_norm": 0.478184312582016,
-      "learning_rate": 8.546627586062346e-06,
-      "loss": 0.4113,
-      "step": 12120
-    },
-    {
-      "epoch": 0.7921704463760538,
-      "grad_norm": 0.4458080232143402,
-      "learning_rate": 8.546381439990847e-06,
-      "loss": 0.3804,
-      "step": 12121
-    },
-    {
-      "epoch": 0.792235801581596,
-      "grad_norm": 0.5123405456542969,
-      "learning_rate": 8.546135276622417e-06,
-      "loss": 0.4001,
-      "step": 12122
-    },
-    {
-      "epoch": 0.7923011567871381,
-      "grad_norm": 0.4072600305080414,
-      "learning_rate": 8.545889095958257e-06,
-      "loss": 0.3518,
-      "step": 12123
-    },
-    {
-      "epoch": 0.7923665119926803,
-      "grad_norm": 0.43664419651031494,
-      "learning_rate": 8.545642897999567e-06,
-      "loss": 0.3409,
-      "step": 12124
-    },
-    {
-      "epoch": 0.7924318671982223,
-      "grad_norm": 0.43069079518318176,
-      "learning_rate": 8.545396682747548e-06,
-      "loss": 0.3779,
-      "step": 12125
-    },
-    {
-      "epoch": 0.7924972224037644,
-      "grad_norm": 0.46924296021461487,
-      "learning_rate": 8.545150450203401e-06,
-      "loss": 0.4053,
-      "step": 12126
-    },
-    {
-      "epoch": 0.7925625776093066,
-      "grad_norm": 0.4362129867076874,
-      "learning_rate": 8.544904200368328e-06,
-      "loss": 0.3578,
-      "step": 12127
-    },
-    {
-      "epoch": 0.7926279328148487,
-      "grad_norm": 0.4406747817993164,
-      "learning_rate": 8.544657933243529e-06,
-      "loss": 0.3544,
-      "step": 12128
-    },
-    {
-      "epoch": 0.7926932880203909,
-      "grad_norm": 0.4239201843738556,
-      "learning_rate": 8.544411648830205e-06,
-      "loss": 0.3447,
-      "step": 12129
-    },
-    {
-      "epoch": 0.7927586432259329,
-      "grad_norm": 0.44481077790260315,
-      "learning_rate": 8.544165347129558e-06,
-      "loss": 0.3991,
-      "step": 12130
-    },
-    {
-      "epoch": 0.7928239984314751,
-      "grad_norm": 0.43191221356391907,
-      "learning_rate": 8.543919028142788e-06,
-      "loss": 0.3629,
-      "step": 12131
-    },
-    {
-      "epoch": 0.7928893536370172,
-      "grad_norm": 0.48538875579833984,
-      "learning_rate": 8.543672691871096e-06,
-      "loss": 0.3817,
-      "step": 12132
-    },
-    {
-      "epoch": 0.7929547088425594,
-      "grad_norm": 0.44361257553100586,
-      "learning_rate": 8.543426338315687e-06,
-      "loss": 0.3706,
-      "step": 12133
-    },
-    {
-      "epoch": 0.7930200640481014,
-      "grad_norm": 0.42010653018951416,
-      "learning_rate": 8.54317996747776e-06,
-      "loss": 0.3392,
-      "step": 12134
-    },
-    {
-      "epoch": 0.7930854192536435,
-      "grad_norm": 0.42698851227760315,
-      "learning_rate": 8.542933579358516e-06,
-      "loss": 0.3383,
-      "step": 12135
-    },
-    {
-      "epoch": 0.7931507744591857,
-      "grad_norm": 0.5849870443344116,
-      "learning_rate": 8.542687173959156e-06,
-      "loss": 0.3811,
-      "step": 12136
-    },
-    {
-      "epoch": 0.7932161296647278,
-      "grad_norm": 0.4533376097679138,
-      "learning_rate": 8.542440751280888e-06,
-      "loss": 0.4063,
-      "step": 12137
-    },
-    {
-      "epoch": 0.79328148487027,
-      "grad_norm": 0.4817275404930115,
-      "learning_rate": 8.542194311324904e-06,
-      "loss": 0.4029,
-      "step": 12138
-    },
-    {
-      "epoch": 0.793346840075812,
-      "grad_norm": 0.4313760995864868,
-      "learning_rate": 8.541947854092413e-06,
-      "loss": 0.3845,
-      "step": 12139
-    },
-    {
-      "epoch": 0.7934121952813542,
-      "grad_norm": 0.47342512011528015,
-      "learning_rate": 8.541701379584618e-06,
-      "loss": 0.4543,
-      "step": 12140
-    },
-    {
-      "epoch": 0.7934775504868963,
-      "grad_norm": 0.4508894085884094,
-      "learning_rate": 8.541454887802715e-06,
-      "loss": 0.3724,
-      "step": 12141
-    },
-    {
-      "epoch": 0.7935429056924383,
-      "grad_norm": 0.4701017141342163,
-      "learning_rate": 8.541208378747912e-06,
-      "loss": 0.3684,
-      "step": 12142
-    },
-    {
-      "epoch": 0.7936082608979805,
-      "grad_norm": 0.4481446146965027,
-      "learning_rate": 8.540961852421407e-06,
-      "loss": 0.3518,
-      "step": 12143
-    },
-    {
-      "epoch": 0.7936736161035226,
-      "grad_norm": 0.43516772985458374,
-      "learning_rate": 8.540715308824405e-06,
-      "loss": 0.3384,
-      "step": 12144
-    },
-    {
-      "epoch": 0.7937389713090648,
-      "grad_norm": 0.43795207142829895,
-      "learning_rate": 8.540468747958107e-06,
-      "loss": 0.3831,
-      "step": 12145
-    },
-    {
-      "epoch": 0.7938043265146069,
-      "grad_norm": 0.4666205048561096,
-      "learning_rate": 8.540222169823718e-06,
-      "loss": 0.4061,
-      "step": 12146
-    },
-    {
-      "epoch": 0.793869681720149,
-      "grad_norm": 0.4302486777305603,
-      "learning_rate": 8.539975574422438e-06,
-      "loss": 0.3591,
-      "step": 12147
-    },
-    {
-      "epoch": 0.7939350369256911,
-      "grad_norm": 0.432462602853775,
-      "learning_rate": 8.539728961755471e-06,
-      "loss": 0.3907,
-      "step": 12148
-    },
-    {
-      "epoch": 0.7940003921312333,
-      "grad_norm": 0.42387595772743225,
-      "learning_rate": 8.539482331824018e-06,
-      "loss": 0.3565,
-      "step": 12149
-    },
-    {
-      "epoch": 0.7940657473367754,
-      "grad_norm": 0.4779075086116791,
-      "learning_rate": 8.539235684629286e-06,
-      "loss": 0.4376,
-      "step": 12150
-    },
-    {
-      "epoch": 0.7941311025423174,
-      "grad_norm": 0.4292527735233307,
-      "learning_rate": 8.538989020172475e-06,
-      "loss": 0.352,
-      "step": 12151
-    },
-    {
-      "epoch": 0.7941964577478596,
-      "grad_norm": 0.4452821910381317,
-      "learning_rate": 8.538742338454789e-06,
-      "loss": 0.4291,
-      "step": 12152
-    },
-    {
-      "epoch": 0.7942618129534017,
-      "grad_norm": 0.5022962093353271,
-      "learning_rate": 8.538495639477428e-06,
-      "loss": 0.4611,
-      "step": 12153
-    },
-    {
-      "epoch": 0.7943271681589439,
-      "grad_norm": 0.44315487146377563,
-      "learning_rate": 8.538248923241601e-06,
-      "loss": 0.3611,
-      "step": 12154
-    },
-    {
-      "epoch": 0.794392523364486,
-      "grad_norm": 0.4711770713329315,
-      "learning_rate": 8.538002189748507e-06,
-      "loss": 0.4083,
-      "step": 12155
-    },
-    {
-      "epoch": 0.7944578785700281,
-      "grad_norm": 0.4432166516780853,
-      "learning_rate": 8.537755438999348e-06,
-      "loss": 0.4045,
-      "step": 12156
-    },
-    {
-      "epoch": 0.7945232337755702,
-      "grad_norm": 0.46643519401550293,
-      "learning_rate": 8.537508670995334e-06,
-      "loss": 0.39,
-      "step": 12157
-    },
-    {
-      "epoch": 0.7945885889811124,
-      "grad_norm": 0.41968482732772827,
-      "learning_rate": 8.537261885737662e-06,
-      "loss": 0.3605,
-      "step": 12158
-    },
-    {
-      "epoch": 0.7946539441866545,
-      "grad_norm": 0.4558030664920807,
-      "learning_rate": 8.53701508322754e-06,
-      "loss": 0.3832,
-      "step": 12159
-    },
-    {
-      "epoch": 0.7947192993921965,
-      "grad_norm": 0.44898343086242676,
-      "learning_rate": 8.536768263466171e-06,
-      "loss": 0.3798,
-      "step": 12160
-    },
-    {
-      "epoch": 0.7947846545977387,
-      "grad_norm": 0.45856231451034546,
-      "learning_rate": 8.536521426454758e-06,
-      "loss": 0.3989,
-      "step": 12161
-    },
-    {
-      "epoch": 0.7948500098032808,
-      "grad_norm": 0.41914859414100647,
-      "learning_rate": 8.536274572194502e-06,
-      "loss": 0.3207,
-      "step": 12162
-    },
-    {
-      "epoch": 0.794915365008823,
-      "grad_norm": 0.4341502785682678,
-      "learning_rate": 8.536027700686613e-06,
-      "loss": 0.3604,
-      "step": 12163
-    },
-    {
-      "epoch": 0.794980720214365,
-      "grad_norm": 0.4363396167755127,
-      "learning_rate": 8.535780811932289e-06,
-      "loss": 0.3688,
-      "step": 12164
-    },
-    {
-      "epoch": 0.7950460754199072,
-      "grad_norm": 0.44624993205070496,
-      "learning_rate": 8.535533905932739e-06,
-      "loss": 0.4038,
-      "step": 12165
-    },
-    {
-      "epoch": 0.7951114306254493,
-      "grad_norm": 0.42349255084991455,
-      "learning_rate": 8.535286982689164e-06,
-      "loss": 0.3669,
-      "step": 12166
-    },
-    {
-      "epoch": 0.7951767858309914,
-      "grad_norm": 0.4911760687828064,
-      "learning_rate": 8.53504004220277e-06,
-      "loss": 0.4371,
-      "step": 12167
-    },
-    {
-      "epoch": 0.7952421410365336,
-      "grad_norm": 0.4804944097995758,
-      "learning_rate": 8.534793084474761e-06,
-      "loss": 0.3773,
-      "step": 12168
-    },
-    {
-      "epoch": 0.7953074962420756,
-      "grad_norm": 0.4469417631626129,
-      "learning_rate": 8.534546109506342e-06,
-      "loss": 0.384,
-      "step": 12169
-    },
-    {
-      "epoch": 0.7953728514476178,
-      "grad_norm": 0.4474070966243744,
-      "learning_rate": 8.534299117298718e-06,
-      "loss": 0.3865,
-      "step": 12170
-    },
-    {
-      "epoch": 0.7954382066531599,
-      "grad_norm": 0.4352922737598419,
-      "learning_rate": 8.534052107853092e-06,
-      "loss": 0.3551,
-      "step": 12171
-    },
-    {
-      "epoch": 0.7955035618587021,
-      "grad_norm": 0.45055779814720154,
-      "learning_rate": 8.533805081170669e-06,
-      "loss": 0.3833,
-      "step": 12172
-    },
-    {
-      "epoch": 0.7955689170642442,
-      "grad_norm": 0.3985026180744171,
-      "learning_rate": 8.533558037252654e-06,
-      "loss": 0.3392,
-      "step": 12173
-    },
-    {
-      "epoch": 0.7956342722697863,
-      "grad_norm": 0.40868639945983887,
-      "learning_rate": 8.533310976100252e-06,
-      "loss": 0.3638,
-      "step": 12174
-    },
-    {
-      "epoch": 0.7956996274753284,
-      "grad_norm": 0.4311105012893677,
-      "learning_rate": 8.53306389771467e-06,
-      "loss": 0.3356,
-      "step": 12175
-    },
-    {
-      "epoch": 0.7957649826808705,
-      "grad_norm": 0.4652247130870819,
-      "learning_rate": 8.532816802097112e-06,
-      "loss": 0.3835,
-      "step": 12176
-    },
-    {
-      "epoch": 0.7958303378864127,
-      "grad_norm": 0.44967931509017944,
-      "learning_rate": 8.532569689248782e-06,
-      "loss": 0.3794,
-      "step": 12177
-    },
-    {
-      "epoch": 0.7958956930919547,
-      "grad_norm": 0.4602295160293579,
-      "learning_rate": 8.532322559170885e-06,
-      "loss": 0.3753,
-      "step": 12178
-    },
-    {
-      "epoch": 0.7959610482974969,
-      "grad_norm": 0.4297156035900116,
-      "learning_rate": 8.532075411864628e-06,
-      "loss": 0.3654,
-      "step": 12179
-    },
-    {
-      "epoch": 0.796026403503039,
-      "grad_norm": 0.46110835671424866,
-      "learning_rate": 8.531828247331213e-06,
-      "loss": 0.4192,
-      "step": 12180
-    },
-    {
-      "epoch": 0.7960917587085812,
-      "grad_norm": 0.43538546562194824,
-      "learning_rate": 8.531581065571852e-06,
-      "loss": 0.3274,
-      "step": 12181
-    },
-    {
-      "epoch": 0.7961571139141232,
-      "grad_norm": 0.46276628971099854,
-      "learning_rate": 8.531333866587744e-06,
-      "loss": 0.38,
-      "step": 12182
-    },
-    {
-      "epoch": 0.7962224691196654,
-      "grad_norm": 0.4322505593299866,
-      "learning_rate": 8.531086650380098e-06,
-      "loss": 0.344,
-      "step": 12183
-    },
-    {
-      "epoch": 0.7962878243252075,
-      "grad_norm": 0.43760770559310913,
-      "learning_rate": 8.53083941695012e-06,
-      "loss": 0.3951,
-      "step": 12184
-    },
-    {
-      "epoch": 0.7963531795307496,
-      "grad_norm": 0.45469361543655396,
-      "learning_rate": 8.530592166299014e-06,
-      "loss": 0.4405,
-      "step": 12185
-    },
-    {
-      "epoch": 0.7964185347362918,
-      "grad_norm": 0.41518306732177734,
-      "learning_rate": 8.530344898427987e-06,
-      "loss": 0.3175,
-      "step": 12186
-    },
-    {
-      "epoch": 0.7964838899418338,
-      "grad_norm": 0.4440505802631378,
-      "learning_rate": 8.530097613338244e-06,
-      "loss": 0.3986,
-      "step": 12187
-    },
-    {
-      "epoch": 0.796549245147376,
-      "grad_norm": 0.4459461569786072,
-      "learning_rate": 8.529850311030994e-06,
-      "loss": 0.3866,
-      "step": 12188
-    },
-    {
-      "epoch": 0.7966146003529181,
-      "grad_norm": 0.5046373009681702,
-      "learning_rate": 8.52960299150744e-06,
-      "loss": 0.3608,
-      "step": 12189
-    },
-    {
-      "epoch": 0.7966799555584603,
-      "grad_norm": 0.44640427827835083,
-      "learning_rate": 8.52935565476879e-06,
-      "loss": 0.3942,
-      "step": 12190
-    },
-    {
-      "epoch": 0.7967453107640023,
-      "grad_norm": 0.4735654592514038,
-      "learning_rate": 8.52910830081625e-06,
-      "loss": 0.4165,
-      "step": 12191
-    },
-    {
-      "epoch": 0.7968106659695445,
-      "grad_norm": 0.47037145495414734,
-      "learning_rate": 8.528860929651024e-06,
-      "loss": 0.3997,
-      "step": 12192
-    },
-    {
-      "epoch": 0.7968760211750866,
-      "grad_norm": 0.41389429569244385,
-      "learning_rate": 8.528613541274323e-06,
-      "loss": 0.3523,
-      "step": 12193
-    },
-    {
-      "epoch": 0.7969413763806287,
-      "grad_norm": 0.44254520535469055,
-      "learning_rate": 8.528366135687351e-06,
-      "loss": 0.3442,
-      "step": 12194
-    },
-    {
-      "epoch": 0.7970067315861709,
-      "grad_norm": 0.4370190501213074,
-      "learning_rate": 8.528118712891314e-06,
-      "loss": 0.3524,
-      "step": 12195
-    },
-    {
-      "epoch": 0.7970720867917129,
-      "grad_norm": 0.423596054315567,
-      "learning_rate": 8.52787127288742e-06,
-      "loss": 0.382,
-      "step": 12196
-    },
-    {
-      "epoch": 0.7971374419972551,
-      "grad_norm": 0.4484865367412567,
-      "learning_rate": 8.527623815676878e-06,
-      "loss": 0.3859,
-      "step": 12197
-    },
-    {
-      "epoch": 0.7972027972027972,
-      "grad_norm": 0.4844169020652771,
-      "learning_rate": 8.52737634126089e-06,
-      "loss": 0.4199,
-      "step": 12198
-    },
-    {
-      "epoch": 0.7972681524083394,
-      "grad_norm": 0.4097655117511749,
-      "learning_rate": 8.527128849640667e-06,
-      "loss": 0.3088,
-      "step": 12199
-    },
-    {
-      "epoch": 0.7973335076138814,
-      "grad_norm": 0.4613126516342163,
-      "learning_rate": 8.526881340817414e-06,
-      "loss": 0.3832,
-      "step": 12200
-    },
-    {
-      "epoch": 0.7973988628194235,
-      "grad_norm": 0.4373398423194885,
-      "learning_rate": 8.526633814792338e-06,
-      "loss": 0.3604,
-      "step": 12201
-    },
-    {
-      "epoch": 0.7974642180249657,
-      "grad_norm": 0.43399253487586975,
-      "learning_rate": 8.526386271566647e-06,
-      "loss": 0.39,
-      "step": 12202
-    },
-    {
-      "epoch": 0.7975295732305078,
-      "grad_norm": 0.4530579447746277,
-      "learning_rate": 8.526138711141551e-06,
-      "loss": 0.3794,
-      "step": 12203
-    },
-    {
-      "epoch": 0.79759492843605,
-      "grad_norm": 0.4073675870895386,
-      "learning_rate": 8.525891133518252e-06,
-      "loss": 0.3299,
-      "step": 12204
-    },
-    {
-      "epoch": 0.797660283641592,
-      "grad_norm": 0.4187154173851013,
-      "learning_rate": 8.525643538697963e-06,
-      "loss": 0.3465,
-      "step": 12205
-    },
-    {
-      "epoch": 0.7977256388471342,
-      "grad_norm": 0.4881112575531006,
-      "learning_rate": 8.525395926681887e-06,
-      "loss": 0.3826,
-      "step": 12206
-    },
-    {
-      "epoch": 0.7977909940526763,
-      "grad_norm": 0.4807402491569519,
-      "learning_rate": 8.525148297471236e-06,
-      "loss": 0.4153,
-      "step": 12207
-    },
-    {
-      "epoch": 0.7978563492582185,
-      "grad_norm": 0.4205600917339325,
-      "learning_rate": 8.524900651067213e-06,
-      "loss": 0.3446,
-      "step": 12208
-    },
-    {
-      "epoch": 0.7979217044637605,
-      "grad_norm": 0.47125667333602905,
-      "learning_rate": 8.524652987471029e-06,
-      "loss": 0.3948,
-      "step": 12209
-    },
-    {
-      "epoch": 0.7979870596693026,
-      "grad_norm": 0.40634217858314514,
-      "learning_rate": 8.524405306683892e-06,
-      "loss": 0.3124,
-      "step": 12210
-    },
-    {
-      "epoch": 0.7980524148748448,
-      "grad_norm": 0.4541061818599701,
-      "learning_rate": 8.52415760870701e-06,
-      "loss": 0.3901,
-      "step": 12211
-    },
-    {
-      "epoch": 0.7981177700803869,
-      "grad_norm": 0.45898061990737915,
-      "learning_rate": 8.52390989354159e-06,
-      "loss": 0.3824,
-      "step": 12212
-    },
-    {
-      "epoch": 0.798183125285929,
-      "grad_norm": 0.3998471796512604,
-      "learning_rate": 8.523662161188839e-06,
-      "loss": 0.3108,
-      "step": 12213
-    },
-    {
-      "epoch": 0.7982484804914711,
-      "grad_norm": 0.4205678701400757,
-      "learning_rate": 8.523414411649968e-06,
-      "loss": 0.3818,
-      "step": 12214
-    },
-    {
-      "epoch": 0.7983138356970133,
-      "grad_norm": 0.4338730275630951,
-      "learning_rate": 8.523166644926185e-06,
-      "loss": 0.3599,
-      "step": 12215
-    },
-    {
-      "epoch": 0.7983791909025554,
-      "grad_norm": 0.4278663396835327,
-      "learning_rate": 8.522918861018698e-06,
-      "loss": 0.3531,
-      "step": 12216
-    },
-    {
-      "epoch": 0.7984445461080976,
-      "grad_norm": 0.488136887550354,
-      "learning_rate": 8.522671059928714e-06,
-      "loss": 0.4025,
-      "step": 12217
-    },
-    {
-      "epoch": 0.7985099013136396,
-      "grad_norm": 0.4601065516471863,
-      "learning_rate": 8.522423241657446e-06,
-      "loss": 0.4304,
-      "step": 12218
-    },
-    {
-      "epoch": 0.7985752565191817,
-      "grad_norm": 0.4728713631629944,
-      "learning_rate": 8.522175406206096e-06,
-      "loss": 0.4518,
-      "step": 12219
-    },
-    {
-      "epoch": 0.7986406117247239,
-      "grad_norm": 0.6645670533180237,
-      "learning_rate": 8.521927553575877e-06,
-      "loss": 0.4378,
-      "step": 12220
-    },
-    {
-      "epoch": 0.798705966930266,
-      "grad_norm": 0.4364054203033447,
-      "learning_rate": 8.521679683767996e-06,
-      "loss": 0.3364,
-      "step": 12221
-    },
-    {
-      "epoch": 0.7987713221358081,
-      "grad_norm": 0.46858781576156616,
-      "learning_rate": 8.521431796783666e-06,
-      "loss": 0.3816,
-      "step": 12222
-    },
-    {
-      "epoch": 0.7988366773413502,
-      "grad_norm": 0.4912148118019104,
-      "learning_rate": 8.521183892624092e-06,
-      "loss": 0.4333,
-      "step": 12223
-    },
-    {
-      "epoch": 0.7989020325468924,
-      "grad_norm": 0.4223543405532837,
-      "learning_rate": 8.520935971290484e-06,
-      "loss": 0.3715,
-      "step": 12224
-    },
-    {
-      "epoch": 0.7989673877524345,
-      "grad_norm": 0.47039949893951416,
-      "learning_rate": 8.520688032784051e-06,
-      "loss": 0.4214,
-      "step": 12225
-    },
-    {
-      "epoch": 0.7990327429579765,
-      "grad_norm": 0.43631458282470703,
-      "learning_rate": 8.520440077106004e-06,
-      "loss": 0.3829,
-      "step": 12226
-    },
-    {
-      "epoch": 0.7990980981635187,
-      "grad_norm": 0.4213436543941498,
-      "learning_rate": 8.52019210425755e-06,
-      "loss": 0.351,
-      "step": 12227
-    },
-    {
-      "epoch": 0.7991634533690608,
-      "grad_norm": 0.4542752802371979,
-      "learning_rate": 8.519944114239901e-06,
-      "loss": 0.4028,
-      "step": 12228
-    },
-    {
-      "epoch": 0.799228808574603,
-      "grad_norm": 0.4376654624938965,
-      "learning_rate": 8.519696107054265e-06,
-      "loss": 0.358,
-      "step": 12229
-    },
-    {
-      "epoch": 0.7992941637801451,
-      "grad_norm": 0.461787611246109,
-      "learning_rate": 8.51944808270185e-06,
-      "loss": 0.4355,
-      "step": 12230
-    },
-    {
-      "epoch": 0.7993595189856872,
-      "grad_norm": 0.49870526790618896,
-      "learning_rate": 8.519200041183868e-06,
-      "loss": 0.3983,
-      "step": 12231
-    },
-    {
-      "epoch": 0.7994248741912293,
-      "grad_norm": 0.4470604360103607,
-      "learning_rate": 8.518951982501529e-06,
-      "loss": 0.3696,
-      "step": 12232
-    },
-    {
-      "epoch": 0.7994902293967715,
-      "grad_norm": 0.46619316935539246,
-      "learning_rate": 8.518703906656042e-06,
-      "loss": 0.4443,
-      "step": 12233
-    },
-    {
-      "epoch": 0.7995555846023136,
-      "grad_norm": 0.47135692834854126,
-      "learning_rate": 8.518455813648616e-06,
-      "loss": 0.3962,
-      "step": 12234
-    },
-    {
-      "epoch": 0.7996209398078556,
-      "grad_norm": 0.4475511908531189,
-      "learning_rate": 8.518207703480463e-06,
-      "loss": 0.3908,
-      "step": 12235
-    },
-    {
-      "epoch": 0.7996862950133978,
-      "grad_norm": 0.4283462166786194,
-      "learning_rate": 8.517959576152793e-06,
-      "loss": 0.3298,
-      "step": 12236
-    },
-    {
-      "epoch": 0.7997516502189399,
-      "grad_norm": 0.41792038083076477,
-      "learning_rate": 8.517711431666816e-06,
-      "loss": 0.3231,
-      "step": 12237
-    },
-    {
-      "epoch": 0.7998170054244821,
-      "grad_norm": 0.4252481758594513,
-      "learning_rate": 8.517463270023737e-06,
-      "loss": 0.3798,
-      "step": 12238
-    },
-    {
-      "epoch": 0.7998823606300242,
-      "grad_norm": 0.42247527837753296,
-      "learning_rate": 8.517215091224777e-06,
-      "loss": 0.3541,
-      "step": 12239
-    },
-    {
-      "epoch": 0.7999477158355663,
-      "grad_norm": 0.43758153915405273,
-      "learning_rate": 8.516966895271137e-06,
-      "loss": 0.3501,
-      "step": 12240
-    },
-    {
-      "epoch": 0.8000130710411084,
-      "grad_norm": 0.4436210095882416,
-      "learning_rate": 8.516718682164032e-06,
-      "loss": 0.3677,
-      "step": 12241
-    },
-    {
-      "epoch": 0.8000784262466506,
-      "grad_norm": 0.4343346655368805,
-      "learning_rate": 8.516470451904673e-06,
-      "loss": 0.4103,
-      "step": 12242
-    },
-    {
-      "epoch": 0.8001437814521927,
-      "grad_norm": 0.41750410199165344,
-      "learning_rate": 8.516222204494267e-06,
-      "loss": 0.3427,
-      "step": 12243
-    },
-    {
-      "epoch": 0.8002091366577347,
-      "grad_norm": 0.4535248279571533,
-      "learning_rate": 8.515973939934027e-06,
-      "loss": 0.4254,
-      "step": 12244
-    },
-    {
-      "epoch": 0.8002744918632769,
-      "grad_norm": 0.44574499130249023,
-      "learning_rate": 8.515725658225167e-06,
-      "loss": 0.3841,
-      "step": 12245
-    },
-    {
-      "epoch": 0.800339847068819,
-      "grad_norm": 0.4469446539878845,
-      "learning_rate": 8.515477359368894e-06,
-      "loss": 0.3697,
-      "step": 12246
-    },
-    {
-      "epoch": 0.8004052022743612,
-      "grad_norm": 0.403382807970047,
-      "learning_rate": 8.515229043366417e-06,
-      "loss": 0.3328,
-      "step": 12247
-    },
-    {
-      "epoch": 0.8004705574799033,
-      "grad_norm": 0.42954376339912415,
-      "learning_rate": 8.514980710218955e-06,
-      "loss": 0.3752,
-      "step": 12248
-    },
-    {
-      "epoch": 0.8005359126854454,
-      "grad_norm": 0.4429025650024414,
-      "learning_rate": 8.51473235992771e-06,
-      "loss": 0.3573,
-      "step": 12249
-    },
-    {
-      "epoch": 0.8006012678909875,
-      "grad_norm": 0.43553319573402405,
-      "learning_rate": 8.5144839924939e-06,
-      "loss": 0.3615,
-      "step": 12250
-    },
-    {
-      "epoch": 0.8006666230965296,
-      "grad_norm": 0.44023191928863525,
-      "learning_rate": 8.514235607918732e-06,
-      "loss": 0.3924,
-      "step": 12251
-    },
-    {
-      "epoch": 0.8007319783020718,
-      "grad_norm": 0.46684056520462036,
-      "learning_rate": 8.513987206203423e-06,
-      "loss": 0.4431,
-      "step": 12252
-    },
-    {
-      "epoch": 0.8007973335076138,
-      "grad_norm": 0.42626672983169556,
-      "learning_rate": 8.513738787349178e-06,
-      "loss": 0.3234,
-      "step": 12253
-    },
-    {
-      "epoch": 0.800862688713156,
-      "grad_norm": 0.4428284168243408,
-      "learning_rate": 8.513490351357212e-06,
-      "loss": 0.3766,
-      "step": 12254
-    },
-    {
-      "epoch": 0.8009280439186981,
-      "grad_norm": 0.43598613142967224,
-      "learning_rate": 8.513241898228737e-06,
-      "loss": 0.3352,
-      "step": 12255
-    },
-    {
-      "epoch": 0.8009933991242403,
-      "grad_norm": 0.45488062500953674,
-      "learning_rate": 8.512993427964964e-06,
-      "loss": 0.3913,
-      "step": 12256
-    },
-    {
-      "epoch": 0.8010587543297824,
-      "grad_norm": 0.42344170808792114,
-      "learning_rate": 8.512744940567105e-06,
-      "loss": 0.3673,
-      "step": 12257
-    },
-    {
-      "epoch": 0.8011241095353245,
-      "grad_norm": 0.4180864095687866,
-      "learning_rate": 8.512496436036372e-06,
-      "loss": 0.3367,
-      "step": 12258
-    },
-    {
-      "epoch": 0.8011894647408666,
-      "grad_norm": 0.4518779218196869,
-      "learning_rate": 8.512247914373979e-06,
-      "loss": 0.38,
-      "step": 12259
-    },
-    {
-      "epoch": 0.8012548199464087,
-      "grad_norm": 0.4913984537124634,
-      "learning_rate": 8.511999375581134e-06,
-      "loss": 0.439,
-      "step": 12260
-    },
-    {
-      "epoch": 0.8013201751519509,
-      "grad_norm": 0.4122963845729828,
-      "learning_rate": 8.511750819659052e-06,
-      "loss": 0.3594,
-      "step": 12261
-    },
-    {
-      "epoch": 0.8013855303574929,
-      "grad_norm": 0.4171193242073059,
-      "learning_rate": 8.511502246608945e-06,
-      "loss": 0.3445,
-      "step": 12262
-    },
-    {
-      "epoch": 0.8014508855630351,
-      "grad_norm": 0.43022075295448303,
-      "learning_rate": 8.511253656432024e-06,
-      "loss": 0.3554,
-      "step": 12263
-    },
-    {
-      "epoch": 0.8015162407685772,
-      "grad_norm": 0.4417416453361511,
-      "learning_rate": 8.511005049129502e-06,
-      "loss": 0.3807,
-      "step": 12264
-    },
-    {
-      "epoch": 0.8015815959741194,
-      "grad_norm": 0.45779842138290405,
-      "learning_rate": 8.510756424702594e-06,
-      "loss": 0.3854,
-      "step": 12265
-    },
-    {
-      "epoch": 0.8016469511796614,
-      "grad_norm": 0.43774542212486267,
-      "learning_rate": 8.51050778315251e-06,
-      "loss": 0.3775,
-      "step": 12266
-    },
-    {
-      "epoch": 0.8017123063852036,
-      "grad_norm": 0.44353967905044556,
-      "learning_rate": 8.510259124480462e-06,
-      "loss": 0.3792,
-      "step": 12267
-    },
-    {
-      "epoch": 0.8017776615907457,
-      "grad_norm": 0.4183558523654938,
-      "learning_rate": 8.510010448687666e-06,
-      "loss": 0.3253,
-      "step": 12268
-    },
-    {
-      "epoch": 0.8018430167962878,
-      "grad_norm": 0.426309734582901,
-      "learning_rate": 8.509761755775331e-06,
-      "loss": 0.3579,
-      "step": 12269
-    },
-    {
-      "epoch": 0.80190837200183,
-      "grad_norm": 0.43674948811531067,
-      "learning_rate": 8.509513045744674e-06,
-      "loss": 0.4057,
-      "step": 12270
-    },
-    {
-      "epoch": 0.801973727207372,
-      "grad_norm": 0.41553187370300293,
-      "learning_rate": 8.509264318596906e-06,
-      "loss": 0.3532,
-      "step": 12271
-    },
-    {
-      "epoch": 0.8020390824129142,
-      "grad_norm": 0.4216391146183014,
-      "learning_rate": 8.50901557433324e-06,
-      "loss": 0.3515,
-      "step": 12272
-    },
-    {
-      "epoch": 0.8021044376184563,
-      "grad_norm": 0.48671314120292664,
-      "learning_rate": 8.50876681295489e-06,
-      "loss": 0.4558,
-      "step": 12273
-    },
-    {
-      "epoch": 0.8021697928239985,
-      "grad_norm": 0.4296746850013733,
-      "learning_rate": 8.508518034463066e-06,
-      "loss": 0.3793,
-      "step": 12274
-    },
-    {
-      "epoch": 0.8022351480295405,
-      "grad_norm": 0.45327967405319214,
-      "learning_rate": 8.508269238858986e-06,
-      "loss": 0.4051,
-      "step": 12275
-    },
-    {
-      "epoch": 0.8023005032350827,
-      "grad_norm": 0.4724697172641754,
-      "learning_rate": 8.508020426143862e-06,
-      "loss": 0.3397,
-      "step": 12276
-    },
-    {
-      "epoch": 0.8023658584406248,
-      "grad_norm": 0.44395411014556885,
-      "learning_rate": 8.507771596318905e-06,
-      "loss": 0.3966,
-      "step": 12277
-    },
-    {
-      "epoch": 0.8024312136461669,
-      "grad_norm": 0.4271918535232544,
-      "learning_rate": 8.507522749385334e-06,
-      "loss": 0.3414,
-      "step": 12278
-    },
-    {
-      "epoch": 0.8024965688517091,
-      "grad_norm": 0.47543302178382874,
-      "learning_rate": 8.507273885344356e-06,
-      "loss": 0.3468,
-      "step": 12279
-    },
-    {
-      "epoch": 0.8025619240572511,
-      "grad_norm": 0.4737512469291687,
-      "learning_rate": 8.50702500419719e-06,
-      "loss": 0.4015,
-      "step": 12280
-    },
-    {
-      "epoch": 0.8026272792627933,
-      "grad_norm": 0.45727089047431946,
-      "learning_rate": 8.506776105945049e-06,
-      "loss": 0.3922,
-      "step": 12281
-    },
-    {
-      "epoch": 0.8026926344683354,
-      "grad_norm": 0.4550285041332245,
-      "learning_rate": 8.506527190589145e-06,
-      "loss": 0.3911,
-      "step": 12282
-    },
-    {
-      "epoch": 0.8027579896738776,
-      "grad_norm": 0.4296853542327881,
-      "learning_rate": 8.506278258130692e-06,
-      "loss": 0.4053,
-      "step": 12283
-    },
-    {
-      "epoch": 0.8028233448794196,
-      "grad_norm": 0.42667803168296814,
-      "learning_rate": 8.506029308570907e-06,
-      "loss": 0.3255,
-      "step": 12284
-    },
-    {
-      "epoch": 0.8028887000849617,
-      "grad_norm": 0.4762926995754242,
-      "learning_rate": 8.505780341911001e-06,
-      "loss": 0.4228,
-      "step": 12285
-    },
-    {
-      "epoch": 0.8029540552905039,
-      "grad_norm": 0.43546655774116516,
-      "learning_rate": 8.505531358152191e-06,
-      "loss": 0.3599,
-      "step": 12286
-    },
-    {
-      "epoch": 0.803019410496046,
-      "grad_norm": 0.47526538372039795,
-      "learning_rate": 8.50528235729569e-06,
-      "loss": 0.4114,
-      "step": 12287
-    },
-    {
-      "epoch": 0.8030847657015882,
-      "grad_norm": 0.452122300863266,
-      "learning_rate": 8.505033339342713e-06,
-      "loss": 0.3833,
-      "step": 12288
-    },
-    {
-      "epoch": 0.8031501209071302,
-      "grad_norm": 0.4374464750289917,
-      "learning_rate": 8.504784304294474e-06,
-      "loss": 0.3714,
-      "step": 12289
-    },
-    {
-      "epoch": 0.8032154761126724,
-      "grad_norm": 0.4244959354400635,
-      "learning_rate": 8.504535252152186e-06,
-      "loss": 0.3376,
-      "step": 12290
-    },
-    {
-      "epoch": 0.8032808313182145,
-      "grad_norm": 0.47145622968673706,
-      "learning_rate": 8.504286182917066e-06,
-      "loss": 0.4154,
-      "step": 12291
-    },
-    {
-      "epoch": 0.8033461865237567,
-      "grad_norm": 0.45672184228897095,
-      "learning_rate": 8.50403709659033e-06,
-      "loss": 0.3833,
-      "step": 12292
-    },
-    {
-      "epoch": 0.8034115417292987,
-      "grad_norm": 0.4509463608264923,
-      "learning_rate": 8.50378799317319e-06,
-      "loss": 0.394,
-      "step": 12293
-    },
-    {
-      "epoch": 0.8034768969348408,
-      "grad_norm": 0.44769760966300964,
-      "learning_rate": 8.503538872666861e-06,
-      "loss": 0.4273,
-      "step": 12294
-    },
-    {
-      "epoch": 0.803542252140383,
-      "grad_norm": 0.4521794021129608,
-      "learning_rate": 8.503289735072561e-06,
-      "loss": 0.3701,
-      "step": 12295
-    },
-    {
-      "epoch": 0.8036076073459251,
-      "grad_norm": 0.46313586831092834,
-      "learning_rate": 8.503040580391503e-06,
-      "loss": 0.386,
-      "step": 12296
-    },
-    {
-      "epoch": 0.8036729625514673,
-      "grad_norm": 0.47603359818458557,
-      "learning_rate": 8.502791408624902e-06,
-      "loss": 0.4026,
-      "step": 12297
-    },
-    {
-      "epoch": 0.8037383177570093,
-      "grad_norm": 0.485567569732666,
-      "learning_rate": 8.502542219773974e-06,
-      "loss": 0.4434,
-      "step": 12298
-    },
-    {
-      "epoch": 0.8038036729625515,
-      "grad_norm": 0.4439026117324829,
-      "learning_rate": 8.502293013839935e-06,
-      "loss": 0.3615,
-      "step": 12299
-    },
-    {
-      "epoch": 0.8038690281680936,
-      "grad_norm": 0.42811155319213867,
-      "learning_rate": 8.502043790823999e-06,
-      "loss": 0.3823,
-      "step": 12300
-    },
-    {
-      "epoch": 0.8039343833736358,
-      "grad_norm": 0.45521873235702515,
-      "learning_rate": 8.50179455072738e-06,
-      "loss": 0.4057,
-      "step": 12301
-    },
-    {
-      "epoch": 0.8039997385791778,
-      "grad_norm": 0.4294131100177765,
-      "learning_rate": 8.5015452935513e-06,
-      "loss": 0.3587,
-      "step": 12302
-    },
-    {
-      "epoch": 0.8040650937847199,
-      "grad_norm": 0.45609045028686523,
-      "learning_rate": 8.50129601929697e-06,
-      "loss": 0.3966,
-      "step": 12303
-    },
-    {
-      "epoch": 0.8041304489902621,
-      "grad_norm": 0.43618375062942505,
-      "learning_rate": 8.501046727965603e-06,
-      "loss": 0.3732,
-      "step": 12304
-    },
-    {
-      "epoch": 0.8041958041958042,
-      "grad_norm": 0.4643310010433197,
-      "learning_rate": 8.50079741955842e-06,
-      "loss": 0.4294,
-      "step": 12305
-    },
-    {
-      "epoch": 0.8042611594013463,
-      "grad_norm": 0.40513744950294495,
-      "learning_rate": 8.500548094076635e-06,
-      "loss": 0.3319,
-      "step": 12306
-    },
-    {
-      "epoch": 0.8043265146068884,
-      "grad_norm": 0.4380747675895691,
-      "learning_rate": 8.500298751521465e-06,
-      "loss": 0.3565,
-      "step": 12307
-    },
-    {
-      "epoch": 0.8043918698124306,
-      "grad_norm": 0.4322327971458435,
-      "learning_rate": 8.500049391894125e-06,
-      "loss": 0.3767,
-      "step": 12308
-    },
-    {
-      "epoch": 0.8044572250179727,
-      "grad_norm": 0.41715767979621887,
-      "learning_rate": 8.499800015195832e-06,
-      "loss": 0.3245,
-      "step": 12309
-    },
-    {
-      "epoch": 0.8045225802235147,
-      "grad_norm": 0.41645169258117676,
-      "learning_rate": 8.499550621427801e-06,
-      "loss": 0.3221,
-      "step": 12310
-    },
-    {
-      "epoch": 0.8045879354290569,
-      "grad_norm": 0.42277106642723083,
-      "learning_rate": 8.49930121059125e-06,
-      "loss": 0.3426,
-      "step": 12311
-    },
-    {
-      "epoch": 0.804653290634599,
-      "grad_norm": 0.4513515830039978,
-      "learning_rate": 8.499051782687394e-06,
-      "loss": 0.3751,
-      "step": 12312
-    },
-    {
-      "epoch": 0.8047186458401412,
-      "grad_norm": 0.41784802079200745,
-      "learning_rate": 8.498802337717451e-06,
-      "loss": 0.3433,
-      "step": 12313
-    },
-    {
-      "epoch": 0.8047840010456833,
-      "grad_norm": 0.4818885028362274,
-      "learning_rate": 8.498552875682635e-06,
-      "loss": 0.4179,
-      "step": 12314
-    },
-    {
-      "epoch": 0.8048493562512254,
-      "grad_norm": 0.43679431080818176,
-      "learning_rate": 8.498303396584166e-06,
-      "loss": 0.3568,
-      "step": 12315
-    },
-    {
-      "epoch": 0.8049147114567675,
-      "grad_norm": 0.4772944450378418,
-      "learning_rate": 8.498053900423259e-06,
-      "loss": 0.4143,
-      "step": 12316
-    },
-    {
-      "epoch": 0.8049800666623097,
-      "grad_norm": 0.47681671380996704,
-      "learning_rate": 8.497804387201133e-06,
-      "loss": 0.4233,
-      "step": 12317
-    },
-    {
-      "epoch": 0.8050454218678518,
-      "grad_norm": 0.47561073303222656,
-      "learning_rate": 8.497554856919001e-06,
-      "loss": 0.376,
-      "step": 12318
-    },
-    {
-      "epoch": 0.8051107770733938,
-      "grad_norm": 0.4355849027633667,
-      "learning_rate": 8.497305309578085e-06,
-      "loss": 0.3985,
-      "step": 12319
-    },
-    {
-      "epoch": 0.805176132278936,
-      "grad_norm": 0.43489521741867065,
-      "learning_rate": 8.497055745179597e-06,
-      "loss": 0.315,
-      "step": 12320
-    },
-    {
-      "epoch": 0.8052414874844781,
-      "grad_norm": 0.45800769329071045,
-      "learning_rate": 8.496806163724758e-06,
-      "loss": 0.3793,
-      "step": 12321
-    },
-    {
-      "epoch": 0.8053068426900203,
-      "grad_norm": 0.47356700897216797,
-      "learning_rate": 8.496556565214783e-06,
-      "loss": 0.4409,
-      "step": 12322
-    },
-    {
-      "epoch": 0.8053721978955624,
-      "grad_norm": 0.44182270765304565,
-      "learning_rate": 8.496306949650892e-06,
-      "loss": 0.3491,
-      "step": 12323
-    },
-    {
-      "epoch": 0.8054375531011045,
-      "grad_norm": 0.4367234408855438,
-      "learning_rate": 8.496057317034299e-06,
-      "loss": 0.3939,
-      "step": 12324
-    },
-    {
-      "epoch": 0.8055029083066466,
-      "grad_norm": 0.45257240533828735,
-      "learning_rate": 8.495807667366222e-06,
-      "loss": 0.3617,
-      "step": 12325
-    },
-    {
-      "epoch": 0.8055682635121888,
-      "grad_norm": 0.4408845007419586,
-      "learning_rate": 8.495558000647883e-06,
-      "loss": 0.3758,
-      "step": 12326
-    },
-    {
-      "epoch": 0.8056336187177309,
-      "grad_norm": 0.47732558846473694,
-      "learning_rate": 8.495308316880496e-06,
-      "loss": 0.3908,
-      "step": 12327
-    },
-    {
-      "epoch": 0.8056989739232729,
-      "grad_norm": 0.41833725571632385,
-      "learning_rate": 8.495058616065278e-06,
-      "loss": 0.3457,
-      "step": 12328
-    },
-    {
-      "epoch": 0.8057643291288151,
-      "grad_norm": 0.43821439146995544,
-      "learning_rate": 8.494808898203448e-06,
-      "loss": 0.366,
-      "step": 12329
-    },
-    {
-      "epoch": 0.8058296843343572,
-      "grad_norm": 0.47947388887405396,
-      "learning_rate": 8.494559163296227e-06,
-      "loss": 0.3653,
-      "step": 12330
-    },
-    {
-      "epoch": 0.8058950395398994,
-      "grad_norm": 0.400860071182251,
-      "learning_rate": 8.494309411344828e-06,
-      "loss": 0.3411,
-      "step": 12331
-    },
-    {
-      "epoch": 0.8059603947454415,
-      "grad_norm": 0.4695069193840027,
-      "learning_rate": 8.494059642350471e-06,
-      "loss": 0.4497,
-      "step": 12332
-    },
-    {
-      "epoch": 0.8060257499509836,
-      "grad_norm": 0.4215098023414612,
-      "learning_rate": 8.493809856314376e-06,
-      "loss": 0.3389,
-      "step": 12333
-    },
-    {
-      "epoch": 0.8060911051565257,
-      "grad_norm": 0.4937513470649719,
-      "learning_rate": 8.493560053237762e-06,
-      "loss": 0.4805,
-      "step": 12334
-    },
-    {
-      "epoch": 0.8061564603620678,
-      "grad_norm": 0.43248605728149414,
-      "learning_rate": 8.493310233121842e-06,
-      "loss": 0.3614,
-      "step": 12335
-    },
-    {
-      "epoch": 0.80622181556761,
-      "grad_norm": 0.4014105796813965,
-      "learning_rate": 8.493060395967839e-06,
-      "loss": 0.3035,
-      "step": 12336
-    },
-    {
-      "epoch": 0.806287170773152,
-      "grad_norm": 0.43782347440719604,
-      "learning_rate": 8.492810541776971e-06,
-      "loss": 0.3547,
-      "step": 12337
-    },
-    {
-      "epoch": 0.8063525259786942,
-      "grad_norm": 0.4088995158672333,
-      "learning_rate": 8.492560670550454e-06,
-      "loss": 0.3,
-      "step": 12338
-    },
-    {
-      "epoch": 0.8064178811842363,
-      "grad_norm": 0.4330845773220062,
-      "learning_rate": 8.49231078228951e-06,
-      "loss": 0.3511,
-      "step": 12339
-    },
-    {
-      "epoch": 0.8064832363897785,
-      "grad_norm": 0.502913773059845,
-      "learning_rate": 8.492060876995356e-06,
-      "loss": 0.4662,
-      "step": 12340
-    },
-    {
-      "epoch": 0.8065485915953206,
-      "grad_norm": 0.4259401559829712,
-      "learning_rate": 8.491810954669213e-06,
-      "loss": 0.3502,
-      "step": 12341
-    },
-    {
-      "epoch": 0.8066139468008627,
-      "grad_norm": 0.4137631058692932,
-      "learning_rate": 8.491561015312296e-06,
-      "loss": 0.3505,
-      "step": 12342
-    },
-    {
-      "epoch": 0.8066793020064048,
-      "grad_norm": 0.4844277501106262,
-      "learning_rate": 8.491311058925827e-06,
-      "loss": 0.4683,
-      "step": 12343
-    },
-    {
-      "epoch": 0.8067446572119469,
-      "grad_norm": 0.40809574723243713,
-      "learning_rate": 8.491061085511026e-06,
-      "loss": 0.3349,
-      "step": 12344
-    },
-    {
-      "epoch": 0.8068100124174891,
-      "grad_norm": 0.4811766743659973,
-      "learning_rate": 8.49081109506911e-06,
-      "loss": 0.4193,
-      "step": 12345
-    },
-    {
-      "epoch": 0.8068753676230311,
-      "grad_norm": 0.5202414393424988,
-      "learning_rate": 8.490561087601297e-06,
-      "loss": 0.4854,
-      "step": 12346
-    },
-    {
-      "epoch": 0.8069407228285733,
-      "grad_norm": 0.4288029074668884,
-      "learning_rate": 8.49031106310881e-06,
-      "loss": 0.3517,
-      "step": 12347
-    },
-    {
-      "epoch": 0.8070060780341154,
-      "grad_norm": 0.4680384695529938,
-      "learning_rate": 8.490061021592867e-06,
-      "loss": 0.3673,
-      "step": 12348
-    },
-    {
-      "epoch": 0.8070714332396576,
-      "grad_norm": 0.5072221755981445,
-      "learning_rate": 8.489810963054687e-06,
-      "loss": 0.4978,
-      "step": 12349
-    },
-    {
-      "epoch": 0.8071367884451996,
-      "grad_norm": 0.4739428460597992,
-      "learning_rate": 8.489560887495489e-06,
-      "loss": 0.3841,
-      "step": 12350
-    },
-    {
-      "epoch": 0.8072021436507418,
-      "grad_norm": 0.4591532051563263,
-      "learning_rate": 8.489310794916495e-06,
-      "loss": 0.3838,
-      "step": 12351
-    },
-    {
-      "epoch": 0.8072674988562839,
-      "grad_norm": 0.49754616618156433,
-      "learning_rate": 8.489060685318923e-06,
-      "loss": 0.4498,
-      "step": 12352
-    },
-    {
-      "epoch": 0.807332854061826,
-      "grad_norm": 0.4266905188560486,
-      "learning_rate": 8.488810558703992e-06,
-      "loss": 0.354,
-      "step": 12353
-    },
-    {
-      "epoch": 0.8073982092673682,
-      "grad_norm": 0.43139177560806274,
-      "learning_rate": 8.488560415072925e-06,
-      "loss": 0.3461,
-      "step": 12354
-    },
-    {
-      "epoch": 0.8074635644729102,
-      "grad_norm": 0.5105013847351074,
-      "learning_rate": 8.48831025442694e-06,
-      "loss": 0.4323,
-      "step": 12355
-    },
-    {
-      "epoch": 0.8075289196784524,
-      "grad_norm": 0.43758299946784973,
-      "learning_rate": 8.488060076767257e-06,
-      "loss": 0.3361,
-      "step": 12356
-    },
-    {
-      "epoch": 0.8075942748839945,
-      "grad_norm": 0.4944319725036621,
-      "learning_rate": 8.487809882095097e-06,
-      "loss": 0.4287,
-      "step": 12357
-    },
-    {
-      "epoch": 0.8076596300895367,
-      "grad_norm": 0.4521756172180176,
-      "learning_rate": 8.48755967041168e-06,
-      "loss": 0.3709,
-      "step": 12358
-    },
-    {
-      "epoch": 0.8077249852950787,
-      "grad_norm": 0.45124170184135437,
-      "learning_rate": 8.487309441718226e-06,
-      "loss": 0.355,
-      "step": 12359
-    },
-    {
-      "epoch": 0.8077903405006209,
-      "grad_norm": 0.4594513177871704,
-      "learning_rate": 8.487059196015955e-06,
-      "loss": 0.4258,
-      "step": 12360
-    },
-    {
-      "epoch": 0.807855695706163,
-      "grad_norm": 0.412546843290329,
-      "learning_rate": 8.48680893330609e-06,
-      "loss": 0.3401,
-      "step": 12361
-    },
-    {
-      "epoch": 0.8079210509117051,
-      "grad_norm": 0.40831243991851807,
-      "learning_rate": 8.486558653589848e-06,
-      "loss": 0.3259,
-      "step": 12362
-    },
-    {
-      "epoch": 0.8079864061172473,
-      "grad_norm": 0.4548529088497162,
-      "learning_rate": 8.486308356868452e-06,
-      "loss": 0.3863,
-      "step": 12363
-    },
-    {
-      "epoch": 0.8080517613227893,
-      "grad_norm": 0.4557829201221466,
-      "learning_rate": 8.486058043143123e-06,
-      "loss": 0.4039,
-      "step": 12364
-    },
-    {
-      "epoch": 0.8081171165283315,
-      "grad_norm": 0.45605432987213135,
-      "learning_rate": 8.485807712415082e-06,
-      "loss": 0.4013,
-      "step": 12365
-    },
-    {
-      "epoch": 0.8081824717338736,
-      "grad_norm": 0.4350353181362152,
-      "learning_rate": 8.485557364685547e-06,
-      "loss": 0.4002,
-      "step": 12366
-    },
-    {
-      "epoch": 0.8082478269394158,
-      "grad_norm": 0.4219260811805725,
-      "learning_rate": 8.485306999955743e-06,
-      "loss": 0.3429,
-      "step": 12367
-    },
-    {
-      "epoch": 0.8083131821449578,
-      "grad_norm": 0.4718751907348633,
-      "learning_rate": 8.48505661822689e-06,
-      "loss": 0.3948,
-      "step": 12368
-    },
-    {
-      "epoch": 0.8083785373504999,
-      "grad_norm": 0.43288910388946533,
-      "learning_rate": 8.484806219500208e-06,
-      "loss": 0.3871,
-      "step": 12369
-    },
-    {
-      "epoch": 0.8084438925560421,
-      "grad_norm": 0.4838804006576538,
-      "learning_rate": 8.484555803776916e-06,
-      "loss": 0.4673,
-      "step": 12370
-    },
-    {
-      "epoch": 0.8085092477615842,
-      "grad_norm": 0.4964217245578766,
-      "learning_rate": 8.48430537105824e-06,
-      "loss": 0.4269,
-      "step": 12371
-    },
-    {
-      "epoch": 0.8085746029671264,
-      "grad_norm": 0.46102991700172424,
-      "learning_rate": 8.484054921345402e-06,
-      "loss": 0.4195,
-      "step": 12372
-    },
-    {
-      "epoch": 0.8086399581726684,
-      "grad_norm": 0.447543203830719,
-      "learning_rate": 8.48380445463962e-06,
-      "loss": 0.3753,
-      "step": 12373
-    },
-    {
-      "epoch": 0.8087053133782106,
-      "grad_norm": 0.41645434498786926,
-      "learning_rate": 8.483553970942115e-06,
-      "loss": 0.3441,
-      "step": 12374
-    },
-    {
-      "epoch": 0.8087706685837527,
-      "grad_norm": 0.47619786858558655,
-      "learning_rate": 8.48330347025411e-06,
-      "loss": 0.4637,
-      "step": 12375
-    },
-    {
-      "epoch": 0.8088360237892949,
-      "grad_norm": 0.45063620805740356,
-      "learning_rate": 8.48305295257683e-06,
-      "loss": 0.3661,
-      "step": 12376
-    },
-    {
-      "epoch": 0.8089013789948369,
-      "grad_norm": 0.42111918330192566,
-      "learning_rate": 8.482802417911492e-06,
-      "loss": 0.3705,
-      "step": 12377
-    },
-    {
-      "epoch": 0.808966734200379,
-      "grad_norm": 0.44826894998550415,
-      "learning_rate": 8.482551866259321e-06,
-      "loss": 0.3673,
-      "step": 12378
-    },
-    {
-      "epoch": 0.8090320894059212,
-      "grad_norm": 0.4359002411365509,
-      "learning_rate": 8.482301297621538e-06,
-      "loss": 0.3374,
-      "step": 12379
-    },
-    {
-      "epoch": 0.8090974446114633,
-      "grad_norm": 0.4520050585269928,
-      "learning_rate": 8.482050711999364e-06,
-      "loss": 0.3802,
-      "step": 12380
-    },
-    {
-      "epoch": 0.8091627998170055,
-      "grad_norm": 0.46814998984336853,
-      "learning_rate": 8.481800109394025e-06,
-      "loss": 0.3661,
-      "step": 12381
-    },
-    {
-      "epoch": 0.8092281550225475,
-      "grad_norm": 0.4591486155986786,
-      "learning_rate": 8.481549489806738e-06,
-      "loss": 0.4094,
-      "step": 12382
-    },
-    {
-      "epoch": 0.8092935102280897,
-      "grad_norm": 0.47207266092300415,
-      "learning_rate": 8.481298853238728e-06,
-      "loss": 0.3787,
-      "step": 12383
-    },
-    {
-      "epoch": 0.8093588654336318,
-      "grad_norm": 0.46365997195243835,
-      "learning_rate": 8.48104819969122e-06,
-      "loss": 0.409,
-      "step": 12384
-    },
-    {
-      "epoch": 0.809424220639174,
-      "grad_norm": 0.44643640518188477,
-      "learning_rate": 8.480797529165431e-06,
-      "loss": 0.3811,
-      "step": 12385
-    },
-    {
-      "epoch": 0.809489575844716,
-      "grad_norm": 0.46963217854499817,
-      "learning_rate": 8.48054684166259e-06,
-      "loss": 0.4256,
-      "step": 12386
-    },
-    {
-      "epoch": 0.8095549310502581,
-      "grad_norm": 0.45074307918548584,
-      "learning_rate": 8.480296137183914e-06,
-      "loss": 0.3761,
-      "step": 12387
-    },
-    {
-      "epoch": 0.8096202862558003,
-      "grad_norm": 0.4560701251029968,
-      "learning_rate": 8.48004541573063e-06,
-      "loss": 0.4014,
-      "step": 12388
-    },
-    {
-      "epoch": 0.8096856414613424,
-      "grad_norm": 0.42646080255508423,
-      "learning_rate": 8.479794677303957e-06,
-      "loss": 0.3746,
-      "step": 12389
-    },
-    {
-      "epoch": 0.8097509966668845,
-      "grad_norm": 0.45829257369041443,
-      "learning_rate": 8.479543921905119e-06,
-      "loss": 0.3924,
-      "step": 12390
-    },
-    {
-      "epoch": 0.8098163518724266,
-      "grad_norm": 0.46567419171333313,
-      "learning_rate": 8.479293149535342e-06,
-      "loss": 0.3624,
-      "step": 12391
-    },
-    {
-      "epoch": 0.8098817070779688,
-      "grad_norm": 0.4683837592601776,
-      "learning_rate": 8.479042360195844e-06,
-      "loss": 0.3922,
-      "step": 12392
-    },
-    {
-      "epoch": 0.8099470622835109,
-      "grad_norm": 0.4206649661064148,
-      "learning_rate": 8.478791553887854e-06,
-      "loss": 0.3616,
-      "step": 12393
-    },
-    {
-      "epoch": 0.810012417489053,
-      "grad_norm": 0.45548540353775024,
-      "learning_rate": 8.478540730612592e-06,
-      "loss": 0.395,
-      "step": 12394
-    },
-    {
-      "epoch": 0.8100777726945951,
-      "grad_norm": 0.4326566755771637,
-      "learning_rate": 8.478289890371281e-06,
-      "loss": 0.3296,
-      "step": 12395
-    },
-    {
-      "epoch": 0.8101431279001372,
-      "grad_norm": 0.4756647050380707,
-      "learning_rate": 8.478039033165146e-06,
-      "loss": 0.4634,
-      "step": 12396
-    },
-    {
-      "epoch": 0.8102084831056794,
-      "grad_norm": 0.41559094190597534,
-      "learning_rate": 8.477788158995409e-06,
-      "loss": 0.3115,
-      "step": 12397
-    },
-    {
-      "epoch": 0.8102738383112215,
-      "grad_norm": 0.45956042408943176,
-      "learning_rate": 8.477537267863295e-06,
-      "loss": 0.4063,
-      "step": 12398
-    },
-    {
-      "epoch": 0.8103391935167636,
-      "grad_norm": 0.41898688673973083,
-      "learning_rate": 8.477286359770025e-06,
-      "loss": 0.33,
-      "step": 12399
-    },
-    {
-      "epoch": 0.8104045487223057,
-      "grad_norm": 0.4533012807369232,
-      "learning_rate": 8.477035434716827e-06,
-      "loss": 0.4039,
-      "step": 12400
-    },
-    {
-      "epoch": 0.8104699039278479,
-      "grad_norm": 0.47679269313812256,
-      "learning_rate": 8.476784492704921e-06,
-      "loss": 0.4135,
-      "step": 12401
-    },
-    {
-      "epoch": 0.81053525913339,
-      "grad_norm": 0.41880643367767334,
-      "learning_rate": 8.476533533735534e-06,
-      "loss": 0.357,
-      "step": 12402
-    },
-    {
-      "epoch": 0.810600614338932,
-      "grad_norm": 0.49020689725875854,
-      "learning_rate": 8.47628255780989e-06,
-      "loss": 0.4069,
-      "step": 12403
-    },
-    {
-      "epoch": 0.8106659695444742,
-      "grad_norm": 0.4441443979740143,
-      "learning_rate": 8.476031564929208e-06,
-      "loss": 0.3994,
-      "step": 12404
-    },
-    {
-      "epoch": 0.8107313247500163,
-      "grad_norm": 0.45281073451042175,
-      "learning_rate": 8.475780555094719e-06,
-      "loss": 0.4222,
-      "step": 12405
-    },
-    {
-      "epoch": 0.8107966799555585,
-      "grad_norm": 0.4089464843273163,
-      "learning_rate": 8.475529528307642e-06,
-      "loss": 0.3225,
-      "step": 12406
-    },
-    {
-      "epoch": 0.8108620351611006,
-      "grad_norm": 0.4631759822368622,
-      "learning_rate": 8.475278484569206e-06,
-      "loss": 0.3728,
-      "step": 12407
-    },
-    {
-      "epoch": 0.8109273903666427,
-      "grad_norm": 0.4352424740791321,
-      "learning_rate": 8.47502742388063e-06,
-      "loss": 0.3604,
-      "step": 12408
-    },
-    {
-      "epoch": 0.8109927455721848,
-      "grad_norm": 0.4908069372177124,
-      "learning_rate": 8.474776346243143e-06,
-      "loss": 0.403,
-      "step": 12409
-    },
-    {
-      "epoch": 0.811058100777727,
-      "grad_norm": 0.4476626217365265,
-      "learning_rate": 8.474525251657966e-06,
-      "loss": 0.3652,
-      "step": 12410
-    },
-    {
-      "epoch": 0.8111234559832691,
-      "grad_norm": 0.4301026165485382,
-      "learning_rate": 8.47427414012633e-06,
-      "loss": 0.3927,
-      "step": 12411
-    },
-    {
-      "epoch": 0.8111888111888111,
-      "grad_norm": 0.4547356367111206,
-      "learning_rate": 8.474023011649451e-06,
-      "loss": 0.4077,
-      "step": 12412
-    },
-    {
-      "epoch": 0.8112541663943533,
-      "grad_norm": 0.4329404830932617,
-      "learning_rate": 8.47377186622856e-06,
-      "loss": 0.3508,
-      "step": 12413
-    },
-    {
-      "epoch": 0.8113195215998954,
-      "grad_norm": 0.44247904419898987,
-      "learning_rate": 8.47352070386488e-06,
-      "loss": 0.3645,
-      "step": 12414
-    },
-    {
-      "epoch": 0.8113848768054376,
-      "grad_norm": 0.4779067635536194,
-      "learning_rate": 8.473269524559634e-06,
-      "loss": 0.4328,
-      "step": 12415
-    },
-    {
-      "epoch": 0.8114502320109797,
-      "grad_norm": 0.4541876018047333,
-      "learning_rate": 8.473018328314054e-06,
-      "loss": 0.3194,
-      "step": 12416
-    },
-    {
-      "epoch": 0.8115155872165218,
-      "grad_norm": 0.41460004448890686,
-      "learning_rate": 8.472767115129356e-06,
-      "loss": 0.3238,
-      "step": 12417
-    },
-    {
-      "epoch": 0.8115809424220639,
-      "grad_norm": 0.4341851770877838,
-      "learning_rate": 8.472515885006771e-06,
-      "loss": 0.3722,
-      "step": 12418
-    },
-    {
-      "epoch": 0.811646297627606,
-      "grad_norm": 0.4967648684978485,
-      "learning_rate": 8.472264637947525e-06,
-      "loss": 0.449,
-      "step": 12419
-    },
-    {
-      "epoch": 0.8117116528331482,
-      "grad_norm": 0.46293848752975464,
-      "learning_rate": 8.472013373952839e-06,
-      "loss": 0.4007,
-      "step": 12420
-    },
-    {
-      "epoch": 0.8117770080386902,
-      "grad_norm": 0.4613957405090332,
-      "learning_rate": 8.471762093023943e-06,
-      "loss": 0.3926,
-      "step": 12421
-    },
-    {
-      "epoch": 0.8118423632442324,
-      "grad_norm": 0.469547301530838,
-      "learning_rate": 8.471510795162058e-06,
-      "loss": 0.3919,
-      "step": 12422
-    },
-    {
-      "epoch": 0.8119077184497745,
-      "grad_norm": 0.4456419050693512,
-      "learning_rate": 8.471259480368415e-06,
-      "loss": 0.379,
-      "step": 12423
-    },
-    {
-      "epoch": 0.8119730736553167,
-      "grad_norm": 0.42810481786727905,
-      "learning_rate": 8.471008148644236e-06,
-      "loss": 0.3436,
-      "step": 12424
-    },
-    {
-      "epoch": 0.8120384288608588,
-      "grad_norm": 0.4358358681201935,
-      "learning_rate": 8.470756799990746e-06,
-      "loss": 0.3955,
-      "step": 12425
-    },
-    {
-      "epoch": 0.8121037840664009,
-      "grad_norm": 0.4390687346458435,
-      "learning_rate": 8.470505434409175e-06,
-      "loss": 0.3886,
-      "step": 12426
-    },
-    {
-      "epoch": 0.812169139271943,
-      "grad_norm": 0.4392316937446594,
-      "learning_rate": 8.470254051900746e-06,
-      "loss": 0.3729,
-      "step": 12427
-    },
-    {
-      "epoch": 0.8122344944774851,
-      "grad_norm": 0.43075257539749146,
-      "learning_rate": 8.470002652466686e-06,
-      "loss": 0.3687,
-      "step": 12428
-    },
-    {
-      "epoch": 0.8122998496830273,
-      "grad_norm": 0.43682315945625305,
-      "learning_rate": 8.46975123610822e-06,
-      "loss": 0.347,
-      "step": 12429
-    },
-    {
-      "epoch": 0.8123652048885693,
-      "grad_norm": 0.4574038088321686,
-      "learning_rate": 8.469499802826577e-06,
-      "loss": 0.3743,
-      "step": 12430
-    },
-    {
-      "epoch": 0.8124305600941115,
-      "grad_norm": 0.4498804211616516,
-      "learning_rate": 8.46924835262298e-06,
-      "loss": 0.3911,
-      "step": 12431
-    },
-    {
-      "epoch": 0.8124959152996536,
-      "grad_norm": 0.44626232981681824,
-      "learning_rate": 8.468996885498657e-06,
-      "loss": 0.3355,
-      "step": 12432
-    },
-    {
-      "epoch": 0.8125612705051958,
-      "grad_norm": 0.4404367208480835,
-      "learning_rate": 8.468745401454834e-06,
-      "loss": 0.3445,
-      "step": 12433
-    },
-    {
-      "epoch": 0.8126266257107378,
-      "grad_norm": 0.5081859827041626,
-      "learning_rate": 8.468493900492738e-06,
-      "loss": 0.481,
-      "step": 12434
-    },
-    {
-      "epoch": 0.81269198091628,
-      "grad_norm": 0.43948298692703247,
-      "learning_rate": 8.468242382613598e-06,
-      "loss": 0.3587,
-      "step": 12435
-    },
-    {
-      "epoch": 0.8127573361218221,
-      "grad_norm": 0.46249765157699585,
-      "learning_rate": 8.467990847818637e-06,
-      "loss": 0.3974,
-      "step": 12436
-    },
-    {
-      "epoch": 0.8128226913273642,
-      "grad_norm": 0.4180997908115387,
-      "learning_rate": 8.467739296109081e-06,
-      "loss": 0.3066,
-      "step": 12437
-    },
-    {
-      "epoch": 0.8128880465329064,
-      "grad_norm": 0.48867717385292053,
-      "learning_rate": 8.46748772748616e-06,
-      "loss": 0.3398,
-      "step": 12438
-    },
-    {
-      "epoch": 0.8129534017384484,
-      "grad_norm": 0.46018847823143005,
-      "learning_rate": 8.467236141951103e-06,
-      "loss": 0.3739,
-      "step": 12439
-    },
-    {
-      "epoch": 0.8130187569439906,
-      "grad_norm": 0.4579855799674988,
-      "learning_rate": 8.466984539505132e-06,
-      "loss": 0.4262,
-      "step": 12440
-    },
-    {
-      "epoch": 0.8130841121495327,
-      "grad_norm": 0.4060579240322113,
-      "learning_rate": 8.466732920149476e-06,
-      "loss": 0.329,
-      "step": 12441
-    },
-    {
-      "epoch": 0.8131494673550749,
-      "grad_norm": 0.44370123744010925,
-      "learning_rate": 8.466481283885363e-06,
-      "loss": 0.3779,
-      "step": 12442
-    },
-    {
-      "epoch": 0.813214822560617,
-      "grad_norm": 0.435161828994751,
-      "learning_rate": 8.46622963071402e-06,
-      "loss": 0.3696,
-      "step": 12443
-    },
-    {
-      "epoch": 0.8132801777661591,
-      "grad_norm": 0.46991202235221863,
-      "learning_rate": 8.465977960636676e-06,
-      "loss": 0.4029,
-      "step": 12444
-    },
-    {
-      "epoch": 0.8133455329717012,
-      "grad_norm": 0.4920562207698822,
-      "learning_rate": 8.465726273654555e-06,
-      "loss": 0.4524,
-      "step": 12445
-    },
-    {
-      "epoch": 0.8134108881772433,
-      "grad_norm": 0.4491201639175415,
-      "learning_rate": 8.465474569768885e-06,
-      "loss": 0.36,
-      "step": 12446
-    },
-    {
-      "epoch": 0.8134762433827855,
-      "grad_norm": 0.5497409105300903,
-      "learning_rate": 8.465222848980896e-06,
-      "loss": 0.4177,
-      "step": 12447
-    },
-    {
-      "epoch": 0.8135415985883275,
-      "grad_norm": 0.48143699765205383,
-      "learning_rate": 8.464971111291815e-06,
-      "loss": 0.415,
-      "step": 12448
-    },
-    {
-      "epoch": 0.8136069537938697,
-      "grad_norm": 0.42294394969940186,
-      "learning_rate": 8.464719356702871e-06,
-      "loss": 0.3423,
-      "step": 12449
-    },
-    {
-      "epoch": 0.8136723089994118,
-      "grad_norm": 0.45985540747642517,
-      "learning_rate": 8.464467585215288e-06,
-      "loss": 0.3896,
-      "step": 12450
-    },
-    {
-      "epoch": 0.813737664204954,
-      "grad_norm": 0.4446311295032501,
-      "learning_rate": 8.464215796830298e-06,
-      "loss": 0.3767,
-      "step": 12451
-    },
-    {
-      "epoch": 0.813803019410496,
-      "grad_norm": 0.4525317847728729,
-      "learning_rate": 8.463963991549127e-06,
-      "loss": 0.3662,
-      "step": 12452
-    },
-    {
-      "epoch": 0.8138683746160381,
-      "grad_norm": 0.44176629185676575,
-      "learning_rate": 8.463712169373002e-06,
-      "loss": 0.3943,
-      "step": 12453
-    },
-    {
-      "epoch": 0.8139337298215803,
-      "grad_norm": 0.45090174674987793,
-      "learning_rate": 8.463460330303154e-06,
-      "loss": 0.4166,
-      "step": 12454
-    },
-    {
-      "epoch": 0.8139990850271224,
-      "grad_norm": 0.4586697816848755,
-      "learning_rate": 8.463208474340811e-06,
-      "loss": 0.3981,
-      "step": 12455
-    },
-    {
-      "epoch": 0.8140644402326646,
-      "grad_norm": 0.44277840852737427,
-      "learning_rate": 8.4629566014872e-06,
-      "loss": 0.3811,
-      "step": 12456
-    },
-    {
-      "epoch": 0.8141297954382066,
-      "grad_norm": 0.46791690587997437,
-      "learning_rate": 8.46270471174355e-06,
-      "loss": 0.3989,
-      "step": 12457
-    },
-    {
-      "epoch": 0.8141951506437488,
-      "grad_norm": 0.41792604327201843,
-      "learning_rate": 8.462452805111089e-06,
-      "loss": 0.3242,
-      "step": 12458
-    },
-    {
-      "epoch": 0.8142605058492909,
-      "grad_norm": 0.45778384804725647,
-      "learning_rate": 8.462200881591046e-06,
-      "loss": 0.4405,
-      "step": 12459
-    },
-    {
-      "epoch": 0.8143258610548331,
-      "grad_norm": 0.42692655324935913,
-      "learning_rate": 8.46194894118465e-06,
-      "loss": 0.3709,
-      "step": 12460
-    },
-    {
-      "epoch": 0.8143912162603751,
-      "grad_norm": 0.4614044427871704,
-      "learning_rate": 8.46169698389313e-06,
-      "loss": 0.4372,
-      "step": 12461
-    },
-    {
-      "epoch": 0.8144565714659172,
-      "grad_norm": 0.47064560651779175,
-      "learning_rate": 8.461445009717714e-06,
-      "loss": 0.4273,
-      "step": 12462
-    },
-    {
-      "epoch": 0.8145219266714594,
-      "grad_norm": 0.46201393008232117,
-      "learning_rate": 8.461193018659633e-06,
-      "loss": 0.3968,
-      "step": 12463
-    },
-    {
-      "epoch": 0.8145872818770015,
-      "grad_norm": 0.46701836585998535,
-      "learning_rate": 8.460941010720114e-06,
-      "loss": 0.3761,
-      "step": 12464
-    },
-    {
-      "epoch": 0.8146526370825437,
-      "grad_norm": 0.45106998085975647,
-      "learning_rate": 8.460688985900386e-06,
-      "loss": 0.4021,
-      "step": 12465
-    },
-    {
-      "epoch": 0.8147179922880857,
-      "grad_norm": 0.4375004470348358,
-      "learning_rate": 8.460436944201678e-06,
-      "loss": 0.394,
-      "step": 12466
-    },
-    {
-      "epoch": 0.8147833474936279,
-      "grad_norm": 0.4488752484321594,
-      "learning_rate": 8.460184885625222e-06,
-      "loss": 0.3968,
-      "step": 12467
-    },
-    {
-      "epoch": 0.81484870269917,
-      "grad_norm": 0.44091296195983887,
-      "learning_rate": 8.459932810172246e-06,
-      "loss": 0.3347,
-      "step": 12468
-    },
-    {
-      "epoch": 0.8149140579047122,
-      "grad_norm": 0.42159703373908997,
-      "learning_rate": 8.459680717843978e-06,
-      "loss": 0.3616,
-      "step": 12469
-    },
-    {
-      "epoch": 0.8149794131102542,
-      "grad_norm": 0.47401347756385803,
-      "learning_rate": 8.459428608641649e-06,
-      "loss": 0.3785,
-      "step": 12470
-    },
-    {
-      "epoch": 0.8150447683157963,
-      "grad_norm": 0.4911261200904846,
-      "learning_rate": 8.459176482566487e-06,
-      "loss": 0.334,
-      "step": 12471
-    },
-    {
-      "epoch": 0.8151101235213385,
-      "grad_norm": 0.4237688183784485,
-      "learning_rate": 8.458924339619726e-06,
-      "loss": 0.3871,
-      "step": 12472
-    },
-    {
-      "epoch": 0.8151754787268806,
-      "grad_norm": 0.468522310256958,
-      "learning_rate": 8.45867217980259e-06,
-      "loss": 0.4178,
-      "step": 12473
-    },
-    {
-      "epoch": 0.8152408339324227,
-      "grad_norm": 0.4498448669910431,
-      "learning_rate": 8.45842000311631e-06,
-      "loss": 0.3646,
-      "step": 12474
-    },
-    {
-      "epoch": 0.8153061891379648,
-      "grad_norm": 0.4376983642578125,
-      "learning_rate": 8.458167809562122e-06,
-      "loss": 0.3605,
-      "step": 12475
-    },
-    {
-      "epoch": 0.815371544343507,
-      "grad_norm": 0.4309103786945343,
-      "learning_rate": 8.457915599141249e-06,
-      "loss": 0.3235,
-      "step": 12476
-    },
-    {
-      "epoch": 0.8154368995490491,
-      "grad_norm": 0.43921563029289246,
-      "learning_rate": 8.457663371854924e-06,
-      "loss": 0.3385,
-      "step": 12477
-    },
-    {
-      "epoch": 0.8155022547545911,
-      "grad_norm": 0.44701018929481506,
-      "learning_rate": 8.457411127704376e-06,
-      "loss": 0.3906,
-      "step": 12478
-    },
-    {
-      "epoch": 0.8155676099601333,
-      "grad_norm": 0.46366724371910095,
-      "learning_rate": 8.457158866690836e-06,
-      "loss": 0.4068,
-      "step": 12479
-    },
-    {
-      "epoch": 0.8156329651656754,
-      "grad_norm": 0.43142345547676086,
-      "learning_rate": 8.456906588815536e-06,
-      "loss": 0.3306,
-      "step": 12480
-    },
-    {
-      "epoch": 0.8156983203712176,
-      "grad_norm": 0.4284942150115967,
-      "learning_rate": 8.456654294079704e-06,
-      "loss": 0.334,
-      "step": 12481
-    },
-    {
-      "epoch": 0.8157636755767597,
-      "grad_norm": 0.45439621806144714,
-      "learning_rate": 8.456401982484573e-06,
-      "loss": 0.4044,
-      "step": 12482
-    },
-    {
-      "epoch": 0.8158290307823018,
-      "grad_norm": 0.42266830801963806,
-      "learning_rate": 8.45614965403137e-06,
-      "loss": 0.3675,
-      "step": 12483
-    },
-    {
-      "epoch": 0.8158943859878439,
-      "grad_norm": 0.4663552939891815,
-      "learning_rate": 8.455897308721329e-06,
-      "loss": 0.3888,
-      "step": 12484
-    },
-    {
-      "epoch": 0.8159597411933861,
-      "grad_norm": 0.453592449426651,
-      "learning_rate": 8.45564494655568e-06,
-      "loss": 0.3982,
-      "step": 12485
-    },
-    {
-      "epoch": 0.8160250963989282,
-      "grad_norm": 0.4378078877925873,
-      "learning_rate": 8.455392567535651e-06,
-      "loss": 0.3343,
-      "step": 12486
-    },
-    {
-      "epoch": 0.8160904516044702,
-      "grad_norm": 0.47541993856430054,
-      "learning_rate": 8.45514017166248e-06,
-      "loss": 0.4538,
-      "step": 12487
-    },
-    {
-      "epoch": 0.8161558068100124,
-      "grad_norm": 0.4201316833496094,
-      "learning_rate": 8.45488775893739e-06,
-      "loss": 0.3422,
-      "step": 12488
-    },
-    {
-      "epoch": 0.8162211620155545,
-      "grad_norm": 0.44259822368621826,
-      "learning_rate": 8.454635329361615e-06,
-      "loss": 0.3678,
-      "step": 12489
-    },
-    {
-      "epoch": 0.8162865172210967,
-      "grad_norm": 0.42908021807670593,
-      "learning_rate": 8.45438288293639e-06,
-      "loss": 0.3764,
-      "step": 12490
-    },
-    {
-      "epoch": 0.8163518724266388,
-      "grad_norm": 0.5003162622451782,
-      "learning_rate": 8.454130419662941e-06,
-      "loss": 0.4751,
-      "step": 12491
-    },
-    {
-      "epoch": 0.8164172276321809,
-      "grad_norm": 0.48089399933815,
-      "learning_rate": 8.4538779395425e-06,
-      "loss": 0.4612,
-      "step": 12492
-    },
-    {
-      "epoch": 0.816482582837723,
-      "grad_norm": 0.4475057125091553,
-      "learning_rate": 8.4536254425763e-06,
-      "loss": 0.3528,
-      "step": 12493
-    },
-    {
-      "epoch": 0.8165479380432652,
-      "grad_norm": 0.4424389600753784,
-      "learning_rate": 8.453372928765575e-06,
-      "loss": 0.3877,
-      "step": 12494
-    },
-    {
-      "epoch": 0.8166132932488073,
-      "grad_norm": 0.47818002104759216,
-      "learning_rate": 8.453120398111552e-06,
-      "loss": 0.4147,
-      "step": 12495
-    },
-    {
-      "epoch": 0.8166786484543493,
-      "grad_norm": 0.44536691904067993,
-      "learning_rate": 8.452867850615464e-06,
-      "loss": 0.3688,
-      "step": 12496
-    },
-    {
-      "epoch": 0.8167440036598915,
-      "grad_norm": 0.47052431106567383,
-      "learning_rate": 8.452615286278544e-06,
-      "loss": 0.3977,
-      "step": 12497
-    },
-    {
-      "epoch": 0.8168093588654336,
-      "grad_norm": 0.4413776099681854,
-      "learning_rate": 8.452362705102025e-06,
-      "loss": 0.3607,
-      "step": 12498
-    },
-    {
-      "epoch": 0.8168747140709758,
-      "grad_norm": 0.46936506032943726,
-      "learning_rate": 8.452110107087134e-06,
-      "loss": 0.4001,
-      "step": 12499
-    },
-    {
-      "epoch": 0.8169400692765179,
-      "grad_norm": 0.42547914385795593,
-      "learning_rate": 8.451857492235107e-06,
-      "loss": 0.3593,
-      "step": 12500
-    },
-    {
-      "epoch": 0.81700542448206,
-      "grad_norm": 0.44651374220848083,
-      "learning_rate": 8.451604860547176e-06,
-      "loss": 0.3998,
-      "step": 12501
-    },
-    {
-      "epoch": 0.8170707796876021,
-      "grad_norm": 0.472301721572876,
-      "learning_rate": 8.45135221202457e-06,
-      "loss": 0.3869,
-      "step": 12502
-    },
-    {
-      "epoch": 0.8171361348931442,
-      "grad_norm": 0.4374332129955292,
-      "learning_rate": 8.451099546668527e-06,
-      "loss": 0.3615,
-      "step": 12503
-    },
-    {
-      "epoch": 0.8172014900986864,
-      "grad_norm": 0.4162733852863312,
-      "learning_rate": 8.450846864480274e-06,
-      "loss": 0.34,
-      "step": 12504
-    },
-    {
-      "epoch": 0.8172668453042284,
-      "grad_norm": 0.4110073745250702,
-      "learning_rate": 8.450594165461046e-06,
-      "loss": 0.3359,
-      "step": 12505
-    },
-    {
-      "epoch": 0.8173322005097706,
-      "grad_norm": 0.41492143273353577,
-      "learning_rate": 8.450341449612075e-06,
-      "loss": 0.3471,
-      "step": 12506
-    },
-    {
-      "epoch": 0.8173975557153127,
-      "grad_norm": 0.42728742957115173,
-      "learning_rate": 8.450088716934593e-06,
-      "loss": 0.3163,
-      "step": 12507
-    },
-    {
-      "epoch": 0.8174629109208549,
-      "grad_norm": 0.4918995499610901,
-      "learning_rate": 8.449835967429832e-06,
-      "loss": 0.371,
-      "step": 12508
-    },
-    {
-      "epoch": 0.817528266126397,
-      "grad_norm": 0.4346361458301544,
-      "learning_rate": 8.449583201099027e-06,
-      "loss": 0.3674,
-      "step": 12509
-    },
-    {
-      "epoch": 0.8175936213319391,
-      "grad_norm": 0.42874497175216675,
-      "learning_rate": 8.44933041794341e-06,
-      "loss": 0.3638,
-      "step": 12510
-    },
-    {
-      "epoch": 0.8176589765374812,
-      "grad_norm": 0.430454283952713,
-      "learning_rate": 8.449077617964212e-06,
-      "loss": 0.3258,
-      "step": 12511
-    },
-    {
-      "epoch": 0.8177243317430233,
-      "grad_norm": 0.46780121326446533,
-      "learning_rate": 8.44882480116267e-06,
-      "loss": 0.4165,
-      "step": 12512
-    },
-    {
-      "epoch": 0.8177896869485655,
-      "grad_norm": 0.47941410541534424,
-      "learning_rate": 8.448571967540014e-06,
-      "loss": 0.3929,
-      "step": 12513
-    },
-    {
-      "epoch": 0.8178550421541075,
-      "grad_norm": 0.4879732131958008,
-      "learning_rate": 8.448319117097477e-06,
-      "loss": 0.4137,
-      "step": 12514
-    },
-    {
-      "epoch": 0.8179203973596497,
-      "grad_norm": 0.42019909620285034,
-      "learning_rate": 8.448066249836293e-06,
-      "loss": 0.3325,
-      "step": 12515
-    },
-    {
-      "epoch": 0.8179857525651918,
-      "grad_norm": 0.42211028933525085,
-      "learning_rate": 8.447813365757696e-06,
-      "loss": 0.3542,
-      "step": 12516
-    },
-    {
-      "epoch": 0.818051107770734,
-      "grad_norm": 0.43780139088630676,
-      "learning_rate": 8.447560464862917e-06,
-      "loss": 0.3282,
-      "step": 12517
-    },
-    {
-      "epoch": 0.818116462976276,
-      "grad_norm": 0.4416830539703369,
-      "learning_rate": 8.447307547153194e-06,
-      "loss": 0.3729,
-      "step": 12518
-    },
-    {
-      "epoch": 0.8181818181818182,
-      "grad_norm": 0.42733025550842285,
-      "learning_rate": 8.447054612629756e-06,
-      "loss": 0.3661,
-      "step": 12519
-    },
-    {
-      "epoch": 0.8182471733873603,
-      "grad_norm": 0.4293109178543091,
-      "learning_rate": 8.44680166129384e-06,
-      "loss": 0.3666,
-      "step": 12520
-    },
-    {
-      "epoch": 0.8183125285929024,
-      "grad_norm": 0.4422782063484192,
-      "learning_rate": 8.446548693146675e-06,
-      "loss": 0.3762,
-      "step": 12521
-    },
-    {
-      "epoch": 0.8183778837984446,
-      "grad_norm": 0.4182201027870178,
-      "learning_rate": 8.4462957081895e-06,
-      "loss": 0.3298,
-      "step": 12522
-    },
-    {
-      "epoch": 0.8184432390039866,
-      "grad_norm": 0.4257052540779114,
-      "learning_rate": 8.446042706423547e-06,
-      "loss": 0.3414,
-      "step": 12523
-    },
-    {
-      "epoch": 0.8185085942095288,
-      "grad_norm": 0.4190555810928345,
-      "learning_rate": 8.445789687850052e-06,
-      "loss": 0.3553,
-      "step": 12524
-    },
-    {
-      "epoch": 0.8185739494150709,
-      "grad_norm": 0.45774173736572266,
-      "learning_rate": 8.445536652470244e-06,
-      "loss": 0.3719,
-      "step": 12525
-    },
-    {
-      "epoch": 0.8186393046206131,
-      "grad_norm": 0.43524685502052307,
-      "learning_rate": 8.44528360028536e-06,
-      "loss": 0.3561,
-      "step": 12526
-    },
-    {
-      "epoch": 0.8187046598261551,
-      "grad_norm": 0.440887987613678,
-      "learning_rate": 8.445030531296636e-06,
-      "loss": 0.395,
-      "step": 12527
-    },
-    {
-      "epoch": 0.8187700150316973,
-      "grad_norm": 0.423302561044693,
-      "learning_rate": 8.444777445505306e-06,
-      "loss": 0.3353,
-      "step": 12528
-    },
-    {
-      "epoch": 0.8188353702372394,
-      "grad_norm": 0.5028983354568481,
-      "learning_rate": 8.4445243429126e-06,
-      "loss": 0.4601,
-      "step": 12529
-    },
-    {
-      "epoch": 0.8189007254427815,
-      "grad_norm": 0.44041121006011963,
-      "learning_rate": 8.444271223519756e-06,
-      "loss": 0.355,
-      "step": 12530
-    },
-    {
-      "epoch": 0.8189660806483237,
-      "grad_norm": 0.4212939441204071,
-      "learning_rate": 8.44401808732801e-06,
-      "loss": 0.3588,
-      "step": 12531
-    },
-    {
-      "epoch": 0.8190314358538657,
-      "grad_norm": 0.42517799139022827,
-      "learning_rate": 8.443764934338592e-06,
-      "loss": 0.3378,
-      "step": 12532
-    },
-    {
-      "epoch": 0.8190967910594079,
-      "grad_norm": 0.43304795026779175,
-      "learning_rate": 8.443511764552741e-06,
-      "loss": 0.3646,
-      "step": 12533
-    },
-    {
-      "epoch": 0.81916214626495,
-      "grad_norm": 0.44127559661865234,
-      "learning_rate": 8.443258577971691e-06,
-      "loss": 0.373,
-      "step": 12534
-    },
-    {
-      "epoch": 0.8192275014704922,
-      "grad_norm": 0.4303012192249298,
-      "learning_rate": 8.443005374596673e-06,
-      "loss": 0.3365,
-      "step": 12535
-    },
-    {
-      "epoch": 0.8192928566760342,
-      "grad_norm": 0.44123363494873047,
-      "learning_rate": 8.442752154428928e-06,
-      "loss": 0.3838,
-      "step": 12536
-    },
-    {
-      "epoch": 0.8193582118815763,
-      "grad_norm": 0.4430181086063385,
-      "learning_rate": 8.442498917469687e-06,
-      "loss": 0.3588,
-      "step": 12537
-    },
-    {
-      "epoch": 0.8194235670871185,
-      "grad_norm": 0.443930447101593,
-      "learning_rate": 8.442245663720186e-06,
-      "loss": 0.3757,
-      "step": 12538
-    },
-    {
-      "epoch": 0.8194889222926606,
-      "grad_norm": 0.4582585096359253,
-      "learning_rate": 8.44199239318166e-06,
-      "loss": 0.4024,
-      "step": 12539
-    },
-    {
-      "epoch": 0.8195542774982028,
-      "grad_norm": 0.42140451073646545,
-      "learning_rate": 8.441739105855345e-06,
-      "loss": 0.3613,
-      "step": 12540
-    },
-    {
-      "epoch": 0.8196196327037448,
-      "grad_norm": 0.43065759539604187,
-      "learning_rate": 8.441485801742478e-06,
-      "loss": 0.3625,
-      "step": 12541
-    },
-    {
-      "epoch": 0.819684987909287,
-      "grad_norm": 0.4550173878669739,
-      "learning_rate": 8.44123248084429e-06,
-      "loss": 0.3965,
-      "step": 12542
-    },
-    {
-      "epoch": 0.8197503431148291,
-      "grad_norm": 0.4012031555175781,
-      "learning_rate": 8.44097914316202e-06,
-      "loss": 0.3358,
-      "step": 12543
-    },
-    {
-      "epoch": 0.8198156983203713,
-      "grad_norm": 0.439313679933548,
-      "learning_rate": 8.440725788696903e-06,
-      "loss": 0.3399,
-      "step": 12544
-    },
-    {
-      "epoch": 0.8198810535259133,
-      "grad_norm": 0.45871588587760925,
-      "learning_rate": 8.440472417450171e-06,
-      "loss": 0.3961,
-      "step": 12545
-    },
-    {
-      "epoch": 0.8199464087314554,
-      "grad_norm": 0.4367266595363617,
-      "learning_rate": 8.440219029423066e-06,
-      "loss": 0.4182,
-      "step": 12546
-    },
-    {
-      "epoch": 0.8200117639369976,
-      "grad_norm": 0.4367128610610962,
-      "learning_rate": 8.439965624616822e-06,
-      "loss": 0.3719,
-      "step": 12547
-    },
-    {
-      "epoch": 0.8200771191425397,
-      "grad_norm": 0.4566749930381775,
-      "learning_rate": 8.439712203032674e-06,
-      "loss": 0.3619,
-      "step": 12548
-    },
-    {
-      "epoch": 0.8201424743480819,
-      "grad_norm": 0.4530434310436249,
-      "learning_rate": 8.439458764671856e-06,
-      "loss": 0.383,
-      "step": 12549
-    },
-    {
-      "epoch": 0.8202078295536239,
-      "grad_norm": 0.40557971596717834,
-      "learning_rate": 8.439205309535606e-06,
-      "loss": 0.3006,
-      "step": 12550
-    },
-    {
-      "epoch": 0.8202731847591661,
-      "grad_norm": 0.42845040559768677,
-      "learning_rate": 8.438951837625162e-06,
-      "loss": 0.3518,
-      "step": 12551
-    },
-    {
-      "epoch": 0.8203385399647082,
-      "grad_norm": 0.4427779018878937,
-      "learning_rate": 8.438698348941756e-06,
-      "loss": 0.3612,
-      "step": 12552
-    },
-    {
-      "epoch": 0.8204038951702504,
-      "grad_norm": 0.4268951117992401,
-      "learning_rate": 8.43844484348663e-06,
-      "loss": 0.3638,
-      "step": 12553
-    },
-    {
-      "epoch": 0.8204692503757924,
-      "grad_norm": 0.4337146580219269,
-      "learning_rate": 8.438191321261015e-06,
-      "loss": 0.3393,
-      "step": 12554
-    },
-    {
-      "epoch": 0.8205346055813345,
-      "grad_norm": 0.44662535190582275,
-      "learning_rate": 8.43793778226615e-06,
-      "loss": 0.3739,
-      "step": 12555
-    },
-    {
-      "epoch": 0.8205999607868767,
-      "grad_norm": 0.4049738645553589,
-      "learning_rate": 8.437684226503273e-06,
-      "loss": 0.2878,
-      "step": 12556
-    },
-    {
-      "epoch": 0.8206653159924188,
-      "grad_norm": 0.40649691224098206,
-      "learning_rate": 8.437430653973619e-06,
-      "loss": 0.3269,
-      "step": 12557
-    },
-    {
-      "epoch": 0.820730671197961,
-      "grad_norm": 0.4747774302959442,
-      "learning_rate": 8.437177064678423e-06,
-      "loss": 0.4154,
-      "step": 12558
-    },
-    {
-      "epoch": 0.820796026403503,
-      "grad_norm": 0.45044398307800293,
-      "learning_rate": 8.436923458618925e-06,
-      "loss": 0.418,
-      "step": 12559
-    },
-    {
-      "epoch": 0.8208613816090452,
-      "grad_norm": 0.4268885552883148,
-      "learning_rate": 8.436669835796361e-06,
-      "loss": 0.3599,
-      "step": 12560
-    },
-    {
-      "epoch": 0.8209267368145873,
-      "grad_norm": 0.4495190680027008,
-      "learning_rate": 8.436416196211967e-06,
-      "loss": 0.4057,
-      "step": 12561
-    },
-    {
-      "epoch": 0.8209920920201293,
-      "grad_norm": 0.5106366872787476,
-      "learning_rate": 8.43616253986698e-06,
-      "loss": 0.39,
-      "step": 12562
-    },
-    {
-      "epoch": 0.8210574472256715,
-      "grad_norm": 0.444840669631958,
-      "learning_rate": 8.435908866762639e-06,
-      "loss": 0.3786,
-      "step": 12563
-    },
-    {
-      "epoch": 0.8211228024312136,
-      "grad_norm": 0.45178380608558655,
-      "learning_rate": 8.43565517690018e-06,
-      "loss": 0.3901,
-      "step": 12564
-    },
-    {
-      "epoch": 0.8211881576367558,
-      "grad_norm": 0.4797780215740204,
-      "learning_rate": 8.435401470280839e-06,
-      "loss": 0.4795,
-      "step": 12565
-    },
-    {
-      "epoch": 0.8212535128422979,
-      "grad_norm": 0.4234144985675812,
-      "learning_rate": 8.435147746905857e-06,
-      "loss": 0.3544,
-      "step": 12566
-    },
-    {
-      "epoch": 0.82131886804784,
-      "grad_norm": 0.441814124584198,
-      "learning_rate": 8.434894006776468e-06,
-      "loss": 0.38,
-      "step": 12567
-    },
-    {
-      "epoch": 0.8213842232533821,
-      "grad_norm": 0.463226854801178,
-      "learning_rate": 8.434640249893911e-06,
-      "loss": 0.4182,
-      "step": 12568
-    },
-    {
-      "epoch": 0.8214495784589243,
-      "grad_norm": 0.4492681622505188,
-      "learning_rate": 8.434386476259425e-06,
-      "loss": 0.3647,
-      "step": 12569
-    },
-    {
-      "epoch": 0.8215149336644664,
-      "grad_norm": 0.4281264543533325,
-      "learning_rate": 8.434132685874245e-06,
-      "loss": 0.3621,
-      "step": 12570
-    },
-    {
-      "epoch": 0.8215802888700084,
-      "grad_norm": 0.47149139642715454,
-      "learning_rate": 8.43387887873961e-06,
-      "loss": 0.3658,
-      "step": 12571
-    },
-    {
-      "epoch": 0.8216456440755506,
-      "grad_norm": 0.42672228813171387,
-      "learning_rate": 8.433625054856759e-06,
-      "loss": 0.3755,
-      "step": 12572
-    },
-    {
-      "epoch": 0.8217109992810927,
-      "grad_norm": 0.4492327570915222,
-      "learning_rate": 8.433371214226928e-06,
-      "loss": 0.378,
-      "step": 12573
-    },
-    {
-      "epoch": 0.8217763544866349,
-      "grad_norm": 0.46517860889434814,
-      "learning_rate": 8.433117356851358e-06,
-      "loss": 0.3939,
-      "step": 12574
-    },
-    {
-      "epoch": 0.821841709692177,
-      "grad_norm": 0.4429529905319214,
-      "learning_rate": 8.432863482731284e-06,
-      "loss": 0.3642,
-      "step": 12575
-    },
-    {
-      "epoch": 0.8219070648977191,
-      "grad_norm": 0.43571212887763977,
-      "learning_rate": 8.432609591867945e-06,
-      "loss": 0.3885,
-      "step": 12576
-    },
-    {
-      "epoch": 0.8219724201032612,
-      "grad_norm": 0.4329453110694885,
-      "learning_rate": 8.432355684262582e-06,
-      "loss": 0.3936,
-      "step": 12577
-    },
-    {
-      "epoch": 0.8220377753088034,
-      "grad_norm": 0.4580669701099396,
-      "learning_rate": 8.43210175991643e-06,
-      "loss": 0.3492,
-      "step": 12578
-    },
-    {
-      "epoch": 0.8221031305143455,
-      "grad_norm": 0.5206211805343628,
-      "learning_rate": 8.43184781883073e-06,
-      "loss": 0.4663,
-      "step": 12579
-    },
-    {
-      "epoch": 0.8221684857198875,
-      "grad_norm": 0.44426101446151733,
-      "learning_rate": 8.431593861006716e-06,
-      "loss": 0.3826,
-      "step": 12580
-    },
-    {
-      "epoch": 0.8222338409254297,
-      "grad_norm": 0.4304702579975128,
-      "learning_rate": 8.431339886445633e-06,
-      "loss": 0.3713,
-      "step": 12581
-    },
-    {
-      "epoch": 0.8222991961309718,
-      "grad_norm": 0.46631917357444763,
-      "learning_rate": 8.431085895148713e-06,
-      "loss": 0.4133,
-      "step": 12582
-    },
-    {
-      "epoch": 0.822364551336514,
-      "grad_norm": 0.4335556626319885,
-      "learning_rate": 8.430831887117201e-06,
-      "loss": 0.3605,
-      "step": 12583
-    },
-    {
-      "epoch": 0.822429906542056,
-      "grad_norm": 0.4497116506099701,
-      "learning_rate": 8.430577862352333e-06,
-      "loss": 0.359,
-      "step": 12584
-    },
-    {
-      "epoch": 0.8224952617475982,
-      "grad_norm": 0.4470299482345581,
-      "learning_rate": 8.43032382085535e-06,
-      "loss": 0.3743,
-      "step": 12585
-    },
-    {
-      "epoch": 0.8225606169531403,
-      "grad_norm": 0.4260590672492981,
-      "learning_rate": 8.430069762627489e-06,
-      "loss": 0.3585,
-      "step": 12586
-    },
-    {
-      "epoch": 0.8226259721586824,
-      "grad_norm": 0.4832247495651245,
-      "learning_rate": 8.429815687669986e-06,
-      "loss": 0.4748,
-      "step": 12587
-    },
-    {
-      "epoch": 0.8226913273642246,
-      "grad_norm": 0.4708634912967682,
-      "learning_rate": 8.429561595984087e-06,
-      "loss": 0.3798,
-      "step": 12588
-    },
-    {
-      "epoch": 0.8227566825697666,
-      "grad_norm": 0.4397064447402954,
-      "learning_rate": 8.429307487571028e-06,
-      "loss": 0.3797,
-      "step": 12589
-    },
-    {
-      "epoch": 0.8228220377753088,
-      "grad_norm": 0.4195902645587921,
-      "learning_rate": 8.429053362432047e-06,
-      "loss": 0.3741,
-      "step": 12590
-    },
-    {
-      "epoch": 0.8228873929808509,
-      "grad_norm": 0.4073476791381836,
-      "learning_rate": 8.428799220568384e-06,
-      "loss": 0.344,
-      "step": 12591
-    },
-    {
-      "epoch": 0.8229527481863931,
-      "grad_norm": 0.4689033031463623,
-      "learning_rate": 8.42854506198128e-06,
-      "loss": 0.4245,
-      "step": 12592
-    },
-    {
-      "epoch": 0.8230181033919352,
-      "grad_norm": 0.4563535451889038,
-      "learning_rate": 8.428290886671976e-06,
-      "loss": 0.384,
-      "step": 12593
-    },
-    {
-      "epoch": 0.8230834585974773,
-      "grad_norm": 0.45254579186439514,
-      "learning_rate": 8.428036694641708e-06,
-      "loss": 0.413,
-      "step": 12594
-    },
-    {
-      "epoch": 0.8231488138030194,
-      "grad_norm": 0.46842119097709656,
-      "learning_rate": 8.427782485891717e-06,
-      "loss": 0.4436,
-      "step": 12595
-    },
-    {
-      "epoch": 0.8232141690085615,
-      "grad_norm": 0.4607884883880615,
-      "learning_rate": 8.427528260423246e-06,
-      "loss": 0.4048,
-      "step": 12596
-    },
-    {
-      "epoch": 0.8232795242141037,
-      "grad_norm": 0.42940106987953186,
-      "learning_rate": 8.42727401823753e-06,
-      "loss": 0.3517,
-      "step": 12597
-    },
-    {
-      "epoch": 0.8233448794196457,
-      "grad_norm": 0.4498884975910187,
-      "learning_rate": 8.42701975933581e-06,
-      "loss": 0.3729,
-      "step": 12598
-    },
-    {
-      "epoch": 0.8234102346251879,
-      "grad_norm": 0.49875614047050476,
-      "learning_rate": 8.42676548371933e-06,
-      "loss": 0.4688,
-      "step": 12599
-    },
-    {
-      "epoch": 0.82347558983073,
-      "grad_norm": 0.4211624562740326,
-      "learning_rate": 8.426511191389325e-06,
-      "loss": 0.3682,
-      "step": 12600
-    },
-    {
-      "epoch": 0.8235409450362722,
-      "grad_norm": 0.4365617334842682,
-      "learning_rate": 8.42625688234704e-06,
-      "loss": 0.3668,
-      "step": 12601
-    },
-    {
-      "epoch": 0.8236063002418142,
-      "grad_norm": 0.40886780619621277,
-      "learning_rate": 8.426002556593712e-06,
-      "loss": 0.3267,
-      "step": 12602
-    },
-    {
-      "epoch": 0.8236716554473564,
-      "grad_norm": 0.3838039040565491,
-      "learning_rate": 8.425748214130584e-06,
-      "loss": 0.2832,
-      "step": 12603
-    },
-    {
-      "epoch": 0.8237370106528985,
-      "grad_norm": 0.4402284622192383,
-      "learning_rate": 8.425493854958895e-06,
-      "loss": 0.3838,
-      "step": 12604
-    },
-    {
-      "epoch": 0.8238023658584406,
-      "grad_norm": 0.460764616727829,
-      "learning_rate": 8.425239479079885e-06,
-      "loss": 0.4122,
-      "step": 12605
-    },
-    {
-      "epoch": 0.8238677210639828,
-      "grad_norm": 0.4580022096633911,
-      "learning_rate": 8.424985086494795e-06,
-      "loss": 0.3883,
-      "step": 12606
-    },
-    {
-      "epoch": 0.8239330762695248,
-      "grad_norm": 0.41545212268829346,
-      "learning_rate": 8.424730677204867e-06,
-      "loss": 0.3249,
-      "step": 12607
-    },
-    {
-      "epoch": 0.823998431475067,
-      "grad_norm": 0.46373510360717773,
-      "learning_rate": 8.42447625121134e-06,
-      "loss": 0.4204,
-      "step": 12608
-    },
-    {
-      "epoch": 0.8240637866806091,
-      "grad_norm": 0.4270155727863312,
-      "learning_rate": 8.424221808515458e-06,
-      "loss": 0.3435,
-      "step": 12609
-    },
-    {
-      "epoch": 0.8241291418861513,
-      "grad_norm": 0.4649266302585602,
-      "learning_rate": 8.423967349118459e-06,
-      "loss": 0.3716,
-      "step": 12610
-    },
-    {
-      "epoch": 0.8241944970916933,
-      "grad_norm": 0.42872345447540283,
-      "learning_rate": 8.423712873021585e-06,
-      "loss": 0.328,
-      "step": 12611
-    },
-    {
-      "epoch": 0.8242598522972355,
-      "grad_norm": 0.47963541746139526,
-      "learning_rate": 8.423458380226077e-06,
-      "loss": 0.422,
-      "step": 12612
-    },
-    {
-      "epoch": 0.8243252075027776,
-      "grad_norm": 0.46212321519851685,
-      "learning_rate": 8.423203870733177e-06,
-      "loss": 0.3602,
-      "step": 12613
-    },
-    {
-      "epoch": 0.8243905627083197,
-      "grad_norm": 0.43106013536453247,
-      "learning_rate": 8.422949344544125e-06,
-      "loss": 0.3455,
-      "step": 12614
-    },
-    {
-      "epoch": 0.8244559179138619,
-      "grad_norm": 0.4852938950061798,
-      "learning_rate": 8.422694801660162e-06,
-      "loss": 0.4061,
-      "step": 12615
-    },
-    {
-      "epoch": 0.8245212731194039,
-      "grad_norm": 0.4699553847312927,
-      "learning_rate": 8.422440242082533e-06,
-      "loss": 0.4368,
-      "step": 12616
-    },
-    {
-      "epoch": 0.8245866283249461,
-      "grad_norm": 0.4388345777988434,
-      "learning_rate": 8.422185665812479e-06,
-      "loss": 0.3468,
-      "step": 12617
-    },
-    {
-      "epoch": 0.8246519835304882,
-      "grad_norm": 0.47370830178260803,
-      "learning_rate": 8.421931072851237e-06,
-      "loss": 0.3819,
-      "step": 12618
-    },
-    {
-      "epoch": 0.8247173387360304,
-      "grad_norm": 0.4548191726207733,
-      "learning_rate": 8.42167646320005e-06,
-      "loss": 0.3617,
-      "step": 12619
-    },
-    {
-      "epoch": 0.8247826939415724,
-      "grad_norm": 0.43634673953056335,
-      "learning_rate": 8.421421836860166e-06,
-      "loss": 0.3671,
-      "step": 12620
-    },
-    {
-      "epoch": 0.8248480491471145,
-      "grad_norm": 0.4693472981452942,
-      "learning_rate": 8.42116719383282e-06,
-      "loss": 0.3635,
-      "step": 12621
-    },
-    {
-      "epoch": 0.8249134043526567,
-      "grad_norm": 0.4220568537712097,
-      "learning_rate": 8.420912534119256e-06,
-      "loss": 0.3292,
-      "step": 12622
-    },
-    {
-      "epoch": 0.8249787595581988,
-      "grad_norm": 0.45013129711151123,
-      "learning_rate": 8.420657857720717e-06,
-      "loss": 0.3635,
-      "step": 12623
-    },
-    {
-      "epoch": 0.825044114763741,
-      "grad_norm": 0.45930156111717224,
-      "learning_rate": 8.420403164638444e-06,
-      "loss": 0.3796,
-      "step": 12624
-    },
-    {
-      "epoch": 0.825109469969283,
-      "grad_norm": 0.44628220796585083,
-      "learning_rate": 8.420148454873681e-06,
-      "loss": 0.3437,
-      "step": 12625
-    },
-    {
-      "epoch": 0.8251748251748252,
-      "grad_norm": 0.4620745778083801,
-      "learning_rate": 8.419893728427668e-06,
-      "loss": 0.3779,
-      "step": 12626
-    },
-    {
-      "epoch": 0.8252401803803673,
-      "grad_norm": 0.4369930922985077,
-      "learning_rate": 8.41963898530165e-06,
-      "loss": 0.3178,
-      "step": 12627
-    },
-    {
-      "epoch": 0.8253055355859095,
-      "grad_norm": 0.47579482197761536,
-      "learning_rate": 8.419384225496868e-06,
-      "loss": 0.4016,
-      "step": 12628
-    },
-    {
-      "epoch": 0.8253708907914515,
-      "grad_norm": 0.400603324174881,
-      "learning_rate": 8.419129449014562e-06,
-      "loss": 0.324,
-      "step": 12629
-    },
-    {
-      "epoch": 0.8254362459969936,
-      "grad_norm": 0.44178307056427,
-      "learning_rate": 8.41887465585598e-06,
-      "loss": 0.3445,
-      "step": 12630
-    },
-    {
-      "epoch": 0.8255016012025358,
-      "grad_norm": 0.4656490981578827,
-      "learning_rate": 8.41861984602236e-06,
-      "loss": 0.4028,
-      "step": 12631
-    },
-    {
-      "epoch": 0.8255669564080779,
-      "grad_norm": 0.4748069941997528,
-      "learning_rate": 8.418365019514946e-06,
-      "loss": 0.4132,
-      "step": 12632
-    },
-    {
-      "epoch": 0.82563231161362,
-      "grad_norm": 0.42328181862831116,
-      "learning_rate": 8.418110176334984e-06,
-      "loss": 0.3367,
-      "step": 12633
-    },
-    {
-      "epoch": 0.8256976668191621,
-      "grad_norm": 0.49895450472831726,
-      "learning_rate": 8.417855316483715e-06,
-      "loss": 0.4207,
-      "step": 12634
-    },
-    {
-      "epoch": 0.8257630220247043,
-      "grad_norm": 0.46849435567855835,
-      "learning_rate": 8.41760043996238e-06,
-      "loss": 0.3787,
-      "step": 12635
-    },
-    {
-      "epoch": 0.8258283772302464,
-      "grad_norm": 0.4596104919910431,
-      "learning_rate": 8.417345546772222e-06,
-      "loss": 0.3656,
-      "step": 12636
-    },
-    {
-      "epoch": 0.8258937324357886,
-      "grad_norm": 0.43287724256515503,
-      "learning_rate": 8.417090636914487e-06,
-      "loss": 0.3826,
-      "step": 12637
-    },
-    {
-      "epoch": 0.8259590876413306,
-      "grad_norm": 0.4663711190223694,
-      "learning_rate": 8.416835710390418e-06,
-      "loss": 0.4286,
-      "step": 12638
-    },
-    {
-      "epoch": 0.8260244428468727,
-      "grad_norm": 0.4456733167171478,
-      "learning_rate": 8.41658076720126e-06,
-      "loss": 0.3693,
-      "step": 12639
-    },
-    {
-      "epoch": 0.8260897980524149,
-      "grad_norm": 0.4583789110183716,
-      "learning_rate": 8.416325807348249e-06,
-      "loss": 0.3734,
-      "step": 12640
-    },
-    {
-      "epoch": 0.826155153257957,
-      "grad_norm": 0.49686723947525024,
-      "learning_rate": 8.416070830832636e-06,
-      "loss": 0.4909,
-      "step": 12641
-    },
-    {
-      "epoch": 0.8262205084634991,
-      "grad_norm": 0.43204590678215027,
-      "learning_rate": 8.415815837655663e-06,
-      "loss": 0.3754,
-      "step": 12642
-    },
-    {
-      "epoch": 0.8262858636690412,
-      "grad_norm": 0.45484909415245056,
-      "learning_rate": 8.415560827818572e-06,
-      "loss": 0.3711,
-      "step": 12643
-    },
-    {
-      "epoch": 0.8263512188745834,
-      "grad_norm": 0.46242430806159973,
-      "learning_rate": 8.415305801322607e-06,
-      "loss": 0.3748,
-      "step": 12644
-    },
-    {
-      "epoch": 0.8264165740801255,
-      "grad_norm": 0.42559534311294556,
-      "learning_rate": 8.415050758169011e-06,
-      "loss": 0.3543,
-      "step": 12645
-    },
-    {
-      "epoch": 0.8264819292856675,
-      "grad_norm": 0.45685240626335144,
-      "learning_rate": 8.414795698359033e-06,
-      "loss": 0.4031,
-      "step": 12646
-    },
-    {
-      "epoch": 0.8265472844912097,
-      "grad_norm": 0.4635372757911682,
-      "learning_rate": 8.41454062189391e-06,
-      "loss": 0.3985,
-      "step": 12647
-    },
-    {
-      "epoch": 0.8266126396967518,
-      "grad_norm": 0.4519408941268921,
-      "learning_rate": 8.414285528774892e-06,
-      "loss": 0.3706,
-      "step": 12648
-    },
-    {
-      "epoch": 0.826677994902294,
-      "grad_norm": 0.41850516200065613,
-      "learning_rate": 8.41403041900322e-06,
-      "loss": 0.3429,
-      "step": 12649
-    },
-    {
-      "epoch": 0.8267433501078361,
-      "grad_norm": 0.4194774925708771,
-      "learning_rate": 8.413775292580137e-06,
-      "loss": 0.3302,
-      "step": 12650
-    },
-    {
-      "epoch": 0.8268087053133782,
-      "grad_norm": 0.4301159679889679,
-      "learning_rate": 8.413520149506892e-06,
-      "loss": 0.3525,
-      "step": 12651
-    },
-    {
-      "epoch": 0.8268740605189203,
-      "grad_norm": 0.4377503991127014,
-      "learning_rate": 8.413264989784726e-06,
-      "loss": 0.355,
-      "step": 12652
-    },
-    {
-      "epoch": 0.8269394157244625,
-      "grad_norm": 0.46422407031059265,
-      "learning_rate": 8.413009813414882e-06,
-      "loss": 0.3483,
-      "step": 12653
-    },
-    {
-      "epoch": 0.8270047709300046,
-      "grad_norm": 0.568018913269043,
-      "learning_rate": 8.412754620398609e-06,
-      "loss": 0.3467,
-      "step": 12654
-    },
-    {
-      "epoch": 0.8270701261355466,
-      "grad_norm": 0.4812344014644623,
-      "learning_rate": 8.412499410737149e-06,
-      "loss": 0.4122,
-      "step": 12655
-    },
-    {
-      "epoch": 0.8271354813410888,
-      "grad_norm": 0.4537135362625122,
-      "learning_rate": 8.412244184431745e-06,
-      "loss": 0.3408,
-      "step": 12656
-    },
-    {
-      "epoch": 0.8272008365466309,
-      "grad_norm": 0.45598137378692627,
-      "learning_rate": 8.411988941483649e-06,
-      "loss": 0.3643,
-      "step": 12657
-    },
-    {
-      "epoch": 0.8272661917521731,
-      "grad_norm": 0.5274432897567749,
-      "learning_rate": 8.411733681894096e-06,
-      "loss": 0.2587,
-      "step": 12658
-    },
-    {
-      "epoch": 0.8273315469577152,
-      "grad_norm": 0.4589974284172058,
-      "learning_rate": 8.41147840566434e-06,
-      "loss": 0.3845,
-      "step": 12659
-    },
-    {
-      "epoch": 0.8273969021632573,
-      "grad_norm": 0.4544839560985565,
-      "learning_rate": 8.41122311279562e-06,
-      "loss": 0.3663,
-      "step": 12660
-    },
-    {
-      "epoch": 0.8274622573687994,
-      "grad_norm": 0.4455678164958954,
-      "learning_rate": 8.410967803289182e-06,
-      "loss": 0.3912,
-      "step": 12661
-    },
-    {
-      "epoch": 0.8275276125743416,
-      "grad_norm": 0.5165808796882629,
-      "learning_rate": 8.410712477146272e-06,
-      "loss": 0.443,
-      "step": 12662
-    },
-    {
-      "epoch": 0.8275929677798837,
-      "grad_norm": 0.5062906742095947,
-      "learning_rate": 8.410457134368137e-06,
-      "loss": 0.4557,
-      "step": 12663
-    },
-    {
-      "epoch": 0.8276583229854257,
-      "grad_norm": 0.4476670026779175,
-      "learning_rate": 8.410201774956021e-06,
-      "loss": 0.3941,
-      "step": 12664
-    },
-    {
-      "epoch": 0.8277236781909679,
-      "grad_norm": 0.4280488193035126,
-      "learning_rate": 8.40994639891117e-06,
-      "loss": 0.3442,
-      "step": 12665
-    },
-    {
-      "epoch": 0.82778903339651,
-      "grad_norm": 0.467495858669281,
-      "learning_rate": 8.409691006234829e-06,
-      "loss": 0.4366,
-      "step": 12666
-    },
-    {
-      "epoch": 0.8278543886020522,
-      "grad_norm": 0.43454861640930176,
-      "learning_rate": 8.409435596928243e-06,
-      "loss": 0.3333,
-      "step": 12667
-    },
-    {
-      "epoch": 0.8279197438075943,
-      "grad_norm": 0.4636070728302002,
-      "learning_rate": 8.40918017099266e-06,
-      "loss": 0.3907,
-      "step": 12668
-    },
-    {
-      "epoch": 0.8279850990131364,
-      "grad_norm": 0.44613513350486755,
-      "learning_rate": 8.408924728429321e-06,
-      "loss": 0.3979,
-      "step": 12669
-    },
-    {
-      "epoch": 0.8280504542186785,
-      "grad_norm": 0.423446387052536,
-      "learning_rate": 8.408669269239478e-06,
-      "loss": 0.3373,
-      "step": 12670
-    },
-    {
-      "epoch": 0.8281158094242207,
-      "grad_norm": 0.4337618052959442,
-      "learning_rate": 8.408413793424372e-06,
-      "loss": 0.352,
-      "step": 12671
-    },
-    {
-      "epoch": 0.8281811646297628,
-      "grad_norm": 0.4586661458015442,
-      "learning_rate": 8.408158300985254e-06,
-      "loss": 0.3742,
-      "step": 12672
-    },
-    {
-      "epoch": 0.8282465198353048,
-      "grad_norm": 0.41534069180488586,
-      "learning_rate": 8.407902791923366e-06,
-      "loss": 0.3567,
-      "step": 12673
-    },
-    {
-      "epoch": 0.828311875040847,
-      "grad_norm": 0.4235283136367798,
-      "learning_rate": 8.407647266239954e-06,
-      "loss": 0.3254,
-      "step": 12674
-    },
-    {
-      "epoch": 0.8283772302463891,
-      "grad_norm": 0.4674146771430969,
-      "learning_rate": 8.407391723936267e-06,
-      "loss": 0.4166,
-      "step": 12675
-    },
-    {
-      "epoch": 0.8284425854519313,
-      "grad_norm": 0.4396532475948334,
-      "learning_rate": 8.40713616501355e-06,
-      "loss": 0.3555,
-      "step": 12676
-    },
-    {
-      "epoch": 0.8285079406574734,
-      "grad_norm": 0.4806467294692993,
-      "learning_rate": 8.40688058947305e-06,
-      "loss": 0.4116,
-      "step": 12677
-    },
-    {
-      "epoch": 0.8285732958630155,
-      "grad_norm": 0.46783992648124695,
-      "learning_rate": 8.406624997316014e-06,
-      "loss": 0.3982,
-      "step": 12678
-    },
-    {
-      "epoch": 0.8286386510685576,
-      "grad_norm": 0.4396979808807373,
-      "learning_rate": 8.406369388543684e-06,
-      "loss": 0.3676,
-      "step": 12679
-    },
-    {
-      "epoch": 0.8287040062740997,
-      "grad_norm": 0.40401342511177063,
-      "learning_rate": 8.406113763157313e-06,
-      "loss": 0.3376,
-      "step": 12680
-    },
-    {
-      "epoch": 0.8287693614796419,
-      "grad_norm": 0.46379831433296204,
-      "learning_rate": 8.405858121158146e-06,
-      "loss": 0.4036,
-      "step": 12681
-    },
-    {
-      "epoch": 0.8288347166851839,
-      "grad_norm": 0.45665764808654785,
-      "learning_rate": 8.405602462547428e-06,
-      "loss": 0.4325,
-      "step": 12682
-    },
-    {
-      "epoch": 0.8289000718907261,
-      "grad_norm": 0.42489299178123474,
-      "learning_rate": 8.405346787326408e-06,
-      "loss": 0.3322,
-      "step": 12683
-    },
-    {
-      "epoch": 0.8289654270962682,
-      "grad_norm": 0.43218982219696045,
-      "learning_rate": 8.40509109549633e-06,
-      "loss": 0.3358,
-      "step": 12684
-    },
-    {
-      "epoch": 0.8290307823018104,
-      "grad_norm": 0.4428415596485138,
-      "learning_rate": 8.404835387058445e-06,
-      "loss": 0.3758,
-      "step": 12685
-    },
-    {
-      "epoch": 0.8290961375073524,
-      "grad_norm": 0.4392610192298889,
-      "learning_rate": 8.404579662013996e-06,
-      "loss": 0.3538,
-      "step": 12686
-    },
-    {
-      "epoch": 0.8291614927128946,
-      "grad_norm": 0.43328729271888733,
-      "learning_rate": 8.404323920364235e-06,
-      "loss": 0.3634,
-      "step": 12687
-    },
-    {
-      "epoch": 0.8292268479184367,
-      "grad_norm": 0.47693005204200745,
-      "learning_rate": 8.404068162110406e-06,
-      "loss": 0.4126,
-      "step": 12688
-    },
-    {
-      "epoch": 0.8292922031239788,
-      "grad_norm": 0.4226832389831543,
-      "learning_rate": 8.403812387253755e-06,
-      "loss": 0.343,
-      "step": 12689
-    },
-    {
-      "epoch": 0.829357558329521,
-      "grad_norm": 0.44714322686195374,
-      "learning_rate": 8.403556595795536e-06,
-      "loss": 0.3805,
-      "step": 12690
-    },
-    {
-      "epoch": 0.829422913535063,
-      "grad_norm": 0.44101881980895996,
-      "learning_rate": 8.40330078773699e-06,
-      "loss": 0.3581,
-      "step": 12691
-    },
-    {
-      "epoch": 0.8294882687406052,
-      "grad_norm": 0.4675297439098358,
-      "learning_rate": 8.403044963079367e-06,
-      "loss": 0.3865,
-      "step": 12692
-    },
-    {
-      "epoch": 0.8295536239461473,
-      "grad_norm": 0.44370898604393005,
-      "learning_rate": 8.402789121823916e-06,
-      "loss": 0.3907,
-      "step": 12693
-    },
-    {
-      "epoch": 0.8296189791516895,
-      "grad_norm": 0.4831228256225586,
-      "learning_rate": 8.402533263971882e-06,
-      "loss": 0.4117,
-      "step": 12694
-    },
-    {
-      "epoch": 0.8296843343572315,
-      "grad_norm": 0.42903903126716614,
-      "learning_rate": 8.402277389524516e-06,
-      "loss": 0.3316,
-      "step": 12695
-    },
-    {
-      "epoch": 0.8297496895627737,
-      "grad_norm": 0.43575528264045715,
-      "learning_rate": 8.402021498483063e-06,
-      "loss": 0.3866,
-      "step": 12696
-    },
-    {
-      "epoch": 0.8298150447683158,
-      "grad_norm": 0.4131913185119629,
-      "learning_rate": 8.401765590848773e-06,
-      "loss": 0.3395,
-      "step": 12697
-    },
-    {
-      "epoch": 0.8298803999738579,
-      "grad_norm": 0.4388498067855835,
-      "learning_rate": 8.401509666622894e-06,
-      "loss": 0.3918,
-      "step": 12698
-    },
-    {
-      "epoch": 0.8299457551794001,
-      "grad_norm": 0.43172523379325867,
-      "learning_rate": 8.401253725806674e-06,
-      "loss": 0.3843,
-      "step": 12699
-    },
-    {
-      "epoch": 0.8300111103849421,
-      "grad_norm": 0.45876553654670715,
-      "learning_rate": 8.40099776840136e-06,
-      "loss": 0.3912,
-      "step": 12700
-    },
-    {
-      "epoch": 0.8300764655904843,
-      "grad_norm": 0.41548505425453186,
-      "learning_rate": 8.400741794408204e-06,
-      "loss": 0.3754,
-      "step": 12701
-    },
-    {
-      "epoch": 0.8301418207960264,
-      "grad_norm": 0.4330996572971344,
-      "learning_rate": 8.40048580382845e-06,
-      "loss": 0.3474,
-      "step": 12702
-    },
-    {
-      "epoch": 0.8302071760015686,
-      "grad_norm": 0.4715730547904968,
-      "learning_rate": 8.400229796663351e-06,
-      "loss": 0.4227,
-      "step": 12703
-    },
-    {
-      "epoch": 0.8302725312071106,
-      "grad_norm": 0.44692686200141907,
-      "learning_rate": 8.399973772914151e-06,
-      "loss": 0.3728,
-      "step": 12704
-    },
-    {
-      "epoch": 0.8303378864126527,
-      "grad_norm": 0.4354611337184906,
-      "learning_rate": 8.399717732582103e-06,
-      "loss": 0.3529,
-      "step": 12705
-    },
-    {
-      "epoch": 0.8304032416181949,
-      "grad_norm": 0.43565553426742554,
-      "learning_rate": 8.399461675668454e-06,
-      "loss": 0.3681,
-      "step": 12706
-    },
-    {
-      "epoch": 0.830468596823737,
-      "grad_norm": 0.4220907688140869,
-      "learning_rate": 8.399205602174451e-06,
-      "loss": 0.3278,
-      "step": 12707
-    },
-    {
-      "epoch": 0.8305339520292792,
-      "grad_norm": 0.45381301641464233,
-      "learning_rate": 8.398949512101345e-06,
-      "loss": 0.4048,
-      "step": 12708
-    },
-    {
-      "epoch": 0.8305993072348212,
-      "grad_norm": 0.43763771653175354,
-      "learning_rate": 8.398693405450385e-06,
-      "loss": 0.3559,
-      "step": 12709
-    },
-    {
-      "epoch": 0.8306646624403634,
-      "grad_norm": 0.45486941933631897,
-      "learning_rate": 8.39843728222282e-06,
-      "loss": 0.4211,
-      "step": 12710
-    },
-    {
-      "epoch": 0.8307300176459055,
-      "grad_norm": 0.44698017835617065,
-      "learning_rate": 8.3981811424199e-06,
-      "loss": 0.362,
-      "step": 12711
-    },
-    {
-      "epoch": 0.8307953728514477,
-      "grad_norm": 0.4658489525318146,
-      "learning_rate": 8.397924986042872e-06,
-      "loss": 0.3659,
-      "step": 12712
-    },
-    {
-      "epoch": 0.8308607280569897,
-      "grad_norm": 0.4626188278198242,
-      "learning_rate": 8.397668813092988e-06,
-      "loss": 0.3945,
-      "step": 12713
-    },
-    {
-      "epoch": 0.8309260832625318,
-      "grad_norm": 0.4650971293449402,
-      "learning_rate": 8.397412623571495e-06,
-      "loss": 0.431,
-      "step": 12714
-    },
-    {
-      "epoch": 0.830991438468074,
-      "grad_norm": 0.4560522437095642,
-      "learning_rate": 8.397156417479642e-06,
-      "loss": 0.4491,
-      "step": 12715
-    },
-    {
-      "epoch": 0.8310567936736161,
-      "grad_norm": 0.4106829762458801,
-      "learning_rate": 8.396900194818682e-06,
-      "loss": 0.34,
-      "step": 12716
-    },
-    {
-      "epoch": 0.8311221488791583,
-      "grad_norm": 0.4265146553516388,
-      "learning_rate": 8.396643955589863e-06,
-      "loss": 0.3529,
-      "step": 12717
-    },
-    {
-      "epoch": 0.8311875040847003,
-      "grad_norm": 0.45510897040367126,
-      "learning_rate": 8.396387699794436e-06,
-      "loss": 0.3871,
-      "step": 12718
-    },
-    {
-      "epoch": 0.8312528592902425,
-      "grad_norm": 0.4605911076068878,
-      "learning_rate": 8.396131427433648e-06,
-      "loss": 0.3943,
-      "step": 12719
-    },
-    {
-      "epoch": 0.8313182144957846,
-      "grad_norm": 0.46505945920944214,
-      "learning_rate": 8.39587513850875e-06,
-      "loss": 0.3951,
-      "step": 12720
-    },
-    {
-      "epoch": 0.8313835697013268,
-      "grad_norm": 0.43580713868141174,
-      "learning_rate": 8.395618833020993e-06,
-      "loss": 0.3512,
-      "step": 12721
-    },
-    {
-      "epoch": 0.8314489249068688,
-      "grad_norm": 0.4356127679347992,
-      "learning_rate": 8.395362510971628e-06,
-      "loss": 0.3679,
-      "step": 12722
-    },
-    {
-      "epoch": 0.8315142801124109,
-      "grad_norm": 0.47002744674682617,
-      "learning_rate": 8.395106172361903e-06,
-      "loss": 0.4242,
-      "step": 12723
-    },
-    {
-      "epoch": 0.8315796353179531,
-      "grad_norm": 0.44220444560050964,
-      "learning_rate": 8.394849817193068e-06,
-      "loss": 0.3727,
-      "step": 12724
-    },
-    {
-      "epoch": 0.8316449905234952,
-      "grad_norm": 0.43452540040016174,
-      "learning_rate": 8.394593445466375e-06,
-      "loss": 0.3664,
-      "step": 12725
-    },
-    {
-      "epoch": 0.8317103457290373,
-      "grad_norm": 0.47207656502723694,
-      "learning_rate": 8.394337057183074e-06,
-      "loss": 0.4063,
-      "step": 12726
-    },
-    {
-      "epoch": 0.8317757009345794,
-      "grad_norm": 0.4225301146507263,
-      "learning_rate": 8.394080652344415e-06,
-      "loss": 0.3163,
-      "step": 12727
-    },
-    {
-      "epoch": 0.8318410561401216,
-      "grad_norm": 0.42666900157928467,
-      "learning_rate": 8.393824230951647e-06,
-      "loss": 0.348,
-      "step": 12728
-    },
-    {
-      "epoch": 0.8319064113456637,
-      "grad_norm": 0.4660142958164215,
-      "learning_rate": 8.393567793006025e-06,
-      "loss": 0.3963,
-      "step": 12729
-    },
-    {
-      "epoch": 0.8319717665512057,
-      "grad_norm": 0.4486709535121918,
-      "learning_rate": 8.393311338508799e-06,
-      "loss": 0.3684,
-      "step": 12730
-    },
-    {
-      "epoch": 0.8320371217567479,
-      "grad_norm": 0.38883456587791443,
-      "learning_rate": 8.393054867461214e-06,
-      "loss": 0.275,
-      "step": 12731
-    },
-    {
-      "epoch": 0.83210247696229,
-      "grad_norm": 0.4592151939868927,
-      "learning_rate": 8.392798379864526e-06,
-      "loss": 0.3827,
-      "step": 12732
-    },
-    {
-      "epoch": 0.8321678321678322,
-      "grad_norm": 0.4490208327770233,
-      "learning_rate": 8.392541875719987e-06,
-      "loss": 0.3601,
-      "step": 12733
-    },
-    {
-      "epoch": 0.8322331873733743,
-      "grad_norm": 0.46575114130973816,
-      "learning_rate": 8.392285355028844e-06,
-      "loss": 0.4081,
-      "step": 12734
-    },
-    {
-      "epoch": 0.8322985425789164,
-      "grad_norm": 0.45290032029151917,
-      "learning_rate": 8.39202881779235e-06,
-      "loss": 0.3864,
-      "step": 12735
-    },
-    {
-      "epoch": 0.8323638977844585,
-      "grad_norm": 0.48589763045310974,
-      "learning_rate": 8.391772264011757e-06,
-      "loss": 0.4194,
-      "step": 12736
-    },
-    {
-      "epoch": 0.8324292529900007,
-      "grad_norm": 0.47278133034706116,
-      "learning_rate": 8.391515693688317e-06,
-      "loss": 0.3597,
-      "step": 12737
-    },
-    {
-      "epoch": 0.8324946081955428,
-      "grad_norm": 0.4128890931606293,
-      "learning_rate": 8.391259106823277e-06,
-      "loss": 0.3481,
-      "step": 12738
-    },
-    {
-      "epoch": 0.8325599634010848,
-      "grad_norm": 0.45492124557495117,
-      "learning_rate": 8.391002503417893e-06,
-      "loss": 0.3701,
-      "step": 12739
-    },
-    {
-      "epoch": 0.832625318606627,
-      "grad_norm": 0.46682876348495483,
-      "learning_rate": 8.390745883473417e-06,
-      "loss": 0.3607,
-      "step": 12740
-    },
-    {
-      "epoch": 0.8326906738121691,
-      "grad_norm": 0.4522010087966919,
-      "learning_rate": 8.390489246991096e-06,
-      "loss": 0.3826,
-      "step": 12741
-    },
-    {
-      "epoch": 0.8327560290177113,
-      "grad_norm": 0.4297303259372711,
-      "learning_rate": 8.390232593972185e-06,
-      "loss": 0.3733,
-      "step": 12742
-    },
-    {
-      "epoch": 0.8328213842232534,
-      "grad_norm": 0.4380183517932892,
-      "learning_rate": 8.389975924417936e-06,
-      "loss": 0.3735,
-      "step": 12743
-    },
-    {
-      "epoch": 0.8328867394287955,
-      "grad_norm": 0.4422387480735779,
-      "learning_rate": 8.389719238329598e-06,
-      "loss": 0.4107,
-      "step": 12744
-    },
-    {
-      "epoch": 0.8329520946343376,
-      "grad_norm": 0.46929383277893066,
-      "learning_rate": 8.389462535708428e-06,
-      "loss": 0.386,
-      "step": 12745
-    },
-    {
-      "epoch": 0.8330174498398798,
-      "grad_norm": 0.44553861021995544,
-      "learning_rate": 8.389205816555673e-06,
-      "loss": 0.3709,
-      "step": 12746
-    },
-    {
-      "epoch": 0.8330828050454219,
-      "grad_norm": 0.5020803213119507,
-      "learning_rate": 8.388949080872588e-06,
-      "loss": 0.3955,
-      "step": 12747
-    },
-    {
-      "epoch": 0.8331481602509639,
-      "grad_norm": 0.45638465881347656,
-      "learning_rate": 8.388692328660423e-06,
-      "loss": 0.3936,
-      "step": 12748
-    },
-    {
-      "epoch": 0.8332135154565061,
-      "grad_norm": 0.46880820393562317,
-      "learning_rate": 8.388435559920433e-06,
-      "loss": 0.4002,
-      "step": 12749
-    },
-    {
-      "epoch": 0.8332788706620482,
-      "grad_norm": 0.4836241602897644,
-      "learning_rate": 8.388178774653869e-06,
-      "loss": 0.4489,
-      "step": 12750
-    },
-    {
-      "epoch": 0.8333442258675904,
-      "grad_norm": 0.5328258275985718,
-      "learning_rate": 8.38792197286198e-06,
-      "loss": 0.5287,
-      "step": 12751
-    },
-    {
-      "epoch": 0.8334095810731325,
-      "grad_norm": 0.4341478645801544,
-      "learning_rate": 8.387665154546025e-06,
-      "loss": 0.3451,
-      "step": 12752
-    },
-    {
-      "epoch": 0.8334749362786746,
-      "grad_norm": 0.4613303542137146,
-      "learning_rate": 8.387408319707254e-06,
-      "loss": 0.4009,
-      "step": 12753
-    },
-    {
-      "epoch": 0.8335402914842167,
-      "grad_norm": 0.4375651776790619,
-      "learning_rate": 8.387151468346916e-06,
-      "loss": 0.3779,
-      "step": 12754
-    },
-    {
-      "epoch": 0.8336056466897589,
-      "grad_norm": 0.43719443678855896,
-      "learning_rate": 8.38689460046627e-06,
-      "loss": 0.3549,
-      "step": 12755
-    },
-    {
-      "epoch": 0.833671001895301,
-      "grad_norm": 0.43977174162864685,
-      "learning_rate": 8.386637716066563e-06,
-      "loss": 0.3449,
-      "step": 12756
-    },
-    {
-      "epoch": 0.833736357100843,
-      "grad_norm": 0.43269723653793335,
-      "learning_rate": 8.386380815149053e-06,
-      "loss": 0.3469,
-      "step": 12757
-    },
-    {
-      "epoch": 0.8338017123063852,
-      "grad_norm": 0.4300509989261627,
-      "learning_rate": 8.386123897714991e-06,
-      "loss": 0.3808,
-      "step": 12758
-    },
-    {
-      "epoch": 0.8338670675119273,
-      "grad_norm": 0.43225598335266113,
-      "learning_rate": 8.385866963765628e-06,
-      "loss": 0.3836,
-      "step": 12759
-    },
-    {
-      "epoch": 0.8339324227174695,
-      "grad_norm": 0.4335007071495056,
-      "learning_rate": 8.38561001330222e-06,
-      "loss": 0.373,
-      "step": 12760
-    },
-    {
-      "epoch": 0.8339977779230116,
-      "grad_norm": 0.4306311011314392,
-      "learning_rate": 8.385353046326017e-06,
-      "loss": 0.3542,
-      "step": 12761
-    },
-    {
-      "epoch": 0.8340631331285537,
-      "grad_norm": 0.4562755823135376,
-      "learning_rate": 8.385096062838275e-06,
-      "loss": 0.403,
-      "step": 12762
-    },
-    {
-      "epoch": 0.8341284883340958,
-      "grad_norm": 0.5411015748977661,
-      "learning_rate": 8.38483906284025e-06,
-      "loss": 0.4526,
-      "step": 12763
-    },
-    {
-      "epoch": 0.8341938435396379,
-      "grad_norm": 0.4221334755420685,
-      "learning_rate": 8.384582046333188e-06,
-      "loss": 0.3668,
-      "step": 12764
-    },
-    {
-      "epoch": 0.8342591987451801,
-      "grad_norm": 0.46379244327545166,
-      "learning_rate": 8.384325013318348e-06,
-      "loss": 0.3976,
-      "step": 12765
-    },
-    {
-      "epoch": 0.8343245539507221,
-      "grad_norm": 0.46996960043907166,
-      "learning_rate": 8.384067963796984e-06,
-      "loss": 0.3805,
-      "step": 12766
-    },
-    {
-      "epoch": 0.8343899091562643,
-      "grad_norm": 0.45651906728744507,
-      "learning_rate": 8.383810897770348e-06,
-      "loss": 0.3634,
-      "step": 12767
-    },
-    {
-      "epoch": 0.8344552643618064,
-      "grad_norm": 0.6424210667610168,
-      "learning_rate": 8.383553815239693e-06,
-      "loss": 0.3589,
-      "step": 12768
-    },
-    {
-      "epoch": 0.8345206195673486,
-      "grad_norm": 0.43621474504470825,
-      "learning_rate": 8.383296716206273e-06,
-      "loss": 0.3912,
-      "step": 12769
-    },
-    {
-      "epoch": 0.8345859747728906,
-      "grad_norm": 0.43640974164009094,
-      "learning_rate": 8.383039600671344e-06,
-      "loss": 0.3523,
-      "step": 12770
-    },
-    {
-      "epoch": 0.8346513299784328,
-      "grad_norm": 0.4584968090057373,
-      "learning_rate": 8.38278246863616e-06,
-      "loss": 0.3949,
-      "step": 12771
-    },
-    {
-      "epoch": 0.8347166851839749,
-      "grad_norm": 0.4630500376224518,
-      "learning_rate": 8.382525320101972e-06,
-      "loss": 0.4189,
-      "step": 12772
-    },
-    {
-      "epoch": 0.834782040389517,
-      "grad_norm": 0.4346463978290558,
-      "learning_rate": 8.382268155070037e-06,
-      "loss": 0.423,
-      "step": 12773
-    },
-    {
-      "epoch": 0.8348473955950592,
-      "grad_norm": 0.4622276723384857,
-      "learning_rate": 8.382010973541608e-06,
-      "loss": 0.4342,
-      "step": 12774
-    },
-    {
-      "epoch": 0.8349127508006012,
-      "grad_norm": 0.43050628900527954,
-      "learning_rate": 8.38175377551794e-06,
-      "loss": 0.3645,
-      "step": 12775
-    },
-    {
-      "epoch": 0.8349781060061434,
-      "grad_norm": 0.4691722095012665,
-      "learning_rate": 8.381496561000289e-06,
-      "loss": 0.4378,
-      "step": 12776
-    },
-    {
-      "epoch": 0.8350434612116855,
-      "grad_norm": 0.45498543977737427,
-      "learning_rate": 8.381239329989905e-06,
-      "loss": 0.419,
-      "step": 12777
-    },
-    {
-      "epoch": 0.8351088164172277,
-      "grad_norm": 0.47149309515953064,
-      "learning_rate": 8.380982082488047e-06,
-      "loss": 0.4234,
-      "step": 12778
-    },
-    {
-      "epoch": 0.8351741716227697,
-      "grad_norm": 0.43489494919776917,
-      "learning_rate": 8.380724818495968e-06,
-      "loss": 0.3675,
-      "step": 12779
-    },
-    {
-      "epoch": 0.8352395268283119,
-      "grad_norm": 0.43044498562812805,
-      "learning_rate": 8.380467538014923e-06,
-      "loss": 0.3623,
-      "step": 12780
-    },
-    {
-      "epoch": 0.835304882033854,
-      "grad_norm": 0.41972649097442627,
-      "learning_rate": 8.380210241046167e-06,
-      "loss": 0.3121,
-      "step": 12781
-    },
-    {
-      "epoch": 0.8353702372393961,
-      "grad_norm": 0.4362650513648987,
-      "learning_rate": 8.379952927590952e-06,
-      "loss": 0.3709,
-      "step": 12782
-    },
-    {
-      "epoch": 0.8354355924449383,
-      "grad_norm": 0.4583292305469513,
-      "learning_rate": 8.379695597650539e-06,
-      "loss": 0.3699,
-      "step": 12783
-    },
-    {
-      "epoch": 0.8355009476504803,
-      "grad_norm": 0.465030699968338,
-      "learning_rate": 8.37943825122618e-06,
-      "loss": 0.3931,
-      "step": 12784
-    },
-    {
-      "epoch": 0.8355663028560225,
-      "grad_norm": 0.4656769037246704,
-      "learning_rate": 8.379180888319127e-06,
-      "loss": 0.4428,
-      "step": 12785
-    },
-    {
-      "epoch": 0.8356316580615646,
-      "grad_norm": 0.43806612491607666,
-      "learning_rate": 8.37892350893064e-06,
-      "loss": 0.3307,
-      "step": 12786
-    },
-    {
-      "epoch": 0.8356970132671068,
-      "grad_norm": 0.4317433536052704,
-      "learning_rate": 8.378666113061973e-06,
-      "loss": 0.3668,
-      "step": 12787
-    },
-    {
-      "epoch": 0.8357623684726488,
-      "grad_norm": 0.4749669134616852,
-      "learning_rate": 8.378408700714378e-06,
-      "loss": 0.4327,
-      "step": 12788
-    },
-    {
-      "epoch": 0.8358277236781909,
-      "grad_norm": 0.48393934965133667,
-      "learning_rate": 8.378151271889117e-06,
-      "loss": 0.4455,
-      "step": 12789
-    },
-    {
-      "epoch": 0.8358930788837331,
-      "grad_norm": 0.4799225330352783,
-      "learning_rate": 8.37789382658744e-06,
-      "loss": 0.3969,
-      "step": 12790
-    },
-    {
-      "epoch": 0.8359584340892752,
-      "grad_norm": 0.4521721303462982,
-      "learning_rate": 8.377636364810605e-06,
-      "loss": 0.3794,
-      "step": 12791
-    },
-    {
-      "epoch": 0.8360237892948174,
-      "grad_norm": 0.4709000885486603,
-      "learning_rate": 8.377378886559865e-06,
-      "loss": 0.4029,
-      "step": 12792
-    },
-    {
-      "epoch": 0.8360891445003594,
-      "grad_norm": 0.4529739320278168,
-      "learning_rate": 8.377121391836483e-06,
-      "loss": 0.3792,
-      "step": 12793
-    },
-    {
-      "epoch": 0.8361544997059016,
-      "grad_norm": 0.4557335078716278,
-      "learning_rate": 8.376863880641705e-06,
-      "loss": 0.3751,
-      "step": 12794
-    },
-    {
-      "epoch": 0.8362198549114437,
-      "grad_norm": 0.4018639326095581,
-      "learning_rate": 8.376606352976795e-06,
-      "loss": 0.3067,
-      "step": 12795
-    },
-    {
-      "epoch": 0.8362852101169859,
-      "grad_norm": 0.429113507270813,
-      "learning_rate": 8.376348808843006e-06,
-      "loss": 0.3596,
-      "step": 12796
-    },
-    {
-      "epoch": 0.8363505653225279,
-      "grad_norm": 0.43608278036117554,
-      "learning_rate": 8.376091248241594e-06,
-      "loss": 0.3448,
-      "step": 12797
-    },
-    {
-      "epoch": 0.83641592052807,
-      "grad_norm": 0.4282897114753723,
-      "learning_rate": 8.375833671173814e-06,
-      "loss": 0.3537,
-      "step": 12798
-    },
-    {
-      "epoch": 0.8364812757336122,
-      "grad_norm": 0.44926175475120544,
-      "learning_rate": 8.375576077640925e-06,
-      "loss": 0.3979,
-      "step": 12799
-    },
-    {
-      "epoch": 0.8365466309391543,
-      "grad_norm": 0.45824864506721497,
-      "learning_rate": 8.375318467644182e-06,
-      "loss": 0.3742,
-      "step": 12800
-    },
-    {
-      "epoch": 0.8366119861446965,
-      "grad_norm": 0.43586090207099915,
-      "learning_rate": 8.375060841184841e-06,
-      "loss": 0.3422,
-      "step": 12801
-    },
-    {
-      "epoch": 0.8366773413502385,
-      "grad_norm": 0.4497327208518982,
-      "learning_rate": 8.374803198264158e-06,
-      "loss": 0.3646,
-      "step": 12802
-    },
-    {
-      "epoch": 0.8367426965557807,
-      "grad_norm": 0.45988771319389343,
-      "learning_rate": 8.374545538883392e-06,
-      "loss": 0.3795,
-      "step": 12803
-    },
-    {
-      "epoch": 0.8368080517613228,
-      "grad_norm": 0.4657236933708191,
-      "learning_rate": 8.374287863043798e-06,
-      "loss": 0.4016,
-      "step": 12804
-    },
-    {
-      "epoch": 0.836873406966865,
-      "grad_norm": 0.4531661570072174,
-      "learning_rate": 8.374030170746635e-06,
-      "loss": 0.3946,
-      "step": 12805
-    },
-    {
-      "epoch": 0.836938762172407,
-      "grad_norm": 0.46240153908729553,
-      "learning_rate": 8.373772461993156e-06,
-      "loss": 0.3895,
-      "step": 12806
-    },
-    {
-      "epoch": 0.8370041173779491,
-      "grad_norm": 0.45432382822036743,
-      "learning_rate": 8.37351473678462e-06,
-      "loss": 0.375,
-      "step": 12807
-    },
-    {
-      "epoch": 0.8370694725834913,
-      "grad_norm": 0.4486439526081085,
-      "learning_rate": 8.373256995122284e-06,
-      "loss": 0.3807,
-      "step": 12808
-    },
-    {
-      "epoch": 0.8371348277890334,
-      "grad_norm": 0.4164700508117676,
-      "learning_rate": 8.372999237007405e-06,
-      "loss": 0.3043,
-      "step": 12809
-    },
-    {
-      "epoch": 0.8372001829945755,
-      "grad_norm": 0.44789445400238037,
-      "learning_rate": 8.37274146244124e-06,
-      "loss": 0.3774,
-      "step": 12810
-    },
-    {
-      "epoch": 0.8372655382001176,
-      "grad_norm": 0.442098468542099,
-      "learning_rate": 8.372483671425047e-06,
-      "loss": 0.3628,
-      "step": 12811
-    },
-    {
-      "epoch": 0.8373308934056598,
-      "grad_norm": 0.44000622630119324,
-      "learning_rate": 8.372225863960083e-06,
-      "loss": 0.3874,
-      "step": 12812
-    },
-    {
-      "epoch": 0.8373962486112019,
-      "grad_norm": 0.46801048517227173,
-      "learning_rate": 8.371968040047604e-06,
-      "loss": 0.3724,
-      "step": 12813
-    },
-    {
-      "epoch": 0.837461603816744,
-      "grad_norm": 0.45521214604377747,
-      "learning_rate": 8.37171019968887e-06,
-      "loss": 0.3803,
-      "step": 12814
-    },
-    {
-      "epoch": 0.8375269590222861,
-      "grad_norm": 0.44060996174812317,
-      "learning_rate": 8.371452342885139e-06,
-      "loss": 0.3878,
-      "step": 12815
-    },
-    {
-      "epoch": 0.8375923142278282,
-      "grad_norm": 0.4750652313232422,
-      "learning_rate": 8.371194469637662e-06,
-      "loss": 0.3971,
-      "step": 12816
-    },
-    {
-      "epoch": 0.8376576694333704,
-      "grad_norm": 0.4220958352088928,
-      "learning_rate": 8.370936579947706e-06,
-      "loss": 0.355,
-      "step": 12817
-    },
-    {
-      "epoch": 0.8377230246389125,
-      "grad_norm": 0.44775012135505676,
-      "learning_rate": 8.370678673816523e-06,
-      "loss": 0.3658,
-      "step": 12818
-    },
-    {
-      "epoch": 0.8377883798444546,
-      "grad_norm": 0.43990033864974976,
-      "learning_rate": 8.370420751245371e-06,
-      "loss": 0.3252,
-      "step": 12819
-    },
-    {
-      "epoch": 0.8378537350499967,
-      "grad_norm": 0.4598587453365326,
-      "learning_rate": 8.370162812235512e-06,
-      "loss": 0.402,
-      "step": 12820
-    },
-    {
-      "epoch": 0.8379190902555389,
-      "grad_norm": 0.45981690287590027,
-      "learning_rate": 8.3699048567882e-06,
-      "loss": 0.3977,
-      "step": 12821
-    },
-    {
-      "epoch": 0.837984445461081,
-      "grad_norm": 0.44870638847351074,
-      "learning_rate": 8.369646884904694e-06,
-      "loss": 0.3702,
-      "step": 12822
-    },
-    {
-      "epoch": 0.838049800666623,
-      "grad_norm": 0.4393848478794098,
-      "learning_rate": 8.369388896586254e-06,
-      "loss": 0.3926,
-      "step": 12823
-    },
-    {
-      "epoch": 0.8381151558721652,
-      "grad_norm": 0.45588740706443787,
-      "learning_rate": 8.369130891834136e-06,
-      "loss": 0.4076,
-      "step": 12824
-    },
-    {
-      "epoch": 0.8381805110777073,
-      "grad_norm": 0.4453178644180298,
-      "learning_rate": 8.3688728706496e-06,
-      "loss": 0.3758,
-      "step": 12825
-    },
-    {
-      "epoch": 0.8382458662832495,
-      "grad_norm": 0.4394017457962036,
-      "learning_rate": 8.368614833033906e-06,
-      "loss": 0.3672,
-      "step": 12826
-    },
-    {
-      "epoch": 0.8383112214887916,
-      "grad_norm": 0.48031318187713623,
-      "learning_rate": 8.368356778988306e-06,
-      "loss": 0.4246,
-      "step": 12827
-    },
-    {
-      "epoch": 0.8383765766943337,
-      "grad_norm": 0.4832472503185272,
-      "learning_rate": 8.368098708514068e-06,
-      "loss": 0.3781,
-      "step": 12828
-    },
-    {
-      "epoch": 0.8384419318998758,
-      "grad_norm": 0.4725753962993622,
-      "learning_rate": 8.367840621612443e-06,
-      "loss": 0.4146,
-      "step": 12829
-    },
-    {
-      "epoch": 0.838507287105418,
-      "grad_norm": 0.489314466714859,
-      "learning_rate": 8.367582518284692e-06,
-      "loss": 0.4495,
-      "step": 12830
-    },
-    {
-      "epoch": 0.8385726423109601,
-      "grad_norm": 0.4142928421497345,
-      "learning_rate": 8.367324398532076e-06,
-      "loss": 0.3103,
-      "step": 12831
-    },
-    {
-      "epoch": 0.8386379975165021,
-      "grad_norm": 0.43661683797836304,
-      "learning_rate": 8.36706626235585e-06,
-      "loss": 0.3608,
-      "step": 12832
-    },
-    {
-      "epoch": 0.8387033527220443,
-      "grad_norm": 0.45844224095344543,
-      "learning_rate": 8.366808109757279e-06,
-      "loss": 0.3807,
-      "step": 12833
-    },
-    {
-      "epoch": 0.8387687079275864,
-      "grad_norm": 0.4577506482601166,
-      "learning_rate": 8.366549940737615e-06,
-      "loss": 0.3649,
-      "step": 12834
-    },
-    {
-      "epoch": 0.8388340631331286,
-      "grad_norm": 0.5160108804702759,
-      "learning_rate": 8.366291755298122e-06,
-      "loss": 0.4036,
-      "step": 12835
-    },
-    {
-      "epoch": 0.8388994183386707,
-      "grad_norm": 0.4346136748790741,
-      "learning_rate": 8.366033553440058e-06,
-      "loss": 0.3832,
-      "step": 12836
-    },
-    {
-      "epoch": 0.8389647735442128,
-      "grad_norm": 0.4836990535259247,
-      "learning_rate": 8.365775335164683e-06,
-      "loss": 0.4562,
-      "step": 12837
-    },
-    {
-      "epoch": 0.8390301287497549,
-      "grad_norm": 0.4245145916938782,
-      "learning_rate": 8.365517100473255e-06,
-      "loss": 0.3644,
-      "step": 12838
-    },
-    {
-      "epoch": 0.8390954839552971,
-      "grad_norm": 0.43731689453125,
-      "learning_rate": 8.365258849367034e-06,
-      "loss": 0.3737,
-      "step": 12839
-    },
-    {
-      "epoch": 0.8391608391608392,
-      "grad_norm": 0.47260788083076477,
-      "learning_rate": 8.365000581847281e-06,
-      "loss": 0.4177,
-      "step": 12840
-    },
-    {
-      "epoch": 0.8392261943663812,
-      "grad_norm": 0.4614366590976715,
-      "learning_rate": 8.364742297915251e-06,
-      "loss": 0.4336,
-      "step": 12841
-    },
-    {
-      "epoch": 0.8392915495719234,
-      "grad_norm": 0.5142934918403625,
-      "learning_rate": 8.364483997572211e-06,
-      "loss": 0.4161,
-      "step": 12842
-    },
-    {
-      "epoch": 0.8393569047774655,
-      "grad_norm": 0.43350541591644287,
-      "learning_rate": 8.364225680819415e-06,
-      "loss": 0.3439,
-      "step": 12843
-    },
-    {
-      "epoch": 0.8394222599830077,
-      "grad_norm": 0.4446878433227539,
-      "learning_rate": 8.363967347658124e-06,
-      "loss": 0.3702,
-      "step": 12844
-    },
-    {
-      "epoch": 0.8394876151885498,
-      "grad_norm": 0.43523842096328735,
-      "learning_rate": 8.3637089980896e-06,
-      "loss": 0.3728,
-      "step": 12845
-    },
-    {
-      "epoch": 0.8395529703940919,
-      "grad_norm": 0.4645645022392273,
-      "learning_rate": 8.363450632115103e-06,
-      "loss": 0.3806,
-      "step": 12846
-    },
-    {
-      "epoch": 0.839618325599634,
-      "grad_norm": 0.42271658778190613,
-      "learning_rate": 8.363192249735892e-06,
-      "loss": 0.3416,
-      "step": 12847
-    },
-    {
-      "epoch": 0.8396836808051761,
-      "grad_norm": 0.4454825222492218,
-      "learning_rate": 8.362933850953227e-06,
-      "loss": 0.3334,
-      "step": 12848
-    },
-    {
-      "epoch": 0.8397490360107183,
-      "grad_norm": 0.42654910683631897,
-      "learning_rate": 8.362675435768369e-06,
-      "loss": 0.34,
-      "step": 12849
-    },
-    {
-      "epoch": 0.8398143912162603,
-      "grad_norm": 0.45451635122299194,
-      "learning_rate": 8.362417004182575e-06,
-      "loss": 0.3877,
-      "step": 12850
-    },
-    {
-      "epoch": 0.8398797464218025,
-      "grad_norm": 0.4170895218849182,
-      "learning_rate": 8.362158556197112e-06,
-      "loss": 0.3222,
-      "step": 12851
-    },
-    {
-      "epoch": 0.8399451016273446,
-      "grad_norm": 0.4565311074256897,
-      "learning_rate": 8.361900091813234e-06,
-      "loss": 0.3909,
-      "step": 12852
-    },
-    {
-      "epoch": 0.8400104568328868,
-      "grad_norm": 0.4371476173400879,
-      "learning_rate": 8.361641611032206e-06,
-      "loss": 0.3945,
-      "step": 12853
-    },
-    {
-      "epoch": 0.8400758120384288,
-      "grad_norm": 0.44827279448509216,
-      "learning_rate": 8.361383113855287e-06,
-      "loss": 0.413,
-      "step": 12854
-    },
-    {
-      "epoch": 0.840141167243971,
-      "grad_norm": 0.4154590666294098,
-      "learning_rate": 8.361124600283738e-06,
-      "loss": 0.3703,
-      "step": 12855
-    },
-    {
-      "epoch": 0.8402065224495131,
-      "grad_norm": 0.44536155462265015,
-      "learning_rate": 8.36086607031882e-06,
-      "loss": 0.4032,
-      "step": 12856
-    },
-    {
-      "epoch": 0.8402718776550552,
-      "grad_norm": 0.44725316762924194,
-      "learning_rate": 8.360607523961794e-06,
-      "loss": 0.4169,
-      "step": 12857
-    },
-    {
-      "epoch": 0.8403372328605974,
-      "grad_norm": 0.43830201029777527,
-      "learning_rate": 8.360348961213922e-06,
-      "loss": 0.3712,
-      "step": 12858
-    },
-    {
-      "epoch": 0.8404025880661394,
-      "grad_norm": 0.4728457033634186,
-      "learning_rate": 8.360090382076462e-06,
-      "loss": 0.3956,
-      "step": 12859
-    },
-    {
-      "epoch": 0.8404679432716816,
-      "grad_norm": 0.40700605511665344,
-      "learning_rate": 8.359831786550679e-06,
-      "loss": 0.3399,
-      "step": 12860
-    },
-    {
-      "epoch": 0.8405332984772237,
-      "grad_norm": 0.4341380000114441,
-      "learning_rate": 8.35957317463783e-06,
-      "loss": 0.4009,
-      "step": 12861
-    },
-    {
-      "epoch": 0.8405986536827659,
-      "grad_norm": 0.4405883252620697,
-      "learning_rate": 8.359314546339181e-06,
-      "loss": 0.3622,
-      "step": 12862
-    },
-    {
-      "epoch": 0.840664008888308,
-      "grad_norm": 0.41906869411468506,
-      "learning_rate": 8.35905590165599e-06,
-      "loss": 0.347,
-      "step": 12863
-    },
-    {
-      "epoch": 0.8407293640938501,
-      "grad_norm": 0.44313690066337585,
-      "learning_rate": 8.35879724058952e-06,
-      "loss": 0.3953,
-      "step": 12864
-    },
-    {
-      "epoch": 0.8407947192993922,
-      "grad_norm": 0.41211745142936707,
-      "learning_rate": 8.358538563141033e-06,
-      "loss": 0.3221,
-      "step": 12865
-    },
-    {
-      "epoch": 0.8408600745049343,
-      "grad_norm": 0.4759785830974579,
-      "learning_rate": 8.358279869311788e-06,
-      "loss": 0.3467,
-      "step": 12866
-    },
-    {
-      "epoch": 0.8409254297104765,
-      "grad_norm": 0.4384678602218628,
-      "learning_rate": 8.35802115910305e-06,
-      "loss": 0.4046,
-      "step": 12867
-    },
-    {
-      "epoch": 0.8409907849160185,
-      "grad_norm": 0.4619692265987396,
-      "learning_rate": 8.357762432516081e-06,
-      "loss": 0.3903,
-      "step": 12868
-    },
-    {
-      "epoch": 0.8410561401215607,
-      "grad_norm": 0.4471484422683716,
-      "learning_rate": 8.35750368955214e-06,
-      "loss": 0.416,
-      "step": 12869
-    },
-    {
-      "epoch": 0.8411214953271028,
-      "grad_norm": 0.463309109210968,
-      "learning_rate": 8.35724493021249e-06,
-      "loss": 0.3772,
-      "step": 12870
-    },
-    {
-      "epoch": 0.841186850532645,
-      "grad_norm": 0.4475812613964081,
-      "learning_rate": 8.356986154498393e-06,
-      "loss": 0.3611,
-      "step": 12871
-    },
-    {
-      "epoch": 0.841252205738187,
-      "grad_norm": 0.41159477829933167,
-      "learning_rate": 8.356727362411112e-06,
-      "loss": 0.3442,
-      "step": 12872
-    },
-    {
-      "epoch": 0.8413175609437291,
-      "grad_norm": 0.4358052909374237,
-      "learning_rate": 8.356468553951908e-06,
-      "loss": 0.3527,
-      "step": 12873
-    },
-    {
-      "epoch": 0.8413829161492713,
-      "grad_norm": 0.4975724220275879,
-      "learning_rate": 8.356209729122045e-06,
-      "loss": 0.4439,
-      "step": 12874
-    },
-    {
-      "epoch": 0.8414482713548134,
-      "grad_norm": 0.41729211807250977,
-      "learning_rate": 8.355950887922786e-06,
-      "loss": 0.3382,
-      "step": 12875
-    },
-    {
-      "epoch": 0.8415136265603556,
-      "grad_norm": 0.4276593029499054,
-      "learning_rate": 8.35569203035539e-06,
-      "loss": 0.3649,
-      "step": 12876
-    },
-    {
-      "epoch": 0.8415789817658976,
-      "grad_norm": 0.4574490785598755,
-      "learning_rate": 8.35543315642112e-06,
-      "loss": 0.38,
-      "step": 12877
-    },
-    {
-      "epoch": 0.8416443369714398,
-      "grad_norm": 0.47274157404899597,
-      "learning_rate": 8.355174266121241e-06,
-      "loss": 0.4263,
-      "step": 12878
-    },
-    {
-      "epoch": 0.8417096921769819,
-      "grad_norm": 0.42799681425094604,
-      "learning_rate": 8.354915359457016e-06,
-      "loss": 0.3276,
-      "step": 12879
-    },
-    {
-      "epoch": 0.8417750473825241,
-      "grad_norm": 0.43358737230300903,
-      "learning_rate": 8.354656436429707e-06,
-      "loss": 0.3579,
-      "step": 12880
-    },
-    {
-      "epoch": 0.8418404025880661,
-      "grad_norm": 0.4326688051223755,
-      "learning_rate": 8.354397497040576e-06,
-      "loss": 0.3788,
-      "step": 12881
-    },
-    {
-      "epoch": 0.8419057577936082,
-      "grad_norm": 0.4247555732727051,
-      "learning_rate": 8.354138541290885e-06,
-      "loss": 0.3503,
-      "step": 12882
-    },
-    {
-      "epoch": 0.8419711129991504,
-      "grad_norm": 0.40687647461891174,
-      "learning_rate": 8.353879569181899e-06,
-      "loss": 0.3005,
-      "step": 12883
-    },
-    {
-      "epoch": 0.8420364682046925,
-      "grad_norm": 0.44955679774284363,
-      "learning_rate": 8.353620580714881e-06,
-      "loss": 0.3785,
-      "step": 12884
-    },
-    {
-      "epoch": 0.8421018234102347,
-      "grad_norm": 0.4818287789821625,
-      "learning_rate": 8.353361575891094e-06,
-      "loss": 0.3912,
-      "step": 12885
-    },
-    {
-      "epoch": 0.8421671786157767,
-      "grad_norm": 0.464008629322052,
-      "learning_rate": 8.3531025547118e-06,
-      "loss": 0.3708,
-      "step": 12886
-    },
-    {
-      "epoch": 0.8422325338213189,
-      "grad_norm": 0.4107102155685425,
-      "learning_rate": 8.352843517178262e-06,
-      "loss": 0.352,
-      "step": 12887
-    },
-    {
-      "epoch": 0.842297889026861,
-      "grad_norm": 0.45311489701271057,
-      "learning_rate": 8.352584463291746e-06,
-      "loss": 0.3707,
-      "step": 12888
-    },
-    {
-      "epoch": 0.8423632442324032,
-      "grad_norm": 0.4310702979564667,
-      "learning_rate": 8.352325393053516e-06,
-      "loss": 0.3502,
-      "step": 12889
-    },
-    {
-      "epoch": 0.8424285994379452,
-      "grad_norm": 0.4197126030921936,
-      "learning_rate": 8.352066306464831e-06,
-      "loss": 0.373,
-      "step": 12890
-    },
-    {
-      "epoch": 0.8424939546434873,
-      "grad_norm": 0.4376210868358612,
-      "learning_rate": 8.351807203526958e-06,
-      "loss": 0.3484,
-      "step": 12891
-    },
-    {
-      "epoch": 0.8425593098490295,
-      "grad_norm": 0.4755789041519165,
-      "learning_rate": 8.35154808424116e-06,
-      "loss": 0.4205,
-      "step": 12892
-    },
-    {
-      "epoch": 0.8426246650545716,
-      "grad_norm": 0.4546374976634979,
-      "learning_rate": 8.351288948608701e-06,
-      "loss": 0.3911,
-      "step": 12893
-    },
-    {
-      "epoch": 0.8426900202601137,
-      "grad_norm": 0.47475185990333557,
-      "learning_rate": 8.351029796630846e-06,
-      "loss": 0.4237,
-      "step": 12894
-    },
-    {
-      "epoch": 0.8427553754656558,
-      "grad_norm": 0.495795875787735,
-      "learning_rate": 8.350770628308857e-06,
-      "loss": 0.4562,
-      "step": 12895
-    },
-    {
-      "epoch": 0.842820730671198,
-      "grad_norm": 0.43273425102233887,
-      "learning_rate": 8.350511443643998e-06,
-      "loss": 0.3165,
-      "step": 12896
-    },
-    {
-      "epoch": 0.8428860858767401,
-      "grad_norm": 0.41151395440101624,
-      "learning_rate": 8.350252242637533e-06,
-      "loss": 0.311,
-      "step": 12897
-    },
-    {
-      "epoch": 0.8429514410822821,
-      "grad_norm": 0.4324301481246948,
-      "learning_rate": 8.34999302529073e-06,
-      "loss": 0.3442,
-      "step": 12898
-    },
-    {
-      "epoch": 0.8430167962878243,
-      "grad_norm": 0.4583076536655426,
-      "learning_rate": 8.349733791604849e-06,
-      "loss": 0.4233,
-      "step": 12899
-    },
-    {
-      "epoch": 0.8430821514933664,
-      "grad_norm": 0.45870479941368103,
-      "learning_rate": 8.349474541581155e-06,
-      "loss": 0.4224,
-      "step": 12900
-    },
-    {
-      "epoch": 0.8431475066989086,
-      "grad_norm": 0.4472617208957672,
-      "learning_rate": 8.349215275220914e-06,
-      "loss": 0.4033,
-      "step": 12901
-    },
-    {
-      "epoch": 0.8432128619044507,
-      "grad_norm": 0.4329296946525574,
-      "learning_rate": 8.348955992525392e-06,
-      "loss": 0.3575,
-      "step": 12902
-    },
-    {
-      "epoch": 0.8432782171099928,
-      "grad_norm": 0.48093709349632263,
-      "learning_rate": 8.348696693495848e-06,
-      "loss": 0.4019,
-      "step": 12903
-    },
-    {
-      "epoch": 0.8433435723155349,
-      "grad_norm": 0.458065390586853,
-      "learning_rate": 8.348437378133552e-06,
-      "loss": 0.4211,
-      "step": 12904
-    },
-    {
-      "epoch": 0.8434089275210771,
-      "grad_norm": 0.447160542011261,
-      "learning_rate": 8.348178046439766e-06,
-      "loss": 0.3533,
-      "step": 12905
-    },
-    {
-      "epoch": 0.8434742827266192,
-      "grad_norm": 0.5594804286956787,
-      "learning_rate": 8.347918698415756e-06,
-      "loss": 0.4342,
-      "step": 12906
-    },
-    {
-      "epoch": 0.8435396379321612,
-      "grad_norm": 0.45463624596595764,
-      "learning_rate": 8.347659334062787e-06,
-      "loss": 0.3642,
-      "step": 12907
-    },
-    {
-      "epoch": 0.8436049931377034,
-      "grad_norm": 0.4449283182621002,
-      "learning_rate": 8.347399953382125e-06,
-      "loss": 0.3833,
-      "step": 12908
-    },
-    {
-      "epoch": 0.8436703483432455,
-      "grad_norm": 0.4590797424316406,
-      "learning_rate": 8.347140556375031e-06,
-      "loss": 0.3704,
-      "step": 12909
-    },
-    {
-      "epoch": 0.8437357035487877,
-      "grad_norm": 0.46438759565353394,
-      "learning_rate": 8.346881143042775e-06,
-      "loss": 0.3612,
-      "step": 12910
-    },
-    {
-      "epoch": 0.8438010587543298,
-      "grad_norm": 0.4726444482803345,
-      "learning_rate": 8.34662171338662e-06,
-      "loss": 0.4013,
-      "step": 12911
-    },
-    {
-      "epoch": 0.8438664139598719,
-      "grad_norm": 0.41788357496261597,
-      "learning_rate": 8.34636226740783e-06,
-      "loss": 0.3407,
-      "step": 12912
-    },
-    {
-      "epoch": 0.843931769165414,
-      "grad_norm": 0.45723414421081543,
-      "learning_rate": 8.346102805107674e-06,
-      "loss": 0.3949,
-      "step": 12913
-    },
-    {
-      "epoch": 0.8439971243709562,
-      "grad_norm": 0.44049227237701416,
-      "learning_rate": 8.345843326487415e-06,
-      "loss": 0.3829,
-      "step": 12914
-    },
-    {
-      "epoch": 0.8440624795764983,
-      "grad_norm": 0.4747612774372101,
-      "learning_rate": 8.345583831548318e-06,
-      "loss": 0.3722,
-      "step": 12915
-    },
-    {
-      "epoch": 0.8441278347820403,
-      "grad_norm": 0.44940876960754395,
-      "learning_rate": 8.34532432029165e-06,
-      "loss": 0.4181,
-      "step": 12916
-    },
-    {
-      "epoch": 0.8441931899875825,
-      "grad_norm": 0.4739570915699005,
-      "learning_rate": 8.345064792718676e-06,
-      "loss": 0.4567,
-      "step": 12917
-    },
-    {
-      "epoch": 0.8442585451931246,
-      "grad_norm": 0.4361695945262909,
-      "learning_rate": 8.344805248830664e-06,
-      "loss": 0.3475,
-      "step": 12918
-    },
-    {
-      "epoch": 0.8443239003986668,
-      "grad_norm": 0.4131276309490204,
-      "learning_rate": 8.344545688628876e-06,
-      "loss": 0.3658,
-      "step": 12919
-    },
-    {
-      "epoch": 0.8443892556042089,
-      "grad_norm": 0.45152372121810913,
-      "learning_rate": 8.344286112114581e-06,
-      "loss": 0.3904,
-      "step": 12920
-    },
-    {
-      "epoch": 0.844454610809751,
-      "grad_norm": 0.43002691864967346,
-      "learning_rate": 8.344026519289043e-06,
-      "loss": 0.3846,
-      "step": 12921
-    },
-    {
-      "epoch": 0.8445199660152931,
-      "grad_norm": 0.4692656397819519,
-      "learning_rate": 8.34376691015353e-06,
-      "loss": 0.4316,
-      "step": 12922
-    },
-    {
-      "epoch": 0.8445853212208353,
-      "grad_norm": 0.4222685694694519,
-      "learning_rate": 8.343507284709307e-06,
-      "loss": 0.346,
-      "step": 12923
-    },
-    {
-      "epoch": 0.8446506764263774,
-      "grad_norm": 0.4289930462837219,
-      "learning_rate": 8.343247642957642e-06,
-      "loss": 0.3635,
-      "step": 12924
-    },
-    {
-      "epoch": 0.8447160316319194,
-      "grad_norm": 0.4378208816051483,
-      "learning_rate": 8.342987984899798e-06,
-      "loss": 0.3636,
-      "step": 12925
-    },
-    {
-      "epoch": 0.8447813868374616,
-      "grad_norm": 0.4417200982570648,
-      "learning_rate": 8.342728310537044e-06,
-      "loss": 0.3965,
-      "step": 12926
-    },
-    {
-      "epoch": 0.8448467420430037,
-      "grad_norm": 0.4177982211112976,
-      "learning_rate": 8.342468619870646e-06,
-      "loss": 0.3686,
-      "step": 12927
-    },
-    {
-      "epoch": 0.8449120972485459,
-      "grad_norm": 0.4426078498363495,
-      "learning_rate": 8.342208912901873e-06,
-      "loss": 0.3696,
-      "step": 12928
-    },
-    {
-      "epoch": 0.844977452454088,
-      "grad_norm": 0.43379274010658264,
-      "learning_rate": 8.341949189631986e-06,
-      "loss": 0.3437,
-      "step": 12929
-    },
-    {
-      "epoch": 0.8450428076596301,
-      "grad_norm": 0.48187845945358276,
-      "learning_rate": 8.341689450062258e-06,
-      "loss": 0.4527,
-      "step": 12930
-    },
-    {
-      "epoch": 0.8451081628651722,
-      "grad_norm": 0.45894575119018555,
-      "learning_rate": 8.34142969419395e-06,
-      "loss": 0.381,
-      "step": 12931
-    },
-    {
-      "epoch": 0.8451735180707143,
-      "grad_norm": 0.4325745701789856,
-      "learning_rate": 8.341169922028334e-06,
-      "loss": 0.3602,
-      "step": 12932
-    },
-    {
-      "epoch": 0.8452388732762565,
-      "grad_norm": 0.4661005437374115,
-      "learning_rate": 8.340910133566673e-06,
-      "loss": 0.4117,
-      "step": 12933
-    },
-    {
-      "epoch": 0.8453042284817985,
-      "grad_norm": 0.46226686239242554,
-      "learning_rate": 8.340650328810238e-06,
-      "loss": 0.3866,
-      "step": 12934
-    },
-    {
-      "epoch": 0.8453695836873407,
-      "grad_norm": 0.47056636214256287,
-      "learning_rate": 8.340390507760292e-06,
-      "loss": 0.4109,
-      "step": 12935
-    },
-    {
-      "epoch": 0.8454349388928828,
-      "grad_norm": 0.42317309975624084,
-      "learning_rate": 8.340130670418104e-06,
-      "loss": 0.3643,
-      "step": 12936
-    },
-    {
-      "epoch": 0.845500294098425,
-      "grad_norm": 0.48881959915161133,
-      "learning_rate": 8.339870816784942e-06,
-      "loss": 0.447,
-      "step": 12937
-    },
-    {
-      "epoch": 0.845565649303967,
-      "grad_norm": 0.43261072039604187,
-      "learning_rate": 8.339610946862075e-06,
-      "loss": 0.3463,
-      "step": 12938
-    },
-    {
-      "epoch": 0.8456310045095092,
-      "grad_norm": 0.479663610458374,
-      "learning_rate": 8.339351060650767e-06,
-      "loss": 0.4781,
-      "step": 12939
-    },
-    {
-      "epoch": 0.8456963597150513,
-      "grad_norm": 0.43825334310531616,
-      "learning_rate": 8.339091158152288e-06,
-      "loss": 0.3867,
-      "step": 12940
-    },
-    {
-      "epoch": 0.8457617149205934,
-      "grad_norm": 0.43406298756599426,
-      "learning_rate": 8.338831239367903e-06,
-      "loss": 0.3708,
-      "step": 12941
-    },
-    {
-      "epoch": 0.8458270701261356,
-      "grad_norm": 0.4622132480144501,
-      "learning_rate": 8.33857130429888e-06,
-      "loss": 0.4289,
-      "step": 12942
-    },
-    {
-      "epoch": 0.8458924253316776,
-      "grad_norm": 0.42942553758621216,
-      "learning_rate": 8.338311352946492e-06,
-      "loss": 0.3265,
-      "step": 12943
-    },
-    {
-      "epoch": 0.8459577805372198,
-      "grad_norm": 0.45094743371009827,
-      "learning_rate": 8.338051385312001e-06,
-      "loss": 0.3818,
-      "step": 12944
-    },
-    {
-      "epoch": 0.8460231357427619,
-      "grad_norm": 0.4672107398509979,
-      "learning_rate": 8.337791401396678e-06,
-      "loss": 0.4254,
-      "step": 12945
-    },
-    {
-      "epoch": 0.8460884909483041,
-      "grad_norm": 0.43508535623550415,
-      "learning_rate": 8.337531401201788e-06,
-      "loss": 0.3795,
-      "step": 12946
-    },
-    {
-      "epoch": 0.8461538461538461,
-      "grad_norm": 0.46768462657928467,
-      "learning_rate": 8.337271384728602e-06,
-      "loss": 0.4143,
-      "step": 12947
-    },
-    {
-      "epoch": 0.8462192013593883,
-      "grad_norm": 0.4383191764354706,
-      "learning_rate": 8.337011351978388e-06,
-      "loss": 0.3866,
-      "step": 12948
-    },
-    {
-      "epoch": 0.8462845565649304,
-      "grad_norm": 0.47301405668258667,
-      "learning_rate": 8.336751302952413e-06,
-      "loss": 0.3686,
-      "step": 12949
-    },
-    {
-      "epoch": 0.8463499117704725,
-      "grad_norm": 0.4595911502838135,
-      "learning_rate": 8.336491237651947e-06,
-      "loss": 0.4079,
-      "step": 12950
-    },
-    {
-      "epoch": 0.8464152669760147,
-      "grad_norm": 0.43553805351257324,
-      "learning_rate": 8.336231156078256e-06,
-      "loss": 0.3164,
-      "step": 12951
-    },
-    {
-      "epoch": 0.8464806221815567,
-      "grad_norm": 0.45041853189468384,
-      "learning_rate": 8.335971058232612e-06,
-      "loss": 0.3903,
-      "step": 12952
-    },
-    {
-      "epoch": 0.8465459773870989,
-      "grad_norm": 0.4387279152870178,
-      "learning_rate": 8.33571094411628e-06,
-      "loss": 0.3634,
-      "step": 12953
-    },
-    {
-      "epoch": 0.846611332592641,
-      "grad_norm": 0.4489176273345947,
-      "learning_rate": 8.33545081373053e-06,
-      "loss": 0.3724,
-      "step": 12954
-    },
-    {
-      "epoch": 0.8466766877981832,
-      "grad_norm": 0.3960930109024048,
-      "learning_rate": 8.33519066707663e-06,
-      "loss": 0.3207,
-      "step": 12955
-    },
-    {
-      "epoch": 0.8467420430037252,
-      "grad_norm": 0.47986871004104614,
-      "learning_rate": 8.33493050415585e-06,
-      "loss": 0.4222,
-      "step": 12956
-    },
-    {
-      "epoch": 0.8468073982092673,
-      "grad_norm": 0.42395836114883423,
-      "learning_rate": 8.33467032496946e-06,
-      "loss": 0.3493,
-      "step": 12957
-    },
-    {
-      "epoch": 0.8468727534148095,
-      "grad_norm": 0.43092361092567444,
-      "learning_rate": 8.334410129518726e-06,
-      "loss": 0.3499,
-      "step": 12958
-    },
-    {
-      "epoch": 0.8469381086203516,
-      "grad_norm": 0.47599807381629944,
-      "learning_rate": 8.334149917804921e-06,
-      "loss": 0.3793,
-      "step": 12959
-    },
-    {
-      "epoch": 0.8470034638258938,
-      "grad_norm": 0.6613618731498718,
-      "learning_rate": 8.33388968982931e-06,
-      "loss": 0.3604,
-      "step": 12960
-    },
-    {
-      "epoch": 0.8470688190314358,
-      "grad_norm": 0.46646255254745483,
-      "learning_rate": 8.333629445593165e-06,
-      "loss": 0.3681,
-      "step": 12961
-    },
-    {
-      "epoch": 0.847134174236978,
-      "grad_norm": 0.3952161371707916,
-      "learning_rate": 8.333369185097752e-06,
-      "loss": 0.3101,
-      "step": 12962
-    },
-    {
-      "epoch": 0.8471995294425201,
-      "grad_norm": 0.44239893555641174,
-      "learning_rate": 8.333108908344345e-06,
-      "loss": 0.4,
-      "step": 12963
-    },
-    {
-      "epoch": 0.8472648846480623,
-      "grad_norm": 0.4682541489601135,
-      "learning_rate": 8.33284861533421e-06,
-      "loss": 0.3764,
-      "step": 12964
-    },
-    {
-      "epoch": 0.8473302398536043,
-      "grad_norm": 0.43617960810661316,
-      "learning_rate": 8.33258830606862e-06,
-      "loss": 0.3537,
-      "step": 12965
-    },
-    {
-      "epoch": 0.8473955950591464,
-      "grad_norm": 0.4275916814804077,
-      "learning_rate": 8.332327980548838e-06,
-      "loss": 0.3557,
-      "step": 12966
-    },
-    {
-      "epoch": 0.8474609502646886,
-      "grad_norm": 0.5054025650024414,
-      "learning_rate": 8.33206763877614e-06,
-      "loss": 0.3797,
-      "step": 12967
-    },
-    {
-      "epoch": 0.8475263054702307,
-      "grad_norm": 0.4242538809776306,
-      "learning_rate": 8.331807280751796e-06,
-      "loss": 0.3303,
-      "step": 12968
-    },
-    {
-      "epoch": 0.8475916606757729,
-      "grad_norm": 0.45625555515289307,
-      "learning_rate": 8.33154690647707e-06,
-      "loss": 0.3952,
-      "step": 12969
-    },
-    {
-      "epoch": 0.8476570158813149,
-      "grad_norm": 0.4328351318836212,
-      "learning_rate": 8.331286515953238e-06,
-      "loss": 0.3684,
-      "step": 12970
-    },
-    {
-      "epoch": 0.8477223710868571,
-      "grad_norm": 0.45689383149147034,
-      "learning_rate": 8.331026109181568e-06,
-      "loss": 0.3839,
-      "step": 12971
-    },
-    {
-      "epoch": 0.8477877262923992,
-      "grad_norm": 0.44383302330970764,
-      "learning_rate": 8.330765686163328e-06,
-      "loss": 0.3954,
-      "step": 12972
-    },
-    {
-      "epoch": 0.8478530814979414,
-      "grad_norm": 0.44091150164604187,
-      "learning_rate": 8.330505246899792e-06,
-      "loss": 0.3801,
-      "step": 12973
-    },
-    {
-      "epoch": 0.8479184367034834,
-      "grad_norm": 0.47562843561172485,
-      "learning_rate": 8.330244791392226e-06,
-      "loss": 0.4355,
-      "step": 12974
-    },
-    {
-      "epoch": 0.8479837919090255,
-      "grad_norm": 0.45469504594802856,
-      "learning_rate": 8.329984319641902e-06,
-      "loss": 0.343,
-      "step": 12975
-    },
-    {
-      "epoch": 0.8480491471145677,
-      "grad_norm": 0.4577144384384155,
-      "learning_rate": 8.329723831650092e-06,
-      "loss": 0.3717,
-      "step": 12976
-    },
-    {
-      "epoch": 0.8481145023201098,
-      "grad_norm": 0.437324196100235,
-      "learning_rate": 8.329463327418066e-06,
-      "loss": 0.3834,
-      "step": 12977
-    },
-    {
-      "epoch": 0.848179857525652,
-      "grad_norm": 0.4720422327518463,
-      "learning_rate": 8.329202806947093e-06,
-      "loss": 0.4333,
-      "step": 12978
-    },
-    {
-      "epoch": 0.848245212731194,
-      "grad_norm": 0.4393203556537628,
-      "learning_rate": 8.328942270238444e-06,
-      "loss": 0.3896,
-      "step": 12979
-    },
-    {
-      "epoch": 0.8483105679367362,
-      "grad_norm": 0.4401390254497528,
-      "learning_rate": 8.328681717293392e-06,
-      "loss": 0.3921,
-      "step": 12980
-    },
-    {
-      "epoch": 0.8483759231422783,
-      "grad_norm": 0.43827110528945923,
-      "learning_rate": 8.328421148113207e-06,
-      "loss": 0.3831,
-      "step": 12981
-    },
-    {
-      "epoch": 0.8484412783478203,
-      "grad_norm": 0.4346137046813965,
-      "learning_rate": 8.328160562699155e-06,
-      "loss": 0.3519,
-      "step": 12982
-    },
-    {
-      "epoch": 0.8485066335533625,
-      "grad_norm": 0.4775579571723938,
-      "learning_rate": 8.327899961052514e-06,
-      "loss": 0.4441,
-      "step": 12983
-    },
-    {
-      "epoch": 0.8485719887589046,
-      "grad_norm": 0.4290394186973572,
-      "learning_rate": 8.327639343174551e-06,
-      "loss": 0.3536,
-      "step": 12984
-    },
-    {
-      "epoch": 0.8486373439644468,
-      "grad_norm": 0.41752007603645325,
-      "learning_rate": 8.327378709066538e-06,
-      "loss": 0.3457,
-      "step": 12985
-    },
-    {
-      "epoch": 0.8487026991699889,
-      "grad_norm": 0.4521631896495819,
-      "learning_rate": 8.327118058729745e-06,
-      "loss": 0.4116,
-      "step": 12986
-    },
-    {
-      "epoch": 0.848768054375531,
-      "grad_norm": 0.45426443219184875,
-      "learning_rate": 8.326857392165449e-06,
-      "loss": 0.4414,
-      "step": 12987
-    },
-    {
-      "epoch": 0.8488334095810731,
-      "grad_norm": 0.43946558237075806,
-      "learning_rate": 8.326596709374913e-06,
-      "loss": 0.3719,
-      "step": 12988
-    },
-    {
-      "epoch": 0.8488987647866153,
-      "grad_norm": 0.42910999059677124,
-      "learning_rate": 8.326336010359413e-06,
-      "loss": 0.3541,
-      "step": 12989
-    },
-    {
-      "epoch": 0.8489641199921574,
-      "grad_norm": 0.4289521872997284,
-      "learning_rate": 8.326075295120222e-06,
-      "loss": 0.3593,
-      "step": 12990
-    },
-    {
-      "epoch": 0.8490294751976994,
-      "grad_norm": 0.43595993518829346,
-      "learning_rate": 8.325814563658607e-06,
-      "loss": 0.3625,
-      "step": 12991
-    },
-    {
-      "epoch": 0.8490948304032416,
-      "grad_norm": 0.577158510684967,
-      "learning_rate": 8.325553815975842e-06,
-      "loss": 0.3838,
-      "step": 12992
-    },
-    {
-      "epoch": 0.8491601856087837,
-      "grad_norm": 0.4579011797904968,
-      "learning_rate": 8.325293052073201e-06,
-      "loss": 0.4171,
-      "step": 12993
-    },
-    {
-      "epoch": 0.8492255408143259,
-      "grad_norm": 0.44544026255607605,
-      "learning_rate": 8.325032271951954e-06,
-      "loss": 0.4117,
-      "step": 12994
-    },
-    {
-      "epoch": 0.849290896019868,
-      "grad_norm": 0.4383991062641144,
-      "learning_rate": 8.324771475613372e-06,
-      "loss": 0.3807,
-      "step": 12995
-    },
-    {
-      "epoch": 0.8493562512254101,
-      "grad_norm": 0.41317757964134216,
-      "learning_rate": 8.324510663058726e-06,
-      "loss": 0.3656,
-      "step": 12996
-    },
-    {
-      "epoch": 0.8494216064309522,
-      "grad_norm": 0.4217069149017334,
-      "learning_rate": 8.32424983428929e-06,
-      "loss": 0.3476,
-      "step": 12997
-    },
-    {
-      "epoch": 0.8494869616364944,
-      "grad_norm": 0.47838321328163147,
-      "learning_rate": 8.32398898930634e-06,
-      "loss": 0.3787,
-      "step": 12998
-    },
-    {
-      "epoch": 0.8495523168420365,
-      "grad_norm": 0.40194016695022583,
-      "learning_rate": 8.323728128111141e-06,
-      "loss": 0.3327,
-      "step": 12999
-    },
-    {
-      "epoch": 0.8496176720475785,
-      "grad_norm": 0.4378214478492737,
-      "learning_rate": 8.323467250704968e-06,
-      "loss": 0.38,
-      "step": 13000
-    },
-    {
-      "epoch": 0.8496830272531207,
-      "grad_norm": 0.40850594639778137,
-      "learning_rate": 8.323206357089094e-06,
-      "loss": 0.3331,
-      "step": 13001
-    },
-    {
-      "epoch": 0.8497483824586628,
-      "grad_norm": 0.4367838203907013,
-      "learning_rate": 8.322945447264792e-06,
-      "loss": 0.4046,
-      "step": 13002
-    },
-    {
-      "epoch": 0.849813737664205,
-      "grad_norm": 0.46336469054222107,
-      "learning_rate": 8.322684521233332e-06,
-      "loss": 0.3661,
-      "step": 13003
-    },
-    {
-      "epoch": 0.849879092869747,
-      "grad_norm": 0.4387741684913635,
-      "learning_rate": 8.322423578995991e-06,
-      "loss": 0.3521,
-      "step": 13004
-    },
-    {
-      "epoch": 0.8499444480752892,
-      "grad_norm": 0.4483397901058197,
-      "learning_rate": 8.32216262055404e-06,
-      "loss": 0.3372,
-      "step": 13005
-    },
-    {
-      "epoch": 0.8500098032808313,
-      "grad_norm": 0.47723543643951416,
-      "learning_rate": 8.321901645908748e-06,
-      "loss": 0.4397,
-      "step": 13006
-    },
-    {
-      "epoch": 0.8500751584863735,
-      "grad_norm": 0.42409658432006836,
-      "learning_rate": 8.321640655061394e-06,
-      "loss": 0.3949,
-      "step": 13007
-    },
-    {
-      "epoch": 0.8501405136919156,
-      "grad_norm": 0.4384118914604187,
-      "learning_rate": 8.321379648013246e-06,
-      "loss": 0.3864,
-      "step": 13008
-    },
-    {
-      "epoch": 0.8502058688974576,
-      "grad_norm": 0.44307631254196167,
-      "learning_rate": 8.321118624765578e-06,
-      "loss": 0.3899,
-      "step": 13009
-    },
-    {
-      "epoch": 0.8502712241029998,
-      "grad_norm": 0.41734904050827026,
-      "learning_rate": 8.320857585319664e-06,
-      "loss": 0.346,
-      "step": 13010
-    },
-    {
-      "epoch": 0.8503365793085419,
-      "grad_norm": 0.4242594540119171,
-      "learning_rate": 8.320596529676778e-06,
-      "loss": 0.3722,
-      "step": 13011
-    },
-    {
-      "epoch": 0.8504019345140841,
-      "grad_norm": 0.41269031167030334,
-      "learning_rate": 8.320335457838194e-06,
-      "loss": 0.3617,
-      "step": 13012
-    },
-    {
-      "epoch": 0.8504672897196262,
-      "grad_norm": 0.4592479467391968,
-      "learning_rate": 8.320074369805182e-06,
-      "loss": 0.3988,
-      "step": 13013
-    },
-    {
-      "epoch": 0.8505326449251683,
-      "grad_norm": 0.41280412673950195,
-      "learning_rate": 8.319813265579017e-06,
-      "loss": 0.3548,
-      "step": 13014
-    },
-    {
-      "epoch": 0.8505980001307104,
-      "grad_norm": 0.46887463331222534,
-      "learning_rate": 8.319552145160972e-06,
-      "loss": 0.3939,
-      "step": 13015
-    },
-    {
-      "epoch": 0.8506633553362525,
-      "grad_norm": 0.41253677010536194,
-      "learning_rate": 8.319291008552321e-06,
-      "loss": 0.3155,
-      "step": 13016
-    },
-    {
-      "epoch": 0.8507287105417947,
-      "grad_norm": 0.4332970082759857,
-      "learning_rate": 8.31902985575434e-06,
-      "loss": 0.3999,
-      "step": 13017
-    },
-    {
-      "epoch": 0.8507940657473367,
-      "grad_norm": 0.4350966215133667,
-      "learning_rate": 8.3187686867683e-06,
-      "loss": 0.4045,
-      "step": 13018
-    },
-    {
-      "epoch": 0.8508594209528789,
-      "grad_norm": 0.44339117407798767,
-      "learning_rate": 8.318507501595474e-06,
-      "loss": 0.3705,
-      "step": 13019
-    },
-    {
-      "epoch": 0.850924776158421,
-      "grad_norm": 0.4180876612663269,
-      "learning_rate": 8.318246300237139e-06,
-      "loss": 0.3499,
-      "step": 13020
-    },
-    {
-      "epoch": 0.8509901313639632,
-      "grad_norm": 0.4269477128982544,
-      "learning_rate": 8.317985082694566e-06,
-      "loss": 0.3773,
-      "step": 13021
-    },
-    {
-      "epoch": 0.8510554865695052,
-      "grad_norm": 0.4391435384750366,
-      "learning_rate": 8.317723848969029e-06,
-      "loss": 0.3592,
-      "step": 13022
-    },
-    {
-      "epoch": 0.8511208417750474,
-      "grad_norm": 0.4249681532382965,
-      "learning_rate": 8.317462599061805e-06,
-      "loss": 0.3507,
-      "step": 13023
-    },
-    {
-      "epoch": 0.8511861969805895,
-      "grad_norm": 0.41892850399017334,
-      "learning_rate": 8.317201332974167e-06,
-      "loss": 0.3491,
-      "step": 13024
-    },
-    {
-      "epoch": 0.8512515521861316,
-      "grad_norm": 0.42302414774894714,
-      "learning_rate": 8.31694005070739e-06,
-      "loss": 0.3122,
-      "step": 13025
-    },
-    {
-      "epoch": 0.8513169073916738,
-      "grad_norm": 0.5472805500030518,
-      "learning_rate": 8.316678752262743e-06,
-      "loss": 0.3863,
-      "step": 13026
-    },
-    {
-      "epoch": 0.8513822625972158,
-      "grad_norm": 0.43365710973739624,
-      "learning_rate": 8.316417437641509e-06,
-      "loss": 0.3487,
-      "step": 13027
-    },
-    {
-      "epoch": 0.851447617802758,
-      "grad_norm": 0.46539852023124695,
-      "learning_rate": 8.316156106844958e-06,
-      "loss": 0.4255,
-      "step": 13028
-    },
-    {
-      "epoch": 0.8515129730083001,
-      "grad_norm": 0.478466272354126,
-      "learning_rate": 8.315894759874361e-06,
-      "loss": 0.4772,
-      "step": 13029
-    },
-    {
-      "epoch": 0.8515783282138423,
-      "grad_norm": 0.45742014050483704,
-      "learning_rate": 8.315633396731e-06,
-      "loss": 0.4016,
-      "step": 13030
-    },
-    {
-      "epoch": 0.8516436834193843,
-      "grad_norm": 0.4491812288761139,
-      "learning_rate": 8.315372017416146e-06,
-      "loss": 0.3737,
-      "step": 13031
-    },
-    {
-      "epoch": 0.8517090386249265,
-      "grad_norm": 0.4290905296802521,
-      "learning_rate": 8.315110621931074e-06,
-      "loss": 0.352,
-      "step": 13032
-    },
-    {
-      "epoch": 0.8517743938304686,
-      "grad_norm": 0.46106651425361633,
-      "learning_rate": 8.314849210277057e-06,
-      "loss": 0.3846,
-      "step": 13033
-    },
-    {
-      "epoch": 0.8518397490360107,
-      "grad_norm": 0.5297082662582397,
-      "learning_rate": 8.314587782455373e-06,
-      "loss": 0.4311,
-      "step": 13034
-    },
-    {
-      "epoch": 0.8519051042415529,
-      "grad_norm": 0.42537423968315125,
-      "learning_rate": 8.314326338467297e-06,
-      "loss": 0.3878,
-      "step": 13035
-    },
-    {
-      "epoch": 0.8519704594470949,
-      "grad_norm": 0.4676016867160797,
-      "learning_rate": 8.314064878314103e-06,
-      "loss": 0.4051,
-      "step": 13036
-    },
-    {
-      "epoch": 0.8520358146526371,
-      "grad_norm": 0.44034287333488464,
-      "learning_rate": 8.313803401997068e-06,
-      "loss": 0.3257,
-      "step": 13037
-    },
-    {
-      "epoch": 0.8521011698581792,
-      "grad_norm": 0.712868869304657,
-      "learning_rate": 8.313541909517463e-06,
-      "loss": 0.3715,
-      "step": 13038
-    },
-    {
-      "epoch": 0.8521665250637214,
-      "grad_norm": 0.4472745358943939,
-      "learning_rate": 8.313280400876566e-06,
-      "loss": 0.3826,
-      "step": 13039
-    },
-    {
-      "epoch": 0.8522318802692634,
-      "grad_norm": 0.4714382588863373,
-      "learning_rate": 8.313018876075656e-06,
-      "loss": 0.4292,
-      "step": 13040
-    },
-    {
-      "epoch": 0.8522972354748055,
-      "grad_norm": 0.4770013689994812,
-      "learning_rate": 8.312757335116002e-06,
-      "loss": 0.4185,
-      "step": 13041
-    },
-    {
-      "epoch": 0.8523625906803477,
-      "grad_norm": 0.42746424674987793,
-      "learning_rate": 8.312495777998883e-06,
-      "loss": 0.3549,
-      "step": 13042
-    },
-    {
-      "epoch": 0.8524279458858898,
-      "grad_norm": 0.452098548412323,
-      "learning_rate": 8.312234204725576e-06,
-      "loss": 0.4103,
-      "step": 13043
-    },
-    {
-      "epoch": 0.852493301091432,
-      "grad_norm": 0.45813503861427307,
-      "learning_rate": 8.311972615297356e-06,
-      "loss": 0.4049,
-      "step": 13044
-    },
-    {
-      "epoch": 0.852558656296974,
-      "grad_norm": 0.4434608519077301,
-      "learning_rate": 8.311711009715497e-06,
-      "loss": 0.3839,
-      "step": 13045
-    },
-    {
-      "epoch": 0.8526240115025162,
-      "grad_norm": 0.4325890839099884,
-      "learning_rate": 8.311449387981277e-06,
-      "loss": 0.398,
-      "step": 13046
-    },
-    {
-      "epoch": 0.8526893667080583,
-      "grad_norm": 0.44462740421295166,
-      "learning_rate": 8.31118775009597e-06,
-      "loss": 0.3865,
-      "step": 13047
-    },
-    {
-      "epoch": 0.8527547219136005,
-      "grad_norm": 0.44148850440979004,
-      "learning_rate": 8.310926096060851e-06,
-      "loss": 0.3735,
-      "step": 13048
-    },
-    {
-      "epoch": 0.8528200771191425,
-      "grad_norm": 0.40927594900131226,
-      "learning_rate": 8.310664425877202e-06,
-      "loss": 0.353,
-      "step": 13049
-    },
-    {
-      "epoch": 0.8528854323246846,
-      "grad_norm": 0.4303451478481293,
-      "learning_rate": 8.310402739546296e-06,
-      "loss": 0.3371,
-      "step": 13050
-    },
-    {
-      "epoch": 0.8529507875302268,
-      "grad_norm": 0.4334465563297272,
-      "learning_rate": 8.310141037069405e-06,
-      "loss": 0.333,
-      "step": 13051
-    },
-    {
-      "epoch": 0.8530161427357689,
-      "grad_norm": 0.3735904097557068,
-      "learning_rate": 8.309879318447814e-06,
-      "loss": 0.266,
-      "step": 13052
-    },
-    {
-      "epoch": 0.853081497941311,
-      "grad_norm": 0.4237179458141327,
-      "learning_rate": 8.309617583682792e-06,
-      "loss": 0.3713,
-      "step": 13053
-    },
-    {
-      "epoch": 0.8531468531468531,
-      "grad_norm": 0.47589799761772156,
-      "learning_rate": 8.30935583277562e-06,
-      "loss": 0.4209,
-      "step": 13054
-    },
-    {
-      "epoch": 0.8532122083523953,
-      "grad_norm": 0.46652454137802124,
-      "learning_rate": 8.309094065727571e-06,
-      "loss": 0.3755,
-      "step": 13055
-    },
-    {
-      "epoch": 0.8532775635579374,
-      "grad_norm": 0.45973077416419983,
-      "learning_rate": 8.308832282539927e-06,
-      "loss": 0.3731,
-      "step": 13056
-    },
-    {
-      "epoch": 0.8533429187634796,
-      "grad_norm": 0.43615809082984924,
-      "learning_rate": 8.30857048321396e-06,
-      "loss": 0.3577,
-      "step": 13057
-    },
-    {
-      "epoch": 0.8534082739690216,
-      "grad_norm": 0.4429352283477783,
-      "learning_rate": 8.30830866775095e-06,
-      "loss": 0.3624,
-      "step": 13058
-    },
-    {
-      "epoch": 0.8534736291745637,
-      "grad_norm": 0.42994409799575806,
-      "learning_rate": 8.30804683615217e-06,
-      "loss": 0.3623,
-      "step": 13059
-    },
-    {
-      "epoch": 0.8535389843801059,
-      "grad_norm": 0.44007644057273865,
-      "learning_rate": 8.3077849884189e-06,
-      "loss": 0.3665,
-      "step": 13060
-    },
-    {
-      "epoch": 0.853604339585648,
-      "grad_norm": 0.4082667827606201,
-      "learning_rate": 8.30752312455242e-06,
-      "loss": 0.3338,
-      "step": 13061
-    },
-    {
-      "epoch": 0.8536696947911901,
-      "grad_norm": 0.44889017939567566,
-      "learning_rate": 8.307261244554e-06,
-      "loss": 0.3684,
-      "step": 13062
-    },
-    {
-      "epoch": 0.8537350499967322,
-      "grad_norm": 0.4447444975376129,
-      "learning_rate": 8.306999348424922e-06,
-      "loss": 0.3814,
-      "step": 13063
-    },
-    {
-      "epoch": 0.8538004052022744,
-      "grad_norm": 0.446433424949646,
-      "learning_rate": 8.306737436166463e-06,
-      "loss": 0.3823,
-      "step": 13064
-    },
-    {
-      "epoch": 0.8538657604078165,
-      "grad_norm": 0.4314079284667969,
-      "learning_rate": 8.306475507779902e-06,
-      "loss": 0.3573,
-      "step": 13065
-    },
-    {
-      "epoch": 0.8539311156133585,
-      "grad_norm": 0.4226812720298767,
-      "learning_rate": 8.30621356326651e-06,
-      "loss": 0.3445,
-      "step": 13066
-    },
-    {
-      "epoch": 0.8539964708189007,
-      "grad_norm": 0.45340994000434875,
-      "learning_rate": 8.305951602627573e-06,
-      "loss": 0.4014,
-      "step": 13067
-    },
-    {
-      "epoch": 0.8540618260244428,
-      "grad_norm": 0.4504512548446655,
-      "learning_rate": 8.305689625864361e-06,
-      "loss": 0.3886,
-      "step": 13068
-    },
-    {
-      "epoch": 0.854127181229985,
-      "grad_norm": 0.45493951439857483,
-      "learning_rate": 8.305427632978159e-06,
-      "loss": 0.3873,
-      "step": 13069
-    },
-    {
-      "epoch": 0.8541925364355271,
-      "grad_norm": 0.48128390312194824,
-      "learning_rate": 8.30516562397024e-06,
-      "loss": 0.4406,
-      "step": 13070
-    },
-    {
-      "epoch": 0.8542578916410692,
-      "grad_norm": 0.4014410972595215,
-      "learning_rate": 8.304903598841884e-06,
-      "loss": 0.3239,
-      "step": 13071
-    },
-    {
-      "epoch": 0.8543232468466113,
-      "grad_norm": 0.43254926800727844,
-      "learning_rate": 8.304641557594366e-06,
-      "loss": 0.3368,
-      "step": 13072
-    },
-    {
-      "epoch": 0.8543886020521535,
-      "grad_norm": 0.4835487902164459,
-      "learning_rate": 8.304379500228968e-06,
-      "loss": 0.3558,
-      "step": 13073
-    },
-    {
-      "epoch": 0.8544539572576956,
-      "grad_norm": 0.44744396209716797,
-      "learning_rate": 8.304117426746966e-06,
-      "loss": 0.3908,
-      "step": 13074
-    },
-    {
-      "epoch": 0.8545193124632376,
-      "grad_norm": 0.43773603439331055,
-      "learning_rate": 8.30385533714964e-06,
-      "loss": 0.3436,
-      "step": 13075
-    },
-    {
-      "epoch": 0.8545846676687798,
-      "grad_norm": 0.45900505781173706,
-      "learning_rate": 8.303593231438265e-06,
-      "loss": 0.4142,
-      "step": 13076
-    },
-    {
-      "epoch": 0.8546500228743219,
-      "grad_norm": 0.43883055448532104,
-      "learning_rate": 8.303331109614122e-06,
-      "loss": 0.3875,
-      "step": 13077
-    },
-    {
-      "epoch": 0.8547153780798641,
-      "grad_norm": 0.4569999873638153,
-      "learning_rate": 8.30306897167849e-06,
-      "loss": 0.3787,
-      "step": 13078
-    },
-    {
-      "epoch": 0.8547807332854062,
-      "grad_norm": 0.4314734935760498,
-      "learning_rate": 8.302806817632645e-06,
-      "loss": 0.3593,
-      "step": 13079
-    },
-    {
-      "epoch": 0.8548460884909483,
-      "grad_norm": 0.4136468172073364,
-      "learning_rate": 8.302544647477868e-06,
-      "loss": 0.355,
-      "step": 13080
-    },
-    {
-      "epoch": 0.8549114436964904,
-      "grad_norm": 0.48597288131713867,
-      "learning_rate": 8.302282461215436e-06,
-      "loss": 0.4367,
-      "step": 13081
-    },
-    {
-      "epoch": 0.8549767989020326,
-      "grad_norm": 0.41474828124046326,
-      "learning_rate": 8.302020258846629e-06,
-      "loss": 0.356,
-      "step": 13082
-    },
-    {
-      "epoch": 0.8550421541075747,
-      "grad_norm": 0.3947749137878418,
-      "learning_rate": 8.301758040372723e-06,
-      "loss": 0.3259,
-      "step": 13083
-    },
-    {
-      "epoch": 0.8551075093131167,
-      "grad_norm": 0.4721922278404236,
-      "learning_rate": 8.301495805795e-06,
-      "loss": 0.4015,
-      "step": 13084
-    },
-    {
-      "epoch": 0.8551728645186589,
-      "grad_norm": 0.4693576991558075,
-      "learning_rate": 8.301233555114741e-06,
-      "loss": 0.4217,
-      "step": 13085
-    },
-    {
-      "epoch": 0.855238219724201,
-      "grad_norm": 0.42758774757385254,
-      "learning_rate": 8.30097128833322e-06,
-      "loss": 0.3216,
-      "step": 13086
-    },
-    {
-      "epoch": 0.8553035749297432,
-      "grad_norm": 0.4395238757133484,
-      "learning_rate": 8.300709005451719e-06,
-      "loss": 0.3893,
-      "step": 13087
-    },
-    {
-      "epoch": 0.8553689301352853,
-      "grad_norm": 0.429995596408844,
-      "learning_rate": 8.300446706471516e-06,
-      "loss": 0.4013,
-      "step": 13088
-    },
-    {
-      "epoch": 0.8554342853408274,
-      "grad_norm": 0.4792253077030182,
-      "learning_rate": 8.300184391393891e-06,
-      "loss": 0.421,
-      "step": 13089
-    },
-    {
-      "epoch": 0.8554996405463695,
-      "grad_norm": 0.49514615535736084,
-      "learning_rate": 8.299922060220124e-06,
-      "loss": 0.455,
-      "step": 13090
-    },
-    {
-      "epoch": 0.8555649957519117,
-      "grad_norm": 0.4884258210659027,
-      "learning_rate": 8.299659712951493e-06,
-      "loss": 0.3791,
-      "step": 13091
-    },
-    {
-      "epoch": 0.8556303509574538,
-      "grad_norm": 0.4589777886867523,
-      "learning_rate": 8.29939734958928e-06,
-      "loss": 0.3846,
-      "step": 13092
-    },
-    {
-      "epoch": 0.8556957061629958,
-      "grad_norm": 0.4226579964160919,
-      "learning_rate": 8.299134970134762e-06,
-      "loss": 0.3427,
-      "step": 13093
-    },
-    {
-      "epoch": 0.855761061368538,
-      "grad_norm": 0.44223830103874207,
-      "learning_rate": 8.29887257458922e-06,
-      "loss": 0.3543,
-      "step": 13094
-    },
-    {
-      "epoch": 0.8558264165740801,
-      "grad_norm": 0.4701899588108063,
-      "learning_rate": 8.298610162953934e-06,
-      "loss": 0.389,
-      "step": 13095
-    },
-    {
-      "epoch": 0.8558917717796223,
-      "grad_norm": 0.44645747542381287,
-      "learning_rate": 8.298347735230184e-06,
-      "loss": 0.3791,
-      "step": 13096
-    },
-    {
-      "epoch": 0.8559571269851644,
-      "grad_norm": 0.45808687806129456,
-      "learning_rate": 8.298085291419248e-06,
-      "loss": 0.3744,
-      "step": 13097
-    },
-    {
-      "epoch": 0.8560224821907065,
-      "grad_norm": 0.4933817982673645,
-      "learning_rate": 8.29782283152241e-06,
-      "loss": 0.4361,
-      "step": 13098
-    },
-    {
-      "epoch": 0.8560878373962486,
-      "grad_norm": 0.39033064246177673,
-      "learning_rate": 8.297560355540945e-06,
-      "loss": 0.3279,
-      "step": 13099
-    },
-    {
-      "epoch": 0.8561531926017907,
-      "grad_norm": 0.42793524265289307,
-      "learning_rate": 8.297297863476136e-06,
-      "loss": 0.3582,
-      "step": 13100
-    },
-    {
-      "epoch": 0.8562185478073329,
-      "grad_norm": 0.4983679950237274,
-      "learning_rate": 8.297035355329264e-06,
-      "loss": 0.4492,
-      "step": 13101
-    },
-    {
-      "epoch": 0.8562839030128749,
-      "grad_norm": 0.4555116295814514,
-      "learning_rate": 8.296772831101608e-06,
-      "loss": 0.4207,
-      "step": 13102
-    },
-    {
-      "epoch": 0.8563492582184171,
-      "grad_norm": 0.443266898393631,
-      "learning_rate": 8.296510290794449e-06,
-      "loss": 0.3645,
-      "step": 13103
-    },
-    {
-      "epoch": 0.8564146134239592,
-      "grad_norm": 0.4548618495464325,
-      "learning_rate": 8.296247734409067e-06,
-      "loss": 0.4258,
-      "step": 13104
-    },
-    {
-      "epoch": 0.8564799686295014,
-      "grad_norm": 0.47024303674697876,
-      "learning_rate": 8.295985161946742e-06,
-      "loss": 0.3809,
-      "step": 13105
-    },
-    {
-      "epoch": 0.8565453238350434,
-      "grad_norm": 0.4740467965602875,
-      "learning_rate": 8.295722573408757e-06,
-      "loss": 0.3743,
-      "step": 13106
-    },
-    {
-      "epoch": 0.8566106790405856,
-      "grad_norm": 0.4315088987350464,
-      "learning_rate": 8.29545996879639e-06,
-      "loss": 0.3476,
-      "step": 13107
-    },
-    {
-      "epoch": 0.8566760342461277,
-      "grad_norm": 0.44808855652809143,
-      "learning_rate": 8.295197348110924e-06,
-      "loss": 0.4072,
-      "step": 13108
-    },
-    {
-      "epoch": 0.8567413894516698,
-      "grad_norm": 0.4371505379676819,
-      "learning_rate": 8.29493471135364e-06,
-      "loss": 0.401,
-      "step": 13109
-    },
-    {
-      "epoch": 0.856806744657212,
-      "grad_norm": 0.4380919933319092,
-      "learning_rate": 8.294672058525815e-06,
-      "loss": 0.3667,
-      "step": 13110
-    },
-    {
-      "epoch": 0.856872099862754,
-      "grad_norm": 0.45407402515411377,
-      "learning_rate": 8.294409389628735e-06,
-      "loss": 0.3687,
-      "step": 13111
-    },
-    {
-      "epoch": 0.8569374550682962,
-      "grad_norm": 0.46568161249160767,
-      "learning_rate": 8.29414670466368e-06,
-      "loss": 0.378,
-      "step": 13112
-    },
-    {
-      "epoch": 0.8570028102738383,
-      "grad_norm": 0.46108999848365784,
-      "learning_rate": 8.293884003631928e-06,
-      "loss": 0.3927,
-      "step": 13113
-    },
-    {
-      "epoch": 0.8570681654793805,
-      "grad_norm": 0.44488200545310974,
-      "learning_rate": 8.293621286534763e-06,
-      "loss": 0.3885,
-      "step": 13114
-    },
-    {
-      "epoch": 0.8571335206849225,
-      "grad_norm": 0.4455205500125885,
-      "learning_rate": 8.293358553373468e-06,
-      "loss": 0.3902,
-      "step": 13115
-    },
-    {
-      "epoch": 0.8571988758904647,
-      "grad_norm": 0.45098504424095154,
-      "learning_rate": 8.29309580414932e-06,
-      "loss": 0.399,
-      "step": 13116
-    },
-    {
-      "epoch": 0.8572642310960068,
-      "grad_norm": 0.4192991256713867,
-      "learning_rate": 8.292833038863603e-06,
-      "loss": 0.3511,
-      "step": 13117
-    },
-    {
-      "epoch": 0.8573295863015489,
-      "grad_norm": 0.4548405408859253,
-      "learning_rate": 8.2925702575176e-06,
-      "loss": 0.3941,
-      "step": 13118
-    },
-    {
-      "epoch": 0.8573949415070911,
-      "grad_norm": 0.4591832458972931,
-      "learning_rate": 8.292307460112592e-06,
-      "loss": 0.3841,
-      "step": 13119
-    },
-    {
-      "epoch": 0.8574602967126331,
-      "grad_norm": 0.4723235070705414,
-      "learning_rate": 8.29204464664986e-06,
-      "loss": 0.4016,
-      "step": 13120
-    },
-    {
-      "epoch": 0.8575256519181753,
-      "grad_norm": 0.4260718524456024,
-      "learning_rate": 8.291781817130682e-06,
-      "loss": 0.3348,
-      "step": 13121
-    },
-    {
-      "epoch": 0.8575910071237174,
-      "grad_norm": 0.43534210324287415,
-      "learning_rate": 8.291518971556348e-06,
-      "loss": 0.3474,
-      "step": 13122
-    },
-    {
-      "epoch": 0.8576563623292596,
-      "grad_norm": 0.4349510073661804,
-      "learning_rate": 8.291256109928133e-06,
-      "loss": 0.3444,
-      "step": 13123
-    },
-    {
-      "epoch": 0.8577217175348016,
-      "grad_norm": 0.44429323077201843,
-      "learning_rate": 8.290993232247322e-06,
-      "loss": 0.3352,
-      "step": 13124
-    },
-    {
-      "epoch": 0.8577870727403437,
-      "grad_norm": 0.48946329951286316,
-      "learning_rate": 8.290730338515198e-06,
-      "loss": 0.4274,
-      "step": 13125
-    },
-    {
-      "epoch": 0.8578524279458859,
-      "grad_norm": 0.4377772808074951,
-      "learning_rate": 8.29046742873304e-06,
-      "loss": 0.3861,
-      "step": 13126
-    },
-    {
-      "epoch": 0.857917783151428,
-      "grad_norm": 0.4502428472042084,
-      "learning_rate": 8.290204502902134e-06,
-      "loss": 0.3836,
-      "step": 13127
-    },
-    {
-      "epoch": 0.8579831383569702,
-      "grad_norm": 0.4187738001346588,
-      "learning_rate": 8.289941561023762e-06,
-      "loss": 0.3425,
-      "step": 13128
-    },
-    {
-      "epoch": 0.8580484935625122,
-      "grad_norm": 0.44563737511634827,
-      "learning_rate": 8.289678603099205e-06,
-      "loss": 0.3748,
-      "step": 13129
-    },
-    {
-      "epoch": 0.8581138487680544,
-      "grad_norm": 0.41198623180389404,
-      "learning_rate": 8.289415629129744e-06,
-      "loss": 0.3382,
-      "step": 13130
-    },
-    {
-      "epoch": 0.8581792039735965,
-      "grad_norm": 0.43811723589897156,
-      "learning_rate": 8.289152639116664e-06,
-      "loss": 0.3676,
-      "step": 13131
-    },
-    {
-      "epoch": 0.8582445591791387,
-      "grad_norm": 0.4321269690990448,
-      "learning_rate": 8.288889633061248e-06,
-      "loss": 0.338,
-      "step": 13132
-    },
-    {
-      "epoch": 0.8583099143846807,
-      "grad_norm": 0.4447796046733856,
-      "learning_rate": 8.288626610964777e-06,
-      "loss": 0.4187,
-      "step": 13133
-    },
-    {
-      "epoch": 0.8583752695902228,
-      "grad_norm": 0.4782610535621643,
-      "learning_rate": 8.288363572828535e-06,
-      "loss": 0.4608,
-      "step": 13134
-    },
-    {
-      "epoch": 0.858440624795765,
-      "grad_norm": 0.481644868850708,
-      "learning_rate": 8.288100518653804e-06,
-      "loss": 0.4104,
-      "step": 13135
-    },
-    {
-      "epoch": 0.8585059800013071,
-      "grad_norm": 0.4687405824661255,
-      "learning_rate": 8.287837448441868e-06,
-      "loss": 0.3788,
-      "step": 13136
-    },
-    {
-      "epoch": 0.8585713352068493,
-      "grad_norm": 0.4344463646411896,
-      "learning_rate": 8.287574362194011e-06,
-      "loss": 0.3625,
-      "step": 13137
-    },
-    {
-      "epoch": 0.8586366904123913,
-      "grad_norm": 0.4399704933166504,
-      "learning_rate": 8.287311259911514e-06,
-      "loss": 0.3852,
-      "step": 13138
-    },
-    {
-      "epoch": 0.8587020456179335,
-      "grad_norm": 0.4202140271663666,
-      "learning_rate": 8.287048141595662e-06,
-      "loss": 0.3669,
-      "step": 13139
-    },
-    {
-      "epoch": 0.8587674008234756,
-      "grad_norm": 0.4349260926246643,
-      "learning_rate": 8.286785007247735e-06,
-      "loss": 0.3761,
-      "step": 13140
-    },
-    {
-      "epoch": 0.8588327560290178,
-      "grad_norm": 0.45881718397140503,
-      "learning_rate": 8.286521856869021e-06,
-      "loss": 0.3957,
-      "step": 13141
-    },
-    {
-      "epoch": 0.8588981112345598,
-      "grad_norm": 0.45289260149002075,
-      "learning_rate": 8.286258690460802e-06,
-      "loss": 0.3853,
-      "step": 13142
-    },
-    {
-      "epoch": 0.8589634664401019,
-      "grad_norm": 0.462878555059433,
-      "learning_rate": 8.285995508024357e-06,
-      "loss": 0.4019,
-      "step": 13143
-    },
-    {
-      "epoch": 0.8590288216456441,
-      "grad_norm": 0.44377198815345764,
-      "learning_rate": 8.28573230956098e-06,
-      "loss": 0.3537,
-      "step": 13144
-    },
-    {
-      "epoch": 0.8590941768511862,
-      "grad_norm": 0.4682895541191101,
-      "learning_rate": 8.285469095071943e-06,
-      "loss": 0.3954,
-      "step": 13145
-    },
-    {
-      "epoch": 0.8591595320567283,
-      "grad_norm": 0.49238693714141846,
-      "learning_rate": 8.285205864558537e-06,
-      "loss": 0.4497,
-      "step": 13146
-    },
-    {
-      "epoch": 0.8592248872622704,
-      "grad_norm": 0.4370424449443817,
-      "learning_rate": 8.284942618022043e-06,
-      "loss": 0.3772,
-      "step": 13147
-    },
-    {
-      "epoch": 0.8592902424678126,
-      "grad_norm": 0.4467172920703888,
-      "learning_rate": 8.284679355463746e-06,
-      "loss": 0.3624,
-      "step": 13148
-    },
-    {
-      "epoch": 0.8593555976733547,
-      "grad_norm": 0.4331428110599518,
-      "learning_rate": 8.284416076884928e-06,
-      "loss": 0.3635,
-      "step": 13149
-    },
-    {
-      "epoch": 0.8594209528788967,
-      "grad_norm": 0.4427790939807892,
-      "learning_rate": 8.284152782286878e-06,
-      "loss": 0.3956,
-      "step": 13150
-    },
-    {
-      "epoch": 0.8594863080844389,
-      "grad_norm": 0.4413788914680481,
-      "learning_rate": 8.283889471670876e-06,
-      "loss": 0.4163,
-      "step": 13151
-    },
-    {
-      "epoch": 0.859551663289981,
-      "grad_norm": 0.46344879269599915,
-      "learning_rate": 8.283626145038208e-06,
-      "loss": 0.3845,
-      "step": 13152
-    },
-    {
-      "epoch": 0.8596170184955232,
-      "grad_norm": 0.4417228698730469,
-      "learning_rate": 8.283362802390157e-06,
-      "loss": 0.363,
-      "step": 13153
-    },
-    {
-      "epoch": 0.8596823737010653,
-      "grad_norm": 0.43309471011161804,
-      "learning_rate": 8.283099443728009e-06,
-      "loss": 0.3351,
-      "step": 13154
-    },
-    {
-      "epoch": 0.8597477289066074,
-      "grad_norm": 0.4852140247821808,
-      "learning_rate": 8.282836069053048e-06,
-      "loss": 0.474,
-      "step": 13155
-    },
-    {
-      "epoch": 0.8598130841121495,
-      "grad_norm": 0.4612131714820862,
-      "learning_rate": 8.282572678366556e-06,
-      "loss": 0.4347,
-      "step": 13156
-    },
-    {
-      "epoch": 0.8598784393176917,
-      "grad_norm": 0.43646934628486633,
-      "learning_rate": 8.282309271669822e-06,
-      "loss": 0.4069,
-      "step": 13157
-    },
-    {
-      "epoch": 0.8599437945232338,
-      "grad_norm": 0.42238733172416687,
-      "learning_rate": 8.282045848964128e-06,
-      "loss": 0.3631,
-      "step": 13158
-    },
-    {
-      "epoch": 0.8600091497287758,
-      "grad_norm": 0.44889938831329346,
-      "learning_rate": 8.281782410250759e-06,
-      "loss": 0.3734,
-      "step": 13159
-    },
-    {
-      "epoch": 0.860074504934318,
-      "grad_norm": 0.44265878200531006,
-      "learning_rate": 8.281518955531001e-06,
-      "loss": 0.3891,
-      "step": 13160
-    },
-    {
-      "epoch": 0.8601398601398601,
-      "grad_norm": 0.4276835024356842,
-      "learning_rate": 8.281255484806136e-06,
-      "loss": 0.3475,
-      "step": 13161
-    },
-    {
-      "epoch": 0.8602052153454023,
-      "grad_norm": 0.4474591612815857,
-      "learning_rate": 8.280991998077454e-06,
-      "loss": 0.3883,
-      "step": 13162
-    },
-    {
-      "epoch": 0.8602705705509444,
-      "grad_norm": 0.4535478949546814,
-      "learning_rate": 8.280728495346236e-06,
-      "loss": 0.3967,
-      "step": 13163
-    },
-    {
-      "epoch": 0.8603359257564865,
-      "grad_norm": 0.48304906487464905,
-      "learning_rate": 8.280464976613768e-06,
-      "loss": 0.4667,
-      "step": 13164
-    },
-    {
-      "epoch": 0.8604012809620286,
-      "grad_norm": 0.463156133890152,
-      "learning_rate": 8.280201441881337e-06,
-      "loss": 0.3567,
-      "step": 13165
-    },
-    {
-      "epoch": 0.8604666361675708,
-      "grad_norm": 0.4587669372558594,
-      "learning_rate": 8.279937891150228e-06,
-      "loss": 0.3872,
-      "step": 13166
-    },
-    {
-      "epoch": 0.8605319913731129,
-      "grad_norm": 0.44457775354385376,
-      "learning_rate": 8.279674324421725e-06,
-      "loss": 0.3724,
-      "step": 13167
-    },
-    {
-      "epoch": 0.8605973465786549,
-      "grad_norm": 0.4893188774585724,
-      "learning_rate": 8.279410741697114e-06,
-      "loss": 0.4544,
-      "step": 13168
-    },
-    {
-      "epoch": 0.8606627017841971,
-      "grad_norm": 0.43643447756767273,
-      "learning_rate": 8.27914714297768e-06,
-      "loss": 0.357,
-      "step": 13169
-    },
-    {
-      "epoch": 0.8607280569897392,
-      "grad_norm": 0.44617798924446106,
-      "learning_rate": 8.27888352826471e-06,
-      "loss": 0.393,
-      "step": 13170
-    },
-    {
-      "epoch": 0.8607934121952814,
-      "grad_norm": 0.46605929732322693,
-      "learning_rate": 8.278619897559488e-06,
-      "loss": 0.4163,
-      "step": 13171
-    },
-    {
-      "epoch": 0.8608587674008235,
-      "grad_norm": 0.41804397106170654,
-      "learning_rate": 8.278356250863302e-06,
-      "loss": 0.3314,
-      "step": 13172
-    },
-    {
-      "epoch": 0.8609241226063656,
-      "grad_norm": 0.42674967646598816,
-      "learning_rate": 8.278092588177435e-06,
-      "loss": 0.3338,
-      "step": 13173
-    },
-    {
-      "epoch": 0.8609894778119077,
-      "grad_norm": 0.43258965015411377,
-      "learning_rate": 8.277828909503178e-06,
-      "loss": 0.3899,
-      "step": 13174
-    },
-    {
-      "epoch": 0.8610548330174499,
-      "grad_norm": 0.3933675289154053,
-      "learning_rate": 8.277565214841812e-06,
-      "loss": 0.3442,
-      "step": 13175
-    },
-    {
-      "epoch": 0.861120188222992,
-      "grad_norm": 0.4413454830646515,
-      "learning_rate": 8.277301504194626e-06,
-      "loss": 0.3419,
-      "step": 13176
-    },
-    {
-      "epoch": 0.861185543428534,
-      "grad_norm": 0.47428226470947266,
-      "learning_rate": 8.277037777562905e-06,
-      "loss": 0.4339,
-      "step": 13177
-    },
-    {
-      "epoch": 0.8612508986340762,
-      "grad_norm": 0.45410338044166565,
-      "learning_rate": 8.276774034947935e-06,
-      "loss": 0.389,
-      "step": 13178
-    },
-    {
-      "epoch": 0.8613162538396183,
-      "grad_norm": 0.5930436253547668,
-      "learning_rate": 8.276510276351003e-06,
-      "loss": 0.3865,
-      "step": 13179
-    },
-    {
-      "epoch": 0.8613816090451605,
-      "grad_norm": 0.4242371618747711,
-      "learning_rate": 8.276246501773393e-06,
-      "loss": 0.335,
-      "step": 13180
-    },
-    {
-      "epoch": 0.8614469642507026,
-      "grad_norm": 0.41871318221092224,
-      "learning_rate": 8.275982711216397e-06,
-      "loss": 0.3378,
-      "step": 13181
-    },
-    {
-      "epoch": 0.8615123194562447,
-      "grad_norm": 0.43339404463768005,
-      "learning_rate": 8.275718904681297e-06,
-      "loss": 0.3761,
-      "step": 13182
-    },
-    {
-      "epoch": 0.8615776746617868,
-      "grad_norm": 0.4290432035923004,
-      "learning_rate": 8.275455082169381e-06,
-      "loss": 0.3475,
-      "step": 13183
-    },
-    {
-      "epoch": 0.8616430298673289,
-      "grad_norm": 0.41263851523399353,
-      "learning_rate": 8.275191243681937e-06,
-      "loss": 0.3305,
-      "step": 13184
-    },
-    {
-      "epoch": 0.8617083850728711,
-      "grad_norm": 0.4255129396915436,
-      "learning_rate": 8.27492738922025e-06,
-      "loss": 0.3382,
-      "step": 13185
-    },
-    {
-      "epoch": 0.8617737402784131,
-      "grad_norm": 0.4154873788356781,
-      "learning_rate": 8.274663518785608e-06,
-      "loss": 0.3365,
-      "step": 13186
-    },
-    {
-      "epoch": 0.8618390954839553,
-      "grad_norm": 0.41808241605758667,
-      "learning_rate": 8.274399632379298e-06,
-      "loss": 0.3077,
-      "step": 13187
-    },
-    {
-      "epoch": 0.8619044506894974,
-      "grad_norm": 0.4594711661338806,
-      "learning_rate": 8.274135730002605e-06,
-      "loss": 0.4076,
-      "step": 13188
-    },
-    {
-      "epoch": 0.8619698058950396,
-      "grad_norm": 0.4508814811706543,
-      "learning_rate": 8.273871811656817e-06,
-      "loss": 0.3974,
-      "step": 13189
-    },
-    {
-      "epoch": 0.8620351611005816,
-      "grad_norm": 0.40516120195388794,
-      "learning_rate": 8.273607877343224e-06,
-      "loss": 0.3491,
-      "step": 13190
-    },
-    {
-      "epoch": 0.8621005163061238,
-      "grad_norm": 0.4845518171787262,
-      "learning_rate": 8.27334392706311e-06,
-      "loss": 0.4144,
-      "step": 13191
-    },
-    {
-      "epoch": 0.8621658715116659,
-      "grad_norm": 0.4769570231437683,
-      "learning_rate": 8.273079960817764e-06,
-      "loss": 0.4041,
-      "step": 13192
-    },
-    {
-      "epoch": 0.862231226717208,
-      "grad_norm": 0.43818965554237366,
-      "learning_rate": 8.272815978608475e-06,
-      "loss": 0.3578,
-      "step": 13193
-    },
-    {
-      "epoch": 0.8622965819227502,
-      "grad_norm": 0.4566902816295624,
-      "learning_rate": 8.272551980436527e-06,
-      "loss": 0.3759,
-      "step": 13194
-    },
-    {
-      "epoch": 0.8623619371282922,
-      "grad_norm": 0.46521174907684326,
-      "learning_rate": 8.27228796630321e-06,
-      "loss": 0.3969,
-      "step": 13195
-    },
-    {
-      "epoch": 0.8624272923338344,
-      "grad_norm": 0.4606292247772217,
-      "learning_rate": 8.27202393620981e-06,
-      "loss": 0.3811,
-      "step": 13196
-    },
-    {
-      "epoch": 0.8624926475393765,
-      "grad_norm": 0.4627600610256195,
-      "learning_rate": 8.271759890157616e-06,
-      "loss": 0.391,
-      "step": 13197
-    },
-    {
-      "epoch": 0.8625580027449187,
-      "grad_norm": 0.42293885350227356,
-      "learning_rate": 8.271495828147916e-06,
-      "loss": 0.3564,
-      "step": 13198
-    },
-    {
-      "epoch": 0.8626233579504607,
-      "grad_norm": 0.4492563307285309,
-      "learning_rate": 8.271231750181997e-06,
-      "loss": 0.4146,
-      "step": 13199
-    },
-    {
-      "epoch": 0.8626887131560029,
-      "grad_norm": 0.47421994805336,
-      "learning_rate": 8.270967656261148e-06,
-      "loss": 0.422,
-      "step": 13200
-    },
-    {
-      "epoch": 0.862754068361545,
-      "grad_norm": 0.42958301305770874,
-      "learning_rate": 8.270703546386656e-06,
-      "loss": 0.3305,
-      "step": 13201
-    },
-    {
-      "epoch": 0.8628194235670871,
-      "grad_norm": 0.431548148393631,
-      "learning_rate": 8.27043942055981e-06,
-      "loss": 0.3772,
-      "step": 13202
-    },
-    {
-      "epoch": 0.8628847787726293,
-      "grad_norm": 0.40252622961997986,
-      "learning_rate": 8.270175278781898e-06,
-      "loss": 0.3363,
-      "step": 13203
-    },
-    {
-      "epoch": 0.8629501339781713,
-      "grad_norm": 0.4037279784679413,
-      "learning_rate": 8.269911121054209e-06,
-      "loss": 0.3279,
-      "step": 13204
-    },
-    {
-      "epoch": 0.8630154891837135,
-      "grad_norm": 0.48043838143348694,
-      "learning_rate": 8.269646947378029e-06,
-      "loss": 0.4278,
-      "step": 13205
-    },
-    {
-      "epoch": 0.8630808443892556,
-      "grad_norm": 0.4289033114910126,
-      "learning_rate": 8.26938275775465e-06,
-      "loss": 0.3295,
-      "step": 13206
-    },
-    {
-      "epoch": 0.8631461995947978,
-      "grad_norm": 0.4539353847503662,
-      "learning_rate": 8.269118552185358e-06,
-      "loss": 0.3967,
-      "step": 13207
-    },
-    {
-      "epoch": 0.8632115548003398,
-      "grad_norm": 0.45041242241859436,
-      "learning_rate": 8.268854330671441e-06,
-      "loss": 0.3757,
-      "step": 13208
-    },
-    {
-      "epoch": 0.8632769100058819,
-      "grad_norm": 0.4105256497859955,
-      "learning_rate": 8.26859009321419e-06,
-      "loss": 0.3341,
-      "step": 13209
-    },
-    {
-      "epoch": 0.8633422652114241,
-      "grad_norm": 0.4560367465019226,
-      "learning_rate": 8.268325839814892e-06,
-      "loss": 0.3842,
-      "step": 13210
-    },
-    {
-      "epoch": 0.8634076204169662,
-      "grad_norm": 0.42491069436073303,
-      "learning_rate": 8.26806157047484e-06,
-      "loss": 0.3448,
-      "step": 13211
-    },
-    {
-      "epoch": 0.8634729756225084,
-      "grad_norm": 0.39981043338775635,
-      "learning_rate": 8.267797285195317e-06,
-      "loss": 0.2981,
-      "step": 13212
-    },
-    {
-      "epoch": 0.8635383308280504,
-      "grad_norm": 0.4760648310184479,
-      "learning_rate": 8.267532983977613e-06,
-      "loss": 0.4309,
-      "step": 13213
-    },
-    {
-      "epoch": 0.8636036860335926,
-      "grad_norm": 0.44349217414855957,
-      "learning_rate": 8.26726866682302e-06,
-      "loss": 0.3741,
-      "step": 13214
-    },
-    {
-      "epoch": 0.8636690412391347,
-      "grad_norm": 0.46941787004470825,
-      "learning_rate": 8.267004333732826e-06,
-      "loss": 0.395,
-      "step": 13215
-    },
-    {
-      "epoch": 0.8637343964446769,
-      "grad_norm": 0.4467675983905792,
-      "learning_rate": 8.26673998470832e-06,
-      "loss": 0.3858,
-      "step": 13216
-    },
-    {
-      "epoch": 0.8637997516502189,
-      "grad_norm": 0.47093433141708374,
-      "learning_rate": 8.266475619750792e-06,
-      "loss": 0.4136,
-      "step": 13217
-    },
-    {
-      "epoch": 0.863865106855761,
-      "grad_norm": 0.47002628445625305,
-      "learning_rate": 8.26621123886153e-06,
-      "loss": 0.3918,
-      "step": 13218
-    },
-    {
-      "epoch": 0.8639304620613032,
-      "grad_norm": 0.44025665521621704,
-      "learning_rate": 8.265946842041823e-06,
-      "loss": 0.3946,
-      "step": 13219
-    },
-    {
-      "epoch": 0.8639958172668453,
-      "grad_norm": 0.4498584270477295,
-      "learning_rate": 8.265682429292964e-06,
-      "loss": 0.3751,
-      "step": 13220
-    },
-    {
-      "epoch": 0.8640611724723875,
-      "grad_norm": 0.46606284379959106,
-      "learning_rate": 8.265418000616242e-06,
-      "loss": 0.3992,
-      "step": 13221
-    },
-    {
-      "epoch": 0.8641265276779295,
-      "grad_norm": 0.48162829875946045,
-      "learning_rate": 8.265153556012942e-06,
-      "loss": 0.407,
-      "step": 13222
-    },
-    {
-      "epoch": 0.8641918828834717,
-      "grad_norm": 0.4349437654018402,
-      "learning_rate": 8.264889095484357e-06,
-      "loss": 0.3403,
-      "step": 13223
-    },
-    {
-      "epoch": 0.8642572380890138,
-      "grad_norm": 0.45929232239723206,
-      "learning_rate": 8.264624619031777e-06,
-      "loss": 0.4039,
-      "step": 13224
-    },
-    {
-      "epoch": 0.864322593294556,
-      "grad_norm": 0.42121267318725586,
-      "learning_rate": 8.264360126656495e-06,
-      "loss": 0.3597,
-      "step": 13225
-    },
-    {
-      "epoch": 0.864387948500098,
-      "grad_norm": 0.4389696717262268,
-      "learning_rate": 8.264095618359794e-06,
-      "loss": 0.354,
-      "step": 13226
-    },
-    {
-      "epoch": 0.8644533037056401,
-      "grad_norm": 0.48989543318748474,
-      "learning_rate": 8.263831094142969e-06,
-      "loss": 0.4056,
-      "step": 13227
-    },
-    {
-      "epoch": 0.8645186589111823,
-      "grad_norm": 0.46081024408340454,
-      "learning_rate": 8.26356655400731e-06,
-      "loss": 0.3697,
-      "step": 13228
-    },
-    {
-      "epoch": 0.8645840141167244,
-      "grad_norm": 0.47132226824760437,
-      "learning_rate": 8.263301997954104e-06,
-      "loss": 0.4067,
-      "step": 13229
-    },
-    {
-      "epoch": 0.8646493693222665,
-      "grad_norm": 0.43644627928733826,
-      "learning_rate": 8.263037425984646e-06,
-      "loss": 0.3557,
-      "step": 13230
-    },
-    {
-      "epoch": 0.8647147245278086,
-      "grad_norm": 0.4656889736652374,
-      "learning_rate": 8.26277283810022e-06,
-      "loss": 0.4019,
-      "step": 13231
-    },
-    {
-      "epoch": 0.8647800797333508,
-      "grad_norm": 0.4414185583591461,
-      "learning_rate": 8.262508234302127e-06,
-      "loss": 0.368,
-      "step": 13232
-    },
-    {
-      "epoch": 0.8648454349388929,
-      "grad_norm": 0.5081361532211304,
-      "learning_rate": 8.262243614591648e-06,
-      "loss": 0.4635,
-      "step": 13233
-    },
-    {
-      "epoch": 0.864910790144435,
-      "grad_norm": 0.4259694516658783,
-      "learning_rate": 8.261978978970075e-06,
-      "loss": 0.3211,
-      "step": 13234
-    },
-    {
-      "epoch": 0.8649761453499771,
-      "grad_norm": 0.43436840176582336,
-      "learning_rate": 8.261714327438703e-06,
-      "loss": 0.3436,
-      "step": 13235
-    },
-    {
-      "epoch": 0.8650415005555192,
-      "grad_norm": 0.44253620505332947,
-      "learning_rate": 8.261449659998819e-06,
-      "loss": 0.4192,
-      "step": 13236
-    },
-    {
-      "epoch": 0.8651068557610614,
-      "grad_norm": 0.44048523902893066,
-      "learning_rate": 8.261184976651715e-06,
-      "loss": 0.3689,
-      "step": 13237
-    },
-    {
-      "epoch": 0.8651722109666035,
-      "grad_norm": 0.4383637607097626,
-      "learning_rate": 8.260920277398683e-06,
-      "loss": 0.3796,
-      "step": 13238
-    },
-    {
-      "epoch": 0.8652375661721456,
-      "grad_norm": 0.43479683995246887,
-      "learning_rate": 8.260655562241011e-06,
-      "loss": 0.3625,
-      "step": 13239
-    },
-    {
-      "epoch": 0.8653029213776877,
-      "grad_norm": 0.4594433605670929,
-      "learning_rate": 8.260390831179995e-06,
-      "loss": 0.3887,
-      "step": 13240
-    },
-    {
-      "epoch": 0.8653682765832299,
-      "grad_norm": 0.45842957496643066,
-      "learning_rate": 8.260126084216922e-06,
-      "loss": 0.3745,
-      "step": 13241
-    },
-    {
-      "epoch": 0.865433631788772,
-      "grad_norm": 0.40813007950782776,
-      "learning_rate": 8.259861321353084e-06,
-      "loss": 0.3102,
-      "step": 13242
-    },
-    {
-      "epoch": 0.865498986994314,
-      "grad_norm": 0.4562146067619324,
-      "learning_rate": 8.259596542589774e-06,
-      "loss": 0.3692,
-      "step": 13243
-    },
-    {
-      "epoch": 0.8655643421998562,
-      "grad_norm": 0.41048356890678406,
-      "learning_rate": 8.259331747928284e-06,
-      "loss": 0.3489,
-      "step": 13244
-    },
-    {
-      "epoch": 0.8656296974053983,
-      "grad_norm": 0.46980148553848267,
-      "learning_rate": 8.259066937369901e-06,
-      "loss": 0.418,
-      "step": 13245
-    },
-    {
-      "epoch": 0.8656950526109405,
-      "grad_norm": 0.4779313802719116,
-      "learning_rate": 8.258802110915922e-06,
-      "loss": 0.4736,
-      "step": 13246
-    },
-    {
-      "epoch": 0.8657604078164826,
-      "grad_norm": 0.4522343575954437,
-      "learning_rate": 8.258537268567634e-06,
-      "loss": 0.3772,
-      "step": 13247
-    },
-    {
-      "epoch": 0.8658257630220247,
-      "grad_norm": 0.45030477643013,
-      "learning_rate": 8.258272410326331e-06,
-      "loss": 0.3695,
-      "step": 13248
-    },
-    {
-      "epoch": 0.8658911182275668,
-      "grad_norm": 0.45079004764556885,
-      "learning_rate": 8.258007536193306e-06,
-      "loss": 0.4004,
-      "step": 13249
-    },
-    {
-      "epoch": 0.865956473433109,
-      "grad_norm": 0.43524375557899475,
-      "learning_rate": 8.257742646169848e-06,
-      "loss": 0.388,
-      "step": 13250
-    },
-    {
-      "epoch": 0.8660218286386511,
-      "grad_norm": 0.4434921443462372,
-      "learning_rate": 8.257477740257254e-06,
-      "loss": 0.3801,
-      "step": 13251
-    },
-    {
-      "epoch": 0.8660871838441931,
-      "grad_norm": 0.4336143434047699,
-      "learning_rate": 8.257212818456809e-06,
-      "loss": 0.3739,
-      "step": 13252
-    },
-    {
-      "epoch": 0.8661525390497353,
-      "grad_norm": 0.45957279205322266,
-      "learning_rate": 8.25694788076981e-06,
-      "loss": 0.3649,
-      "step": 13253
-    },
-    {
-      "epoch": 0.8662178942552774,
-      "grad_norm": 0.4267975687980652,
-      "learning_rate": 8.256682927197547e-06,
-      "loss": 0.3414,
-      "step": 13254
-    },
-    {
-      "epoch": 0.8662832494608196,
-      "grad_norm": 0.6603761315345764,
-      "learning_rate": 8.256417957741313e-06,
-      "loss": 0.4134,
-      "step": 13255
-    },
-    {
-      "epoch": 0.8663486046663617,
-      "grad_norm": 0.462980180978775,
-      "learning_rate": 8.256152972402403e-06,
-      "loss": 0.347,
-      "step": 13256
-    },
-    {
-      "epoch": 0.8664139598719038,
-      "grad_norm": 0.44443926215171814,
-      "learning_rate": 8.255887971182103e-06,
-      "loss": 0.3384,
-      "step": 13257
-    },
-    {
-      "epoch": 0.8664793150774459,
-      "grad_norm": 0.4551098942756653,
-      "learning_rate": 8.255622954081713e-06,
-      "loss": 0.4131,
-      "step": 13258
-    },
-    {
-      "epoch": 0.8665446702829881,
-      "grad_norm": 0.4636535346508026,
-      "learning_rate": 8.25535792110252e-06,
-      "loss": 0.4575,
-      "step": 13259
-    },
-    {
-      "epoch": 0.8666100254885302,
-      "grad_norm": 0.4483840763568878,
-      "learning_rate": 8.25509287224582e-06,
-      "loss": 0.345,
-      "step": 13260
-    },
-    {
-      "epoch": 0.8666753806940722,
-      "grad_norm": 0.43046480417251587,
-      "learning_rate": 8.254827807512904e-06,
-      "loss": 0.3733,
-      "step": 13261
-    },
-    {
-      "epoch": 0.8667407358996144,
-      "grad_norm": 0.4565449059009552,
-      "learning_rate": 8.254562726905064e-06,
-      "loss": 0.3785,
-      "step": 13262
-    },
-    {
-      "epoch": 0.8668060911051565,
-      "grad_norm": 0.44150421023368835,
-      "learning_rate": 8.254297630423595e-06,
-      "loss": 0.3821,
-      "step": 13263
-    },
-    {
-      "epoch": 0.8668714463106987,
-      "grad_norm": 0.49192196130752563,
-      "learning_rate": 8.25403251806979e-06,
-      "loss": 0.4364,
-      "step": 13264
-    },
-    {
-      "epoch": 0.8669368015162408,
-      "grad_norm": 0.42523103952407837,
-      "learning_rate": 8.253767389844939e-06,
-      "loss": 0.3604,
-      "step": 13265
-    },
-    {
-      "epoch": 0.8670021567217829,
-      "grad_norm": 0.4398192763328552,
-      "learning_rate": 8.253502245750338e-06,
-      "loss": 0.3873,
-      "step": 13266
-    },
-    {
-      "epoch": 0.867067511927325,
-      "grad_norm": 0.39868608117103577,
-      "learning_rate": 8.25323708578728e-06,
-      "loss": 0.2922,
-      "step": 13267
-    },
-    {
-      "epoch": 0.8671328671328671,
-      "grad_norm": 0.4361017048358917,
-      "learning_rate": 8.252971909957058e-06,
-      "loss": 0.3355,
-      "step": 13268
-    },
-    {
-      "epoch": 0.8671982223384093,
-      "grad_norm": 0.4512507915496826,
-      "learning_rate": 8.252706718260964e-06,
-      "loss": 0.4073,
-      "step": 13269
-    },
-    {
-      "epoch": 0.8672635775439513,
-      "grad_norm": 0.4323147237300873,
-      "learning_rate": 8.252441510700294e-06,
-      "loss": 0.3667,
-      "step": 13270
-    },
-    {
-      "epoch": 0.8673289327494935,
-      "grad_norm": 0.43514636158943176,
-      "learning_rate": 8.252176287276338e-06,
-      "loss": 0.3669,
-      "step": 13271
-    },
-    {
-      "epoch": 0.8673942879550356,
-      "grad_norm": 0.4541299343109131,
-      "learning_rate": 8.251911047990393e-06,
-      "loss": 0.3936,
-      "step": 13272
-    },
-    {
-      "epoch": 0.8674596431605778,
-      "grad_norm": 0.477446973323822,
-      "learning_rate": 8.25164579284375e-06,
-      "loss": 0.3783,
-      "step": 13273
-    },
-    {
-      "epoch": 0.8675249983661198,
-      "grad_norm": 0.38905927538871765,
-      "learning_rate": 8.251380521837706e-06,
-      "loss": 0.3131,
-      "step": 13274
-    },
-    {
-      "epoch": 0.867590353571662,
-      "grad_norm": 0.42171740531921387,
-      "learning_rate": 8.25111523497355e-06,
-      "loss": 0.3596,
-      "step": 13275
-    },
-    {
-      "epoch": 0.8676557087772041,
-      "grad_norm": 0.46107205748558044,
-      "learning_rate": 8.250849932252581e-06,
-      "loss": 0.3548,
-      "step": 13276
-    },
-    {
-      "epoch": 0.8677210639827462,
-      "grad_norm": 0.4519311189651489,
-      "learning_rate": 8.25058461367609e-06,
-      "loss": 0.373,
-      "step": 13277
-    },
-    {
-      "epoch": 0.8677864191882884,
-      "grad_norm": 0.49622842669487,
-      "learning_rate": 8.250319279245373e-06,
-      "loss": 0.4223,
-      "step": 13278
-    },
-    {
-      "epoch": 0.8678517743938304,
-      "grad_norm": 0.4517560601234436,
-      "learning_rate": 8.250053928961722e-06,
-      "loss": 0.3535,
-      "step": 13279
-    },
-    {
-      "epoch": 0.8679171295993726,
-      "grad_norm": 0.43382585048675537,
-      "learning_rate": 8.249788562826431e-06,
-      "loss": 0.3795,
-      "step": 13280
-    },
-    {
-      "epoch": 0.8679824848049147,
-      "grad_norm": 0.43188732862472534,
-      "learning_rate": 8.249523180840795e-06,
-      "loss": 0.3755,
-      "step": 13281
-    },
-    {
-      "epoch": 0.8680478400104569,
-      "grad_norm": 0.4784072935581207,
-      "learning_rate": 8.249257783006111e-06,
-      "loss": 0.4162,
-      "step": 13282
-    },
-    {
-      "epoch": 0.868113195215999,
-      "grad_norm": 0.4593978822231293,
-      "learning_rate": 8.24899236932367e-06,
-      "loss": 0.3703,
-      "step": 13283
-    },
-    {
-      "epoch": 0.8681785504215411,
-      "grad_norm": 0.43106651306152344,
-      "learning_rate": 8.248726939794767e-06,
-      "loss": 0.3776,
-      "step": 13284
-    },
-    {
-      "epoch": 0.8682439056270832,
-      "grad_norm": 0.43448102474212646,
-      "learning_rate": 8.248461494420696e-06,
-      "loss": 0.3613,
-      "step": 13285
-    },
-    {
-      "epoch": 0.8683092608326253,
-      "grad_norm": 0.42373764514923096,
-      "learning_rate": 8.248196033202756e-06,
-      "loss": 0.3631,
-      "step": 13286
-    },
-    {
-      "epoch": 0.8683746160381675,
-      "grad_norm": 0.45077648758888245,
-      "learning_rate": 8.247930556142238e-06,
-      "loss": 0.3562,
-      "step": 13287
-    },
-    {
-      "epoch": 0.8684399712437095,
-      "grad_norm": 0.4450948238372803,
-      "learning_rate": 8.247665063240437e-06,
-      "loss": 0.4043,
-      "step": 13288
-    },
-    {
-      "epoch": 0.8685053264492517,
-      "grad_norm": 0.45834505558013916,
-      "learning_rate": 8.247399554498647e-06,
-      "loss": 0.372,
-      "step": 13289
-    },
-    {
-      "epoch": 0.8685706816547938,
-      "grad_norm": 0.44911620020866394,
-      "learning_rate": 8.247134029918167e-06,
-      "loss": 0.4059,
-      "step": 13290
-    },
-    {
-      "epoch": 0.868636036860336,
-      "grad_norm": 0.4145478308200836,
-      "learning_rate": 8.246868489500287e-06,
-      "loss": 0.3053,
-      "step": 13291
-    },
-    {
-      "epoch": 0.868701392065878,
-      "grad_norm": 0.44386187195777893,
-      "learning_rate": 8.246602933246306e-06,
-      "loss": 0.3635,
-      "step": 13292
-    },
-    {
-      "epoch": 0.8687667472714201,
-      "grad_norm": 0.42866528034210205,
-      "learning_rate": 8.246337361157517e-06,
-      "loss": 0.3544,
-      "step": 13293
-    },
-    {
-      "epoch": 0.8688321024769623,
-      "grad_norm": 0.4357426166534424,
-      "learning_rate": 8.246071773235217e-06,
-      "loss": 0.3562,
-      "step": 13294
-    },
-    {
-      "epoch": 0.8688974576825044,
-      "grad_norm": 0.462495356798172,
-      "learning_rate": 8.2458061694807e-06,
-      "loss": 0.4126,
-      "step": 13295
-    },
-    {
-      "epoch": 0.8689628128880466,
-      "grad_norm": 0.43929940462112427,
-      "learning_rate": 8.245540549895262e-06,
-      "loss": 0.351,
-      "step": 13296
-    },
-    {
-      "epoch": 0.8690281680935886,
-      "grad_norm": 0.4430977404117584,
-      "learning_rate": 8.245274914480196e-06,
-      "loss": 0.4129,
-      "step": 13297
-    },
-    {
-      "epoch": 0.8690935232991308,
-      "grad_norm": 0.49655118584632874,
-      "learning_rate": 8.245009263236803e-06,
-      "loss": 0.4958,
-      "step": 13298
-    },
-    {
-      "epoch": 0.8691588785046729,
-      "grad_norm": 0.39067956805229187,
-      "learning_rate": 8.244743596166374e-06,
-      "loss": 0.2936,
-      "step": 13299
-    },
-    {
-      "epoch": 0.8692242337102151,
-      "grad_norm": 0.41374117136001587,
-      "learning_rate": 8.244477913270208e-06,
-      "loss": 0.3257,
-      "step": 13300
-    },
-    {
-      "epoch": 0.8692895889157571,
-      "grad_norm": 0.44752123951911926,
-      "learning_rate": 8.244212214549598e-06,
-      "loss": 0.3663,
-      "step": 13301
-    },
-    {
-      "epoch": 0.8693549441212992,
-      "grad_norm": 0.41446512937545776,
-      "learning_rate": 8.24394650000584e-06,
-      "loss": 0.3404,
-      "step": 13302
-    },
-    {
-      "epoch": 0.8694202993268414,
-      "grad_norm": 0.4655797481536865,
-      "learning_rate": 8.243680769640234e-06,
-      "loss": 0.4057,
-      "step": 13303
-    },
-    {
-      "epoch": 0.8694856545323835,
-      "grad_norm": 0.4714828431606293,
-      "learning_rate": 8.24341502345407e-06,
-      "loss": 0.4201,
-      "step": 13304
-    },
-    {
-      "epoch": 0.8695510097379257,
-      "grad_norm": 0.4187307059764862,
-      "learning_rate": 8.24314926144865e-06,
-      "loss": 0.3593,
-      "step": 13305
-    },
-    {
-      "epoch": 0.8696163649434677,
-      "grad_norm": 0.4411788284778595,
-      "learning_rate": 8.242883483625266e-06,
-      "loss": 0.3757,
-      "step": 13306
-    },
-    {
-      "epoch": 0.8696817201490099,
-      "grad_norm": 0.48096975684165955,
-      "learning_rate": 8.242617689985217e-06,
-      "loss": 0.4154,
-      "step": 13307
-    },
-    {
-      "epoch": 0.869747075354552,
-      "grad_norm": 0.4491497576236725,
-      "learning_rate": 8.242351880529797e-06,
-      "loss": 0.4145,
-      "step": 13308
-    },
-    {
-      "epoch": 0.8698124305600942,
-      "grad_norm": 0.40352532267570496,
-      "learning_rate": 8.242086055260306e-06,
-      "loss": 0.3419,
-      "step": 13309
-    },
-    {
-      "epoch": 0.8698777857656362,
-      "grad_norm": 0.4654463231563568,
-      "learning_rate": 8.241820214178036e-06,
-      "loss": 0.4165,
-      "step": 13310
-    },
-    {
-      "epoch": 0.8699431409711783,
-      "grad_norm": 0.4417929947376251,
-      "learning_rate": 8.241554357284284e-06,
-      "loss": 0.3897,
-      "step": 13311
-    },
-    {
-      "epoch": 0.8700084961767205,
-      "grad_norm": 0.42508944869041443,
-      "learning_rate": 8.241288484580352e-06,
-      "loss": 0.3593,
-      "step": 13312
-    },
-    {
-      "epoch": 0.8700738513822626,
-      "grad_norm": 0.4737236797809601,
-      "learning_rate": 8.24102259606753e-06,
-      "loss": 0.4201,
-      "step": 13313
-    },
-    {
-      "epoch": 0.8701392065878047,
-      "grad_norm": 0.47437816858291626,
-      "learning_rate": 8.24075669174712e-06,
-      "loss": 0.4089,
-      "step": 13314
-    },
-    {
-      "epoch": 0.8702045617933468,
-      "grad_norm": 0.48102623224258423,
-      "learning_rate": 8.240490771620416e-06,
-      "loss": 0.3758,
-      "step": 13315
-    },
-    {
-      "epoch": 0.870269916998889,
-      "grad_norm": 0.46707481145858765,
-      "learning_rate": 8.240224835688716e-06,
-      "loss": 0.3877,
-      "step": 13316
-    },
-    {
-      "epoch": 0.8703352722044311,
-      "grad_norm": 0.4536297023296356,
-      "learning_rate": 8.239958883953319e-06,
-      "loss": 0.3981,
-      "step": 13317
-    },
-    {
-      "epoch": 0.8704006274099733,
-      "grad_norm": 0.42421531677246094,
-      "learning_rate": 8.239692916415518e-06,
-      "loss": 0.3487,
-      "step": 13318
-    },
-    {
-      "epoch": 0.8704659826155153,
-      "grad_norm": 0.43656229972839355,
-      "learning_rate": 8.239426933076613e-06,
-      "loss": 0.3673,
-      "step": 13319
-    },
-    {
-      "epoch": 0.8705313378210574,
-      "grad_norm": 0.4687826633453369,
-      "learning_rate": 8.2391609339379e-06,
-      "loss": 0.42,
-      "step": 13320
-    },
-    {
-      "epoch": 0.8705966930265996,
-      "grad_norm": 0.42849722504615784,
-      "learning_rate": 8.238894919000677e-06,
-      "loss": 0.3645,
-      "step": 13321
-    },
-    {
-      "epoch": 0.8706620482321417,
-      "grad_norm": 0.43855661153793335,
-      "learning_rate": 8.238628888266241e-06,
-      "loss": 0.3674,
-      "step": 13322
-    },
-    {
-      "epoch": 0.8707274034376838,
-      "grad_norm": 0.4502992630004883,
-      "learning_rate": 8.238362841735891e-06,
-      "loss": 0.3542,
-      "step": 13323
-    },
-    {
-      "epoch": 0.8707927586432259,
-      "grad_norm": 0.42601001262664795,
-      "learning_rate": 8.238096779410923e-06,
-      "loss": 0.3577,
-      "step": 13324
-    },
-    {
-      "epoch": 0.8708581138487681,
-      "grad_norm": 0.4329441487789154,
-      "learning_rate": 8.237830701292634e-06,
-      "loss": 0.378,
-      "step": 13325
-    },
-    {
-      "epoch": 0.8709234690543102,
-      "grad_norm": 0.45150476694107056,
-      "learning_rate": 8.237564607382328e-06,
-      "loss": 0.3538,
-      "step": 13326
-    },
-    {
-      "epoch": 0.8709888242598522,
-      "grad_norm": 0.4647296369075775,
-      "learning_rate": 8.237298497681292e-06,
-      "loss": 0.4128,
-      "step": 13327
-    },
-    {
-      "epoch": 0.8710541794653944,
-      "grad_norm": 0.42833244800567627,
-      "learning_rate": 8.237032372190832e-06,
-      "loss": 0.3441,
-      "step": 13328
-    },
-    {
-      "epoch": 0.8711195346709365,
-      "grad_norm": 0.4719427227973938,
-      "learning_rate": 8.236766230912243e-06,
-      "loss": 0.3978,
-      "step": 13329
-    },
-    {
-      "epoch": 0.8711848898764787,
-      "grad_norm": 0.40451669692993164,
-      "learning_rate": 8.236500073846826e-06,
-      "loss": 0.3538,
-      "step": 13330
-    },
-    {
-      "epoch": 0.8712502450820208,
-      "grad_norm": 0.45975998044013977,
-      "learning_rate": 8.236233900995874e-06,
-      "loss": 0.3982,
-      "step": 13331
-    },
-    {
-      "epoch": 0.8713156002875629,
-      "grad_norm": 0.45589467883110046,
-      "learning_rate": 8.23596771236069e-06,
-      "loss": 0.3863,
-      "step": 13332
-    },
-    {
-      "epoch": 0.871380955493105,
-      "grad_norm": 0.44050973653793335,
-      "learning_rate": 8.235701507942571e-06,
-      "loss": 0.3772,
-      "step": 13333
-    },
-    {
-      "epoch": 0.8714463106986472,
-      "grad_norm": 0.47511225938796997,
-      "learning_rate": 8.235435287742813e-06,
-      "loss": 0.4241,
-      "step": 13334
-    },
-    {
-      "epoch": 0.8715116659041893,
-      "grad_norm": 0.45889803767204285,
-      "learning_rate": 8.235169051762718e-06,
-      "loss": 0.381,
-      "step": 13335
-    },
-    {
-      "epoch": 0.8715770211097313,
-      "grad_norm": 0.44785118103027344,
-      "learning_rate": 8.234902800003581e-06,
-      "loss": 0.3952,
-      "step": 13336
-    },
-    {
-      "epoch": 0.8716423763152735,
-      "grad_norm": 0.4250652492046356,
-      "learning_rate": 8.234636532466702e-06,
-      "loss": 0.3512,
-      "step": 13337
-    },
-    {
-      "epoch": 0.8717077315208156,
-      "grad_norm": 0.4104115664958954,
-      "learning_rate": 8.234370249153381e-06,
-      "loss": 0.3504,
-      "step": 13338
-    },
-    {
-      "epoch": 0.8717730867263578,
-      "grad_norm": 0.4380771815776825,
-      "learning_rate": 8.234103950064916e-06,
-      "loss": 0.3525,
-      "step": 13339
-    },
-    {
-      "epoch": 0.8718384419318999,
-      "grad_norm": 0.43101492524147034,
-      "learning_rate": 8.233837635202604e-06,
-      "loss": 0.3409,
-      "step": 13340
-    },
-    {
-      "epoch": 0.871903797137442,
-      "grad_norm": 0.4336869716644287,
-      "learning_rate": 8.23357130456775e-06,
-      "loss": 0.3367,
-      "step": 13341
-    },
-    {
-      "epoch": 0.8719691523429841,
-      "grad_norm": 0.46948251128196716,
-      "learning_rate": 8.233304958161643e-06,
-      "loss": 0.3894,
-      "step": 13342
-    },
-    {
-      "epoch": 0.8720345075485263,
-      "grad_norm": 0.4291236698627472,
-      "learning_rate": 8.233038595985592e-06,
-      "loss": 0.3703,
-      "step": 13343
-    },
-    {
-      "epoch": 0.8720998627540684,
-      "grad_norm": 0.42561033368110657,
-      "learning_rate": 8.23277221804089e-06,
-      "loss": 0.3664,
-      "step": 13344
-    },
-    {
-      "epoch": 0.8721652179596104,
-      "grad_norm": 0.4196188747882843,
-      "learning_rate": 8.232505824328837e-06,
-      "loss": 0.343,
-      "step": 13345
-    },
-    {
-      "epoch": 0.8722305731651526,
-      "grad_norm": 0.4520960748195648,
-      "learning_rate": 8.232239414850734e-06,
-      "loss": 0.3751,
-      "step": 13346
-    },
-    {
-      "epoch": 0.8722959283706947,
-      "grad_norm": 0.4551304578781128,
-      "learning_rate": 8.23197298960788e-06,
-      "loss": 0.3863,
-      "step": 13347
-    },
-    {
-      "epoch": 0.8723612835762369,
-      "grad_norm": 0.4123007655143738,
-      "learning_rate": 8.231706548601572e-06,
-      "loss": 0.355,
-      "step": 13348
-    },
-    {
-      "epoch": 0.872426638781779,
-      "grad_norm": 0.43818527460098267,
-      "learning_rate": 8.231440091833113e-06,
-      "loss": 0.3584,
-      "step": 13349
-    },
-    {
-      "epoch": 0.8724919939873211,
-      "grad_norm": 0.45201465487480164,
-      "learning_rate": 8.231173619303802e-06,
-      "loss": 0.3575,
-      "step": 13350
-    },
-    {
-      "epoch": 0.8725573491928632,
-      "grad_norm": 0.42593809962272644,
-      "learning_rate": 8.23090713101494e-06,
-      "loss": 0.3527,
-      "step": 13351
-    },
-    {
-      "epoch": 0.8726227043984053,
-      "grad_norm": 0.4379211366176605,
-      "learning_rate": 8.230640626967821e-06,
-      "loss": 0.397,
-      "step": 13352
-    },
-    {
-      "epoch": 0.8726880596039475,
-      "grad_norm": 0.4190900921821594,
-      "learning_rate": 8.23037410716375e-06,
-      "loss": 0.3235,
-      "step": 13353
-    },
-    {
-      "epoch": 0.8727534148094895,
-      "grad_norm": 0.4702818691730499,
-      "learning_rate": 8.230107571604025e-06,
-      "loss": 0.4428,
-      "step": 13354
-    },
-    {
-      "epoch": 0.8728187700150317,
-      "grad_norm": 0.48189520835876465,
-      "learning_rate": 8.229841020289947e-06,
-      "loss": 0.3803,
-      "step": 13355
-    },
-    {
-      "epoch": 0.8728841252205738,
-      "grad_norm": 0.4529944360256195,
-      "learning_rate": 8.229574453222812e-06,
-      "loss": 0.354,
-      "step": 13356
-    },
-    {
-      "epoch": 0.872949480426116,
-      "grad_norm": 0.4514579772949219,
-      "learning_rate": 8.229307870403928e-06,
-      "loss": 0.3307,
-      "step": 13357
-    },
-    {
-      "epoch": 0.873014835631658,
-      "grad_norm": 0.4548887312412262,
-      "learning_rate": 8.229041271834588e-06,
-      "loss": 0.4085,
-      "step": 13358
-    },
-    {
-      "epoch": 0.8730801908372002,
-      "grad_norm": 0.4029901325702667,
-      "learning_rate": 8.228774657516097e-06,
-      "loss": 0.3146,
-      "step": 13359
-    },
-    {
-      "epoch": 0.8731455460427423,
-      "grad_norm": 0.4346332252025604,
-      "learning_rate": 8.228508027449752e-06,
-      "loss": 0.3511,
-      "step": 13360
-    },
-    {
-      "epoch": 0.8732109012482844,
-      "grad_norm": 0.4323817491531372,
-      "learning_rate": 8.228241381636855e-06,
-      "loss": 0.3758,
-      "step": 13361
-    },
-    {
-      "epoch": 0.8732762564538266,
-      "grad_norm": 0.43761152029037476,
-      "learning_rate": 8.227974720078708e-06,
-      "loss": 0.383,
-      "step": 13362
-    },
-    {
-      "epoch": 0.8733416116593686,
-      "grad_norm": 0.4516158103942871,
-      "learning_rate": 8.227708042776608e-06,
-      "loss": 0.3827,
-      "step": 13363
-    },
-    {
-      "epoch": 0.8734069668649108,
-      "grad_norm": 0.40703877806663513,
-      "learning_rate": 8.227441349731858e-06,
-      "loss": 0.3404,
-      "step": 13364
-    },
-    {
-      "epoch": 0.8734723220704529,
-      "grad_norm": 0.4412977397441864,
-      "learning_rate": 8.227174640945759e-06,
-      "loss": 0.3778,
-      "step": 13365
-    },
-    {
-      "epoch": 0.8735376772759951,
-      "grad_norm": 0.4496159553527832,
-      "learning_rate": 8.226907916419611e-06,
-      "loss": 0.3676,
-      "step": 13366
-    },
-    {
-      "epoch": 0.8736030324815371,
-      "grad_norm": 0.43719109892845154,
-      "learning_rate": 8.226641176154715e-06,
-      "loss": 0.3709,
-      "step": 13367
-    },
-    {
-      "epoch": 0.8736683876870793,
-      "grad_norm": 0.4351934492588043,
-      "learning_rate": 8.226374420152372e-06,
-      "loss": 0.3732,
-      "step": 13368
-    },
-    {
-      "epoch": 0.8737337428926214,
-      "grad_norm": 0.4426000416278839,
-      "learning_rate": 8.226107648413885e-06,
-      "loss": 0.3802,
-      "step": 13369
-    },
-    {
-      "epoch": 0.8737990980981635,
-      "grad_norm": 0.4504286050796509,
-      "learning_rate": 8.225840860940554e-06,
-      "loss": 0.4121,
-      "step": 13370
-    },
-    {
-      "epoch": 0.8738644533037057,
-      "grad_norm": 0.4409453868865967,
-      "learning_rate": 8.225574057733676e-06,
-      "loss": 0.3669,
-      "step": 13371
-    },
-    {
-      "epoch": 0.8739298085092477,
-      "grad_norm": 0.44252267479896545,
-      "learning_rate": 8.225307238794558e-06,
-      "loss": 0.3754,
-      "step": 13372
-    },
-    {
-      "epoch": 0.8739951637147899,
-      "grad_norm": 0.45822015404701233,
-      "learning_rate": 8.2250404041245e-06,
-      "loss": 0.4182,
-      "step": 13373
-    },
-    {
-      "epoch": 0.874060518920332,
-      "grad_norm": 0.4433598220348358,
-      "learning_rate": 8.224773553724802e-06,
-      "loss": 0.3604,
-      "step": 13374
-    },
-    {
-      "epoch": 0.8741258741258742,
-      "grad_norm": 0.4358460307121277,
-      "learning_rate": 8.224506687596764e-06,
-      "loss": 0.375,
-      "step": 13375
-    },
-    {
-      "epoch": 0.8741912293314162,
-      "grad_norm": 0.4563209116458893,
-      "learning_rate": 8.224239805741692e-06,
-      "loss": 0.4319,
-      "step": 13376
-    },
-    {
-      "epoch": 0.8742565845369583,
-      "grad_norm": 0.4441712498664856,
-      "learning_rate": 8.223972908160884e-06,
-      "loss": 0.3805,
-      "step": 13377
-    },
-    {
-      "epoch": 0.8743219397425005,
-      "grad_norm": 0.4398649036884308,
-      "learning_rate": 8.223705994855646e-06,
-      "loss": 0.3994,
-      "step": 13378
-    },
-    {
-      "epoch": 0.8743872949480426,
-      "grad_norm": 0.4517597258090973,
-      "learning_rate": 8.223439065827274e-06,
-      "loss": 0.4171,
-      "step": 13379
-    },
-    {
-      "epoch": 0.8744526501535848,
-      "grad_norm": 0.42800846695899963,
-      "learning_rate": 8.223172121077074e-06,
-      "loss": 0.3614,
-      "step": 13380
-    },
-    {
-      "epoch": 0.8745180053591268,
-      "grad_norm": 0.4026123881340027,
-      "learning_rate": 8.22290516060635e-06,
-      "loss": 0.3238,
-      "step": 13381
-    },
-    {
-      "epoch": 0.874583360564669,
-      "grad_norm": 0.3890553116798401,
-      "learning_rate": 8.222638184416397e-06,
-      "loss": 0.2857,
-      "step": 13382
-    },
-    {
-      "epoch": 0.8746487157702111,
-      "grad_norm": 0.4353051781654358,
-      "learning_rate": 8.222371192508522e-06,
-      "loss": 0.3355,
-      "step": 13383
-    },
-    {
-      "epoch": 0.8747140709757533,
-      "grad_norm": 0.43043285608291626,
-      "learning_rate": 8.222104184884026e-06,
-      "loss": 0.36,
-      "step": 13384
-    },
-    {
-      "epoch": 0.8747794261812953,
-      "grad_norm": 0.42799270153045654,
-      "learning_rate": 8.221837161544212e-06,
-      "loss": 0.373,
-      "step": 13385
-    },
-    {
-      "epoch": 0.8748447813868374,
-      "grad_norm": 0.4003741443157196,
-      "learning_rate": 8.221570122490384e-06,
-      "loss": 0.3133,
-      "step": 13386
-    },
-    {
-      "epoch": 0.8749101365923796,
-      "grad_norm": 0.4339466989040375,
-      "learning_rate": 8.22130306772384e-06,
-      "loss": 0.3597,
-      "step": 13387
-    },
-    {
-      "epoch": 0.8749754917979217,
-      "grad_norm": 0.45489901304244995,
-      "learning_rate": 8.221035997245886e-06,
-      "loss": 0.412,
-      "step": 13388
-    },
-    {
-      "epoch": 0.8750408470034639,
-      "grad_norm": 0.42823946475982666,
-      "learning_rate": 8.220768911057823e-06,
-      "loss": 0.3535,
-      "step": 13389
-    },
-    {
-      "epoch": 0.8751062022090059,
-      "grad_norm": 0.4632386863231659,
-      "learning_rate": 8.220501809160955e-06,
-      "loss": 0.3883,
-      "step": 13390
-    },
-    {
-      "epoch": 0.8751715574145481,
-      "grad_norm": 0.39702773094177246,
-      "learning_rate": 8.220234691556583e-06,
-      "loss": 0.309,
-      "step": 13391
-    },
-    {
-      "epoch": 0.8752369126200902,
-      "grad_norm": 0.4396913945674896,
-      "learning_rate": 8.219967558246013e-06,
-      "loss": 0.3775,
-      "step": 13392
-    },
-    {
-      "epoch": 0.8753022678256324,
-      "grad_norm": 0.4301755428314209,
-      "learning_rate": 8.219700409230545e-06,
-      "loss": 0.3493,
-      "step": 13393
-    },
-    {
-      "epoch": 0.8753676230311744,
-      "grad_norm": 0.49475932121276855,
-      "learning_rate": 8.219433244511481e-06,
-      "loss": 0.4561,
-      "step": 13394
-    },
-    {
-      "epoch": 0.8754329782367165,
-      "grad_norm": 0.4847348928451538,
-      "learning_rate": 8.219166064090127e-06,
-      "loss": 0.423,
-      "step": 13395
-    },
-    {
-      "epoch": 0.8754983334422587,
-      "grad_norm": 0.45300471782684326,
-      "learning_rate": 8.218898867967785e-06,
-      "loss": 0.3565,
-      "step": 13396
-    },
-    {
-      "epoch": 0.8755636886478008,
-      "grad_norm": 0.4917092025279999,
-      "learning_rate": 8.218631656145757e-06,
-      "loss": 0.4421,
-      "step": 13397
-    },
-    {
-      "epoch": 0.875629043853343,
-      "grad_norm": 0.4750673174858093,
-      "learning_rate": 8.218364428625347e-06,
-      "loss": 0.3823,
-      "step": 13398
-    },
-    {
-      "epoch": 0.875694399058885,
-      "grad_norm": 0.4574526846408844,
-      "learning_rate": 8.21809718540786e-06,
-      "loss": 0.411,
-      "step": 13399
-    },
-    {
-      "epoch": 0.8757597542644272,
-      "grad_norm": 0.43643367290496826,
-      "learning_rate": 8.217829926494598e-06,
-      "loss": 0.3821,
-      "step": 13400
-    },
-    {
-      "epoch": 0.8758251094699693,
-      "grad_norm": 0.4311563968658447,
-      "learning_rate": 8.217562651886866e-06,
-      "loss": 0.326,
-      "step": 13401
-    },
-    {
-      "epoch": 0.8758904646755115,
-      "grad_norm": 0.4255685806274414,
-      "learning_rate": 8.217295361585964e-06,
-      "loss": 0.3526,
-      "step": 13402
-    },
-    {
-      "epoch": 0.8759558198810535,
-      "grad_norm": 0.4884625971317291,
-      "learning_rate": 8.2170280555932e-06,
-      "loss": 0.3865,
-      "step": 13403
-    },
-    {
-      "epoch": 0.8760211750865956,
-      "grad_norm": 0.4539845585823059,
-      "learning_rate": 8.216760733909874e-06,
-      "loss": 0.3797,
-      "step": 13404
-    },
-    {
-      "epoch": 0.8760865302921378,
-      "grad_norm": 0.44040000438690186,
-      "learning_rate": 8.216493396537291e-06,
-      "loss": 0.3672,
-      "step": 13405
-    },
-    {
-      "epoch": 0.8761518854976799,
-      "grad_norm": 0.43532800674438477,
-      "learning_rate": 8.216226043476757e-06,
-      "loss": 0.3606,
-      "step": 13406
-    },
-    {
-      "epoch": 0.876217240703222,
-      "grad_norm": 0.46042540669441223,
-      "learning_rate": 8.215958674729572e-06,
-      "loss": 0.4519,
-      "step": 13407
-    },
-    {
-      "epoch": 0.8762825959087641,
-      "grad_norm": 0.4292353391647339,
-      "learning_rate": 8.215691290297045e-06,
-      "loss": 0.3696,
-      "step": 13408
-    },
-    {
-      "epoch": 0.8763479511143063,
-      "grad_norm": 0.4638332426548004,
-      "learning_rate": 8.215423890180478e-06,
-      "loss": 0.4069,
-      "step": 13409
-    },
-    {
-      "epoch": 0.8764133063198484,
-      "grad_norm": 0.4396990239620209,
-      "learning_rate": 8.215156474381173e-06,
-      "loss": 0.4132,
-      "step": 13410
-    },
-    {
-      "epoch": 0.8764786615253904,
-      "grad_norm": 0.4405803978443146,
-      "learning_rate": 8.214889042900436e-06,
-      "loss": 0.3993,
-      "step": 13411
-    },
-    {
-      "epoch": 0.8765440167309326,
-      "grad_norm": 0.4370073676109314,
-      "learning_rate": 8.214621595739571e-06,
-      "loss": 0.3423,
-      "step": 13412
-    },
-    {
-      "epoch": 0.8766093719364747,
-      "grad_norm": 0.47184428572654724,
-      "learning_rate": 8.214354132899884e-06,
-      "loss": 0.4272,
-      "step": 13413
-    },
-    {
-      "epoch": 0.8766747271420169,
-      "grad_norm": 0.4257681667804718,
-      "learning_rate": 8.214086654382681e-06,
-      "loss": 0.3701,
-      "step": 13414
-    },
-    {
-      "epoch": 0.876740082347559,
-      "grad_norm": 0.4332144558429718,
-      "learning_rate": 8.213819160189261e-06,
-      "loss": 0.3746,
-      "step": 13415
-    },
-    {
-      "epoch": 0.8768054375531011,
-      "grad_norm": 0.4621717929840088,
-      "learning_rate": 8.21355165032093e-06,
-      "loss": 0.3906,
-      "step": 13416
-    },
-    {
-      "epoch": 0.8768707927586432,
-      "grad_norm": 0.40356600284576416,
-      "learning_rate": 8.213284124778994e-06,
-      "loss": 0.3285,
-      "step": 13417
-    },
-    {
-      "epoch": 0.8769361479641854,
-      "grad_norm": 0.43172144889831543,
-      "learning_rate": 8.213016583564761e-06,
-      "loss": 0.381,
-      "step": 13418
-    },
-    {
-      "epoch": 0.8770015031697275,
-      "grad_norm": 0.4825325906276703,
-      "learning_rate": 8.212749026679533e-06,
-      "loss": 0.4114,
-      "step": 13419
-    },
-    {
-      "epoch": 0.8770668583752695,
-      "grad_norm": 0.45453810691833496,
-      "learning_rate": 8.212481454124615e-06,
-      "loss": 0.3914,
-      "step": 13420
-    },
-    {
-      "epoch": 0.8771322135808117,
-      "grad_norm": 0.4427175521850586,
-      "learning_rate": 8.21221386590131e-06,
-      "loss": 0.3682,
-      "step": 13421
-    },
-    {
-      "epoch": 0.8771975687863538,
-      "grad_norm": 0.4396561086177826,
-      "learning_rate": 8.211946262010925e-06,
-      "loss": 0.4073,
-      "step": 13422
-    },
-    {
-      "epoch": 0.877262923991896,
-      "grad_norm": 0.4252471625804901,
-      "learning_rate": 8.211678642454768e-06,
-      "loss": 0.3584,
-      "step": 13423
-    },
-    {
-      "epoch": 0.877328279197438,
-      "grad_norm": 0.45310062170028687,
-      "learning_rate": 8.21141100723414e-06,
-      "loss": 0.3729,
-      "step": 13424
-    },
-    {
-      "epoch": 0.8773936344029802,
-      "grad_norm": 0.44264861941337585,
-      "learning_rate": 8.211143356350348e-06,
-      "loss": 0.3864,
-      "step": 13425
-    },
-    {
-      "epoch": 0.8774589896085223,
-      "grad_norm": 0.41715845465660095,
-      "learning_rate": 8.210875689804699e-06,
-      "loss": 0.3247,
-      "step": 13426
-    },
-    {
-      "epoch": 0.8775243448140645,
-      "grad_norm": 0.4420039653778076,
-      "learning_rate": 8.210608007598495e-06,
-      "loss": 0.363,
-      "step": 13427
-    },
-    {
-      "epoch": 0.8775897000196066,
-      "grad_norm": 0.45903250575065613,
-      "learning_rate": 8.210340309733043e-06,
-      "loss": 0.3951,
-      "step": 13428
-    },
-    {
-      "epoch": 0.8776550552251486,
-      "grad_norm": 0.4627886116504669,
-      "learning_rate": 8.21007259620965e-06,
-      "loss": 0.4165,
-      "step": 13429
-    },
-    {
-      "epoch": 0.8777204104306908,
-      "grad_norm": 0.428143709897995,
-      "learning_rate": 8.209804867029623e-06,
-      "loss": 0.3477,
-      "step": 13430
-    },
-    {
-      "epoch": 0.8777857656362329,
-      "grad_norm": 0.4345650374889374,
-      "learning_rate": 8.209537122194262e-06,
-      "loss": 0.385,
-      "step": 13431
-    },
-    {
-      "epoch": 0.8778511208417751,
-      "grad_norm": 0.44252845644950867,
-      "learning_rate": 8.20926936170488e-06,
-      "loss": 0.3801,
-      "step": 13432
-    },
-    {
-      "epoch": 0.8779164760473172,
-      "grad_norm": 0.43946585059165955,
-      "learning_rate": 8.209001585562777e-06,
-      "loss": 0.3596,
-      "step": 13433
-    },
-    {
-      "epoch": 0.8779818312528593,
-      "grad_norm": 0.4799671471118927,
-      "learning_rate": 8.208733793769261e-06,
-      "loss": 0.4373,
-      "step": 13434
-    },
-    {
-      "epoch": 0.8780471864584014,
-      "grad_norm": 0.4021196663379669,
-      "learning_rate": 8.208465986325642e-06,
-      "loss": 0.3167,
-      "step": 13435
-    },
-    {
-      "epoch": 0.8781125416639435,
-      "grad_norm": 0.44872426986694336,
-      "learning_rate": 8.20819816323322e-06,
-      "loss": 0.3781,
-      "step": 13436
-    },
-    {
-      "epoch": 0.8781778968694857,
-      "grad_norm": 0.43821343779563904,
-      "learning_rate": 8.207930324493304e-06,
-      "loss": 0.3715,
-      "step": 13437
-    },
-    {
-      "epoch": 0.8782432520750277,
-      "grad_norm": 0.4665985107421875,
-      "learning_rate": 8.207662470107203e-06,
-      "loss": 0.4173,
-      "step": 13438
-    },
-    {
-      "epoch": 0.8783086072805699,
-      "grad_norm": 0.42889121174812317,
-      "learning_rate": 8.20739460007622e-06,
-      "loss": 0.3498,
-      "step": 13439
-    },
-    {
-      "epoch": 0.878373962486112,
-      "grad_norm": 0.48825544118881226,
-      "learning_rate": 8.207126714401661e-06,
-      "loss": 0.4491,
-      "step": 13440
-    },
-    {
-      "epoch": 0.8784393176916542,
-      "grad_norm": 0.443212628364563,
-      "learning_rate": 8.206858813084835e-06,
-      "loss": 0.39,
-      "step": 13441
-    },
-    {
-      "epoch": 0.8785046728971962,
-      "grad_norm": 0.417248010635376,
-      "learning_rate": 8.206590896127048e-06,
-      "loss": 0.3144,
-      "step": 13442
-    },
-    {
-      "epoch": 0.8785700281027384,
-      "grad_norm": 0.41721847653388977,
-      "learning_rate": 8.206322963529605e-06,
-      "loss": 0.3571,
-      "step": 13443
-    },
-    {
-      "epoch": 0.8786353833082805,
-      "grad_norm": 0.47829684615135193,
-      "learning_rate": 8.206055015293815e-06,
-      "loss": 0.4379,
-      "step": 13444
-    },
-    {
-      "epoch": 0.8787007385138226,
-      "grad_norm": 0.42701518535614014,
-      "learning_rate": 8.205787051420983e-06,
-      "loss": 0.3682,
-      "step": 13445
-    },
-    {
-      "epoch": 0.8787660937193648,
-      "grad_norm": 0.5073907375335693,
-      "learning_rate": 8.205519071912418e-06,
-      "loss": 0.372,
-      "step": 13446
-    },
-    {
-      "epoch": 0.8788314489249068,
-      "grad_norm": 0.44476792216300964,
-      "learning_rate": 8.205251076769427e-06,
-      "loss": 0.4086,
-      "step": 13447
-    },
-    {
-      "epoch": 0.878896804130449,
-      "grad_norm": 0.4226808547973633,
-      "learning_rate": 8.204983065993315e-06,
-      "loss": 0.3642,
-      "step": 13448
-    },
-    {
-      "epoch": 0.8789621593359911,
-      "grad_norm": 0.44482994079589844,
-      "learning_rate": 8.204715039585389e-06,
-      "loss": 0.3877,
-      "step": 13449
-    },
-    {
-      "epoch": 0.8790275145415333,
-      "grad_norm": 0.4508343040943146,
-      "learning_rate": 8.20444699754696e-06,
-      "loss": 0.3891,
-      "step": 13450
-    },
-    {
-      "epoch": 0.8790928697470753,
-      "grad_norm": 0.4405125081539154,
-      "learning_rate": 8.20417893987933e-06,
-      "loss": 0.3835,
-      "step": 13451
-    },
-    {
-      "epoch": 0.8791582249526175,
-      "grad_norm": 0.41821563243865967,
-      "learning_rate": 8.203910866583811e-06,
-      "loss": 0.3321,
-      "step": 13452
-    },
-    {
-      "epoch": 0.8792235801581596,
-      "grad_norm": 0.43777090311050415,
-      "learning_rate": 8.203642777661708e-06,
-      "loss": 0.3788,
-      "step": 13453
-    },
-    {
-      "epoch": 0.8792889353637017,
-      "grad_norm": 0.4499454200267792,
-      "learning_rate": 8.20337467311433e-06,
-      "loss": 0.3819,
-      "step": 13454
-    },
-    {
-      "epoch": 0.8793542905692439,
-      "grad_norm": 0.44110482931137085,
-      "learning_rate": 8.203106552942985e-06,
-      "loss": 0.3551,
-      "step": 13455
-    },
-    {
-      "epoch": 0.8794196457747859,
-      "grad_norm": 0.4749086797237396,
-      "learning_rate": 8.202838417148979e-06,
-      "loss": 0.4466,
-      "step": 13456
-    },
-    {
-      "epoch": 0.8794850009803281,
-      "grad_norm": 0.4303574860095978,
-      "learning_rate": 8.202570265733619e-06,
-      "loss": 0.3556,
-      "step": 13457
-    },
-    {
-      "epoch": 0.8795503561858702,
-      "grad_norm": 0.46187305450439453,
-      "learning_rate": 8.202302098698215e-06,
-      "loss": 0.4206,
-      "step": 13458
-    },
-    {
-      "epoch": 0.8796157113914124,
-      "grad_norm": 0.4228116571903229,
-      "learning_rate": 8.202033916044076e-06,
-      "loss": 0.3433,
-      "step": 13459
-    },
-    {
-      "epoch": 0.8796810665969544,
-      "grad_norm": 0.49848175048828125,
-      "learning_rate": 8.201765717772507e-06,
-      "loss": 0.3984,
-      "step": 13460
-    },
-    {
-      "epoch": 0.8797464218024965,
-      "grad_norm": 0.4178866147994995,
-      "learning_rate": 8.201497503884816e-06,
-      "loss": 0.3018,
-      "step": 13461
-    },
-    {
-      "epoch": 0.8798117770080387,
-      "grad_norm": 0.46181097626686096,
-      "learning_rate": 8.201229274382315e-06,
-      "loss": 0.4107,
-      "step": 13462
-    },
-    {
-      "epoch": 0.8798771322135808,
-      "grad_norm": 0.44718632102012634,
-      "learning_rate": 8.200961029266308e-06,
-      "loss": 0.366,
-      "step": 13463
-    },
-    {
-      "epoch": 0.879942487419123,
-      "grad_norm": 0.43121084570884705,
-      "learning_rate": 8.200692768538105e-06,
-      "loss": 0.355,
-      "step": 13464
-    },
-    {
-      "epoch": 0.880007842624665,
-      "grad_norm": 0.5150291919708252,
-      "learning_rate": 8.200424492199017e-06,
-      "loss": 0.4489,
-      "step": 13465
-    },
-    {
-      "epoch": 0.8800731978302072,
-      "grad_norm": 0.4343971610069275,
-      "learning_rate": 8.200156200250348e-06,
-      "loss": 0.3467,
-      "step": 13466
-    },
-    {
-      "epoch": 0.8801385530357493,
-      "grad_norm": 0.46228814125061035,
-      "learning_rate": 8.199887892693409e-06,
-      "loss": 0.4217,
-      "step": 13467
-    },
-    {
-      "epoch": 0.8802039082412915,
-      "grad_norm": 0.45131438970565796,
-      "learning_rate": 8.19961956952951e-06,
-      "loss": 0.394,
-      "step": 13468
-    },
-    {
-      "epoch": 0.8802692634468335,
-      "grad_norm": 0.48295700550079346,
-      "learning_rate": 8.199351230759955e-06,
-      "loss": 0.4558,
-      "step": 13469
-    },
-    {
-      "epoch": 0.8803346186523756,
-      "grad_norm": 0.436484158039093,
-      "learning_rate": 8.199082876386056e-06,
-      "loss": 0.3622,
-      "step": 13470
-    },
-    {
-      "epoch": 0.8803999738579178,
-      "grad_norm": 0.39471927285194397,
-      "learning_rate": 8.198814506409122e-06,
-      "loss": 0.3218,
-      "step": 13471
-    },
-    {
-      "epoch": 0.8804653290634599,
-      "grad_norm": 0.4155243933200836,
-      "learning_rate": 8.198546120830462e-06,
-      "loss": 0.366,
-      "step": 13472
-    },
-    {
-      "epoch": 0.880530684269002,
-      "grad_norm": 0.4520508944988251,
-      "learning_rate": 8.198277719651384e-06,
-      "loss": 0.3685,
-      "step": 13473
-    },
-    {
-      "epoch": 0.8805960394745441,
-      "grad_norm": 0.4244966506958008,
-      "learning_rate": 8.198009302873198e-06,
-      "loss": 0.3691,
-      "step": 13474
-    },
-    {
-      "epoch": 0.8806613946800863,
-      "grad_norm": 0.4450370967388153,
-      "learning_rate": 8.197740870497212e-06,
-      "loss": 0.4094,
-      "step": 13475
-    },
-    {
-      "epoch": 0.8807267498856284,
-      "grad_norm": 0.4767009913921356,
-      "learning_rate": 8.197472422524738e-06,
-      "loss": 0.4617,
-      "step": 13476
-    },
-    {
-      "epoch": 0.8807921050911706,
-      "grad_norm": 0.45346152782440186,
-      "learning_rate": 8.197203958957082e-06,
-      "loss": 0.3877,
-      "step": 13477
-    },
-    {
-      "epoch": 0.8808574602967126,
-      "grad_norm": 0.4302191138267517,
-      "learning_rate": 8.196935479795555e-06,
-      "loss": 0.3602,
-      "step": 13478
-    },
-    {
-      "epoch": 0.8809228155022547,
-      "grad_norm": 0.43985727429389954,
-      "learning_rate": 8.196666985041465e-06,
-      "loss": 0.375,
-      "step": 13479
-    },
-    {
-      "epoch": 0.8809881707077969,
-      "grad_norm": 0.4335164725780487,
-      "learning_rate": 8.196398474696123e-06,
-      "loss": 0.3597,
-      "step": 13480
-    },
-    {
-      "epoch": 0.881053525913339,
-      "grad_norm": 0.4480811059474945,
-      "learning_rate": 8.196129948760839e-06,
-      "loss": 0.396,
-      "step": 13481
-    },
-    {
-      "epoch": 0.8811188811188811,
-      "grad_norm": 0.43360355496406555,
-      "learning_rate": 8.195861407236921e-06,
-      "loss": 0.3488,
-      "step": 13482
-    },
-    {
-      "epoch": 0.8811842363244232,
-      "grad_norm": 0.42258724570274353,
-      "learning_rate": 8.195592850125681e-06,
-      "loss": 0.3353,
-      "step": 13483
-    },
-    {
-      "epoch": 0.8812495915299654,
-      "grad_norm": 0.43493205308914185,
-      "learning_rate": 8.195324277428427e-06,
-      "loss": 0.3353,
-      "step": 13484
-    },
-    {
-      "epoch": 0.8813149467355075,
-      "grad_norm": 0.4324595332145691,
-      "learning_rate": 8.195055689146469e-06,
-      "loss": 0.3577,
-      "step": 13485
-    },
-    {
-      "epoch": 0.8813803019410497,
-      "grad_norm": 0.4718596935272217,
-      "learning_rate": 8.194787085281118e-06,
-      "loss": 0.3921,
-      "step": 13486
-    },
-    {
-      "epoch": 0.8814456571465917,
-      "grad_norm": 0.41776734590530396,
-      "learning_rate": 8.194518465833684e-06,
-      "loss": 0.3356,
-      "step": 13487
-    },
-    {
-      "epoch": 0.8815110123521338,
-      "grad_norm": 0.4012083411216736,
-      "learning_rate": 8.194249830805476e-06,
-      "loss": 0.3277,
-      "step": 13488
-    },
-    {
-      "epoch": 0.881576367557676,
-      "grad_norm": 0.45053204894065857,
-      "learning_rate": 8.193981180197806e-06,
-      "loss": 0.3435,
-      "step": 13489
-    },
-    {
-      "epoch": 0.8816417227632181,
-      "grad_norm": 0.43780970573425293,
-      "learning_rate": 8.193712514011982e-06,
-      "loss": 0.3802,
-      "step": 13490
-    },
-    {
-      "epoch": 0.8817070779687602,
-      "grad_norm": 0.445178359746933,
-      "learning_rate": 8.193443832249316e-06,
-      "loss": 0.3613,
-      "step": 13491
-    },
-    {
-      "epoch": 0.8817724331743023,
-      "grad_norm": 0.4471881687641144,
-      "learning_rate": 8.19317513491112e-06,
-      "loss": 0.3854,
-      "step": 13492
-    },
-    {
-      "epoch": 0.8818377883798445,
-      "grad_norm": 0.41607406735420227,
-      "learning_rate": 8.192906421998701e-06,
-      "loss": 0.3689,
-      "step": 13493
-    },
-    {
-      "epoch": 0.8819031435853866,
-      "grad_norm": 0.4103538990020752,
-      "learning_rate": 8.192637693513372e-06,
-      "loss": 0.3649,
-      "step": 13494
-    },
-    {
-      "epoch": 0.8819684987909286,
-      "grad_norm": 0.47198987007141113,
-      "learning_rate": 8.192368949456441e-06,
-      "loss": 0.4608,
-      "step": 13495
-    },
-    {
-      "epoch": 0.8820338539964708,
-      "grad_norm": 0.4210141897201538,
-      "learning_rate": 8.192100189829222e-06,
-      "loss": 0.3392,
-      "step": 13496
-    },
-    {
-      "epoch": 0.8820992092020129,
-      "grad_norm": 0.47057467699050903,
-      "learning_rate": 8.191831414633024e-06,
-      "loss": 0.4058,
-      "step": 13497
-    },
-    {
-      "epoch": 0.8821645644075551,
-      "grad_norm": 0.4252711832523346,
-      "learning_rate": 8.19156262386916e-06,
-      "loss": 0.3404,
-      "step": 13498
-    },
-    {
-      "epoch": 0.8822299196130972,
-      "grad_norm": 0.4368647038936615,
-      "learning_rate": 8.19129381753894e-06,
-      "loss": 0.3897,
-      "step": 13499
-    },
-    {
-      "epoch": 0.8822952748186393,
-      "grad_norm": 0.4264031946659088,
-      "learning_rate": 8.191024995643672e-06,
-      "loss": 0.3994,
-      "step": 13500
-    },
-    {
-      "epoch": 0.8823606300241814,
-      "grad_norm": 0.4451930820941925,
-      "learning_rate": 8.19075615818467e-06,
-      "loss": 0.3871,
-      "step": 13501
-    },
-    {
-      "epoch": 0.8824259852297236,
-      "grad_norm": 0.435319721698761,
-      "learning_rate": 8.190487305163245e-06,
-      "loss": 0.3474,
-      "step": 13502
-    },
-    {
-      "epoch": 0.8824913404352657,
-      "grad_norm": 0.4761326014995575,
-      "learning_rate": 8.190218436580707e-06,
-      "loss": 0.3979,
-      "step": 13503
-    },
-    {
-      "epoch": 0.8825566956408077,
-      "grad_norm": 0.4505530595779419,
-      "learning_rate": 8.189949552438373e-06,
-      "loss": 0.3187,
-      "step": 13504
-    },
-    {
-      "epoch": 0.8826220508463499,
-      "grad_norm": 0.4660145342350006,
-      "learning_rate": 8.189680652737546e-06,
-      "loss": 0.4287,
-      "step": 13505
-    },
-    {
-      "epoch": 0.882687406051892,
-      "grad_norm": 0.41383031010627747,
-      "learning_rate": 8.189411737479542e-06,
-      "loss": 0.3507,
-      "step": 13506
-    },
-    {
-      "epoch": 0.8827527612574342,
-      "grad_norm": 0.46919748187065125,
-      "learning_rate": 8.189142806665672e-06,
-      "loss": 0.4421,
-      "step": 13507
-    },
-    {
-      "epoch": 0.8828181164629763,
-      "grad_norm": 0.4191434383392334,
-      "learning_rate": 8.188873860297248e-06,
-      "loss": 0.3416,
-      "step": 13508
-    },
-    {
-      "epoch": 0.8828834716685184,
-      "grad_norm": 0.4576224088668823,
-      "learning_rate": 8.18860489837558e-06,
-      "loss": 0.4024,
-      "step": 13509
-    },
-    {
-      "epoch": 0.8829488268740605,
-      "grad_norm": 0.4578748047351837,
-      "learning_rate": 8.188335920901984e-06,
-      "loss": 0.4308,
-      "step": 13510
-    },
-    {
-      "epoch": 0.8830141820796027,
-      "grad_norm": 0.4221654236316681,
-      "learning_rate": 8.188066927877769e-06,
-      "loss": 0.3446,
-      "step": 13511
-    },
-    {
-      "epoch": 0.8830795372851448,
-      "grad_norm": 0.43121111392974854,
-      "learning_rate": 8.187797919304246e-06,
-      "loss": 0.3931,
-      "step": 13512
-    },
-    {
-      "epoch": 0.8831448924906868,
-      "grad_norm": 0.49072685837745667,
-      "learning_rate": 8.187528895182727e-06,
-      "loss": 0.4501,
-      "step": 13513
-    },
-    {
-      "epoch": 0.883210247696229,
-      "grad_norm": 0.46443790197372437,
-      "learning_rate": 8.187259855514527e-06,
-      "loss": 0.414,
-      "step": 13514
-    },
-    {
-      "epoch": 0.8832756029017711,
-      "grad_norm": 0.44713345170021057,
-      "learning_rate": 8.186990800300956e-06,
-      "loss": 0.3852,
-      "step": 13515
-    },
-    {
-      "epoch": 0.8833409581073133,
-      "grad_norm": 0.4597233533859253,
-      "learning_rate": 8.186721729543326e-06,
-      "loss": 0.3726,
-      "step": 13516
-    },
-    {
-      "epoch": 0.8834063133128554,
-      "grad_norm": 0.42686402797698975,
-      "learning_rate": 8.186452643242952e-06,
-      "loss": 0.3824,
-      "step": 13517
-    },
-    {
-      "epoch": 0.8834716685183975,
-      "grad_norm": 0.4627307057380676,
-      "learning_rate": 8.186183541401142e-06,
-      "loss": 0.3968,
-      "step": 13518
-    },
-    {
-      "epoch": 0.8835370237239396,
-      "grad_norm": 0.45088204741477966,
-      "learning_rate": 8.185914424019214e-06,
-      "loss": 0.3726,
-      "step": 13519
-    },
-    {
-      "epoch": 0.8836023789294817,
-      "grad_norm": 0.5059247612953186,
-      "learning_rate": 8.185645291098476e-06,
-      "loss": 0.4832,
-      "step": 13520
-    },
-    {
-      "epoch": 0.8836677341350239,
-      "grad_norm": 0.4470418691635132,
-      "learning_rate": 8.185376142640244e-06,
-      "loss": 0.3688,
-      "step": 13521
-    },
-    {
-      "epoch": 0.8837330893405659,
-      "grad_norm": 0.4880659580230713,
-      "learning_rate": 8.185106978645827e-06,
-      "loss": 0.4504,
-      "step": 13522
-    },
-    {
-      "epoch": 0.8837984445461081,
-      "grad_norm": 0.5302170515060425,
-      "learning_rate": 8.18483779911654e-06,
-      "loss": 0.4002,
-      "step": 13523
-    },
-    {
-      "epoch": 0.8838637997516502,
-      "grad_norm": 0.5098953247070312,
-      "learning_rate": 8.184568604053696e-06,
-      "loss": 0.4668,
-      "step": 13524
-    },
-    {
-      "epoch": 0.8839291549571924,
-      "grad_norm": 0.49490153789520264,
-      "learning_rate": 8.184299393458608e-06,
-      "loss": 0.3942,
-      "step": 13525
-    },
-    {
-      "epoch": 0.8839945101627344,
-      "grad_norm": 0.4526436924934387,
-      "learning_rate": 8.184030167332589e-06,
-      "loss": 0.3904,
-      "step": 13526
-    },
-    {
-      "epoch": 0.8840598653682766,
-      "grad_norm": 0.4353998303413391,
-      "learning_rate": 8.183760925676951e-06,
-      "loss": 0.36,
-      "step": 13527
-    },
-    {
-      "epoch": 0.8841252205738187,
-      "grad_norm": 0.4766745865345001,
-      "learning_rate": 8.183491668493009e-06,
-      "loss": 0.4335,
-      "step": 13528
-    },
-    {
-      "epoch": 0.8841905757793608,
-      "grad_norm": 0.45100441575050354,
-      "learning_rate": 8.183222395782074e-06,
-      "loss": 0.366,
-      "step": 13529
-    },
-    {
-      "epoch": 0.884255930984903,
-      "grad_norm": 0.4555814266204834,
-      "learning_rate": 8.182953107545462e-06,
-      "loss": 0.353,
-      "step": 13530
-    },
-    {
-      "epoch": 0.884321286190445,
-      "grad_norm": 0.4688302278518677,
-      "learning_rate": 8.182683803784484e-06,
-      "loss": 0.4396,
-      "step": 13531
-    },
-    {
-      "epoch": 0.8843866413959872,
-      "grad_norm": 0.7728043794631958,
-      "learning_rate": 8.182414484500454e-06,
-      "loss": 0.36,
-      "step": 13532
-    },
-    {
-      "epoch": 0.8844519966015293,
-      "grad_norm": 0.41783252358436584,
-      "learning_rate": 8.182145149694687e-06,
-      "loss": 0.3638,
-      "step": 13533
-    },
-    {
-      "epoch": 0.8845173518070715,
-      "grad_norm": 0.4418119788169861,
-      "learning_rate": 8.181875799368496e-06,
-      "loss": 0.3429,
-      "step": 13534
-    },
-    {
-      "epoch": 0.8845827070126135,
-      "grad_norm": 0.42237022519111633,
-      "learning_rate": 8.181606433523193e-06,
-      "loss": 0.3164,
-      "step": 13535
-    },
-    {
-      "epoch": 0.8846480622181557,
-      "grad_norm": 0.466979444026947,
-      "learning_rate": 8.181337052160094e-06,
-      "loss": 0.3914,
-      "step": 13536
-    },
-    {
-      "epoch": 0.8847134174236978,
-      "grad_norm": 0.49458831548690796,
-      "learning_rate": 8.181067655280512e-06,
-      "loss": 0.4054,
-      "step": 13537
-    },
-    {
-      "epoch": 0.8847787726292399,
-      "grad_norm": 0.44746413826942444,
-      "learning_rate": 8.180798242885762e-06,
-      "loss": 0.4021,
-      "step": 13538
-    },
-    {
-      "epoch": 0.8848441278347821,
-      "grad_norm": 0.44023093581199646,
-      "learning_rate": 8.180528814977157e-06,
-      "loss": 0.3858,
-      "step": 13539
-    },
-    {
-      "epoch": 0.8849094830403241,
-      "grad_norm": 0.41607964038848877,
-      "learning_rate": 8.180259371556011e-06,
-      "loss": 0.3404,
-      "step": 13540
-    },
-    {
-      "epoch": 0.8849748382458663,
-      "grad_norm": 0.44271618127822876,
-      "learning_rate": 8.179989912623638e-06,
-      "loss": 0.3684,
-      "step": 13541
-    },
-    {
-      "epoch": 0.8850401934514084,
-      "grad_norm": 0.43409037590026855,
-      "learning_rate": 8.179720438181352e-06,
-      "loss": 0.3791,
-      "step": 13542
-    },
-    {
-      "epoch": 0.8851055486569506,
-      "grad_norm": 0.4831492602825165,
-      "learning_rate": 8.179450948230467e-06,
-      "loss": 0.3961,
-      "step": 13543
-    },
-    {
-      "epoch": 0.8851709038624926,
-      "grad_norm": 0.4891977310180664,
-      "learning_rate": 8.1791814427723e-06,
-      "loss": 0.4174,
-      "step": 13544
-    },
-    {
-      "epoch": 0.8852362590680347,
-      "grad_norm": 0.41980496048927307,
-      "learning_rate": 8.178911921808164e-06,
-      "loss": 0.3158,
-      "step": 13545
-    },
-    {
-      "epoch": 0.8853016142735769,
-      "grad_norm": 0.46518459916114807,
-      "learning_rate": 8.178642385339372e-06,
-      "loss": 0.3933,
-      "step": 13546
-    },
-    {
-      "epoch": 0.885366969479119,
-      "grad_norm": 0.43595242500305176,
-      "learning_rate": 8.178372833367239e-06,
-      "loss": 0.4049,
-      "step": 13547
-    },
-    {
-      "epoch": 0.8854323246846612,
-      "grad_norm": 0.4373858571052551,
-      "learning_rate": 8.178103265893082e-06,
-      "loss": 0.3689,
-      "step": 13548
-    },
-    {
-      "epoch": 0.8854976798902032,
-      "grad_norm": 0.41803237795829773,
-      "learning_rate": 8.177833682918215e-06,
-      "loss": 0.3161,
-      "step": 13549
-    },
-    {
-      "epoch": 0.8855630350957454,
-      "grad_norm": 0.4319811165332794,
-      "learning_rate": 8.177564084443951e-06,
-      "loss": 0.372,
-      "step": 13550
-    },
-    {
-      "epoch": 0.8856283903012875,
-      "grad_norm": 0.45131462812423706,
-      "learning_rate": 8.177294470471607e-06,
-      "loss": 0.4278,
-      "step": 13551
-    },
-    {
-      "epoch": 0.8856937455068297,
-      "grad_norm": 0.4333389401435852,
-      "learning_rate": 8.177024841002497e-06,
-      "loss": 0.3837,
-      "step": 13552
-    },
-    {
-      "epoch": 0.8857591007123717,
-      "grad_norm": 0.4520132541656494,
-      "learning_rate": 8.176755196037935e-06,
-      "loss": 0.3613,
-      "step": 13553
-    },
-    {
-      "epoch": 0.8858244559179138,
-      "grad_norm": 0.5520606637001038,
-      "learning_rate": 8.17648553557924e-06,
-      "loss": 0.3561,
-      "step": 13554
-    },
-    {
-      "epoch": 0.885889811123456,
-      "grad_norm": 0.4240726828575134,
-      "learning_rate": 8.176215859627722e-06,
-      "loss": 0.3279,
-      "step": 13555
-    },
-    {
-      "epoch": 0.8859551663289981,
-      "grad_norm": 0.44275182485580444,
-      "learning_rate": 8.1759461681847e-06,
-      "loss": 0.3705,
-      "step": 13556
-    },
-    {
-      "epoch": 0.8860205215345403,
-      "grad_norm": 0.4383064806461334,
-      "learning_rate": 8.175676461251488e-06,
-      "loss": 0.3734,
-      "step": 13557
-    },
-    {
-      "epoch": 0.8860858767400823,
-      "grad_norm": 0.4564521312713623,
-      "learning_rate": 8.175406738829402e-06,
-      "loss": 0.3925,
-      "step": 13558
-    },
-    {
-      "epoch": 0.8861512319456245,
-      "grad_norm": 0.45131510496139526,
-      "learning_rate": 8.175137000919758e-06,
-      "loss": 0.4209,
-      "step": 13559
-    },
-    {
-      "epoch": 0.8862165871511666,
-      "grad_norm": 0.43068280816078186,
-      "learning_rate": 8.17486724752387e-06,
-      "loss": 0.3589,
-      "step": 13560
-    },
-    {
-      "epoch": 0.8862819423567088,
-      "grad_norm": 0.4380665123462677,
-      "learning_rate": 8.174597478643055e-06,
-      "loss": 0.3863,
-      "step": 13561
-    },
-    {
-      "epoch": 0.8863472975622508,
-      "grad_norm": 0.44074466824531555,
-      "learning_rate": 8.174327694278627e-06,
-      "loss": 0.3839,
-      "step": 13562
-    },
-    {
-      "epoch": 0.8864126527677929,
-      "grad_norm": 0.43280836939811707,
-      "learning_rate": 8.174057894431904e-06,
-      "loss": 0.3674,
-      "step": 13563
-    },
-    {
-      "epoch": 0.8864780079733351,
-      "grad_norm": 0.4463036060333252,
-      "learning_rate": 8.173788079104202e-06,
-      "loss": 0.3661,
-      "step": 13564
-    },
-    {
-      "epoch": 0.8865433631788772,
-      "grad_norm": 0.43831667304039,
-      "learning_rate": 8.173518248296834e-06,
-      "loss": 0.3903,
-      "step": 13565
-    },
-    {
-      "epoch": 0.8866087183844193,
-      "grad_norm": 0.4397267699241638,
-      "learning_rate": 8.17324840201112e-06,
-      "loss": 0.3919,
-      "step": 13566
-    },
-    {
-      "epoch": 0.8866740735899614,
-      "grad_norm": 0.4563021659851074,
-      "learning_rate": 8.172978540248374e-06,
-      "loss": 0.401,
-      "step": 13567
-    },
-    {
-      "epoch": 0.8867394287955036,
-      "grad_norm": 0.42863720655441284,
-      "learning_rate": 8.17270866300991e-06,
-      "loss": 0.3793,
-      "step": 13568
-    },
-    {
-      "epoch": 0.8868047840010457,
-      "grad_norm": 0.40967488288879395,
-      "learning_rate": 8.17243877029705e-06,
-      "loss": 0.3378,
-      "step": 13569
-    },
-    {
-      "epoch": 0.8868701392065879,
-      "grad_norm": 0.46297669410705566,
-      "learning_rate": 8.172168862111105e-06,
-      "loss": 0.4137,
-      "step": 13570
-    },
-    {
-      "epoch": 0.8869354944121299,
-      "grad_norm": 0.46139878034591675,
-      "learning_rate": 8.171898938453395e-06,
-      "loss": 0.4214,
-      "step": 13571
-    },
-    {
-      "epoch": 0.887000849617672,
-      "grad_norm": 0.5869432091712952,
-      "learning_rate": 8.171628999325234e-06,
-      "loss": 0.3508,
-      "step": 13572
-    },
-    {
-      "epoch": 0.8870662048232142,
-      "grad_norm": 0.44121357798576355,
-      "learning_rate": 8.17135904472794e-06,
-      "loss": 0.3831,
-      "step": 13573
-    },
-    {
-      "epoch": 0.8871315600287563,
-      "grad_norm": 0.44181132316589355,
-      "learning_rate": 8.171089074662827e-06,
-      "loss": 0.3809,
-      "step": 13574
-    },
-    {
-      "epoch": 0.8871969152342984,
-      "grad_norm": 0.4677775502204895,
-      "learning_rate": 8.170819089131217e-06,
-      "loss": 0.436,
-      "step": 13575
-    },
-    {
-      "epoch": 0.8872622704398405,
-      "grad_norm": 0.6737505197525024,
-      "learning_rate": 8.170549088134423e-06,
-      "loss": 0.3741,
-      "step": 13576
-    },
-    {
-      "epoch": 0.8873276256453827,
-      "grad_norm": 0.4607175588607788,
-      "learning_rate": 8.170279071673764e-06,
-      "loss": 0.3698,
-      "step": 13577
-    },
-    {
-      "epoch": 0.8873929808509248,
-      "grad_norm": 0.45459550619125366,
-      "learning_rate": 8.170009039750554e-06,
-      "loss": 0.3706,
-      "step": 13578
-    },
-    {
-      "epoch": 0.8874583360564668,
-      "grad_norm": 0.45948904752731323,
-      "learning_rate": 8.169738992366111e-06,
-      "loss": 0.4077,
-      "step": 13579
-    },
-    {
-      "epoch": 0.887523691262009,
-      "grad_norm": 0.44636574387550354,
-      "learning_rate": 8.169468929521755e-06,
-      "loss": 0.3731,
-      "step": 13580
-    },
-    {
-      "epoch": 0.8875890464675511,
-      "grad_norm": 0.4502796530723572,
-      "learning_rate": 8.169198851218799e-06,
-      "loss": 0.3737,
-      "step": 13581
-    },
-    {
-      "epoch": 0.8876544016730933,
-      "grad_norm": 0.4874180257320404,
-      "learning_rate": 8.168928757458565e-06,
-      "loss": 0.415,
-      "step": 13582
-    },
-    {
-      "epoch": 0.8877197568786354,
-      "grad_norm": 0.43732890486717224,
-      "learning_rate": 8.168658648242365e-06,
-      "loss": 0.366,
-      "step": 13583
-    },
-    {
-      "epoch": 0.8877851120841775,
-      "grad_norm": 0.43505555391311646,
-      "learning_rate": 8.16838852357152e-06,
-      "loss": 0.3621,
-      "step": 13584
-    },
-    {
-      "epoch": 0.8878504672897196,
-      "grad_norm": 0.4564209282398224,
-      "learning_rate": 8.168118383447346e-06,
-      "loss": 0.3748,
-      "step": 13585
-    },
-    {
-      "epoch": 0.8879158224952618,
-      "grad_norm": 0.4783664345741272,
-      "learning_rate": 8.167848227871163e-06,
-      "loss": 0.4131,
-      "step": 13586
-    },
-    {
-      "epoch": 0.8879811777008039,
-      "grad_norm": 0.448490172624588,
-      "learning_rate": 8.167578056844284e-06,
-      "loss": 0.3866,
-      "step": 13587
-    },
-    {
-      "epoch": 0.8880465329063459,
-      "grad_norm": 0.48500195145606995,
-      "learning_rate": 8.16730787036803e-06,
-      "loss": 0.4342,
-      "step": 13588
-    },
-    {
-      "epoch": 0.8881118881118881,
-      "grad_norm": 0.38912174105644226,
-      "learning_rate": 8.16703766844372e-06,
-      "loss": 0.3086,
-      "step": 13589
-    },
-    {
-      "epoch": 0.8881772433174302,
-      "grad_norm": 0.4245205223560333,
-      "learning_rate": 8.166767451072669e-06,
-      "loss": 0.3638,
-      "step": 13590
-    },
-    {
-      "epoch": 0.8882425985229724,
-      "grad_norm": 0.44031450152397156,
-      "learning_rate": 8.166497218256195e-06,
-      "loss": 0.3993,
-      "step": 13591
-    },
-    {
-      "epoch": 0.8883079537285145,
-      "grad_norm": 0.43893659114837646,
-      "learning_rate": 8.166226969995618e-06,
-      "loss": 0.3339,
-      "step": 13592
-    },
-    {
-      "epoch": 0.8883733089340566,
-      "grad_norm": 0.4216693341732025,
-      "learning_rate": 8.165956706292256e-06,
-      "loss": 0.3282,
-      "step": 13593
-    },
-    {
-      "epoch": 0.8884386641395987,
-      "grad_norm": 0.45118266344070435,
-      "learning_rate": 8.165686427147425e-06,
-      "loss": 0.3719,
-      "step": 13594
-    },
-    {
-      "epoch": 0.8885040193451409,
-      "grad_norm": 0.43111103773117065,
-      "learning_rate": 8.165416132562447e-06,
-      "loss": 0.3888,
-      "step": 13595
-    },
-    {
-      "epoch": 0.888569374550683,
-      "grad_norm": 0.4648542106151581,
-      "learning_rate": 8.165145822538635e-06,
-      "loss": 0.4031,
-      "step": 13596
-    },
-    {
-      "epoch": 0.888634729756225,
-      "grad_norm": 0.4463261365890503,
-      "learning_rate": 8.164875497077313e-06,
-      "loss": 0.3744,
-      "step": 13597
-    },
-    {
-      "epoch": 0.8887000849617672,
-      "grad_norm": 0.4441666007041931,
-      "learning_rate": 8.164605156179794e-06,
-      "loss": 0.3784,
-      "step": 13598
-    },
-    {
-      "epoch": 0.8887654401673093,
-      "grad_norm": 0.4290092885494232,
-      "learning_rate": 8.164334799847402e-06,
-      "loss": 0.3232,
-      "step": 13599
-    },
-    {
-      "epoch": 0.8888307953728515,
-      "grad_norm": 0.46337729692459106,
-      "learning_rate": 8.164064428081451e-06,
-      "loss": 0.3923,
-      "step": 13600
-    },
-    {
-      "epoch": 0.8888961505783936,
-      "grad_norm": 0.4588935971260071,
-      "learning_rate": 8.163794040883261e-06,
-      "loss": 0.346,
-      "step": 13601
-    },
-    {
-      "epoch": 0.8889615057839357,
-      "grad_norm": 0.42463478446006775,
-      "learning_rate": 8.163523638254154e-06,
-      "loss": 0.3771,
-      "step": 13602
-    },
-    {
-      "epoch": 0.8890268609894778,
-      "grad_norm": 0.42009827494621277,
-      "learning_rate": 8.163253220195446e-06,
-      "loss": 0.3443,
-      "step": 13603
-    },
-    {
-      "epoch": 0.8890922161950199,
-      "grad_norm": 0.48326075077056885,
-      "learning_rate": 8.162982786708455e-06,
-      "loss": 0.368,
-      "step": 13604
-    },
-    {
-      "epoch": 0.8891575714005621,
-      "grad_norm": 0.4304189682006836,
-      "learning_rate": 8.1627123377945e-06,
-      "loss": 0.3572,
-      "step": 13605
-    },
-    {
-      "epoch": 0.8892229266061041,
-      "grad_norm": 0.4433961510658264,
-      "learning_rate": 8.162441873454904e-06,
-      "loss": 0.3799,
-      "step": 13606
-    },
-    {
-      "epoch": 0.8892882818116463,
-      "grad_norm": 0.433493435382843,
-      "learning_rate": 8.162171393690982e-06,
-      "loss": 0.4022,
-      "step": 13607
-    },
-    {
-      "epoch": 0.8893536370171884,
-      "grad_norm": 0.411797434091568,
-      "learning_rate": 8.161900898504055e-06,
-      "loss": 0.343,
-      "step": 13608
-    },
-    {
-      "epoch": 0.8894189922227306,
-      "grad_norm": 0.4686589539051056,
-      "learning_rate": 8.16163038789544e-06,
-      "loss": 0.4436,
-      "step": 13609
-    },
-    {
-      "epoch": 0.8894843474282726,
-      "grad_norm": 0.40443435311317444,
-      "learning_rate": 8.16135986186646e-06,
-      "loss": 0.3243,
-      "step": 13610
-    },
-    {
-      "epoch": 0.8895497026338148,
-      "grad_norm": 0.44047921895980835,
-      "learning_rate": 8.161089320418434e-06,
-      "loss": 0.3616,
-      "step": 13611
-    },
-    {
-      "epoch": 0.8896150578393569,
-      "grad_norm": 0.43029654026031494,
-      "learning_rate": 8.160818763552677e-06,
-      "loss": 0.3781,
-      "step": 13612
-    },
-    {
-      "epoch": 0.889680413044899,
-      "grad_norm": 0.4242212176322937,
-      "learning_rate": 8.160548191270516e-06,
-      "loss": 0.3692,
-      "step": 13613
-    },
-    {
-      "epoch": 0.8897457682504412,
-      "grad_norm": 0.4410896301269531,
-      "learning_rate": 8.160277603573263e-06,
-      "loss": 0.4013,
-      "step": 13614
-    },
-    {
-      "epoch": 0.8898111234559832,
-      "grad_norm": 0.45493629574775696,
-      "learning_rate": 8.160007000462243e-06,
-      "loss": 0.4028,
-      "step": 13615
-    },
-    {
-      "epoch": 0.8898764786615254,
-      "grad_norm": 0.41715767979621887,
-      "learning_rate": 8.159736381938775e-06,
-      "loss": 0.324,
-      "step": 13616
-    },
-    {
-      "epoch": 0.8899418338670675,
-      "grad_norm": 0.47945287823677063,
-      "learning_rate": 8.159465748004177e-06,
-      "loss": 0.3804,
-      "step": 13617
-    },
-    {
-      "epoch": 0.8900071890726097,
-      "grad_norm": 0.46555617451667786,
-      "learning_rate": 8.15919509865977e-06,
-      "loss": 0.3864,
-      "step": 13618
-    },
-    {
-      "epoch": 0.8900725442781517,
-      "grad_norm": 0.4466501772403717,
-      "learning_rate": 8.158924433906872e-06,
-      "loss": 0.3498,
-      "step": 13619
-    },
-    {
-      "epoch": 0.8901378994836939,
-      "grad_norm": 0.4793967008590698,
-      "learning_rate": 8.158653753746808e-06,
-      "loss": 0.4582,
-      "step": 13620
-    },
-    {
-      "epoch": 0.890203254689236,
-      "grad_norm": 0.4156058728694916,
-      "learning_rate": 8.158383058180894e-06,
-      "loss": 0.3434,
-      "step": 13621
-    },
-    {
-      "epoch": 0.8902686098947781,
-      "grad_norm": 0.41390785574913025,
-      "learning_rate": 8.158112347210452e-06,
-      "loss": 0.3234,
-      "step": 13622
-    },
-    {
-      "epoch": 0.8903339651003203,
-      "grad_norm": 0.4056573808193207,
-      "learning_rate": 8.157841620836802e-06,
-      "loss": 0.3568,
-      "step": 13623
-    },
-    {
-      "epoch": 0.8903993203058623,
-      "grad_norm": 0.4190270006656647,
-      "learning_rate": 8.157570879061265e-06,
-      "loss": 0.3441,
-      "step": 13624
-    },
-    {
-      "epoch": 0.8904646755114045,
-      "grad_norm": 0.4631539583206177,
-      "learning_rate": 8.157300121885162e-06,
-      "loss": 0.4338,
-      "step": 13625
-    },
-    {
-      "epoch": 0.8905300307169466,
-      "grad_norm": 0.47524601221084595,
-      "learning_rate": 8.15702934930981e-06,
-      "loss": 0.3547,
-      "step": 13626
-    },
-    {
-      "epoch": 0.8905953859224888,
-      "grad_norm": 0.4354708194732666,
-      "learning_rate": 8.156758561336533e-06,
-      "loss": 0.3513,
-      "step": 13627
-    },
-    {
-      "epoch": 0.8906607411280308,
-      "grad_norm": 0.42499879002571106,
-      "learning_rate": 8.156487757966652e-06,
-      "loss": 0.3374,
-      "step": 13628
-    },
-    {
-      "epoch": 0.8907260963335729,
-      "grad_norm": 0.42320162057876587,
-      "learning_rate": 8.156216939201484e-06,
-      "loss": 0.3458,
-      "step": 13629
-    },
-    {
-      "epoch": 0.8907914515391151,
-      "grad_norm": 0.4369945526123047,
-      "learning_rate": 8.155946105042355e-06,
-      "loss": 0.3914,
-      "step": 13630
-    },
-    {
-      "epoch": 0.8908568067446572,
-      "grad_norm": 0.46945497393608093,
-      "learning_rate": 8.155675255490582e-06,
-      "loss": 0.4225,
-      "step": 13631
-    },
-    {
-      "epoch": 0.8909221619501994,
-      "grad_norm": 0.43981820344924927,
-      "learning_rate": 8.155404390547489e-06,
-      "loss": 0.3629,
-      "step": 13632
-    },
-    {
-      "epoch": 0.8909875171557414,
-      "grad_norm": 0.4373462200164795,
-      "learning_rate": 8.155133510214395e-06,
-      "loss": 0.3807,
-      "step": 13633
-    },
-    {
-      "epoch": 0.8910528723612836,
-      "grad_norm": 0.5137832164764404,
-      "learning_rate": 8.154862614492623e-06,
-      "loss": 0.396,
-      "step": 13634
-    },
-    {
-      "epoch": 0.8911182275668257,
-      "grad_norm": 0.40821897983551025,
-      "learning_rate": 8.154591703383492e-06,
-      "loss": 0.3357,
-      "step": 13635
-    },
-    {
-      "epoch": 0.8911835827723679,
-      "grad_norm": 0.9074820876121521,
-      "learning_rate": 8.154320776888323e-06,
-      "loss": 0.5001,
-      "step": 13636
-    },
-    {
-      "epoch": 0.8912489379779099,
-      "grad_norm": 0.45739322900772095,
-      "learning_rate": 8.15404983500844e-06,
-      "loss": 0.3331,
-      "step": 13637
-    },
-    {
-      "epoch": 0.891314293183452,
-      "grad_norm": 0.38675034046173096,
-      "learning_rate": 8.153778877745164e-06,
-      "loss": 0.2627,
-      "step": 13638
-    },
-    {
-      "epoch": 0.8913796483889942,
-      "grad_norm": 0.4908689558506012,
-      "learning_rate": 8.153507905099814e-06,
-      "loss": 0.4555,
-      "step": 13639
-    },
-    {
-      "epoch": 0.8914450035945363,
-      "grad_norm": 0.4473533630371094,
-      "learning_rate": 8.153236917073714e-06,
-      "loss": 0.3491,
-      "step": 13640
-    },
-    {
-      "epoch": 0.8915103588000785,
-      "grad_norm": 0.44573792815208435,
-      "learning_rate": 8.152965913668188e-06,
-      "loss": 0.396,
-      "step": 13641
-    },
-    {
-      "epoch": 0.8915757140056205,
-      "grad_norm": 0.43579259514808655,
-      "learning_rate": 8.152694894884552e-06,
-      "loss": 0.3697,
-      "step": 13642
-    },
-    {
-      "epoch": 0.8916410692111627,
-      "grad_norm": 0.4252687692642212,
-      "learning_rate": 8.152423860724132e-06,
-      "loss": 0.3572,
-      "step": 13643
-    },
-    {
-      "epoch": 0.8917064244167048,
-      "grad_norm": 0.42283689975738525,
-      "learning_rate": 8.152152811188248e-06,
-      "loss": 0.3982,
-      "step": 13644
-    },
-    {
-      "epoch": 0.891771779622247,
-      "grad_norm": 0.3983820974826813,
-      "learning_rate": 8.151881746278224e-06,
-      "loss": 0.3204,
-      "step": 13645
-    },
-    {
-      "epoch": 0.891837134827789,
-      "grad_norm": 0.4324599504470825,
-      "learning_rate": 8.15161066599538e-06,
-      "loss": 0.3933,
-      "step": 13646
-    },
-    {
-      "epoch": 0.8919024900333311,
-      "grad_norm": 0.4834281802177429,
-      "learning_rate": 8.151339570341039e-06,
-      "loss": 0.419,
-      "step": 13647
-    },
-    {
-      "epoch": 0.8919678452388733,
-      "grad_norm": 0.44146692752838135,
-      "learning_rate": 8.151068459316522e-06,
-      "loss": 0.3747,
-      "step": 13648
-    },
-    {
-      "epoch": 0.8920332004444154,
-      "grad_norm": 0.4567866027355194,
-      "learning_rate": 8.150797332923154e-06,
-      "loss": 0.4196,
-      "step": 13649
-    },
-    {
-      "epoch": 0.8920985556499575,
-      "grad_norm": 0.4537763297557831,
-      "learning_rate": 8.150526191162255e-06,
-      "loss": 0.3938,
-      "step": 13650
-    },
-    {
-      "epoch": 0.8921639108554996,
-      "grad_norm": 0.430987685918808,
-      "learning_rate": 8.15025503403515e-06,
-      "loss": 0.3707,
-      "step": 13651
-    },
-    {
-      "epoch": 0.8922292660610418,
-      "grad_norm": 0.43445083498954773,
-      "learning_rate": 8.149983861543159e-06,
-      "loss": 0.3291,
-      "step": 13652
-    },
-    {
-      "epoch": 0.8922946212665839,
-      "grad_norm": 0.45490071177482605,
-      "learning_rate": 8.149712673687604e-06,
-      "loss": 0.4088,
-      "step": 13653
-    },
-    {
-      "epoch": 0.8923599764721261,
-      "grad_norm": 0.43494483828544617,
-      "learning_rate": 8.14944147046981e-06,
-      "loss": 0.3474,
-      "step": 13654
-    },
-    {
-      "epoch": 0.8924253316776681,
-      "grad_norm": 0.4581224322319031,
-      "learning_rate": 8.1491702518911e-06,
-      "loss": 0.365,
-      "step": 13655
-    },
-    {
-      "epoch": 0.8924906868832102,
-      "grad_norm": 0.44352850317955017,
-      "learning_rate": 8.148899017952794e-06,
-      "loss": 0.3834,
-      "step": 13656
-    },
-    {
-      "epoch": 0.8925560420887524,
-      "grad_norm": 0.5057024359703064,
-      "learning_rate": 8.148627768656217e-06,
-      "loss": 0.4331,
-      "step": 13657
-    },
-    {
-      "epoch": 0.8926213972942945,
-      "grad_norm": 0.4180637300014496,
-      "learning_rate": 8.148356504002692e-06,
-      "loss": 0.3211,
-      "step": 13658
-    },
-    {
-      "epoch": 0.8926867524998366,
-      "grad_norm": 0.4715318977832794,
-      "learning_rate": 8.148085223993541e-06,
-      "loss": 0.3602,
-      "step": 13659
-    },
-    {
-      "epoch": 0.8927521077053787,
-      "grad_norm": 0.4708700478076935,
-      "learning_rate": 8.147813928630088e-06,
-      "loss": 0.4002,
-      "step": 13660
-    },
-    {
-      "epoch": 0.8928174629109209,
-      "grad_norm": 0.42736631631851196,
-      "learning_rate": 8.147542617913657e-06,
-      "loss": 0.3477,
-      "step": 13661
-    },
-    {
-      "epoch": 0.892882818116463,
-      "grad_norm": 0.4274823069572449,
-      "learning_rate": 8.14727129184557e-06,
-      "loss": 0.3696,
-      "step": 13662
-    },
-    {
-      "epoch": 0.892948173322005,
-      "grad_norm": 0.45414844155311584,
-      "learning_rate": 8.14699995042715e-06,
-      "loss": 0.4161,
-      "step": 13663
-    },
-    {
-      "epoch": 0.8930135285275472,
-      "grad_norm": 0.4576912522315979,
-      "learning_rate": 8.14672859365972e-06,
-      "loss": 0.3728,
-      "step": 13664
-    },
-    {
-      "epoch": 0.8930788837330893,
-      "grad_norm": 0.3824891746044159,
-      "learning_rate": 8.146457221544606e-06,
-      "loss": 0.3073,
-      "step": 13665
-    },
-    {
-      "epoch": 0.8931442389386315,
-      "grad_norm": 0.44985726475715637,
-      "learning_rate": 8.14618583408313e-06,
-      "loss": 0.4023,
-      "step": 13666
-    },
-    {
-      "epoch": 0.8932095941441736,
-      "grad_norm": 0.42375385761260986,
-      "learning_rate": 8.145914431276616e-06,
-      "loss": 0.3415,
-      "step": 13667
-    },
-    {
-      "epoch": 0.8932749493497157,
-      "grad_norm": 0.45005932450294495,
-      "learning_rate": 8.145643013126388e-06,
-      "loss": 0.3987,
-      "step": 13668
-    },
-    {
-      "epoch": 0.8933403045552578,
-      "grad_norm": 0.4389643967151642,
-      "learning_rate": 8.145371579633767e-06,
-      "loss": 0.3559,
-      "step": 13669
-    },
-    {
-      "epoch": 0.8934056597608,
-      "grad_norm": 0.4423743188381195,
-      "learning_rate": 8.14510013080008e-06,
-      "loss": 0.3757,
-      "step": 13670
-    },
-    {
-      "epoch": 0.8934710149663421,
-      "grad_norm": 0.4107572138309479,
-      "learning_rate": 8.144828666626651e-06,
-      "loss": 0.3334,
-      "step": 13671
-    },
-    {
-      "epoch": 0.8935363701718841,
-      "grad_norm": 0.4441332221031189,
-      "learning_rate": 8.144557187114803e-06,
-      "loss": 0.347,
-      "step": 13672
-    },
-    {
-      "epoch": 0.8936017253774263,
-      "grad_norm": 0.45187175273895264,
-      "learning_rate": 8.14428569226586e-06,
-      "loss": 0.3978,
-      "step": 13673
-    },
-    {
-      "epoch": 0.8936670805829684,
-      "grad_norm": 0.46893805265426636,
-      "learning_rate": 8.144014182081145e-06,
-      "loss": 0.4094,
-      "step": 13674
-    },
-    {
-      "epoch": 0.8937324357885106,
-      "grad_norm": 0.4110645651817322,
-      "learning_rate": 8.143742656561986e-06,
-      "loss": 0.3416,
-      "step": 13675
-    },
-    {
-      "epoch": 0.8937977909940527,
-      "grad_norm": 0.4819435477256775,
-      "learning_rate": 8.143471115709703e-06,
-      "loss": 0.4356,
-      "step": 13676
-    },
-    {
-      "epoch": 0.8938631461995948,
-      "grad_norm": 0.43986037373542786,
-      "learning_rate": 8.143199559525624e-06,
-      "loss": 0.3576,
-      "step": 13677
-    },
-    {
-      "epoch": 0.8939285014051369,
-      "grad_norm": 0.4579463005065918,
-      "learning_rate": 8.14292798801107e-06,
-      "loss": 0.418,
-      "step": 13678
-    },
-    {
-      "epoch": 0.8939938566106791,
-      "grad_norm": 0.4032513201236725,
-      "learning_rate": 8.142656401167367e-06,
-      "loss": 0.3178,
-      "step": 13679
-    },
-    {
-      "epoch": 0.8940592118162212,
-      "grad_norm": 0.46665820479393005,
-      "learning_rate": 8.142384798995843e-06,
-      "loss": 0.4178,
-      "step": 13680
-    },
-    {
-      "epoch": 0.8941245670217632,
-      "grad_norm": 0.44561415910720825,
-      "learning_rate": 8.142113181497819e-06,
-      "loss": 0.3954,
-      "step": 13681
-    },
-    {
-      "epoch": 0.8941899222273054,
-      "grad_norm": 0.4361459016799927,
-      "learning_rate": 8.141841548674619e-06,
-      "loss": 0.3518,
-      "step": 13682
-    },
-    {
-      "epoch": 0.8942552774328475,
-      "grad_norm": 0.46624186635017395,
-      "learning_rate": 8.141569900527571e-06,
-      "loss": 0.4021,
-      "step": 13683
-    },
-    {
-      "epoch": 0.8943206326383897,
-      "grad_norm": 0.42903366684913635,
-      "learning_rate": 8.141298237057995e-06,
-      "loss": 0.3499,
-      "step": 13684
-    },
-    {
-      "epoch": 0.8943859878439318,
-      "grad_norm": 0.4311058521270752,
-      "learning_rate": 8.141026558267222e-06,
-      "loss": 0.363,
-      "step": 13685
-    },
-    {
-      "epoch": 0.8944513430494739,
-      "grad_norm": 0.4667481780052185,
-      "learning_rate": 8.140754864156573e-06,
-      "loss": 0.4418,
-      "step": 13686
-    },
-    {
-      "epoch": 0.894516698255016,
-      "grad_norm": 0.4319532811641693,
-      "learning_rate": 8.140483154727376e-06,
-      "loss": 0.3388,
-      "step": 13687
-    },
-    {
-      "epoch": 0.8945820534605581,
-      "grad_norm": 0.4356854259967804,
-      "learning_rate": 8.140211429980955e-06,
-      "loss": 0.3668,
-      "step": 13688
-    },
-    {
-      "epoch": 0.8946474086661003,
-      "grad_norm": 0.4317588806152344,
-      "learning_rate": 8.139939689918634e-06,
-      "loss": 0.3533,
-      "step": 13689
-    },
-    {
-      "epoch": 0.8947127638716423,
-      "grad_norm": 0.4060245156288147,
-      "learning_rate": 8.139667934541738e-06,
-      "loss": 0.2844,
-      "step": 13690
-    },
-    {
-      "epoch": 0.8947781190771845,
-      "grad_norm": 0.4455457925796509,
-      "learning_rate": 8.139396163851596e-06,
-      "loss": 0.3959,
-      "step": 13691
-    },
-    {
-      "epoch": 0.8948434742827266,
-      "grad_norm": 0.464456170797348,
-      "learning_rate": 8.13912437784953e-06,
-      "loss": 0.433,
-      "step": 13692
-    },
-    {
-      "epoch": 0.8949088294882688,
-      "grad_norm": 0.43154770135879517,
-      "learning_rate": 8.138852576536868e-06,
-      "loss": 0.3251,
-      "step": 13693
-    },
-    {
-      "epoch": 0.8949741846938108,
-      "grad_norm": 0.43340033292770386,
-      "learning_rate": 8.138580759914933e-06,
-      "loss": 0.3567,
-      "step": 13694
-    },
-    {
-      "epoch": 0.895039539899353,
-      "grad_norm": 0.42331504821777344,
-      "learning_rate": 8.138308927985053e-06,
-      "loss": 0.3336,
-      "step": 13695
-    },
-    {
-      "epoch": 0.8951048951048951,
-      "grad_norm": 0.4122401177883148,
-      "learning_rate": 8.138037080748552e-06,
-      "loss": 0.3355,
-      "step": 13696
-    },
-    {
-      "epoch": 0.8951702503104372,
-      "grad_norm": 0.4513709247112274,
-      "learning_rate": 8.137765218206759e-06,
-      "loss": 0.393,
-      "step": 13697
-    },
-    {
-      "epoch": 0.8952356055159794,
-      "grad_norm": 0.42458638548851013,
-      "learning_rate": 8.137493340360998e-06,
-      "loss": 0.367,
-      "step": 13698
-    },
-    {
-      "epoch": 0.8953009607215214,
-      "grad_norm": 0.4798518717288971,
-      "learning_rate": 8.137221447212594e-06,
-      "loss": 0.4237,
-      "step": 13699
-    },
-    {
-      "epoch": 0.8953663159270636,
-      "grad_norm": 0.4395330250263214,
-      "learning_rate": 8.136949538762874e-06,
-      "loss": 0.3851,
-      "step": 13700
-    },
-    {
-      "epoch": 0.8954316711326057,
-      "grad_norm": 0.8126868605613708,
-      "learning_rate": 8.136677615013167e-06,
-      "loss": 0.4555,
-      "step": 13701
-    },
-    {
-      "epoch": 0.8954970263381479,
-      "grad_norm": 0.4594406485557556,
-      "learning_rate": 8.136405675964793e-06,
-      "loss": 0.3757,
-      "step": 13702
-    },
-    {
-      "epoch": 0.89556238154369,
-      "grad_norm": 0.4782710373401642,
-      "learning_rate": 8.136133721619084e-06,
-      "loss": 0.4029,
-      "step": 13703
-    },
-    {
-      "epoch": 0.8956277367492321,
-      "grad_norm": 0.4623786509037018,
-      "learning_rate": 8.135861751977363e-06,
-      "loss": 0.4077,
-      "step": 13704
-    },
-    {
-      "epoch": 0.8956930919547742,
-      "grad_norm": 0.4414733946323395,
-      "learning_rate": 8.135589767040959e-06,
-      "loss": 0.393,
-      "step": 13705
-    },
-    {
-      "epoch": 0.8957584471603163,
-      "grad_norm": 0.4439255893230438,
-      "learning_rate": 8.135317766811197e-06,
-      "loss": 0.3723,
-      "step": 13706
-    },
-    {
-      "epoch": 0.8958238023658585,
-      "grad_norm": 0.47301366925239563,
-      "learning_rate": 8.135045751289405e-06,
-      "loss": 0.4238,
-      "step": 13707
-    },
-    {
-      "epoch": 0.8958891575714005,
-      "grad_norm": 0.440405011177063,
-      "learning_rate": 8.13477372047691e-06,
-      "loss": 0.3672,
-      "step": 13708
-    },
-    {
-      "epoch": 0.8959545127769427,
-      "grad_norm": 0.47691601514816284,
-      "learning_rate": 8.134501674375035e-06,
-      "loss": 0.438,
-      "step": 13709
-    },
-    {
-      "epoch": 0.8960198679824848,
-      "grad_norm": 0.41470226645469666,
-      "learning_rate": 8.13422961298511e-06,
-      "loss": 0.3602,
-      "step": 13710
-    },
-    {
-      "epoch": 0.896085223188027,
-      "grad_norm": 0.43777281045913696,
-      "learning_rate": 8.133957536308461e-06,
-      "loss": 0.4027,
-      "step": 13711
-    },
-    {
-      "epoch": 0.896150578393569,
-      "grad_norm": 0.49178236722946167,
-      "learning_rate": 8.133685444346415e-06,
-      "loss": 0.402,
-      "step": 13712
-    },
-    {
-      "epoch": 0.8962159335991111,
-      "grad_norm": 0.41768455505371094,
-      "learning_rate": 8.133413337100302e-06,
-      "loss": 0.3535,
-      "step": 13713
-    },
-    {
-      "epoch": 0.8962812888046533,
-      "grad_norm": 0.47148391604423523,
-      "learning_rate": 8.133141214571444e-06,
-      "loss": 0.4033,
-      "step": 13714
-    },
-    {
-      "epoch": 0.8963466440101954,
-      "grad_norm": 0.45111405849456787,
-      "learning_rate": 8.132869076761171e-06,
-      "loss": 0.3994,
-      "step": 13715
-    },
-    {
-      "epoch": 0.8964119992157376,
-      "grad_norm": 0.41851431131362915,
-      "learning_rate": 8.132596923670811e-06,
-      "loss": 0.3361,
-      "step": 13716
-    },
-    {
-      "epoch": 0.8964773544212796,
-      "grad_norm": 0.451364129781723,
-      "learning_rate": 8.13232475530169e-06,
-      "loss": 0.3664,
-      "step": 13717
-    },
-    {
-      "epoch": 0.8965427096268218,
-      "grad_norm": 0.4423421621322632,
-      "learning_rate": 8.132052571655138e-06,
-      "loss": 0.4072,
-      "step": 13718
-    },
-    {
-      "epoch": 0.8966080648323639,
-      "grad_norm": 0.43970921635627747,
-      "learning_rate": 8.131780372732479e-06,
-      "loss": 0.3657,
-      "step": 13719
-    },
-    {
-      "epoch": 0.8966734200379061,
-      "grad_norm": 0.4629170000553131,
-      "learning_rate": 8.131508158535042e-06,
-      "loss": 0.4172,
-      "step": 13720
-    },
-    {
-      "epoch": 0.8967387752434481,
-      "grad_norm": 0.47206780314445496,
-      "learning_rate": 8.131235929064155e-06,
-      "loss": 0.4114,
-      "step": 13721
-    },
-    {
-      "epoch": 0.8968041304489902,
-      "grad_norm": 0.44379761815071106,
-      "learning_rate": 8.130963684321146e-06,
-      "loss": 0.3913,
-      "step": 13722
-    },
-    {
-      "epoch": 0.8968694856545324,
-      "grad_norm": 0.44583436846733093,
-      "learning_rate": 8.130691424307342e-06,
-      "loss": 0.4011,
-      "step": 13723
-    },
-    {
-      "epoch": 0.8969348408600745,
-      "grad_norm": 0.42743930220603943,
-      "learning_rate": 8.130419149024071e-06,
-      "loss": 0.3323,
-      "step": 13724
-    },
-    {
-      "epoch": 0.8970001960656167,
-      "grad_norm": 0.3742446303367615,
-      "learning_rate": 8.130146858472662e-06,
-      "loss": 0.2674,
-      "step": 13725
-    },
-    {
-      "epoch": 0.8970655512711587,
-      "grad_norm": 0.49502402544021606,
-      "learning_rate": 8.129874552654441e-06,
-      "loss": 0.4736,
-      "step": 13726
-    },
-    {
-      "epoch": 0.8971309064767009,
-      "grad_norm": 0.4888174533843994,
-      "learning_rate": 8.129602231570742e-06,
-      "loss": 0.4525,
-      "step": 13727
-    },
-    {
-      "epoch": 0.897196261682243,
-      "grad_norm": 0.4640890061855316,
-      "learning_rate": 8.129329895222884e-06,
-      "loss": 0.3483,
-      "step": 13728
-    },
-    {
-      "epoch": 0.8972616168877852,
-      "grad_norm": 0.45433375239372253,
-      "learning_rate": 8.129057543612202e-06,
-      "loss": 0.4101,
-      "step": 13729
-    },
-    {
-      "epoch": 0.8973269720933272,
-      "grad_norm": 0.4321691691875458,
-      "learning_rate": 8.128785176740021e-06,
-      "loss": 0.3739,
-      "step": 13730
-    },
-    {
-      "epoch": 0.8973923272988693,
-      "grad_norm": 0.4399643540382385,
-      "learning_rate": 8.128512794607672e-06,
-      "loss": 0.3908,
-      "step": 13731
-    },
-    {
-      "epoch": 0.8974576825044115,
-      "grad_norm": 0.44024062156677246,
-      "learning_rate": 8.128240397216482e-06,
-      "loss": 0.3608,
-      "step": 13732
-    },
-    {
-      "epoch": 0.8975230377099536,
-      "grad_norm": 0.4510750472545624,
-      "learning_rate": 8.12796798456778e-06,
-      "loss": 0.3969,
-      "step": 13733
-    },
-    {
-      "epoch": 0.8975883929154957,
-      "grad_norm": 0.42435726523399353,
-      "learning_rate": 8.127695556662895e-06,
-      "loss": 0.3484,
-      "step": 13734
-    },
-    {
-      "epoch": 0.8976537481210378,
-      "grad_norm": 0.4564155638217926,
-      "learning_rate": 8.127423113503154e-06,
-      "loss": 0.4008,
-      "step": 13735
-    },
-    {
-      "epoch": 0.89771910332658,
-      "grad_norm": 0.46148067712783813,
-      "learning_rate": 8.127150655089886e-06,
-      "loss": 0.3756,
-      "step": 13736
-    },
-    {
-      "epoch": 0.8977844585321221,
-      "grad_norm": 0.44304126501083374,
-      "learning_rate": 8.126878181424423e-06,
-      "loss": 0.3744,
-      "step": 13737
-    },
-    {
-      "epoch": 0.8978498137376643,
-      "grad_norm": 0.44300344586372375,
-      "learning_rate": 8.126605692508091e-06,
-      "loss": 0.338,
-      "step": 13738
-    },
-    {
-      "epoch": 0.8979151689432063,
-      "grad_norm": 0.46930742263793945,
-      "learning_rate": 8.126333188342222e-06,
-      "loss": 0.3926,
-      "step": 13739
-    },
-    {
-      "epoch": 0.8979805241487484,
-      "grad_norm": 0.4636451303958893,
-      "learning_rate": 8.126060668928141e-06,
-      "loss": 0.4222,
-      "step": 13740
-    },
-    {
-      "epoch": 0.8980458793542906,
-      "grad_norm": 0.4638378918170929,
-      "learning_rate": 8.125788134267178e-06,
-      "loss": 0.3767,
-      "step": 13741
-    },
-    {
-      "epoch": 0.8981112345598327,
-      "grad_norm": 0.4696529805660248,
-      "learning_rate": 8.125515584360666e-06,
-      "loss": 0.4268,
-      "step": 13742
-    },
-    {
-      "epoch": 0.8981765897653748,
-      "grad_norm": 0.4246528446674347,
-      "learning_rate": 8.12524301920993e-06,
-      "loss": 0.347,
-      "step": 13743
-    },
-    {
-      "epoch": 0.8982419449709169,
-      "grad_norm": 0.4272640645503998,
-      "learning_rate": 8.124970438816302e-06,
-      "loss": 0.3747,
-      "step": 13744
-    },
-    {
-      "epoch": 0.8983073001764591,
-      "grad_norm": 0.48389050364494324,
-      "learning_rate": 8.124697843181108e-06,
-      "loss": 0.3115,
-      "step": 13745
-    },
-    {
-      "epoch": 0.8983726553820012,
-      "grad_norm": 0.40088316798210144,
-      "learning_rate": 8.124425232305684e-06,
-      "loss": 0.3072,
-      "step": 13746
-    },
-    {
-      "epoch": 0.8984380105875432,
-      "grad_norm": 0.4198073148727417,
-      "learning_rate": 8.124152606191353e-06,
-      "loss": 0.35,
-      "step": 13747
-    },
-    {
-      "epoch": 0.8985033657930854,
-      "grad_norm": 0.40573132038116455,
-      "learning_rate": 8.12387996483945e-06,
-      "loss": 0.336,
-      "step": 13748
-    },
-    {
-      "epoch": 0.8985687209986275,
-      "grad_norm": 0.4379951059818268,
-      "learning_rate": 8.1236073082513e-06,
-      "loss": 0.3724,
-      "step": 13749
-    },
-    {
-      "epoch": 0.8986340762041697,
-      "grad_norm": 0.43794867396354675,
-      "learning_rate": 8.123334636428236e-06,
-      "loss": 0.3534,
-      "step": 13750
-    },
-    {
-      "epoch": 0.8986994314097118,
-      "grad_norm": 0.43171659111976624,
-      "learning_rate": 8.123061949371587e-06,
-      "loss": 0.3549,
-      "step": 13751
-    },
-    {
-      "epoch": 0.8987647866152539,
-      "grad_norm": 0.3995577394962311,
-      "learning_rate": 8.122789247082681e-06,
-      "loss": 0.3258,
-      "step": 13752
-    },
-    {
-      "epoch": 0.898830141820796,
-      "grad_norm": 0.45060256123542786,
-      "learning_rate": 8.122516529562852e-06,
-      "loss": 0.3708,
-      "step": 13753
-    },
-    {
-      "epoch": 0.8988954970263382,
-      "grad_norm": 0.45653200149536133,
-      "learning_rate": 8.122243796813427e-06,
-      "loss": 0.4065,
-      "step": 13754
-    },
-    {
-      "epoch": 0.8989608522318803,
-      "grad_norm": 0.4437415599822998,
-      "learning_rate": 8.121971048835739e-06,
-      "loss": 0.3822,
-      "step": 13755
-    },
-    {
-      "epoch": 0.8990262074374223,
-      "grad_norm": 0.45178404450416565,
-      "learning_rate": 8.121698285631114e-06,
-      "loss": 0.4043,
-      "step": 13756
-    },
-    {
-      "epoch": 0.8990915626429645,
-      "grad_norm": 0.506862998008728,
-      "learning_rate": 8.12142550720089e-06,
-      "loss": 0.4152,
-      "step": 13757
-    },
-    {
-      "epoch": 0.8991569178485066,
-      "grad_norm": 0.4507945477962494,
-      "learning_rate": 8.121152713546387e-06,
-      "loss": 0.3737,
-      "step": 13758
-    },
-    {
-      "epoch": 0.8992222730540488,
-      "grad_norm": 0.4415077865123749,
-      "learning_rate": 8.120879904668943e-06,
-      "loss": 0.3758,
-      "step": 13759
-    },
-    {
-      "epoch": 0.8992876282595909,
-      "grad_norm": 0.4535658061504364,
-      "learning_rate": 8.120607080569886e-06,
-      "loss": 0.4043,
-      "step": 13760
-    },
-    {
-      "epoch": 0.899352983465133,
-      "grad_norm": 0.44528043270111084,
-      "learning_rate": 8.120334241250549e-06,
-      "loss": 0.3303,
-      "step": 13761
-    },
-    {
-      "epoch": 0.8994183386706751,
-      "grad_norm": 0.43991243839263916,
-      "learning_rate": 8.120061386712259e-06,
-      "loss": 0.3726,
-      "step": 13762
-    },
-    {
-      "epoch": 0.8994836938762173,
-      "grad_norm": 0.4297688901424408,
-      "learning_rate": 8.119788516956348e-06,
-      "loss": 0.4037,
-      "step": 13763
-    },
-    {
-      "epoch": 0.8995490490817594,
-      "grad_norm": 0.43185365200042725,
-      "learning_rate": 8.11951563198415e-06,
-      "loss": 0.3607,
-      "step": 13764
-    },
-    {
-      "epoch": 0.8996144042873014,
-      "grad_norm": 0.44259557127952576,
-      "learning_rate": 8.119242731796993e-06,
-      "loss": 0.3392,
-      "step": 13765
-    },
-    {
-      "epoch": 0.8996797594928436,
-      "grad_norm": 0.41821321845054626,
-      "learning_rate": 8.118969816396208e-06,
-      "loss": 0.3211,
-      "step": 13766
-    },
-    {
-      "epoch": 0.8997451146983857,
-      "grad_norm": 0.40150561928749084,
-      "learning_rate": 8.118696885783125e-06,
-      "loss": 0.3271,
-      "step": 13767
-    },
-    {
-      "epoch": 0.8998104699039279,
-      "grad_norm": 0.46457138657569885,
-      "learning_rate": 8.11842393995908e-06,
-      "loss": 0.3877,
-      "step": 13768
-    },
-    {
-      "epoch": 0.89987582510947,
-      "grad_norm": 0.4529329836368561,
-      "learning_rate": 8.118150978925399e-06,
-      "loss": 0.3659,
-      "step": 13769
-    },
-    {
-      "epoch": 0.8999411803150121,
-      "grad_norm": 0.4387982189655304,
-      "learning_rate": 8.117878002683418e-06,
-      "loss": 0.3639,
-      "step": 13770
-    },
-    {
-      "epoch": 0.9000065355205542,
-      "grad_norm": 0.4521043300628662,
-      "learning_rate": 8.117605011234464e-06,
-      "loss": 0.3952,
-      "step": 13771
-    },
-    {
-      "epoch": 0.9000718907260963,
-      "grad_norm": 0.4835241734981537,
-      "learning_rate": 8.11733200457987e-06,
-      "loss": 0.4276,
-      "step": 13772
-    },
-    {
-      "epoch": 0.9001372459316385,
-      "grad_norm": 0.4648262560367584,
-      "learning_rate": 8.117058982720968e-06,
-      "loss": 0.3848,
-      "step": 13773
-    },
-    {
-      "epoch": 0.9002026011371805,
-      "grad_norm": 0.4252789616584778,
-      "learning_rate": 8.11678594565909e-06,
-      "loss": 0.3104,
-      "step": 13774
-    },
-    {
-      "epoch": 0.9002679563427227,
-      "grad_norm": 0.4380900263786316,
-      "learning_rate": 8.116512893395567e-06,
-      "loss": 0.3858,
-      "step": 13775
-    },
-    {
-      "epoch": 0.9003333115482648,
-      "grad_norm": 0.4245436191558838,
-      "learning_rate": 8.11623982593173e-06,
-      "loss": 0.3176,
-      "step": 13776
-    },
-    {
-      "epoch": 0.900398666753807,
-      "grad_norm": 0.4453778862953186,
-      "learning_rate": 8.115966743268914e-06,
-      "loss": 0.3885,
-      "step": 13777
-    },
-    {
-      "epoch": 0.900464021959349,
-      "grad_norm": 0.4539322257041931,
-      "learning_rate": 8.115693645408447e-06,
-      "loss": 0.3565,
-      "step": 13778
-    },
-    {
-      "epoch": 0.9005293771648912,
-      "grad_norm": 0.42480286955833435,
-      "learning_rate": 8.115420532351662e-06,
-      "loss": 0.3322,
-      "step": 13779
-    },
-    {
-      "epoch": 0.9005947323704333,
-      "grad_norm": 0.4224318861961365,
-      "learning_rate": 8.115147404099895e-06,
-      "loss": 0.3568,
-      "step": 13780
-    },
-    {
-      "epoch": 0.9006600875759754,
-      "grad_norm": 0.425488144159317,
-      "learning_rate": 8.114874260654472e-06,
-      "loss": 0.3614,
-      "step": 13781
-    },
-    {
-      "epoch": 0.9007254427815176,
-      "grad_norm": 0.4492007791996002,
-      "learning_rate": 8.114601102016727e-06,
-      "loss": 0.3752,
-      "step": 13782
-    },
-    {
-      "epoch": 0.9007907979870596,
-      "grad_norm": 0.4592430889606476,
-      "learning_rate": 8.114327928187997e-06,
-      "loss": 0.4078,
-      "step": 13783
-    },
-    {
-      "epoch": 0.9008561531926018,
-      "grad_norm": 0.4468863904476166,
-      "learning_rate": 8.11405473916961e-06,
-      "loss": 0.3765,
-      "step": 13784
-    },
-    {
-      "epoch": 0.9009215083981439,
-      "grad_norm": 0.4448256492614746,
-      "learning_rate": 8.113781534962897e-06,
-      "loss": 0.401,
-      "step": 13785
-    },
-    {
-      "epoch": 0.9009868636036861,
-      "grad_norm": 0.44565296173095703,
-      "learning_rate": 8.113508315569195e-06,
-      "loss": 0.3931,
-      "step": 13786
-    },
-    {
-      "epoch": 0.9010522188092281,
-      "grad_norm": 0.46520334482192993,
-      "learning_rate": 8.113235080989834e-06,
-      "loss": 0.4184,
-      "step": 13787
-    },
-    {
-      "epoch": 0.9011175740147703,
-      "grad_norm": 0.42197686433792114,
-      "learning_rate": 8.112961831226145e-06,
-      "loss": 0.3337,
-      "step": 13788
-    },
-    {
-      "epoch": 0.9011829292203124,
-      "grad_norm": 0.4567323327064514,
-      "learning_rate": 8.112688566279465e-06,
-      "loss": 0.4058,
-      "step": 13789
-    },
-    {
-      "epoch": 0.9012482844258545,
-      "grad_norm": 0.43261972069740295,
-      "learning_rate": 8.112415286151123e-06,
-      "loss": 0.3538,
-      "step": 13790
-    },
-    {
-      "epoch": 0.9013136396313967,
-      "grad_norm": 0.4408476650714874,
-      "learning_rate": 8.112141990842455e-06,
-      "loss": 0.329,
-      "step": 13791
-    },
-    {
-      "epoch": 0.9013789948369387,
-      "grad_norm": 0.4256812036037445,
-      "learning_rate": 8.111868680354792e-06,
-      "loss": 0.3502,
-      "step": 13792
-    },
-    {
-      "epoch": 0.9014443500424809,
-      "grad_norm": 0.43711304664611816,
-      "learning_rate": 8.111595354689466e-06,
-      "loss": 0.3636,
-      "step": 13793
-    },
-    {
-      "epoch": 0.901509705248023,
-      "grad_norm": 0.4469901919364929,
-      "learning_rate": 8.111322013847813e-06,
-      "loss": 0.3766,
-      "step": 13794
-    },
-    {
-      "epoch": 0.9015750604535652,
-      "grad_norm": 0.4421667158603668,
-      "learning_rate": 8.111048657831164e-06,
-      "loss": 0.3681,
-      "step": 13795
-    },
-    {
-      "epoch": 0.9016404156591072,
-      "grad_norm": 0.460467666387558,
-      "learning_rate": 8.110775286640852e-06,
-      "loss": 0.4402,
-      "step": 13796
-    },
-    {
-      "epoch": 0.9017057708646493,
-      "grad_norm": 0.44781821966171265,
-      "learning_rate": 8.110501900278213e-06,
-      "loss": 0.3603,
-      "step": 13797
-    },
-    {
-      "epoch": 0.9017711260701915,
-      "grad_norm": 0.429470032453537,
-      "learning_rate": 8.110228498744577e-06,
-      "loss": 0.3647,
-      "step": 13798
-    },
-    {
-      "epoch": 0.9018364812757336,
-      "grad_norm": 0.45834243297576904,
-      "learning_rate": 8.10995508204128e-06,
-      "loss": 0.4131,
-      "step": 13799
-    },
-    {
-      "epoch": 0.9019018364812758,
-      "grad_norm": 0.43037450313568115,
-      "learning_rate": 8.109681650169655e-06,
-      "loss": 0.3253,
-      "step": 13800
-    },
-    {
-      "epoch": 0.9019671916868178,
-      "grad_norm": 0.44325825572013855,
-      "learning_rate": 8.109408203131034e-06,
-      "loss": 0.381,
-      "step": 13801
-    },
-    {
-      "epoch": 0.90203254689236,
-      "grad_norm": 0.47546830773353577,
-      "learning_rate": 8.109134740926754e-06,
-      "loss": 0.4066,
-      "step": 13802
-    },
-    {
-      "epoch": 0.9020979020979021,
-      "grad_norm": 0.42107558250427246,
-      "learning_rate": 8.108861263558145e-06,
-      "loss": 0.3827,
-      "step": 13803
-    },
-    {
-      "epoch": 0.9021632573034443,
-      "grad_norm": 0.4274398386478424,
-      "learning_rate": 8.108587771026543e-06,
-      "loss": 0.3857,
-      "step": 13804
-    },
-    {
-      "epoch": 0.9022286125089863,
-      "grad_norm": 0.4399937689304352,
-      "learning_rate": 8.108314263333283e-06,
-      "loss": 0.34,
-      "step": 13805
-    },
-    {
-      "epoch": 0.9022939677145284,
-      "grad_norm": 0.3941992223262787,
-      "learning_rate": 8.108040740479696e-06,
-      "loss": 0.311,
-      "step": 13806
-    },
-    {
-      "epoch": 0.9023593229200706,
-      "grad_norm": 0.45216676592826843,
-      "learning_rate": 8.107767202467119e-06,
-      "loss": 0.3997,
-      "step": 13807
-    },
-    {
-      "epoch": 0.9024246781256127,
-      "grad_norm": 0.5037730932235718,
-      "learning_rate": 8.107493649296883e-06,
-      "loss": 0.3851,
-      "step": 13808
-    },
-    {
-      "epoch": 0.9024900333311549,
-      "grad_norm": 0.46031659841537476,
-      "learning_rate": 8.107220080970325e-06,
-      "loss": 0.3871,
-      "step": 13809
-    },
-    {
-      "epoch": 0.9025553885366969,
-      "grad_norm": 0.4860627353191376,
-      "learning_rate": 8.106946497488777e-06,
-      "loss": 0.4026,
-      "step": 13810
-    },
-    {
-      "epoch": 0.9026207437422391,
-      "grad_norm": 0.49016475677490234,
-      "learning_rate": 8.106672898853576e-06,
-      "loss": 0.385,
-      "step": 13811
-    },
-    {
-      "epoch": 0.9026860989477812,
-      "grad_norm": 0.4665054976940155,
-      "learning_rate": 8.106399285066053e-06,
-      "loss": 0.4158,
-      "step": 13812
-    },
-    {
-      "epoch": 0.9027514541533234,
-      "grad_norm": 0.4561840891838074,
-      "learning_rate": 8.106125656127547e-06,
-      "loss": 0.3633,
-      "step": 13813
-    },
-    {
-      "epoch": 0.9028168093588654,
-      "grad_norm": 0.4671241343021393,
-      "learning_rate": 8.10585201203939e-06,
-      "loss": 0.3925,
-      "step": 13814
-    },
-    {
-      "epoch": 0.9028821645644075,
-      "grad_norm": 0.42687904834747314,
-      "learning_rate": 8.105578352802915e-06,
-      "loss": 0.3553,
-      "step": 13815
-    },
-    {
-      "epoch": 0.9029475197699497,
-      "grad_norm": 0.4388181269168854,
-      "learning_rate": 8.10530467841946e-06,
-      "loss": 0.3148,
-      "step": 13816
-    },
-    {
-      "epoch": 0.9030128749754918,
-      "grad_norm": 0.467172771692276,
-      "learning_rate": 8.105030988890357e-06,
-      "loss": 0.4019,
-      "step": 13817
-    },
-    {
-      "epoch": 0.903078230181034,
-      "grad_norm": 0.4428863227367401,
-      "learning_rate": 8.104757284216942e-06,
-      "loss": 0.3665,
-      "step": 13818
-    },
-    {
-      "epoch": 0.903143585386576,
-      "grad_norm": 0.4159409999847412,
-      "learning_rate": 8.104483564400552e-06,
-      "loss": 0.3494,
-      "step": 13819
-    },
-    {
-      "epoch": 0.9032089405921182,
-      "grad_norm": 0.42701825499534607,
-      "learning_rate": 8.104209829442518e-06,
-      "loss": 0.3642,
-      "step": 13820
-    },
-    {
-      "epoch": 0.9032742957976603,
-      "grad_norm": 0.43664073944091797,
-      "learning_rate": 8.103936079344179e-06,
-      "loss": 0.3884,
-      "step": 13821
-    },
-    {
-      "epoch": 0.9033396510032025,
-      "grad_norm": 0.4308601915836334,
-      "learning_rate": 8.103662314106869e-06,
-      "loss": 0.3211,
-      "step": 13822
-    },
-    {
-      "epoch": 0.9034050062087445,
-      "grad_norm": 0.50096195936203,
-      "learning_rate": 8.10338853373192e-06,
-      "loss": 0.4162,
-      "step": 13823
-    },
-    {
-      "epoch": 0.9034703614142866,
-      "grad_norm": 0.42874860763549805,
-      "learning_rate": 8.103114738220673e-06,
-      "loss": 0.3999,
-      "step": 13824
-    },
-    {
-      "epoch": 0.9035357166198288,
-      "grad_norm": 0.42832908034324646,
-      "learning_rate": 8.102840927574458e-06,
-      "loss": 0.3604,
-      "step": 13825
-    },
-    {
-      "epoch": 0.9036010718253709,
-      "grad_norm": 0.44836941361427307,
-      "learning_rate": 8.102567101794613e-06,
-      "loss": 0.3981,
-      "step": 13826
-    },
-    {
-      "epoch": 0.903666427030913,
-      "grad_norm": 0.4566391706466675,
-      "learning_rate": 8.102293260882475e-06,
-      "loss": 0.4006,
-      "step": 13827
-    },
-    {
-      "epoch": 0.9037317822364551,
-      "grad_norm": 0.4433964788913727,
-      "learning_rate": 8.102019404839377e-06,
-      "loss": 0.3463,
-      "step": 13828
-    },
-    {
-      "epoch": 0.9037971374419973,
-      "grad_norm": 0.45907121896743774,
-      "learning_rate": 8.101745533666655e-06,
-      "loss": 0.3543,
-      "step": 13829
-    },
-    {
-      "epoch": 0.9038624926475394,
-      "grad_norm": 0.44854235649108887,
-      "learning_rate": 8.101471647365646e-06,
-      "loss": 0.3626,
-      "step": 13830
-    },
-    {
-      "epoch": 0.9039278478530814,
-      "grad_norm": 0.4422043263912201,
-      "learning_rate": 8.101197745937686e-06,
-      "loss": 0.3769,
-      "step": 13831
-    },
-    {
-      "epoch": 0.9039932030586236,
-      "grad_norm": 0.4449627101421356,
-      "learning_rate": 8.10092382938411e-06,
-      "loss": 0.4009,
-      "step": 13832
-    },
-    {
-      "epoch": 0.9040585582641657,
-      "grad_norm": 0.46181216835975647,
-      "learning_rate": 8.100649897706254e-06,
-      "loss": 0.3798,
-      "step": 13833
-    },
-    {
-      "epoch": 0.9041239134697079,
-      "grad_norm": 0.4497895836830139,
-      "learning_rate": 8.100375950905454e-06,
-      "loss": 0.369,
-      "step": 13834
-    },
-    {
-      "epoch": 0.90418926867525,
-      "grad_norm": 0.43777915835380554,
-      "learning_rate": 8.100101988983048e-06,
-      "loss": 0.3561,
-      "step": 13835
-    },
-    {
-      "epoch": 0.9042546238807921,
-      "grad_norm": 0.4683510959148407,
-      "learning_rate": 8.09982801194037e-06,
-      "loss": 0.4394,
-      "step": 13836
-    },
-    {
-      "epoch": 0.9043199790863342,
-      "grad_norm": 0.45694056153297424,
-      "learning_rate": 8.099554019778755e-06,
-      "loss": 0.3922,
-      "step": 13837
-    },
-    {
-      "epoch": 0.9043853342918764,
-      "grad_norm": 0.4695807695388794,
-      "learning_rate": 8.099280012499542e-06,
-      "loss": 0.3461,
-      "step": 13838
-    },
-    {
-      "epoch": 0.9044506894974185,
-      "grad_norm": 0.42546918988227844,
-      "learning_rate": 8.099005990104068e-06,
-      "loss": 0.3417,
-      "step": 13839
-    },
-    {
-      "epoch": 0.9045160447029605,
-      "grad_norm": 0.4493861198425293,
-      "learning_rate": 8.098731952593668e-06,
-      "loss": 0.371,
-      "step": 13840
-    },
-    {
-      "epoch": 0.9045813999085027,
-      "grad_norm": 0.45559778809547424,
-      "learning_rate": 8.098457899969679e-06,
-      "loss": 0.3691,
-      "step": 13841
-    },
-    {
-      "epoch": 0.9046467551140448,
-      "grad_norm": 0.4804634749889374,
-      "learning_rate": 8.098183832233437e-06,
-      "loss": 0.3904,
-      "step": 13842
-    },
-    {
-      "epoch": 0.904712110319587,
-      "grad_norm": 0.4272224009037018,
-      "learning_rate": 8.097909749386276e-06,
-      "loss": 0.3362,
-      "step": 13843
-    },
-    {
-      "epoch": 0.904777465525129,
-      "grad_norm": 0.4513688087463379,
-      "learning_rate": 8.09763565142954e-06,
-      "loss": 0.3895,
-      "step": 13844
-    },
-    {
-      "epoch": 0.9048428207306712,
-      "grad_norm": 0.4479421079158783,
-      "learning_rate": 8.097361538364561e-06,
-      "loss": 0.368,
-      "step": 13845
-    },
-    {
-      "epoch": 0.9049081759362133,
-      "grad_norm": 0.42748603224754333,
-      "learning_rate": 8.097087410192676e-06,
-      "loss": 0.3533,
-      "step": 13846
-    },
-    {
-      "epoch": 0.9049735311417555,
-      "grad_norm": 0.4828256070613861,
-      "learning_rate": 8.096813266915222e-06,
-      "loss": 0.4548,
-      "step": 13847
-    },
-    {
-      "epoch": 0.9050388863472976,
-      "grad_norm": 0.47530660033226013,
-      "learning_rate": 8.09653910853354e-06,
-      "loss": 0.4373,
-      "step": 13848
-    },
-    {
-      "epoch": 0.9051042415528396,
-      "grad_norm": 0.44951948523521423,
-      "learning_rate": 8.096264935048961e-06,
-      "loss": 0.3812,
-      "step": 13849
-    },
-    {
-      "epoch": 0.9051695967583818,
-      "grad_norm": 0.4373512268066406,
-      "learning_rate": 8.095990746462826e-06,
-      "loss": 0.3731,
-      "step": 13850
-    },
-    {
-      "epoch": 0.9052349519639239,
-      "grad_norm": 0.42946070432662964,
-      "learning_rate": 8.095716542776471e-06,
-      "loss": 0.4,
-      "step": 13851
-    },
-    {
-      "epoch": 0.9053003071694661,
-      "grad_norm": 0.46854913234710693,
-      "learning_rate": 8.095442323991236e-06,
-      "loss": 0.3603,
-      "step": 13852
-    },
-    {
-      "epoch": 0.9053656623750082,
-      "grad_norm": 0.48517006635665894,
-      "learning_rate": 8.095168090108453e-06,
-      "loss": 0.4112,
-      "step": 13853
-    },
-    {
-      "epoch": 0.9054310175805503,
-      "grad_norm": 0.4519181251525879,
-      "learning_rate": 8.094893841129468e-06,
-      "loss": 0.3424,
-      "step": 13854
-    },
-    {
-      "epoch": 0.9054963727860924,
-      "grad_norm": 0.45790523290634155,
-      "learning_rate": 8.094619577055609e-06,
-      "loss": 0.3832,
-      "step": 13855
-    },
-    {
-      "epoch": 0.9055617279916345,
-      "grad_norm": 0.44878947734832764,
-      "learning_rate": 8.09434529788822e-06,
-      "loss": 0.3947,
-      "step": 13856
-    },
-    {
-      "epoch": 0.9056270831971767,
-      "grad_norm": 0.4312663972377777,
-      "learning_rate": 8.094071003628637e-06,
-      "loss": 0.362,
-      "step": 13857
-    },
-    {
-      "epoch": 0.9056924384027187,
-      "grad_norm": 0.44874027371406555,
-      "learning_rate": 8.093796694278198e-06,
-      "loss": 0.3727,
-      "step": 13858
-    },
-    {
-      "epoch": 0.9057577936082609,
-      "grad_norm": 0.41829913854599,
-      "learning_rate": 8.09352236983824e-06,
-      "loss": 0.3432,
-      "step": 13859
-    },
-    {
-      "epoch": 0.905823148813803,
-      "grad_norm": 0.45856165885925293,
-      "learning_rate": 8.093248030310102e-06,
-      "loss": 0.4302,
-      "step": 13860
-    },
-    {
-      "epoch": 0.9058885040193452,
-      "grad_norm": 0.5171898007392883,
-      "learning_rate": 8.092973675695122e-06,
-      "loss": 0.3942,
-      "step": 13861
-    },
-    {
-      "epoch": 0.9059538592248872,
-      "grad_norm": 0.4549436867237091,
-      "learning_rate": 8.092699305994639e-06,
-      "loss": 0.3762,
-      "step": 13862
-    },
-    {
-      "epoch": 0.9060192144304294,
-      "grad_norm": 0.4204540252685547,
-      "learning_rate": 8.092424921209989e-06,
-      "loss": 0.3248,
-      "step": 13863
-    },
-    {
-      "epoch": 0.9060845696359715,
-      "grad_norm": 0.41366302967071533,
-      "learning_rate": 8.09215052134251e-06,
-      "loss": 0.3608,
-      "step": 13864
-    },
-    {
-      "epoch": 0.9061499248415136,
-      "grad_norm": 0.3849446177482605,
-      "learning_rate": 8.091876106393544e-06,
-      "loss": 0.2911,
-      "step": 13865
-    },
-    {
-      "epoch": 0.9062152800470558,
-      "grad_norm": 0.4732769727706909,
-      "learning_rate": 8.091601676364424e-06,
-      "loss": 0.3774,
-      "step": 13866
-    },
-    {
-      "epoch": 0.9062806352525978,
-      "grad_norm": 0.5221768617630005,
-      "learning_rate": 8.091327231256495e-06,
-      "loss": 0.5066,
-      "step": 13867
-    },
-    {
-      "epoch": 0.90634599045814,
-      "grad_norm": 0.4407196640968323,
-      "learning_rate": 8.09105277107109e-06,
-      "loss": 0.3817,
-      "step": 13868
-    },
-    {
-      "epoch": 0.9064113456636821,
-      "grad_norm": 0.4109002649784088,
-      "learning_rate": 8.090778295809552e-06,
-      "loss": 0.3026,
-      "step": 13869
-    },
-    {
-      "epoch": 0.9064767008692243,
-      "grad_norm": 0.43740326166152954,
-      "learning_rate": 8.090503805473216e-06,
-      "loss": 0.3809,
-      "step": 13870
-    },
-    {
-      "epoch": 0.9065420560747663,
-      "grad_norm": 0.45587050914764404,
-      "learning_rate": 8.09022930006342e-06,
-      "loss": 0.4207,
-      "step": 13871
-    },
-    {
-      "epoch": 0.9066074112803085,
-      "grad_norm": 0.4603224992752075,
-      "learning_rate": 8.089954779581508e-06,
-      "loss": 0.3795,
-      "step": 13872
-    },
-    {
-      "epoch": 0.9066727664858506,
-      "grad_norm": 0.44550371170043945,
-      "learning_rate": 8.089680244028817e-06,
-      "loss": 0.3529,
-      "step": 13873
-    },
-    {
-      "epoch": 0.9067381216913927,
-      "grad_norm": 0.41442403197288513,
-      "learning_rate": 8.089405693406683e-06,
-      "loss": 0.3249,
-      "step": 13874
-    },
-    {
-      "epoch": 0.9068034768969349,
-      "grad_norm": 0.4551732540130615,
-      "learning_rate": 8.08913112771645e-06,
-      "loss": 0.3902,
-      "step": 13875
-    },
-    {
-      "epoch": 0.9068688321024769,
-      "grad_norm": 0.4341244399547577,
-      "learning_rate": 8.08885654695945e-06,
-      "loss": 0.3342,
-      "step": 13876
-    },
-    {
-      "epoch": 0.9069341873080191,
-      "grad_norm": 0.4735921025276184,
-      "learning_rate": 8.088581951137029e-06,
-      "loss": 0.4375,
-      "step": 13877
-    },
-    {
-      "epoch": 0.9069995425135612,
-      "grad_norm": 0.4558418393135071,
-      "learning_rate": 8.088307340250524e-06,
-      "loss": 0.3953,
-      "step": 13878
-    },
-    {
-      "epoch": 0.9070648977191034,
-      "grad_norm": 0.39925768971443176,
-      "learning_rate": 8.088032714301272e-06,
-      "loss": 0.3412,
-      "step": 13879
-    },
-    {
-      "epoch": 0.9071302529246454,
-      "grad_norm": 0.4197119176387787,
-      "learning_rate": 8.087758073290618e-06,
-      "loss": 0.3374,
-      "step": 13880
-    },
-    {
-      "epoch": 0.9071956081301875,
-      "grad_norm": 0.43608343601226807,
-      "learning_rate": 8.087483417219897e-06,
-      "loss": 0.3535,
-      "step": 13881
-    },
-    {
-      "epoch": 0.9072609633357297,
-      "grad_norm": 0.46259912848472595,
-      "learning_rate": 8.087208746090448e-06,
-      "loss": 0.3699,
-      "step": 13882
-    },
-    {
-      "epoch": 0.9073263185412718,
-      "grad_norm": 0.4515265226364136,
-      "learning_rate": 8.086934059903613e-06,
-      "loss": 0.402,
-      "step": 13883
-    },
-    {
-      "epoch": 0.907391673746814,
-      "grad_norm": 0.43163689970970154,
-      "learning_rate": 8.086659358660734e-06,
-      "loss": 0.3499,
-      "step": 13884
-    },
-    {
-      "epoch": 0.907457028952356,
-      "grad_norm": 0.4312930107116699,
-      "learning_rate": 8.086384642363144e-06,
-      "loss": 0.3928,
-      "step": 13885
-    },
-    {
-      "epoch": 0.9075223841578982,
-      "grad_norm": 0.4166937470436096,
-      "learning_rate": 8.08610991101219e-06,
-      "loss": 0.326,
-      "step": 13886
-    },
-    {
-      "epoch": 0.9075877393634403,
-      "grad_norm": 0.47274652123451233,
-      "learning_rate": 8.085835164609205e-06,
-      "loss": 0.3898,
-      "step": 13887
-    },
-    {
-      "epoch": 0.9076530945689825,
-      "grad_norm": 0.4820028245449066,
-      "learning_rate": 8.085560403155536e-06,
-      "loss": 0.404,
-      "step": 13888
-    },
-    {
-      "epoch": 0.9077184497745245,
-      "grad_norm": 0.40644168853759766,
-      "learning_rate": 8.08528562665252e-06,
-      "loss": 0.3365,
-      "step": 13889
-    },
-    {
-      "epoch": 0.9077838049800666,
-      "grad_norm": 0.4479687809944153,
-      "learning_rate": 8.085010835101496e-06,
-      "loss": 0.41,
-      "step": 13890
-    },
-    {
-      "epoch": 0.9078491601856088,
-      "grad_norm": 0.44872233271598816,
-      "learning_rate": 8.084736028503808e-06,
-      "loss": 0.3813,
-      "step": 13891
-    },
-    {
-      "epoch": 0.9079145153911509,
-      "grad_norm": 0.4451698064804077,
-      "learning_rate": 8.08446120686079e-06,
-      "loss": 0.3877,
-      "step": 13892
-    },
-    {
-      "epoch": 0.907979870596693,
-      "grad_norm": 0.4394383728504181,
-      "learning_rate": 8.084186370173787e-06,
-      "loss": 0.3586,
-      "step": 13893
-    },
-    {
-      "epoch": 0.9080452258022351,
-      "grad_norm": 0.43240517377853394,
-      "learning_rate": 8.083911518444141e-06,
-      "loss": 0.3569,
-      "step": 13894
-    },
-    {
-      "epoch": 0.9081105810077773,
-      "grad_norm": 0.4298829436302185,
-      "learning_rate": 8.083636651673187e-06,
-      "loss": 0.3438,
-      "step": 13895
-    },
-    {
-      "epoch": 0.9081759362133194,
-      "grad_norm": 0.44177982211112976,
-      "learning_rate": 8.083361769862272e-06,
-      "loss": 0.3519,
-      "step": 13896
-    },
-    {
-      "epoch": 0.9082412914188616,
-      "grad_norm": 0.45620620250701904,
-      "learning_rate": 8.083086873012732e-06,
-      "loss": 0.383,
-      "step": 13897
-    },
-    {
-      "epoch": 0.9083066466244036,
-      "grad_norm": 0.42998021841049194,
-      "learning_rate": 8.082811961125908e-06,
-      "loss": 0.3387,
-      "step": 13898
-    },
-    {
-      "epoch": 0.9083720018299457,
-      "grad_norm": 0.4094111919403076,
-      "learning_rate": 8.082537034203145e-06,
-      "loss": 0.3298,
-      "step": 13899
-    },
-    {
-      "epoch": 0.9084373570354879,
-      "grad_norm": 0.46914973855018616,
-      "learning_rate": 8.082262092245779e-06,
-      "loss": 0.4069,
-      "step": 13900
-    },
-    {
-      "epoch": 0.90850271224103,
-      "grad_norm": 0.44797879457473755,
-      "learning_rate": 8.081987135255152e-06,
-      "loss": 0.3753,
-      "step": 13901
-    },
-    {
-      "epoch": 0.9085680674465721,
-      "grad_norm": 0.436219722032547,
-      "learning_rate": 8.081712163232607e-06,
-      "loss": 0.385,
-      "step": 13902
-    },
-    {
-      "epoch": 0.9086334226521142,
-      "grad_norm": 0.4074752628803253,
-      "learning_rate": 8.081437176179485e-06,
-      "loss": 0.3354,
-      "step": 13903
-    },
-    {
-      "epoch": 0.9086987778576564,
-      "grad_norm": 0.43233710527420044,
-      "learning_rate": 8.081162174097125e-06,
-      "loss": 0.3769,
-      "step": 13904
-    },
-    {
-      "epoch": 0.9087641330631985,
-      "grad_norm": 0.4211293160915375,
-      "learning_rate": 8.080887156986873e-06,
-      "loss": 0.3617,
-      "step": 13905
-    },
-    {
-      "epoch": 0.9088294882687407,
-      "grad_norm": 0.46026962995529175,
-      "learning_rate": 8.080612124850062e-06,
-      "loss": 0.4012,
-      "step": 13906
-    },
-    {
-      "epoch": 0.9088948434742827,
-      "grad_norm": 0.4387584328651428,
-      "learning_rate": 8.080337077688042e-06,
-      "loss": 0.3735,
-      "step": 13907
-    },
-    {
-      "epoch": 0.9089601986798248,
-      "grad_norm": 0.43514472246170044,
-      "learning_rate": 8.08006201550215e-06,
-      "loss": 0.3609,
-      "step": 13908
-    },
-    {
-      "epoch": 0.909025553885367,
-      "grad_norm": 0.4435819685459137,
-      "learning_rate": 8.079786938293727e-06,
-      "loss": 0.3997,
-      "step": 13909
-    },
-    {
-      "epoch": 0.9090909090909091,
-      "grad_norm": 0.4352990388870239,
-      "learning_rate": 8.079511846064119e-06,
-      "loss": 0.3341,
-      "step": 13910
-    },
-    {
-      "epoch": 0.9091562642964512,
-      "grad_norm": 0.47828343510627747,
-      "learning_rate": 8.079236738814662e-06,
-      "loss": 0.3977,
-      "step": 13911
-    },
-    {
-      "epoch": 0.9092216195019933,
-      "grad_norm": 0.45470914244651794,
-      "learning_rate": 8.078961616546702e-06,
-      "loss": 0.3826,
-      "step": 13912
-    },
-    {
-      "epoch": 0.9092869747075355,
-      "grad_norm": 0.4326404929161072,
-      "learning_rate": 8.07868647926158e-06,
-      "loss": 0.3691,
-      "step": 13913
-    },
-    {
-      "epoch": 0.9093523299130776,
-      "grad_norm": 0.4797097146511078,
-      "learning_rate": 8.078411326960637e-06,
-      "loss": 0.3779,
-      "step": 13914
-    },
-    {
-      "epoch": 0.9094176851186196,
-      "grad_norm": 0.47024965286254883,
-      "learning_rate": 8.078136159645216e-06,
-      "loss": 0.3828,
-      "step": 13915
-    },
-    {
-      "epoch": 0.9094830403241618,
-      "grad_norm": 0.4575149416923523,
-      "learning_rate": 8.077860977316657e-06,
-      "loss": 0.3733,
-      "step": 13916
-    },
-    {
-      "epoch": 0.9095483955297039,
-      "grad_norm": 0.4415610432624817,
-      "learning_rate": 8.077585779976306e-06,
-      "loss": 0.3381,
-      "step": 13917
-    },
-    {
-      "epoch": 0.9096137507352461,
-      "grad_norm": 0.43210041522979736,
-      "learning_rate": 8.077310567625503e-06,
-      "loss": 0.3489,
-      "step": 13918
-    },
-    {
-      "epoch": 0.9096791059407882,
-      "grad_norm": 0.4586709439754486,
-      "learning_rate": 8.077035340265588e-06,
-      "loss": 0.394,
-      "step": 13919
-    },
-    {
-      "epoch": 0.9097444611463303,
-      "grad_norm": 0.44284117221832275,
-      "learning_rate": 8.076760097897907e-06,
-      "loss": 0.359,
-      "step": 13920
-    },
-    {
-      "epoch": 0.9098098163518724,
-      "grad_norm": 0.49802714586257935,
-      "learning_rate": 8.0764848405238e-06,
-      "loss": 0.4215,
-      "step": 13921
-    },
-    {
-      "epoch": 0.9098751715574146,
-      "grad_norm": 0.4488286077976227,
-      "learning_rate": 8.076209568144612e-06,
-      "loss": 0.4021,
-      "step": 13922
-    },
-    {
-      "epoch": 0.9099405267629567,
-      "grad_norm": 0.48121508955955505,
-      "learning_rate": 8.075934280761684e-06,
-      "loss": 0.3896,
-      "step": 13923
-    },
-    {
-      "epoch": 0.9100058819684987,
-      "grad_norm": 0.45127177238464355,
-      "learning_rate": 8.075658978376358e-06,
-      "loss": 0.3658,
-      "step": 13924
-    },
-    {
-      "epoch": 0.9100712371740409,
-      "grad_norm": 0.4304015338420868,
-      "learning_rate": 8.075383660989978e-06,
-      "loss": 0.3454,
-      "step": 13925
-    },
-    {
-      "epoch": 0.910136592379583,
-      "grad_norm": 0.44155532121658325,
-      "learning_rate": 8.075108328603886e-06,
-      "loss": 0.3974,
-      "step": 13926
-    },
-    {
-      "epoch": 0.9102019475851252,
-      "grad_norm": 0.45621612668037415,
-      "learning_rate": 8.074832981219428e-06,
-      "loss": 0.3569,
-      "step": 13927
-    },
-    {
-      "epoch": 0.9102673027906673,
-      "grad_norm": 0.40090957283973694,
-      "learning_rate": 8.07455761883794e-06,
-      "loss": 0.323,
-      "step": 13928
-    },
-    {
-      "epoch": 0.9103326579962094,
-      "grad_norm": 0.4417174160480499,
-      "learning_rate": 8.074282241460774e-06,
-      "loss": 0.3923,
-      "step": 13929
-    },
-    {
-      "epoch": 0.9103980132017515,
-      "grad_norm": 0.4724140465259552,
-      "learning_rate": 8.074006849089266e-06,
-      "loss": 0.3918,
-      "step": 13930
-    },
-    {
-      "epoch": 0.9104633684072937,
-      "grad_norm": 0.40536820888519287,
-      "learning_rate": 8.073731441724762e-06,
-      "loss": 0.3269,
-      "step": 13931
-    },
-    {
-      "epoch": 0.9105287236128358,
-      "grad_norm": 0.455199658870697,
-      "learning_rate": 8.073456019368604e-06,
-      "loss": 0.394,
-      "step": 13932
-    },
-    {
-      "epoch": 0.9105940788183778,
-      "grad_norm": 0.4606650769710541,
-      "learning_rate": 8.073180582022138e-06,
-      "loss": 0.4054,
-      "step": 13933
-    },
-    {
-      "epoch": 0.91065943402392,
-      "grad_norm": 0.4483288824558258,
-      "learning_rate": 8.072905129686705e-06,
-      "loss": 0.3586,
-      "step": 13934
-    },
-    {
-      "epoch": 0.9107247892294621,
-      "grad_norm": 0.45802754163742065,
-      "learning_rate": 8.072629662363648e-06,
-      "loss": 0.373,
-      "step": 13935
-    },
-    {
-      "epoch": 0.9107901444350043,
-      "grad_norm": 0.4426094591617584,
-      "learning_rate": 8.072354180054312e-06,
-      "loss": 0.3642,
-      "step": 13936
-    },
-    {
-      "epoch": 0.9108554996405464,
-      "grad_norm": 0.46466970443725586,
-      "learning_rate": 8.072078682760042e-06,
-      "loss": 0.4344,
-      "step": 13937
-    },
-    {
-      "epoch": 0.9109208548460885,
-      "grad_norm": 0.46857601404190063,
-      "learning_rate": 8.07180317048218e-06,
-      "loss": 0.4125,
-      "step": 13938
-    },
-    {
-      "epoch": 0.9109862100516306,
-      "grad_norm": 0.5448115468025208,
-      "learning_rate": 8.071527643222068e-06,
-      "loss": 0.4113,
-      "step": 13939
-    },
-    {
-      "epoch": 0.9110515652571727,
-      "grad_norm": 0.43733006715774536,
-      "learning_rate": 8.071252100981053e-06,
-      "loss": 0.3353,
-      "step": 13940
-    },
-    {
-      "epoch": 0.9111169204627149,
-      "grad_norm": 0.4095253348350525,
-      "learning_rate": 8.070976543760475e-06,
-      "loss": 0.3314,
-      "step": 13941
-    },
-    {
-      "epoch": 0.9111822756682569,
-      "grad_norm": 0.4773953855037689,
-      "learning_rate": 8.070700971561682e-06,
-      "loss": 0.4402,
-      "step": 13942
-    },
-    {
-      "epoch": 0.9112476308737991,
-      "grad_norm": 0.4827759861946106,
-      "learning_rate": 8.070425384386018e-06,
-      "loss": 0.455,
-      "step": 13943
-    },
-    {
-      "epoch": 0.9113129860793412,
-      "grad_norm": 0.44401463866233826,
-      "learning_rate": 8.070149782234823e-06,
-      "loss": 0.3515,
-      "step": 13944
-    },
-    {
-      "epoch": 0.9113783412848834,
-      "grad_norm": 0.45008859038352966,
-      "learning_rate": 8.069874165109447e-06,
-      "loss": 0.3595,
-      "step": 13945
-    },
-    {
-      "epoch": 0.9114436964904254,
-      "grad_norm": 0.44510313868522644,
-      "learning_rate": 8.06959853301123e-06,
-      "loss": 0.3815,
-      "step": 13946
-    },
-    {
-      "epoch": 0.9115090516959676,
-      "grad_norm": 0.4350227117538452,
-      "learning_rate": 8.069322885941517e-06,
-      "loss": 0.3498,
-      "step": 13947
-    },
-    {
-      "epoch": 0.9115744069015097,
-      "grad_norm": 0.46352311968803406,
-      "learning_rate": 8.069047223901652e-06,
-      "loss": 0.3966,
-      "step": 13948
-    },
-    {
-      "epoch": 0.9116397621070518,
-      "grad_norm": 0.44037729501724243,
-      "learning_rate": 8.068771546892982e-06,
-      "loss": 0.3924,
-      "step": 13949
-    },
-    {
-      "epoch": 0.911705117312594,
-      "grad_norm": 0.4215856194496155,
-      "learning_rate": 8.068495854916849e-06,
-      "loss": 0.3472,
-      "step": 13950
-    },
-    {
-      "epoch": 0.911770472518136,
-      "grad_norm": 0.44475454092025757,
-      "learning_rate": 8.068220147974599e-06,
-      "loss": 0.3879,
-      "step": 13951
-    },
-    {
-      "epoch": 0.9118358277236782,
-      "grad_norm": 0.4228137731552124,
-      "learning_rate": 8.067944426067577e-06,
-      "loss": 0.3527,
-      "step": 13952
-    },
-    {
-      "epoch": 0.9119011829292203,
-      "grad_norm": 0.44084176421165466,
-      "learning_rate": 8.067668689197128e-06,
-      "loss": 0.4205,
-      "step": 13953
-    },
-    {
-      "epoch": 0.9119665381347625,
-      "grad_norm": 0.42345526814460754,
-      "learning_rate": 8.067392937364594e-06,
-      "loss": 0.3312,
-      "step": 13954
-    },
-    {
-      "epoch": 0.9120318933403045,
-      "grad_norm": 0.4450491964817047,
-      "learning_rate": 8.067117170571323e-06,
-      "loss": 0.364,
-      "step": 13955
-    },
-    {
-      "epoch": 0.9120972485458467,
-      "grad_norm": 0.41773533821105957,
-      "learning_rate": 8.06684138881866e-06,
-      "loss": 0.3477,
-      "step": 13956
-    },
-    {
-      "epoch": 0.9121626037513888,
-      "grad_norm": 0.4562980830669403,
-      "learning_rate": 8.066565592107947e-06,
-      "loss": 0.3764,
-      "step": 13957
-    },
-    {
-      "epoch": 0.9122279589569309,
-      "grad_norm": 0.4346955716609955,
-      "learning_rate": 8.066289780440532e-06,
-      "loss": 0.4004,
-      "step": 13958
-    },
-    {
-      "epoch": 0.9122933141624731,
-      "grad_norm": 0.41704708337783813,
-      "learning_rate": 8.066013953817762e-06,
-      "loss": 0.3318,
-      "step": 13959
-    },
-    {
-      "epoch": 0.9123586693680151,
-      "grad_norm": 0.4251265227794647,
-      "learning_rate": 8.065738112240977e-06,
-      "loss": 0.3644,
-      "step": 13960
-    },
-    {
-      "epoch": 0.9124240245735573,
-      "grad_norm": 0.41447633504867554,
-      "learning_rate": 8.065462255711526e-06,
-      "loss": 0.3574,
-      "step": 13961
-    },
-    {
-      "epoch": 0.9124893797790994,
-      "grad_norm": 0.43733349442481995,
-      "learning_rate": 8.065186384230752e-06,
-      "loss": 0.3472,
-      "step": 13962
-    },
-    {
-      "epoch": 0.9125547349846416,
-      "grad_norm": 0.48272380232810974,
-      "learning_rate": 8.064910497800005e-06,
-      "loss": 0.4234,
-      "step": 13963
-    },
-    {
-      "epoch": 0.9126200901901836,
-      "grad_norm": 0.465209037065506,
-      "learning_rate": 8.064634596420627e-06,
-      "loss": 0.4303,
-      "step": 13964
-    },
-    {
-      "epoch": 0.9126854453957257,
-      "grad_norm": 0.42566102743148804,
-      "learning_rate": 8.064358680093962e-06,
-      "loss": 0.3427,
-      "step": 13965
-    },
-    {
-      "epoch": 0.9127508006012679,
-      "grad_norm": 0.44575703144073486,
-      "learning_rate": 8.06408274882136e-06,
-      "loss": 0.3666,
-      "step": 13966
-    },
-    {
-      "epoch": 0.91281615580681,
-      "grad_norm": 0.46295708417892456,
-      "learning_rate": 8.063806802604164e-06,
-      "loss": 0.3987,
-      "step": 13967
-    },
-    {
-      "epoch": 0.9128815110123522,
-      "grad_norm": 0.48901307582855225,
-      "learning_rate": 8.063530841443721e-06,
-      "loss": 0.3808,
-      "step": 13968
-    },
-    {
-      "epoch": 0.9129468662178942,
-      "grad_norm": 0.47041264176368713,
-      "learning_rate": 8.063254865341378e-06,
-      "loss": 0.4315,
-      "step": 13969
-    },
-    {
-      "epoch": 0.9130122214234364,
-      "grad_norm": 0.4150421917438507,
-      "learning_rate": 8.062978874298479e-06,
-      "loss": 0.3498,
-      "step": 13970
-    },
-    {
-      "epoch": 0.9130775766289785,
-      "grad_norm": 0.45428165793418884,
-      "learning_rate": 8.06270286831637e-06,
-      "loss": 0.3606,
-      "step": 13971
-    },
-    {
-      "epoch": 0.9131429318345207,
-      "grad_norm": 0.435101717710495,
-      "learning_rate": 8.062426847396401e-06,
-      "loss": 0.3515,
-      "step": 13972
-    },
-    {
-      "epoch": 0.9132082870400627,
-      "grad_norm": 0.46090081334114075,
-      "learning_rate": 8.062150811539912e-06,
-      "loss": 0.3965,
-      "step": 13973
-    },
-    {
-      "epoch": 0.9132736422456048,
-      "grad_norm": 0.492020845413208,
-      "learning_rate": 8.061874760748254e-06,
-      "loss": 0.3953,
-      "step": 13974
-    },
-    {
-      "epoch": 0.913338997451147,
-      "grad_norm": 0.4630420207977295,
-      "learning_rate": 8.061598695022772e-06,
-      "loss": 0.3355,
-      "step": 13975
-    },
-    {
-      "epoch": 0.9134043526566891,
-      "grad_norm": 0.42773622274398804,
-      "learning_rate": 8.061322614364813e-06,
-      "loss": 0.336,
-      "step": 13976
-    },
-    {
-      "epoch": 0.9134697078622313,
-      "grad_norm": 0.45446163415908813,
-      "learning_rate": 8.061046518775722e-06,
-      "loss": 0.3332,
-      "step": 13977
-    },
-    {
-      "epoch": 0.9135350630677733,
-      "grad_norm": 0.4189533591270447,
-      "learning_rate": 8.060770408256849e-06,
-      "loss": 0.3306,
-      "step": 13978
-    },
-    {
-      "epoch": 0.9136004182733155,
-      "grad_norm": 0.4757717549800873,
-      "learning_rate": 8.060494282809534e-06,
-      "loss": 0.4246,
-      "step": 13979
-    },
-    {
-      "epoch": 0.9136657734788576,
-      "grad_norm": 0.4250771999359131,
-      "learning_rate": 8.060218142435133e-06,
-      "loss": 0.3579,
-      "step": 13980
-    },
-    {
-      "epoch": 0.9137311286843998,
-      "grad_norm": 0.44313472509384155,
-      "learning_rate": 8.059941987134985e-06,
-      "loss": 0.346,
-      "step": 13981
-    },
-    {
-      "epoch": 0.9137964838899418,
-      "grad_norm": 0.4288141429424286,
-      "learning_rate": 8.05966581691044e-06,
-      "loss": 0.3292,
-      "step": 13982
-    },
-    {
-      "epoch": 0.9138618390954839,
-      "grad_norm": 0.4938511848449707,
-      "learning_rate": 8.059389631762847e-06,
-      "loss": 0.4498,
-      "step": 13983
-    },
-    {
-      "epoch": 0.9139271943010261,
-      "grad_norm": 0.45982861518859863,
-      "learning_rate": 8.05911343169355e-06,
-      "loss": 0.3969,
-      "step": 13984
-    },
-    {
-      "epoch": 0.9139925495065682,
-      "grad_norm": 0.44021573662757874,
-      "learning_rate": 8.058837216703897e-06,
-      "loss": 0.3446,
-      "step": 13985
-    },
-    {
-      "epoch": 0.9140579047121103,
-      "grad_norm": 0.44781169295310974,
-      "learning_rate": 8.058560986795233e-06,
-      "loss": 0.3935,
-      "step": 13986
-    },
-    {
-      "epoch": 0.9141232599176524,
-      "grad_norm": 0.43591830134391785,
-      "learning_rate": 8.05828474196891e-06,
-      "loss": 0.344,
-      "step": 13987
-    },
-    {
-      "epoch": 0.9141886151231946,
-      "grad_norm": 0.47213393449783325,
-      "learning_rate": 8.058008482226272e-06,
-      "loss": 0.3925,
-      "step": 13988
-    },
-    {
-      "epoch": 0.9142539703287367,
-      "grad_norm": 0.43465113639831543,
-      "learning_rate": 8.057732207568666e-06,
-      "loss": 0.3828,
-      "step": 13989
-    },
-    {
-      "epoch": 0.9143193255342789,
-      "grad_norm": 0.4505982995033264,
-      "learning_rate": 8.057455917997443e-06,
-      "loss": 0.4209,
-      "step": 13990
-    },
-    {
-      "epoch": 0.9143846807398209,
-      "grad_norm": 0.43934178352355957,
-      "learning_rate": 8.057179613513945e-06,
-      "loss": 0.3914,
-      "step": 13991
-    },
-    {
-      "epoch": 0.914450035945363,
-      "grad_norm": 0.4296603500843048,
-      "learning_rate": 8.056903294119527e-06,
-      "loss": 0.3319,
-      "step": 13992
-    },
-    {
-      "epoch": 0.9145153911509052,
-      "grad_norm": 0.43655118346214294,
-      "learning_rate": 8.05662695981553e-06,
-      "loss": 0.3668,
-      "step": 13993
-    },
-    {
-      "epoch": 0.9145807463564473,
-      "grad_norm": 0.42281386256217957,
-      "learning_rate": 8.056350610603305e-06,
-      "loss": 0.3508,
-      "step": 13994
-    },
-    {
-      "epoch": 0.9146461015619894,
-      "grad_norm": 0.46946394443511963,
-      "learning_rate": 8.0560742464842e-06,
-      "loss": 0.39,
-      "step": 13995
-    },
-    {
-      "epoch": 0.9147114567675315,
-      "grad_norm": 0.44197699427604675,
-      "learning_rate": 8.05579786745956e-06,
-      "loss": 0.3943,
-      "step": 13996
-    },
-    {
-      "epoch": 0.9147768119730737,
-      "grad_norm": 0.43652769923210144,
-      "learning_rate": 8.055521473530737e-06,
-      "loss": 0.3576,
-      "step": 13997
-    },
-    {
-      "epoch": 0.9148421671786158,
-      "grad_norm": 0.43092644214630127,
-      "learning_rate": 8.055245064699077e-06,
-      "loss": 0.3719,
-      "step": 13998
-    },
-    {
-      "epoch": 0.9149075223841578,
-      "grad_norm": 0.4001566469669342,
-      "learning_rate": 8.054968640965929e-06,
-      "loss": 0.2946,
-      "step": 13999
-    },
-    {
-      "epoch": 0.9149728775897,
-      "grad_norm": 0.4617891311645508,
-      "learning_rate": 8.05469220233264e-06,
-      "loss": 0.423,
-      "step": 14000
-    },
-    {
-      "epoch": 0.9150382327952421,
-      "grad_norm": 0.44813108444213867,
-      "learning_rate": 8.054415748800559e-06,
-      "loss": 0.3614,
-      "step": 14001
-    },
-    {
-      "epoch": 0.9151035880007843,
-      "grad_norm": 0.47789061069488525,
-      "learning_rate": 8.054139280371034e-06,
-      "loss": 0.4192,
-      "step": 14002
-    },
-    {
-      "epoch": 0.9151689432063264,
-      "grad_norm": 0.42289867997169495,
-      "learning_rate": 8.053862797045413e-06,
-      "loss": 0.3332,
-      "step": 14003
-    },
-    {
-      "epoch": 0.9152342984118685,
-      "grad_norm": 0.4325043261051178,
-      "learning_rate": 8.053586298825047e-06,
-      "loss": 0.3664,
-      "step": 14004
-    },
-    {
-      "epoch": 0.9152996536174106,
-      "grad_norm": 0.41553187370300293,
-      "learning_rate": 8.053309785711281e-06,
-      "loss": 0.3513,
-      "step": 14005
-    },
-    {
-      "epoch": 0.9153650088229528,
-      "grad_norm": 0.4572798013687134,
-      "learning_rate": 8.053033257705467e-06,
-      "loss": 0.3855,
-      "step": 14006
-    },
-    {
-      "epoch": 0.9154303640284949,
-      "grad_norm": 0.4481726288795471,
-      "learning_rate": 8.052756714808951e-06,
-      "loss": 0.4085,
-      "step": 14007
-    },
-    {
-      "epoch": 0.9154957192340369,
-      "grad_norm": 0.3968733847141266,
-      "learning_rate": 8.052480157023083e-06,
-      "loss": 0.2995,
-      "step": 14008
-    },
-    {
-      "epoch": 0.9155610744395791,
-      "grad_norm": 0.45097097754478455,
-      "learning_rate": 8.052203584349211e-06,
-      "loss": 0.3681,
-      "step": 14009
-    },
-    {
-      "epoch": 0.9156264296451212,
-      "grad_norm": 0.46446356177330017,
-      "learning_rate": 8.051926996788685e-06,
-      "loss": 0.4181,
-      "step": 14010
-    },
-    {
-      "epoch": 0.9156917848506634,
-      "grad_norm": 0.44969937205314636,
-      "learning_rate": 8.051650394342856e-06,
-      "loss": 0.3817,
-      "step": 14011
-    },
-    {
-      "epoch": 0.9157571400562055,
-      "grad_norm": 0.44434255361557007,
-      "learning_rate": 8.05137377701307e-06,
-      "loss": 0.4061,
-      "step": 14012
-    },
-    {
-      "epoch": 0.9158224952617476,
-      "grad_norm": 0.3985587954521179,
-      "learning_rate": 8.051097144800675e-06,
-      "loss": 0.3089,
-      "step": 14013
-    },
-    {
-      "epoch": 0.9158878504672897,
-      "grad_norm": 0.4738868772983551,
-      "learning_rate": 8.050820497707023e-06,
-      "loss": 0.4415,
-      "step": 14014
-    },
-    {
-      "epoch": 0.9159532056728319,
-      "grad_norm": 0.4262668490409851,
-      "learning_rate": 8.050543835733463e-06,
-      "loss": 0.3278,
-      "step": 14015
-    },
-    {
-      "epoch": 0.916018560878374,
-      "grad_norm": 0.4603838622570038,
-      "learning_rate": 8.050267158881344e-06,
-      "loss": 0.3748,
-      "step": 14016
-    },
-    {
-      "epoch": 0.916083916083916,
-      "grad_norm": 0.4666097164154053,
-      "learning_rate": 8.049990467152016e-06,
-      "loss": 0.4056,
-      "step": 14017
-    },
-    {
-      "epoch": 0.9161492712894582,
-      "grad_norm": 0.46902790665626526,
-      "learning_rate": 8.049713760546827e-06,
-      "loss": 0.4064,
-      "step": 14018
-    },
-    {
-      "epoch": 0.9162146264950003,
-      "grad_norm": 0.4662460684776306,
-      "learning_rate": 8.049437039067127e-06,
-      "loss": 0.4338,
-      "step": 14019
-    },
-    {
-      "epoch": 0.9162799817005425,
-      "grad_norm": 0.4359288811683655,
-      "learning_rate": 8.049160302714267e-06,
-      "loss": 0.3762,
-      "step": 14020
-    },
-    {
-      "epoch": 0.9163453369060846,
-      "grad_norm": 0.4116736650466919,
-      "learning_rate": 8.048883551489595e-06,
-      "loss": 0.3213,
-      "step": 14021
-    },
-    {
-      "epoch": 0.9164106921116267,
-      "grad_norm": 0.4233904480934143,
-      "learning_rate": 8.048606785394464e-06,
-      "loss": 0.349,
-      "step": 14022
-    },
-    {
-      "epoch": 0.9164760473171688,
-      "grad_norm": 0.4174635112285614,
-      "learning_rate": 8.048330004430219e-06,
-      "loss": 0.3442,
-      "step": 14023
-    },
-    {
-      "epoch": 0.9165414025227109,
-      "grad_norm": 0.49676185846328735,
-      "learning_rate": 8.048053208598213e-06,
-      "loss": 0.4151,
-      "step": 14024
-    },
-    {
-      "epoch": 0.9166067577282531,
-      "grad_norm": 0.43644121289253235,
-      "learning_rate": 8.047776397899796e-06,
-      "loss": 0.3533,
-      "step": 14025
-    },
-    {
-      "epoch": 0.9166721129337951,
-      "grad_norm": 0.4449264705181122,
-      "learning_rate": 8.047499572336316e-06,
-      "loss": 0.3655,
-      "step": 14026
-    },
-    {
-      "epoch": 0.9167374681393373,
-      "grad_norm": 0.43979716300964355,
-      "learning_rate": 8.047222731909128e-06,
-      "loss": 0.3836,
-      "step": 14027
-    },
-    {
-      "epoch": 0.9168028233448794,
-      "grad_norm": 0.44416573643684387,
-      "learning_rate": 8.046945876619577e-06,
-      "loss": 0.3648,
-      "step": 14028
-    },
-    {
-      "epoch": 0.9168681785504216,
-      "grad_norm": 0.4326207637786865,
-      "learning_rate": 8.046669006469017e-06,
-      "loss": 0.3672,
-      "step": 14029
-    },
-    {
-      "epoch": 0.9169335337559636,
-      "grad_norm": 0.4313278794288635,
-      "learning_rate": 8.046392121458795e-06,
-      "loss": 0.3228,
-      "step": 14030
-    },
-    {
-      "epoch": 0.9169988889615058,
-      "grad_norm": 0.46649694442749023,
-      "learning_rate": 8.046115221590263e-06,
-      "loss": 0.3869,
-      "step": 14031
-    },
-    {
-      "epoch": 0.9170642441670479,
-      "grad_norm": 0.46140167117118835,
-      "learning_rate": 8.045838306864772e-06,
-      "loss": 0.3821,
-      "step": 14032
-    },
-    {
-      "epoch": 0.91712959937259,
-      "grad_norm": 0.4016384184360504,
-      "learning_rate": 8.045561377283675e-06,
-      "loss": 0.3536,
-      "step": 14033
-    },
-    {
-      "epoch": 0.9171949545781322,
-      "grad_norm": 0.41999879479408264,
-      "learning_rate": 8.045284432848317e-06,
-      "loss": 0.3638,
-      "step": 14034
-    },
-    {
-      "epoch": 0.9172603097836742,
-      "grad_norm": 0.41920387744903564,
-      "learning_rate": 8.045007473560053e-06,
-      "loss": 0.3452,
-      "step": 14035
-    },
-    {
-      "epoch": 0.9173256649892164,
-      "grad_norm": 0.4589771330356598,
-      "learning_rate": 8.044730499420233e-06,
-      "loss": 0.3609,
-      "step": 14036
-    },
-    {
-      "epoch": 0.9173910201947585,
-      "grad_norm": 0.46324291825294495,
-      "learning_rate": 8.044453510430208e-06,
-      "loss": 0.4555,
-      "step": 14037
-    },
-    {
-      "epoch": 0.9174563754003007,
-      "grad_norm": 0.46082064509391785,
-      "learning_rate": 8.044176506591328e-06,
-      "loss": 0.3829,
-      "step": 14038
-    },
-    {
-      "epoch": 0.9175217306058427,
-      "grad_norm": 0.4257607161998749,
-      "learning_rate": 8.043899487904943e-06,
-      "loss": 0.3424,
-      "step": 14039
-    },
-    {
-      "epoch": 0.9175870858113849,
-      "grad_norm": 0.4278295040130615,
-      "learning_rate": 8.043622454372407e-06,
-      "loss": 0.3565,
-      "step": 14040
-    },
-    {
-      "epoch": 0.917652441016927,
-      "grad_norm": 0.4225910007953644,
-      "learning_rate": 8.04334540599507e-06,
-      "loss": 0.3504,
-      "step": 14041
-    },
-    {
-      "epoch": 0.9177177962224691,
-      "grad_norm": 0.4419727623462677,
-      "learning_rate": 8.043068342774283e-06,
-      "loss": 0.3587,
-      "step": 14042
-    },
-    {
-      "epoch": 0.9177831514280113,
-      "grad_norm": 0.4980928897857666,
-      "learning_rate": 8.042791264711398e-06,
-      "loss": 0.4225,
-      "step": 14043
-    },
-    {
-      "epoch": 0.9178485066335533,
-      "grad_norm": 0.44496211409568787,
-      "learning_rate": 8.042514171807767e-06,
-      "loss": 0.4079,
-      "step": 14044
-    },
-    {
-      "epoch": 0.9179138618390955,
-      "grad_norm": 0.4453846514225006,
-      "learning_rate": 8.042237064064737e-06,
-      "loss": 0.3677,
-      "step": 14045
-    },
-    {
-      "epoch": 0.9179792170446376,
-      "grad_norm": 0.4587365388870239,
-      "learning_rate": 8.041959941483666e-06,
-      "loss": 0.401,
-      "step": 14046
-    },
-    {
-      "epoch": 0.9180445722501798,
-      "grad_norm": 0.432699590921402,
-      "learning_rate": 8.0416828040659e-06,
-      "loss": 0.3428,
-      "step": 14047
-    },
-    {
-      "epoch": 0.9181099274557218,
-      "grad_norm": 0.47861722111701965,
-      "learning_rate": 8.041405651812794e-06,
-      "loss": 0.4425,
-      "step": 14048
-    },
-    {
-      "epoch": 0.918175282661264,
-      "grad_norm": 0.457981675863266,
-      "learning_rate": 8.0411284847257e-06,
-      "loss": 0.3687,
-      "step": 14049
-    },
-    {
-      "epoch": 0.9182406378668061,
-      "grad_norm": 0.46564221382141113,
-      "learning_rate": 8.040851302805968e-06,
-      "loss": 0.3863,
-      "step": 14050
-    },
-    {
-      "epoch": 0.9183059930723482,
-      "grad_norm": 0.4485551714897156,
-      "learning_rate": 8.040574106054952e-06,
-      "loss": 0.3785,
-      "step": 14051
-    },
-    {
-      "epoch": 0.9183713482778904,
-      "grad_norm": 0.4502100944519043,
-      "learning_rate": 8.040296894474e-06,
-      "loss": 0.3991,
-      "step": 14052
-    },
-    {
-      "epoch": 0.9184367034834324,
-      "grad_norm": 0.46024709939956665,
-      "learning_rate": 8.04001966806447e-06,
-      "loss": 0.376,
-      "step": 14053
-    },
-    {
-      "epoch": 0.9185020586889746,
-      "grad_norm": 0.42094698548316956,
-      "learning_rate": 8.039742426827709e-06,
-      "loss": 0.3352,
-      "step": 14054
-    },
-    {
-      "epoch": 0.9185674138945167,
-      "grad_norm": 0.4074893593788147,
-      "learning_rate": 8.03946517076507e-06,
-      "loss": 0.3268,
-      "step": 14055
-    },
-    {
-      "epoch": 0.9186327691000589,
-      "grad_norm": 0.4475942552089691,
-      "learning_rate": 8.03918789987791e-06,
-      "loss": 0.4081,
-      "step": 14056
-    },
-    {
-      "epoch": 0.9186981243056009,
-      "grad_norm": 0.44974157214164734,
-      "learning_rate": 8.038910614167574e-06,
-      "loss": 0.405,
-      "step": 14057
-    },
-    {
-      "epoch": 0.918763479511143,
-      "grad_norm": 0.4386823773384094,
-      "learning_rate": 8.03863331363542e-06,
-      "loss": 0.3863,
-      "step": 14058
-    },
-    {
-      "epoch": 0.9188288347166852,
-      "grad_norm": 0.45584502816200256,
-      "learning_rate": 8.038355998282799e-06,
-      "loss": 0.3887,
-      "step": 14059
-    },
-    {
-      "epoch": 0.9188941899222273,
-      "grad_norm": 0.42862972617149353,
-      "learning_rate": 8.038078668111062e-06,
-      "loss": 0.3615,
-      "step": 14060
-    },
-    {
-      "epoch": 0.9189595451277695,
-      "grad_norm": 0.4222893714904785,
-      "learning_rate": 8.037801323121564e-06,
-      "loss": 0.3486,
-      "step": 14061
-    },
-    {
-      "epoch": 0.9190249003333115,
-      "grad_norm": 0.4321688711643219,
-      "learning_rate": 8.037523963315655e-06,
-      "loss": 0.3694,
-      "step": 14062
-    },
-    {
-      "epoch": 0.9190902555388537,
-      "grad_norm": 0.4637817442417145,
-      "learning_rate": 8.037246588694692e-06,
-      "loss": 0.3965,
-      "step": 14063
-    },
-    {
-      "epoch": 0.9191556107443958,
-      "grad_norm": 0.4694924056529999,
-      "learning_rate": 8.036969199260023e-06,
-      "loss": 0.4058,
-      "step": 14064
-    },
-    {
-      "epoch": 0.919220965949938,
-      "grad_norm": 0.4824604392051697,
-      "learning_rate": 8.036691795013004e-06,
-      "loss": 0.446,
-      "step": 14065
-    },
-    {
-      "epoch": 0.91928632115548,
-      "grad_norm": 0.5190874934196472,
-      "learning_rate": 8.036414375954986e-06,
-      "loss": 0.3854,
-      "step": 14066
-    },
-    {
-      "epoch": 0.9193516763610221,
-      "grad_norm": 0.42385995388031006,
-      "learning_rate": 8.036136942087324e-06,
-      "loss": 0.3283,
-      "step": 14067
-    },
-    {
-      "epoch": 0.9194170315665643,
-      "grad_norm": 0.43375396728515625,
-      "learning_rate": 8.03585949341137e-06,
-      "loss": 0.3572,
-      "step": 14068
-    },
-    {
-      "epoch": 0.9194823867721064,
-      "grad_norm": 0.43796199560165405,
-      "learning_rate": 8.03558202992848e-06,
-      "loss": 0.3424,
-      "step": 14069
-    },
-    {
-      "epoch": 0.9195477419776485,
-      "grad_norm": 0.4456939101219177,
-      "learning_rate": 8.035304551640002e-06,
-      "loss": 0.3824,
-      "step": 14070
-    },
-    {
-      "epoch": 0.9196130971831906,
-      "grad_norm": 0.45292073488235474,
-      "learning_rate": 8.035027058547292e-06,
-      "loss": 0.3733,
-      "step": 14071
-    },
-    {
-      "epoch": 0.9196784523887328,
-      "grad_norm": 0.48807671666145325,
-      "learning_rate": 8.034749550651704e-06,
-      "loss": 0.4123,
-      "step": 14072
-    },
-    {
-      "epoch": 0.9197438075942749,
-      "grad_norm": 0.47011175751686096,
-      "learning_rate": 8.034472027954592e-06,
-      "loss": 0.3735,
-      "step": 14073
-    },
-    {
-      "epoch": 0.9198091627998171,
-      "grad_norm": 0.412279337644577,
-      "learning_rate": 8.034194490457308e-06,
-      "loss": 0.3731,
-      "step": 14074
-    },
-    {
-      "epoch": 0.9198745180053591,
-      "grad_norm": 0.45856234431266785,
-      "learning_rate": 8.033916938161205e-06,
-      "loss": 0.371,
-      "step": 14075
-    },
-    {
-      "epoch": 0.9199398732109012,
-      "grad_norm": 0.4508492648601532,
-      "learning_rate": 8.03363937106764e-06,
-      "loss": 0.3777,
-      "step": 14076
-    },
-    {
-      "epoch": 0.9200052284164434,
-      "grad_norm": 0.4206467866897583,
-      "learning_rate": 8.033361789177964e-06,
-      "loss": 0.366,
-      "step": 14077
-    },
-    {
-      "epoch": 0.9200705836219855,
-      "grad_norm": 0.5154498815536499,
-      "learning_rate": 8.033084192493534e-06,
-      "loss": 0.4033,
-      "step": 14078
-    },
-    {
-      "epoch": 0.9201359388275276,
-      "grad_norm": 0.4669542908668518,
-      "learning_rate": 8.0328065810157e-06,
-      "loss": 0.4259,
-      "step": 14079
-    },
-    {
-      "epoch": 0.9202012940330697,
-      "grad_norm": 0.45770663022994995,
-      "learning_rate": 8.032528954745817e-06,
-      "loss": 0.3668,
-      "step": 14080
-    },
-    {
-      "epoch": 0.9202666492386119,
-      "grad_norm": 0.418260782957077,
-      "learning_rate": 8.03225131368524e-06,
-      "loss": 0.3461,
-      "step": 14081
-    },
-    {
-      "epoch": 0.920332004444154,
-      "grad_norm": 0.4325152337551117,
-      "learning_rate": 8.031973657835321e-06,
-      "loss": 0.3519,
-      "step": 14082
-    },
-    {
-      "epoch": 0.920397359649696,
-      "grad_norm": 0.4321635663509369,
-      "learning_rate": 8.03169598719742e-06,
-      "loss": 0.3512,
-      "step": 14083
-    },
-    {
-      "epoch": 0.9204627148552382,
-      "grad_norm": 0.48397862911224365,
-      "learning_rate": 8.031418301772884e-06,
-      "loss": 0.4605,
-      "step": 14084
-    },
-    {
-      "epoch": 0.9205280700607803,
-      "grad_norm": 0.43660688400268555,
-      "learning_rate": 8.031140601563073e-06,
-      "loss": 0.3559,
-      "step": 14085
-    },
-    {
-      "epoch": 0.9205934252663225,
-      "grad_norm": 0.4134843945503235,
-      "learning_rate": 8.030862886569339e-06,
-      "loss": 0.3341,
-      "step": 14086
-    },
-    {
-      "epoch": 0.9206587804718646,
-      "grad_norm": 0.40737178921699524,
-      "learning_rate": 8.030585156793035e-06,
-      "loss": 0.3015,
-      "step": 14087
-    },
-    {
-      "epoch": 0.9207241356774067,
-      "grad_norm": 0.42058470845222473,
-      "learning_rate": 8.030307412235519e-06,
-      "loss": 0.3553,
-      "step": 14088
-    },
-    {
-      "epoch": 0.9207894908829488,
-      "grad_norm": 0.402561753988266,
-      "learning_rate": 8.030029652898144e-06,
-      "loss": 0.3393,
-      "step": 14089
-    },
-    {
-      "epoch": 0.920854846088491,
-      "grad_norm": 0.4387030601501465,
-      "learning_rate": 8.029751878782264e-06,
-      "loss": 0.3579,
-      "step": 14090
-    },
-    {
-      "epoch": 0.9209202012940331,
-      "grad_norm": 0.4189580976963043,
-      "learning_rate": 8.029474089889232e-06,
-      "loss": 0.346,
-      "step": 14091
-    },
-    {
-      "epoch": 0.9209855564995751,
-      "grad_norm": 0.40532082319259644,
-      "learning_rate": 8.029196286220409e-06,
-      "loss": 0.3278,
-      "step": 14092
-    },
-    {
-      "epoch": 0.9210509117051173,
-      "grad_norm": 0.4416372776031494,
-      "learning_rate": 8.028918467777145e-06,
-      "loss": 0.3642,
-      "step": 14093
-    },
-    {
-      "epoch": 0.9211162669106594,
-      "grad_norm": 0.4237021803855896,
-      "learning_rate": 8.028640634560796e-06,
-      "loss": 0.3623,
-      "step": 14094
-    },
-    {
-      "epoch": 0.9211816221162016,
-      "grad_norm": 0.4447159171104431,
-      "learning_rate": 8.028362786572718e-06,
-      "loss": 0.3813,
-      "step": 14095
-    },
-    {
-      "epoch": 0.9212469773217437,
-      "grad_norm": 0.4404435157775879,
-      "learning_rate": 8.028084923814266e-06,
-      "loss": 0.3373,
-      "step": 14096
-    },
-    {
-      "epoch": 0.9213123325272858,
-      "grad_norm": 0.4356805086135864,
-      "learning_rate": 8.027807046286795e-06,
-      "loss": 0.3691,
-      "step": 14097
-    },
-    {
-      "epoch": 0.9213776877328279,
-      "grad_norm": 0.455952525138855,
-      "learning_rate": 8.027529153991659e-06,
-      "loss": 0.355,
-      "step": 14098
-    },
-    {
-      "epoch": 0.9214430429383701,
-      "grad_norm": 0.42456090450286865,
-      "learning_rate": 8.027251246930214e-06,
-      "loss": 0.342,
-      "step": 14099
-    },
-    {
-      "epoch": 0.9215083981439122,
-      "grad_norm": 0.4053103029727936,
-      "learning_rate": 8.026973325103818e-06,
-      "loss": 0.3009,
-      "step": 14100
-    },
-    {
-      "epoch": 0.9215737533494542,
-      "grad_norm": 0.44154906272888184,
-      "learning_rate": 8.026695388513822e-06,
-      "loss": 0.399,
-      "step": 14101
-    },
-    {
-      "epoch": 0.9216391085549964,
-      "grad_norm": 0.41678181290626526,
-      "learning_rate": 8.026417437161585e-06,
-      "loss": 0.3506,
-      "step": 14102
-    },
-    {
-      "epoch": 0.9217044637605385,
-      "grad_norm": 0.41899269819259644,
-      "learning_rate": 8.026139471048462e-06,
-      "loss": 0.3613,
-      "step": 14103
-    },
-    {
-      "epoch": 0.9217698189660807,
-      "grad_norm": 0.4102371037006378,
-      "learning_rate": 8.025861490175809e-06,
-      "loss": 0.3574,
-      "step": 14104
-    },
-    {
-      "epoch": 0.9218351741716228,
-      "grad_norm": 0.415998637676239,
-      "learning_rate": 8.025583494544979e-06,
-      "loss": 0.3489,
-      "step": 14105
-    },
-    {
-      "epoch": 0.9219005293771649,
-      "grad_norm": 0.4359937310218811,
-      "learning_rate": 8.025305484157332e-06,
-      "loss": 0.4027,
-      "step": 14106
-    },
-    {
-      "epoch": 0.921965884582707,
-      "grad_norm": 0.44541677832603455,
-      "learning_rate": 8.025027459014223e-06,
-      "loss": 0.3686,
-      "step": 14107
-    },
-    {
-      "epoch": 0.9220312397882491,
-      "grad_norm": 0.4568682014942169,
-      "learning_rate": 8.024749419117007e-06,
-      "loss": 0.3972,
-      "step": 14108
-    },
-    {
-      "epoch": 0.9220965949937913,
-      "grad_norm": 0.47529077529907227,
-      "learning_rate": 8.024471364467039e-06,
-      "loss": 0.4396,
-      "step": 14109
-    },
-    {
-      "epoch": 0.9221619501993333,
-      "grad_norm": 0.4220034182071686,
-      "learning_rate": 8.024193295065677e-06,
-      "loss": 0.336,
-      "step": 14110
-    },
-    {
-      "epoch": 0.9222273054048755,
-      "grad_norm": 0.45059680938720703,
-      "learning_rate": 8.023915210914274e-06,
-      "loss": 0.3526,
-      "step": 14111
-    },
-    {
-      "epoch": 0.9222926606104176,
-      "grad_norm": 0.4628525674343109,
-      "learning_rate": 8.023637112014192e-06,
-      "loss": 0.4058,
-      "step": 14112
-    },
-    {
-      "epoch": 0.9223580158159598,
-      "grad_norm": 0.42804980278015137,
-      "learning_rate": 8.023358998366783e-06,
-      "loss": 0.3568,
-      "step": 14113
-    },
-    {
-      "epoch": 0.9224233710215018,
-      "grad_norm": 0.4664314091205597,
-      "learning_rate": 8.023080869973405e-06,
-      "loss": 0.408,
-      "step": 14114
-    },
-    {
-      "epoch": 0.922488726227044,
-      "grad_norm": 0.43588483333587646,
-      "learning_rate": 8.022802726835415e-06,
-      "loss": 0.4116,
-      "step": 14115
-    },
-    {
-      "epoch": 0.9225540814325861,
-      "grad_norm": 0.43505632877349854,
-      "learning_rate": 8.022524568954169e-06,
-      "loss": 0.3795,
-      "step": 14116
-    },
-    {
-      "epoch": 0.9226194366381282,
-      "grad_norm": 0.4895874261856079,
-      "learning_rate": 8.022246396331022e-06,
-      "loss": 0.3742,
-      "step": 14117
-    },
-    {
-      "epoch": 0.9226847918436704,
-      "grad_norm": 0.4584488570690155,
-      "learning_rate": 8.021968208967334e-06,
-      "loss": 0.4031,
-      "step": 14118
-    },
-    {
-      "epoch": 0.9227501470492124,
-      "grad_norm": 0.4242425560951233,
-      "learning_rate": 8.021690006864459e-06,
-      "loss": 0.3383,
-      "step": 14119
-    },
-    {
-      "epoch": 0.9228155022547546,
-      "grad_norm": 0.4379027187824249,
-      "learning_rate": 8.021411790023755e-06,
-      "loss": 0.3827,
-      "step": 14120
-    },
-    {
-      "epoch": 0.9228808574602967,
-      "grad_norm": 0.4540347158908844,
-      "learning_rate": 8.02113355844658e-06,
-      "loss": 0.4023,
-      "step": 14121
-    },
-    {
-      "epoch": 0.9229462126658389,
-      "grad_norm": 0.528438925743103,
-      "learning_rate": 8.020855312134289e-06,
-      "loss": 0.4738,
-      "step": 14122
-    },
-    {
-      "epoch": 0.923011567871381,
-      "grad_norm": 0.4343968331813812,
-      "learning_rate": 8.020577051088241e-06,
-      "loss": 0.3494,
-      "step": 14123
-    },
-    {
-      "epoch": 0.9230769230769231,
-      "grad_norm": 0.41011497378349304,
-      "learning_rate": 8.020298775309792e-06,
-      "loss": 0.3222,
-      "step": 14124
-    },
-    {
-      "epoch": 0.9231422782824652,
-      "grad_norm": 0.42649218440055847,
-      "learning_rate": 8.0200204848003e-06,
-      "loss": 0.3851,
-      "step": 14125
-    },
-    {
-      "epoch": 0.9232076334880073,
-      "grad_norm": 0.4764016568660736,
-      "learning_rate": 8.019742179561119e-06,
-      "loss": 0.3977,
-      "step": 14126
-    },
-    {
-      "epoch": 0.9232729886935495,
-      "grad_norm": 0.4466722905635834,
-      "learning_rate": 8.019463859593613e-06,
-      "loss": 0.3403,
-      "step": 14127
-    },
-    {
-      "epoch": 0.9233383438990915,
-      "grad_norm": 0.41392815113067627,
-      "learning_rate": 8.019185524899133e-06,
-      "loss": 0.3264,
-      "step": 14128
-    },
-    {
-      "epoch": 0.9234036991046337,
-      "grad_norm": 0.4437607526779175,
-      "learning_rate": 8.018907175479041e-06,
-      "loss": 0.3684,
-      "step": 14129
-    },
-    {
-      "epoch": 0.9234690543101758,
-      "grad_norm": 0.4070267677307129,
-      "learning_rate": 8.018628811334693e-06,
-      "loss": 0.3307,
-      "step": 14130
-    },
-    {
-      "epoch": 0.923534409515718,
-      "grad_norm": 0.44076064229011536,
-      "learning_rate": 8.018350432467446e-06,
-      "loss": 0.3766,
-      "step": 14131
-    },
-    {
-      "epoch": 0.92359976472126,
-      "grad_norm": 0.4590109884738922,
-      "learning_rate": 8.018072038878657e-06,
-      "loss": 0.3683,
-      "step": 14132
-    },
-    {
-      "epoch": 0.9236651199268022,
-      "grad_norm": 0.4435414671897888,
-      "learning_rate": 8.017793630569689e-06,
-      "loss": 0.3853,
-      "step": 14133
-    },
-    {
-      "epoch": 0.9237304751323443,
-      "grad_norm": 0.44672533869743347,
-      "learning_rate": 8.017515207541892e-06,
-      "loss": 0.4068,
-      "step": 14134
-    },
-    {
-      "epoch": 0.9237958303378864,
-      "grad_norm": 0.41453424096107483,
-      "learning_rate": 8.017236769796628e-06,
-      "loss": 0.3774,
-      "step": 14135
-    },
-    {
-      "epoch": 0.9238611855434286,
-      "grad_norm": 0.4354002773761749,
-      "learning_rate": 8.016958317335257e-06,
-      "loss": 0.3926,
-      "step": 14136
-    },
-    {
-      "epoch": 0.9239265407489706,
-      "grad_norm": 0.43771126866340637,
-      "learning_rate": 8.016679850159134e-06,
-      "loss": 0.3681,
-      "step": 14137
-    },
-    {
-      "epoch": 0.9239918959545128,
-      "grad_norm": 0.3939455449581146,
-      "learning_rate": 8.016401368269618e-06,
-      "loss": 0.3074,
-      "step": 14138
-    },
-    {
-      "epoch": 0.9240572511600549,
-      "grad_norm": 0.5587556958198547,
-      "learning_rate": 8.016122871668068e-06,
-      "loss": 0.3394,
-      "step": 14139
-    },
-    {
-      "epoch": 0.9241226063655971,
-      "grad_norm": 0.4305996894836426,
-      "learning_rate": 8.015844360355841e-06,
-      "loss": 0.3716,
-      "step": 14140
-    },
-    {
-      "epoch": 0.9241879615711391,
-      "grad_norm": 0.45352616906166077,
-      "learning_rate": 8.0155658343343e-06,
-      "loss": 0.3955,
-      "step": 14141
-    },
-    {
-      "epoch": 0.9242533167766812,
-      "grad_norm": 0.4218350052833557,
-      "learning_rate": 8.015287293604796e-06,
-      "loss": 0.3251,
-      "step": 14142
-    },
-    {
-      "epoch": 0.9243186719822234,
-      "grad_norm": 0.44621509313583374,
-      "learning_rate": 8.015008738168692e-06,
-      "loss": 0.3738,
-      "step": 14143
-    },
-    {
-      "epoch": 0.9243840271877655,
-      "grad_norm": 0.4402707815170288,
-      "learning_rate": 8.014730168027345e-06,
-      "loss": 0.3782,
-      "step": 14144
-    },
-    {
-      "epoch": 0.9244493823933077,
-      "grad_norm": 0.4362731873989105,
-      "learning_rate": 8.014451583182117e-06,
-      "loss": 0.3671,
-      "step": 14145
-    },
-    {
-      "epoch": 0.9245147375988497,
-      "grad_norm": 0.4343254566192627,
-      "learning_rate": 8.014172983634363e-06,
-      "loss": 0.36,
-      "step": 14146
-    },
-    {
-      "epoch": 0.9245800928043919,
-      "grad_norm": 0.43919089436531067,
-      "learning_rate": 8.013894369385442e-06,
-      "loss": 0.3773,
-      "step": 14147
-    },
-    {
-      "epoch": 0.924645448009934,
-      "grad_norm": 0.490822970867157,
-      "learning_rate": 8.013615740436717e-06,
-      "loss": 0.4198,
-      "step": 14148
-    },
-    {
-      "epoch": 0.9247108032154762,
-      "grad_norm": 0.48890480399131775,
-      "learning_rate": 8.013337096789541e-06,
-      "loss": 0.4519,
-      "step": 14149
-    },
-    {
-      "epoch": 0.9247761584210182,
-      "grad_norm": 0.4303399324417114,
-      "learning_rate": 8.013058438445278e-06,
-      "loss": 0.3518,
-      "step": 14150
-    },
-    {
-      "epoch": 0.9248415136265603,
-      "grad_norm": 0.4564321041107178,
-      "learning_rate": 8.012779765405285e-06,
-      "loss": 0.4345,
-      "step": 14151
-    },
-    {
-      "epoch": 0.9249068688321025,
-      "grad_norm": 0.4025900661945343,
-      "learning_rate": 8.012501077670922e-06,
-      "loss": 0.3022,
-      "step": 14152
-    },
-    {
-      "epoch": 0.9249722240376446,
-      "grad_norm": 0.5529888272285461,
-      "learning_rate": 8.012222375243545e-06,
-      "loss": 0.565,
-      "step": 14153
-    },
-    {
-      "epoch": 0.9250375792431867,
-      "grad_norm": 0.4156367778778076,
-      "learning_rate": 8.011943658124516e-06,
-      "loss": 0.3465,
-      "step": 14154
-    },
-    {
-      "epoch": 0.9251029344487288,
-      "grad_norm": 0.4275127649307251,
-      "learning_rate": 8.011664926315197e-06,
-      "loss": 0.356,
-      "step": 14155
-    },
-    {
-      "epoch": 0.925168289654271,
-      "grad_norm": 0.47950467467308044,
-      "learning_rate": 8.011386179816944e-06,
-      "loss": 0.3727,
-      "step": 14156
-    },
-    {
-      "epoch": 0.9252336448598131,
-      "grad_norm": 0.4584794044494629,
-      "learning_rate": 8.011107418631117e-06,
-      "loss": 0.4177,
-      "step": 14157
-    },
-    {
-      "epoch": 0.9252990000653553,
-      "grad_norm": 0.4537228047847748,
-      "learning_rate": 8.010828642759076e-06,
-      "loss": 0.3941,
-      "step": 14158
-    },
-    {
-      "epoch": 0.9253643552708973,
-      "grad_norm": 0.46418997645378113,
-      "learning_rate": 8.01054985220218e-06,
-      "loss": 0.4151,
-      "step": 14159
-    },
-    {
-      "epoch": 0.9254297104764394,
-      "grad_norm": 0.42200714349746704,
-      "learning_rate": 8.010271046961791e-06,
-      "loss": 0.3486,
-      "step": 14160
-    },
-    {
-      "epoch": 0.9254950656819816,
-      "grad_norm": 0.4400850236415863,
-      "learning_rate": 8.009992227039264e-06,
-      "loss": 0.3802,
-      "step": 14161
-    },
-    {
-      "epoch": 0.9255604208875237,
-      "grad_norm": 0.4134165346622467,
-      "learning_rate": 8.009713392435966e-06,
-      "loss": 0.3162,
-      "step": 14162
-    },
-    {
-      "epoch": 0.9256257760930658,
-      "grad_norm": 0.4397551119327545,
-      "learning_rate": 8.009434543153252e-06,
-      "loss": 0.3634,
-      "step": 14163
-    },
-    {
-      "epoch": 0.9256911312986079,
-      "grad_norm": 0.43329912424087524,
-      "learning_rate": 8.009155679192482e-06,
-      "loss": 0.3662,
-      "step": 14164
-    },
-    {
-      "epoch": 0.9257564865041501,
-      "grad_norm": 0.44430699944496155,
-      "learning_rate": 8.008876800555018e-06,
-      "loss": 0.3675,
-      "step": 14165
-    },
-    {
-      "epoch": 0.9258218417096922,
-      "grad_norm": 0.424127995967865,
-      "learning_rate": 8.00859790724222e-06,
-      "loss": 0.3598,
-      "step": 14166
-    },
-    {
-      "epoch": 0.9258871969152342,
-      "grad_norm": 0.41890501976013184,
-      "learning_rate": 8.008318999255447e-06,
-      "loss": 0.3285,
-      "step": 14167
-    },
-    {
-      "epoch": 0.9259525521207764,
-      "grad_norm": 0.4419791102409363,
-      "learning_rate": 8.008040076596059e-06,
-      "loss": 0.4035,
-      "step": 14168
-    },
-    {
-      "epoch": 0.9260179073263185,
-      "grad_norm": 0.4523894786834717,
-      "learning_rate": 8.00776113926542e-06,
-      "loss": 0.4088,
-      "step": 14169
-    },
-    {
-      "epoch": 0.9260832625318607,
-      "grad_norm": 0.4440203905105591,
-      "learning_rate": 8.007482187264885e-06,
-      "loss": 0.3575,
-      "step": 14170
-    },
-    {
-      "epoch": 0.9261486177374028,
-      "grad_norm": 0.4431575834751129,
-      "learning_rate": 8.007203220595821e-06,
-      "loss": 0.4295,
-      "step": 14171
-    },
-    {
-      "epoch": 0.9262139729429449,
-      "grad_norm": 0.46677714586257935,
-      "learning_rate": 8.006924239259582e-06,
-      "loss": 0.4087,
-      "step": 14172
-    },
-    {
-      "epoch": 0.926279328148487,
-      "grad_norm": 0.40673959255218506,
-      "learning_rate": 8.006645243257534e-06,
-      "loss": 0.3306,
-      "step": 14173
-    },
-    {
-      "epoch": 0.9263446833540292,
-      "grad_norm": 0.45933017134666443,
-      "learning_rate": 8.006366232591035e-06,
-      "loss": 0.4008,
-      "step": 14174
-    },
-    {
-      "epoch": 0.9264100385595713,
-      "grad_norm": 0.42018696665763855,
-      "learning_rate": 8.006087207261445e-06,
-      "loss": 0.3508,
-      "step": 14175
-    },
-    {
-      "epoch": 0.9264753937651133,
-      "grad_norm": 0.42296692728996277,
-      "learning_rate": 8.005808167270126e-06,
-      "loss": 0.3451,
-      "step": 14176
-    },
-    {
-      "epoch": 0.9265407489706555,
-      "grad_norm": 0.43631839752197266,
-      "learning_rate": 8.00552911261844e-06,
-      "loss": 0.3254,
-      "step": 14177
-    },
-    {
-      "epoch": 0.9266061041761976,
-      "grad_norm": 0.47382402420043945,
-      "learning_rate": 8.005250043307749e-06,
-      "loss": 0.4232,
-      "step": 14178
-    },
-    {
-      "epoch": 0.9266714593817398,
-      "grad_norm": 0.43479087948799133,
-      "learning_rate": 8.004970959339411e-06,
-      "loss": 0.3538,
-      "step": 14179
-    },
-    {
-      "epoch": 0.9267368145872819,
-      "grad_norm": 0.44161510467529297,
-      "learning_rate": 8.004691860714788e-06,
-      "loss": 0.3582,
-      "step": 14180
-    },
-    {
-      "epoch": 0.926802169792824,
-      "grad_norm": 0.4279502332210541,
-      "learning_rate": 8.004412747435244e-06,
-      "loss": 0.3639,
-      "step": 14181
-    },
-    {
-      "epoch": 0.9268675249983661,
-      "grad_norm": 0.432258278131485,
-      "learning_rate": 8.004133619502137e-06,
-      "loss": 0.3585,
-      "step": 14182
-    },
-    {
-      "epoch": 0.9269328802039083,
-      "grad_norm": 0.43750303983688354,
-      "learning_rate": 8.003854476916828e-06,
-      "loss": 0.3829,
-      "step": 14183
-    },
-    {
-      "epoch": 0.9269982354094504,
-      "grad_norm": 0.42415958642959595,
-      "learning_rate": 8.003575319680682e-06,
-      "loss": 0.3415,
-      "step": 14184
-    },
-    {
-      "epoch": 0.9270635906149924,
-      "grad_norm": 0.4484017491340637,
-      "learning_rate": 8.003296147795058e-06,
-      "loss": 0.3768,
-      "step": 14185
-    },
-    {
-      "epoch": 0.9271289458205346,
-      "grad_norm": 0.4493042826652527,
-      "learning_rate": 8.003016961261318e-06,
-      "loss": 0.373,
-      "step": 14186
-    },
-    {
-      "epoch": 0.9271943010260767,
-      "grad_norm": 0.43740928173065186,
-      "learning_rate": 8.002737760080826e-06,
-      "loss": 0.3696,
-      "step": 14187
-    },
-    {
-      "epoch": 0.9272596562316189,
-      "grad_norm": 0.42805278301239014,
-      "learning_rate": 8.002458544254939e-06,
-      "loss": 0.3669,
-      "step": 14188
-    },
-    {
-      "epoch": 0.927325011437161,
-      "grad_norm": 0.4790594279766083,
-      "learning_rate": 8.002179313785023e-06,
-      "loss": 0.4044,
-      "step": 14189
-    },
-    {
-      "epoch": 0.9273903666427031,
-      "grad_norm": 0.5047306418418884,
-      "learning_rate": 8.001900068672438e-06,
-      "loss": 0.4053,
-      "step": 14190
-    },
-    {
-      "epoch": 0.9274557218482452,
-      "grad_norm": 0.490181028842926,
-      "learning_rate": 8.001620808918546e-06,
-      "loss": 0.4357,
-      "step": 14191
-    },
-    {
-      "epoch": 0.9275210770537873,
-      "grad_norm": 0.44902509450912476,
-      "learning_rate": 8.00134153452471e-06,
-      "loss": 0.391,
-      "step": 14192
-    },
-    {
-      "epoch": 0.9275864322593295,
-      "grad_norm": 0.43170663714408875,
-      "learning_rate": 8.001062245492291e-06,
-      "loss": 0.3439,
-      "step": 14193
-    },
-    {
-      "epoch": 0.9276517874648715,
-      "grad_norm": 0.3930788040161133,
-      "learning_rate": 8.000782941822653e-06,
-      "loss": 0.2906,
-      "step": 14194
-    },
-    {
-      "epoch": 0.9277171426704137,
-      "grad_norm": 0.4794231355190277,
-      "learning_rate": 8.000503623517155e-06,
-      "loss": 0.3957,
-      "step": 14195
-    },
-    {
-      "epoch": 0.9277824978759558,
-      "grad_norm": 0.4684624969959259,
-      "learning_rate": 8.000224290577164e-06,
-      "loss": 0.4034,
-      "step": 14196
-    },
-    {
-      "epoch": 0.927847853081498,
-      "grad_norm": 0.4549058675765991,
-      "learning_rate": 7.999944943004036e-06,
-      "loss": 0.4003,
-      "step": 14197
-    },
-    {
-      "epoch": 0.92791320828704,
-      "grad_norm": 0.45974260568618774,
-      "learning_rate": 7.999665580799142e-06,
-      "loss": 0.405,
-      "step": 14198
-    },
-    {
-      "epoch": 0.9279785634925822,
-      "grad_norm": 0.47040218114852905,
-      "learning_rate": 7.999386203963836e-06,
-      "loss": 0.4308,
-      "step": 14199
-    },
-    {
-      "epoch": 0.9280439186981243,
-      "grad_norm": 0.524333119392395,
-      "learning_rate": 7.999106812499486e-06,
-      "loss": 0.5041,
-      "step": 14200
-    },
-    {
-      "epoch": 0.9281092739036664,
-      "grad_norm": 0.4343332350254059,
-      "learning_rate": 7.998827406407453e-06,
-      "loss": 0.3523,
-      "step": 14201
-    },
-    {
-      "epoch": 0.9281746291092086,
-      "grad_norm": 0.46028608083724976,
-      "learning_rate": 7.9985479856891e-06,
-      "loss": 0.3914,
-      "step": 14202
-    },
-    {
-      "epoch": 0.9282399843147506,
-      "grad_norm": 0.4479673206806183,
-      "learning_rate": 7.998268550345788e-06,
-      "loss": 0.3111,
-      "step": 14203
-    },
-    {
-      "epoch": 0.9283053395202928,
-      "grad_norm": 0.46122410893440247,
-      "learning_rate": 7.997989100378883e-06,
-      "loss": 0.4118,
-      "step": 14204
-    },
-    {
-      "epoch": 0.9283706947258349,
-      "grad_norm": 0.4570966362953186,
-      "learning_rate": 7.997709635789746e-06,
-      "loss": 0.39,
-      "step": 14205
-    },
-    {
-      "epoch": 0.9284360499313771,
-      "grad_norm": 0.42992284893989563,
-      "learning_rate": 7.997430156579742e-06,
-      "loss": 0.3462,
-      "step": 14206
-    },
-    {
-      "epoch": 0.9285014051369191,
-      "grad_norm": 0.4584921598434448,
-      "learning_rate": 7.997150662750232e-06,
-      "loss": 0.3759,
-      "step": 14207
-    },
-    {
-      "epoch": 0.9285667603424613,
-      "grad_norm": 0.45437732338905334,
-      "learning_rate": 7.99687115430258e-06,
-      "loss": 0.3729,
-      "step": 14208
-    },
-    {
-      "epoch": 0.9286321155480034,
-      "grad_norm": 0.47542712092399597,
-      "learning_rate": 7.996591631238149e-06,
-      "loss": 0.3741,
-      "step": 14209
-    },
-    {
-      "epoch": 0.9286974707535455,
-      "grad_norm": 0.4160076975822449,
-      "learning_rate": 7.9963120935583e-06,
-      "loss": 0.3752,
-      "step": 14210
-    },
-    {
-      "epoch": 0.9287628259590877,
-      "grad_norm": 0.4273441731929779,
-      "learning_rate": 7.996032541264403e-06,
-      "loss": 0.3467,
-      "step": 14211
-    },
-    {
-      "epoch": 0.9288281811646297,
-      "grad_norm": 0.4041471481323242,
-      "learning_rate": 7.995752974357816e-06,
-      "loss": 0.2815,
-      "step": 14212
-    },
-    {
-      "epoch": 0.9288935363701719,
-      "grad_norm": 0.4183214008808136,
-      "learning_rate": 7.995473392839903e-06,
-      "loss": 0.3351,
-      "step": 14213
-    },
-    {
-      "epoch": 0.928958891575714,
-      "grad_norm": 0.4557066857814789,
-      "learning_rate": 7.995193796712028e-06,
-      "loss": 0.3843,
-      "step": 14214
-    },
-    {
-      "epoch": 0.9290242467812562,
-      "grad_norm": 0.4735686779022217,
-      "learning_rate": 7.994914185975557e-06,
-      "loss": 0.4101,
-      "step": 14215
-    },
-    {
-      "epoch": 0.9290896019867982,
-      "grad_norm": 0.43914374709129333,
-      "learning_rate": 7.99463456063185e-06,
-      "loss": 0.3932,
-      "step": 14216
-    },
-    {
-      "epoch": 0.9291549571923404,
-      "grad_norm": 0.42830193042755127,
-      "learning_rate": 7.994354920682276e-06,
-      "loss": 0.3551,
-      "step": 14217
-    },
-    {
-      "epoch": 0.9292203123978825,
-      "grad_norm": 0.4560054838657379,
-      "learning_rate": 7.994075266128194e-06,
-      "loss": 0.369,
-      "step": 14218
-    },
-    {
-      "epoch": 0.9292856676034246,
-      "grad_norm": 0.4357489347457886,
-      "learning_rate": 7.993795596970968e-06,
-      "loss": 0.3852,
-      "step": 14219
-    },
-    {
-      "epoch": 0.9293510228089668,
-      "grad_norm": 0.42314496636390686,
-      "learning_rate": 7.993515913211967e-06,
-      "loss": 0.3185,
-      "step": 14220
-    },
-    {
-      "epoch": 0.9294163780145088,
-      "grad_norm": 0.4712805151939392,
-      "learning_rate": 7.993236214852548e-06,
-      "loss": 0.3932,
-      "step": 14221
-    },
-    {
-      "epoch": 0.929481733220051,
-      "grad_norm": 0.43544238805770874,
-      "learning_rate": 7.992956501894083e-06,
-      "loss": 0.376,
-      "step": 14222
-    },
-    {
-      "epoch": 0.9295470884255931,
-      "grad_norm": 0.4176951050758362,
-      "learning_rate": 7.992676774337928e-06,
-      "loss": 0.3721,
-      "step": 14223
-    },
-    {
-      "epoch": 0.9296124436311353,
-      "grad_norm": 0.45315301418304443,
-      "learning_rate": 7.992397032185455e-06,
-      "loss": 0.3852,
-      "step": 14224
-    },
-    {
-      "epoch": 0.9296777988366773,
-      "grad_norm": 0.47756463289260864,
-      "learning_rate": 7.992117275438024e-06,
-      "loss": 0.408,
-      "step": 14225
-    },
-    {
-      "epoch": 0.9297431540422194,
-      "grad_norm": 0.5052573084831238,
-      "learning_rate": 7.991837504097e-06,
-      "loss": 0.4338,
-      "step": 14226
-    },
-    {
-      "epoch": 0.9298085092477616,
-      "grad_norm": 0.4030114412307739,
-      "learning_rate": 7.99155771816375e-06,
-      "loss": 0.3056,
-      "step": 14227
-    },
-    {
-      "epoch": 0.9298738644533037,
-      "grad_norm": 0.45975929498672485,
-      "learning_rate": 7.991277917639633e-06,
-      "loss": 0.3831,
-      "step": 14228
-    },
-    {
-      "epoch": 0.9299392196588459,
-      "grad_norm": 0.49597951769828796,
-      "learning_rate": 7.99099810252602e-06,
-      "loss": 0.3914,
-      "step": 14229
-    },
-    {
-      "epoch": 0.9300045748643879,
-      "grad_norm": 0.4525391161441803,
-      "learning_rate": 7.990718272824271e-06,
-      "loss": 0.3737,
-      "step": 14230
-    },
-    {
-      "epoch": 0.9300699300699301,
-      "grad_norm": 0.4771629869937897,
-      "learning_rate": 7.990438428535757e-06,
-      "loss": 0.4152,
-      "step": 14231
-    },
-    {
-      "epoch": 0.9301352852754722,
-      "grad_norm": 0.46415191888809204,
-      "learning_rate": 7.990158569661836e-06,
-      "loss": 0.3592,
-      "step": 14232
-    },
-    {
-      "epoch": 0.9302006404810144,
-      "grad_norm": 0.43217816948890686,
-      "learning_rate": 7.989878696203875e-06,
-      "loss": 0.384,
-      "step": 14233
-    },
-    {
-      "epoch": 0.9302659956865564,
-      "grad_norm": 0.4254879057407379,
-      "learning_rate": 7.98959880816324e-06,
-      "loss": 0.3658,
-      "step": 14234
-    },
-    {
-      "epoch": 0.9303313508920985,
-      "grad_norm": 0.4584832787513733,
-      "learning_rate": 7.989318905541297e-06,
-      "loss": 0.3765,
-      "step": 14235
-    },
-    {
-      "epoch": 0.9303967060976407,
-      "grad_norm": 0.46705707907676697,
-      "learning_rate": 7.98903898833941e-06,
-      "loss": 0.4238,
-      "step": 14236
-    },
-    {
-      "epoch": 0.9304620613031828,
-      "grad_norm": 0.436081200838089,
-      "learning_rate": 7.988759056558945e-06,
-      "loss": 0.3584,
-      "step": 14237
-    },
-    {
-      "epoch": 0.930527416508725,
-      "grad_norm": 0.4634794592857361,
-      "learning_rate": 7.988479110201266e-06,
-      "loss": 0.3619,
-      "step": 14238
-    },
-    {
-      "epoch": 0.930592771714267,
-      "grad_norm": 0.47877487540245056,
-      "learning_rate": 7.988199149267737e-06,
-      "loss": 0.4499,
-      "step": 14239
-    },
-    {
-      "epoch": 0.9306581269198092,
-      "grad_norm": 0.4497900605201721,
-      "learning_rate": 7.987919173759728e-06,
-      "loss": 0.3661,
-      "step": 14240
-    },
-    {
-      "epoch": 0.9307234821253513,
-      "grad_norm": 0.44208231568336487,
-      "learning_rate": 7.987639183678601e-06,
-      "loss": 0.3839,
-      "step": 14241
-    },
-    {
-      "epoch": 0.9307888373308935,
-      "grad_norm": 0.472713440656662,
-      "learning_rate": 7.987359179025725e-06,
-      "loss": 0.412,
-      "step": 14242
-    },
-    {
-      "epoch": 0.9308541925364355,
-      "grad_norm": 0.4360504448413849,
-      "learning_rate": 7.98707915980246e-06,
-      "loss": 0.372,
-      "step": 14243
-    },
-    {
-      "epoch": 0.9309195477419776,
-      "grad_norm": 0.46010735630989075,
-      "learning_rate": 7.986799126010176e-06,
-      "loss": 0.3819,
-      "step": 14244
-    },
-    {
-      "epoch": 0.9309849029475198,
-      "grad_norm": 0.41593727469444275,
-      "learning_rate": 7.986519077650239e-06,
-      "loss": 0.3237,
-      "step": 14245
-    },
-    {
-      "epoch": 0.9310502581530619,
-      "grad_norm": 0.45757660269737244,
-      "learning_rate": 7.986239014724013e-06,
-      "loss": 0.3619,
-      "step": 14246
-    },
-    {
-      "epoch": 0.931115613358604,
-      "grad_norm": 0.4612928330898285,
-      "learning_rate": 7.985958937232865e-06,
-      "loss": 0.3523,
-      "step": 14247
-    },
-    {
-      "epoch": 0.9311809685641461,
-      "grad_norm": 0.4958532452583313,
-      "learning_rate": 7.985678845178162e-06,
-      "loss": 0.4336,
-      "step": 14248
-    },
-    {
-      "epoch": 0.9312463237696883,
-      "grad_norm": 0.43681925535202026,
-      "learning_rate": 7.985398738561267e-06,
-      "loss": 0.3707,
-      "step": 14249
-    },
-    {
-      "epoch": 0.9313116789752304,
-      "grad_norm": 0.44724389910697937,
-      "learning_rate": 7.98511861738355e-06,
-      "loss": 0.3772,
-      "step": 14250
-    },
-    {
-      "epoch": 0.9313770341807724,
-      "grad_norm": 0.4231940805912018,
-      "learning_rate": 7.984838481646374e-06,
-      "loss": 0.3316,
-      "step": 14251
-    },
-    {
-      "epoch": 0.9314423893863146,
-      "grad_norm": 0.4736866056919098,
-      "learning_rate": 7.984558331351106e-06,
-      "loss": 0.3722,
-      "step": 14252
-    },
-    {
-      "epoch": 0.9315077445918567,
-      "grad_norm": 0.41867658495903015,
-      "learning_rate": 7.984278166499116e-06,
-      "loss": 0.3261,
-      "step": 14253
-    },
-    {
-      "epoch": 0.9315730997973989,
-      "grad_norm": 0.4349023401737213,
-      "learning_rate": 7.983997987091765e-06,
-      "loss": 0.3559,
-      "step": 14254
-    },
-    {
-      "epoch": 0.931638455002941,
-      "grad_norm": 0.46389687061309814,
-      "learning_rate": 7.983717793130423e-06,
-      "loss": 0.3973,
-      "step": 14255
-    },
-    {
-      "epoch": 0.9317038102084831,
-      "grad_norm": 0.4258340299129486,
-      "learning_rate": 7.983437584616455e-06,
-      "loss": 0.3873,
-      "step": 14256
-    },
-    {
-      "epoch": 0.9317691654140252,
-      "grad_norm": 0.4512598216533661,
-      "learning_rate": 7.983157361551229e-06,
-      "loss": 0.3656,
-      "step": 14257
-    },
-    {
-      "epoch": 0.9318345206195674,
-      "grad_norm": 0.5446600317955017,
-      "learning_rate": 7.98287712393611e-06,
-      "loss": 0.4037,
-      "step": 14258
-    },
-    {
-      "epoch": 0.9318998758251095,
-      "grad_norm": 0.437092125415802,
-      "learning_rate": 7.982596871772468e-06,
-      "loss": 0.3498,
-      "step": 14259
-    },
-    {
-      "epoch": 0.9319652310306515,
-      "grad_norm": 1.145544409751892,
-      "learning_rate": 7.982316605061665e-06,
-      "loss": 0.3887,
-      "step": 14260
-    },
-    {
-      "epoch": 0.9320305862361937,
-      "grad_norm": 0.42923155426979065,
-      "learning_rate": 7.982036323805074e-06,
-      "loss": 0.3733,
-      "step": 14261
-    },
-    {
-      "epoch": 0.9320959414417358,
-      "grad_norm": 0.43429166078567505,
-      "learning_rate": 7.981756028004054e-06,
-      "loss": 0.3385,
-      "step": 14262
-    },
-    {
-      "epoch": 0.932161296647278,
-      "grad_norm": 0.47522273659706116,
-      "learning_rate": 7.98147571765998e-06,
-      "loss": 0.3935,
-      "step": 14263
-    },
-    {
-      "epoch": 0.93222665185282,
-      "grad_norm": 0.4046888053417206,
-      "learning_rate": 7.981195392774218e-06,
-      "loss": 0.3657,
-      "step": 14264
-    },
-    {
-      "epoch": 0.9322920070583622,
-      "grad_norm": 0.43544861674308777,
-      "learning_rate": 7.98091505334813e-06,
-      "loss": 0.3991,
-      "step": 14265
-    },
-    {
-      "epoch": 0.9323573622639043,
-      "grad_norm": 0.4282855987548828,
-      "learning_rate": 7.98063469938309e-06,
-      "loss": 0.3398,
-      "step": 14266
-    },
-    {
-      "epoch": 0.9324227174694465,
-      "grad_norm": 0.4156448245048523,
-      "learning_rate": 7.980354330880458e-06,
-      "loss": 0.3609,
-      "step": 14267
-    },
-    {
-      "epoch": 0.9324880726749886,
-      "grad_norm": 0.48186618089675903,
-      "learning_rate": 7.980073947841607e-06,
-      "loss": 0.4331,
-      "step": 14268
-    },
-    {
-      "epoch": 0.9325534278805306,
-      "grad_norm": 0.45654037594795227,
-      "learning_rate": 7.979793550267904e-06,
-      "loss": 0.406,
-      "step": 14269
-    },
-    {
-      "epoch": 0.9326187830860728,
-      "grad_norm": 0.4358753561973572,
-      "learning_rate": 7.979513138160716e-06,
-      "loss": 0.3524,
-      "step": 14270
-    },
-    {
-      "epoch": 0.9326841382916149,
-      "grad_norm": 0.4549356698989868,
-      "learning_rate": 7.979232711521407e-06,
-      "loss": 0.3817,
-      "step": 14271
-    },
-    {
-      "epoch": 0.9327494934971571,
-      "grad_norm": 0.41262155771255493,
-      "learning_rate": 7.97895227035135e-06,
-      "loss": 0.3298,
-      "step": 14272
-    },
-    {
-      "epoch": 0.9328148487026992,
-      "grad_norm": 0.40901660919189453,
-      "learning_rate": 7.97867181465191e-06,
-      "loss": 0.3205,
-      "step": 14273
-    },
-    {
-      "epoch": 0.9328802039082413,
-      "grad_norm": 0.4325229525566101,
-      "learning_rate": 7.978391344424457e-06,
-      "loss": 0.3487,
-      "step": 14274
-    },
-    {
-      "epoch": 0.9329455591137834,
-      "grad_norm": 0.42647942900657654,
-      "learning_rate": 7.978110859670358e-06,
-      "loss": 0.3353,
-      "step": 14275
-    },
-    {
-      "epoch": 0.9330109143193255,
-      "grad_norm": 0.4505908489227295,
-      "learning_rate": 7.97783036039098e-06,
-      "loss": 0.4073,
-      "step": 14276
-    },
-    {
-      "epoch": 0.9330762695248677,
-      "grad_norm": 0.4391935467720032,
-      "learning_rate": 7.977549846587691e-06,
-      "loss": 0.3993,
-      "step": 14277
-    },
-    {
-      "epoch": 0.9331416247304097,
-      "grad_norm": 0.45251569151878357,
-      "learning_rate": 7.97726931826186e-06,
-      "loss": 0.3808,
-      "step": 14278
-    },
-    {
-      "epoch": 0.9332069799359519,
-      "grad_norm": 0.4612550735473633,
-      "learning_rate": 7.976988775414855e-06,
-      "loss": 0.383,
-      "step": 14279
-    },
-    {
-      "epoch": 0.933272335141494,
-      "grad_norm": 0.40265682339668274,
-      "learning_rate": 7.976708218048044e-06,
-      "loss": 0.3279,
-      "step": 14280
-    },
-    {
-      "epoch": 0.9333376903470362,
-      "grad_norm": 0.46680349111557007,
-      "learning_rate": 7.976427646162796e-06,
-      "loss": 0.4127,
-      "step": 14281
-    },
-    {
-      "epoch": 0.9334030455525782,
-      "grad_norm": 0.4239867031574249,
-      "learning_rate": 7.97614705976048e-06,
-      "loss": 0.341,
-      "step": 14282
-    },
-    {
-      "epoch": 0.9334684007581204,
-      "grad_norm": 0.48194319009780884,
-      "learning_rate": 7.975866458842463e-06,
-      "loss": 0.4154,
-      "step": 14283
-    },
-    {
-      "epoch": 0.9335337559636625,
-      "grad_norm": 0.5396852493286133,
-      "learning_rate": 7.975585843410115e-06,
-      "loss": 0.3535,
-      "step": 14284
-    },
-    {
-      "epoch": 0.9335991111692046,
-      "grad_norm": 0.45082712173461914,
-      "learning_rate": 7.975305213464805e-06,
-      "loss": 0.3234,
-      "step": 14285
-    },
-    {
-      "epoch": 0.9336644663747468,
-      "grad_norm": 0.4681273102760315,
-      "learning_rate": 7.975024569007899e-06,
-      "loss": 0.4297,
-      "step": 14286
-    },
-    {
-      "epoch": 0.9337298215802888,
-      "grad_norm": 0.4718344211578369,
-      "learning_rate": 7.974743910040768e-06,
-      "loss": 0.4079,
-      "step": 14287
-    },
-    {
-      "epoch": 0.933795176785831,
-      "grad_norm": 0.4437362849712372,
-      "learning_rate": 7.97446323656478e-06,
-      "loss": 0.3928,
-      "step": 14288
-    },
-    {
-      "epoch": 0.9338605319913731,
-      "grad_norm": 0.44397616386413574,
-      "learning_rate": 7.974182548581305e-06,
-      "loss": 0.3746,
-      "step": 14289
-    },
-    {
-      "epoch": 0.9339258871969153,
-      "grad_norm": 0.44536322355270386,
-      "learning_rate": 7.973901846091712e-06,
-      "loss": 0.3745,
-      "step": 14290
-    },
-    {
-      "epoch": 0.9339912424024573,
-      "grad_norm": 0.47594279050827026,
-      "learning_rate": 7.973621129097369e-06,
-      "loss": 0.3749,
-      "step": 14291
-    },
-    {
-      "epoch": 0.9340565976079995,
-      "grad_norm": 0.464814692735672,
-      "learning_rate": 7.973340397599644e-06,
-      "loss": 0.379,
-      "step": 14292
-    },
-    {
-      "epoch": 0.9341219528135416,
-      "grad_norm": 0.4069635570049286,
-      "learning_rate": 7.973059651599908e-06,
-      "loss": 0.2935,
-      "step": 14293
-    },
-    {
-      "epoch": 0.9341873080190837,
-      "grad_norm": 0.45391401648521423,
-      "learning_rate": 7.972778891099532e-06,
-      "loss": 0.3521,
-      "step": 14294
-    },
-    {
-      "epoch": 0.9342526632246259,
-      "grad_norm": 0.48291680216789246,
-      "learning_rate": 7.972498116099882e-06,
-      "loss": 0.4188,
-      "step": 14295
-    },
-    {
-      "epoch": 0.9343180184301679,
-      "grad_norm": 0.4733894169330597,
-      "learning_rate": 7.972217326602331e-06,
-      "loss": 0.4372,
-      "step": 14296
-    },
-    {
-      "epoch": 0.9343833736357101,
-      "grad_norm": 0.4554188549518585,
-      "learning_rate": 7.971936522608245e-06,
-      "loss": 0.3554,
-      "step": 14297
-    },
-    {
-      "epoch": 0.9344487288412522,
-      "grad_norm": 0.48150956630706787,
-      "learning_rate": 7.971655704118994e-06,
-      "loss": 0.4032,
-      "step": 14298
-    },
-    {
-      "epoch": 0.9345140840467944,
-      "grad_norm": 0.42098093032836914,
-      "learning_rate": 7.971374871135951e-06,
-      "loss": 0.3273,
-      "step": 14299
-    },
-    {
-      "epoch": 0.9345794392523364,
-      "grad_norm": 0.44666042923927307,
-      "learning_rate": 7.97109402366048e-06,
-      "loss": 0.3735,
-      "step": 14300
-    },
-    {
-      "epoch": 0.9346447944578786,
-      "grad_norm": 0.42981797456741333,
-      "learning_rate": 7.970813161693957e-06,
-      "loss": 0.3409,
-      "step": 14301
-    },
-    {
-      "epoch": 0.9347101496634207,
-      "grad_norm": 0.4457097053527832,
-      "learning_rate": 7.970532285237749e-06,
-      "loss": 0.3904,
-      "step": 14302
-    },
-    {
-      "epoch": 0.9347755048689628,
-      "grad_norm": 0.4529748558998108,
-      "learning_rate": 7.970251394293225e-06,
-      "loss": 0.3798,
-      "step": 14303
-    },
-    {
-      "epoch": 0.934840860074505,
-      "grad_norm": 0.4386380910873413,
-      "learning_rate": 7.969970488861757e-06,
-      "loss": 0.352,
-      "step": 14304
-    },
-    {
-      "epoch": 0.934906215280047,
-      "grad_norm": 0.4380452334880829,
-      "learning_rate": 7.969689568944711e-06,
-      "loss": 0.3498,
-      "step": 14305
-    },
-    {
-      "epoch": 0.9349715704855892,
-      "grad_norm": 0.45035868883132935,
-      "learning_rate": 7.969408634543462e-06,
-      "loss": 0.389,
-      "step": 14306
-    },
-    {
-      "epoch": 0.9350369256911313,
-      "grad_norm": 0.45483487844467163,
-      "learning_rate": 7.96912768565938e-06,
-      "loss": 0.3827,
-      "step": 14307
-    },
-    {
-      "epoch": 0.9351022808966735,
-      "grad_norm": 0.41082796454429626,
-      "learning_rate": 7.96884672229383e-06,
-      "loss": 0.3463,
-      "step": 14308
-    },
-    {
-      "epoch": 0.9351676361022155,
-      "grad_norm": 0.59320068359375,
-      "learning_rate": 7.968565744448188e-06,
-      "loss": 0.3686,
-      "step": 14309
-    },
-    {
-      "epoch": 0.9352329913077576,
-      "grad_norm": 0.4542233347892761,
-      "learning_rate": 7.968284752123825e-06,
-      "loss": 0.4166,
-      "step": 14310
-    },
-    {
-      "epoch": 0.9352983465132998,
-      "grad_norm": 0.4471418261528015,
-      "learning_rate": 7.968003745322106e-06,
-      "loss": 0.3917,
-      "step": 14311
-    },
-    {
-      "epoch": 0.9353637017188419,
-      "grad_norm": 0.47591525316238403,
-      "learning_rate": 7.967722724044404e-06,
-      "loss": 0.4491,
-      "step": 14312
-    },
-    {
-      "epoch": 0.935429056924384,
-      "grad_norm": 0.4212082624435425,
-      "learning_rate": 7.967441688292093e-06,
-      "loss": 0.3488,
-      "step": 14313
-    },
-    {
-      "epoch": 0.9354944121299261,
-      "grad_norm": 0.4605009853839874,
-      "learning_rate": 7.967160638066537e-06,
-      "loss": 0.4503,
-      "step": 14314
-    },
-    {
-      "epoch": 0.9355597673354683,
-      "grad_norm": 0.4623265564441681,
-      "learning_rate": 7.966879573369115e-06,
-      "loss": 0.4164,
-      "step": 14315
-    },
-    {
-      "epoch": 0.9356251225410104,
-      "grad_norm": 0.42452770471572876,
-      "learning_rate": 7.966598494201192e-06,
-      "loss": 0.3894,
-      "step": 14316
-    },
-    {
-      "epoch": 0.9356904777465526,
-      "grad_norm": 0.42500030994415283,
-      "learning_rate": 7.966317400564139e-06,
-      "loss": 0.362,
-      "step": 14317
-    },
-    {
-      "epoch": 0.9357558329520946,
-      "grad_norm": 0.4604122042655945,
-      "learning_rate": 7.96603629245933e-06,
-      "loss": 0.4283,
-      "step": 14318
-    },
-    {
-      "epoch": 0.9358211881576367,
-      "grad_norm": 0.4024207890033722,
-      "learning_rate": 7.965755169888132e-06,
-      "loss": 0.3342,
-      "step": 14319
-    },
-    {
-      "epoch": 0.9358865433631789,
-      "grad_norm": 0.4626029133796692,
-      "learning_rate": 7.965474032851921e-06,
-      "loss": 0.3919,
-      "step": 14320
-    },
-    {
-      "epoch": 0.935951898568721,
-      "grad_norm": 0.46209120750427246,
-      "learning_rate": 7.965192881352067e-06,
-      "loss": 0.3819,
-      "step": 14321
-    },
-    {
-      "epoch": 0.9360172537742631,
-      "grad_norm": 0.4297623038291931,
-      "learning_rate": 7.964911715389937e-06,
-      "loss": 0.3824,
-      "step": 14322
-    },
-    {
-      "epoch": 0.9360826089798052,
-      "grad_norm": 0.4633261561393738,
-      "learning_rate": 7.964630534966906e-06,
-      "loss": 0.4345,
-      "step": 14323
-    },
-    {
-      "epoch": 0.9361479641853474,
-      "grad_norm": 0.444621205329895,
-      "learning_rate": 7.964349340084345e-06,
-      "loss": 0.3714,
-      "step": 14324
-    },
-    {
-      "epoch": 0.9362133193908895,
-      "grad_norm": 0.43743178248405457,
-      "learning_rate": 7.964068130743626e-06,
-      "loss": 0.3333,
-      "step": 14325
-    },
-    {
-      "epoch": 0.9362786745964317,
-      "grad_norm": 0.45553866028785706,
-      "learning_rate": 7.96378690694612e-06,
-      "loss": 0.3631,
-      "step": 14326
-    },
-    {
-      "epoch": 0.9363440298019737,
-      "grad_norm": 0.40102672576904297,
-      "learning_rate": 7.963505668693197e-06,
-      "loss": 0.3029,
-      "step": 14327
-    },
-    {
-      "epoch": 0.9364093850075158,
-      "grad_norm": 0.42924097180366516,
-      "learning_rate": 7.963224415986233e-06,
-      "loss": 0.3497,
-      "step": 14328
-    },
-    {
-      "epoch": 0.936474740213058,
-      "grad_norm": 0.46029043197631836,
-      "learning_rate": 7.962943148826595e-06,
-      "loss": 0.3981,
-      "step": 14329
-    },
-    {
-      "epoch": 0.9365400954186001,
-      "grad_norm": 0.41947486996650696,
-      "learning_rate": 7.962661867215658e-06,
-      "loss": 0.3515,
-      "step": 14330
-    },
-    {
-      "epoch": 0.9366054506241422,
-      "grad_norm": 0.44270771741867065,
-      "learning_rate": 7.962380571154791e-06,
-      "loss": 0.3921,
-      "step": 14331
-    },
-    {
-      "epoch": 0.9366708058296843,
-      "grad_norm": 0.43892335891723633,
-      "learning_rate": 7.962099260645369e-06,
-      "loss": 0.3395,
-      "step": 14332
-    },
-    {
-      "epoch": 0.9367361610352265,
-      "grad_norm": 0.4416338801383972,
-      "learning_rate": 7.96181793568876e-06,
-      "loss": 0.3705,
-      "step": 14333
-    },
-    {
-      "epoch": 0.9368015162407686,
-      "grad_norm": 0.43199318647384644,
-      "learning_rate": 7.961536596286343e-06,
-      "loss": 0.3922,
-      "step": 14334
-    },
-    {
-      "epoch": 0.9368668714463106,
-      "grad_norm": 0.44440150260925293,
-      "learning_rate": 7.961255242439483e-06,
-      "loss": 0.3825,
-      "step": 14335
-    },
-    {
-      "epoch": 0.9369322266518528,
-      "grad_norm": 0.4097616374492645,
-      "learning_rate": 7.960973874149555e-06,
-      "loss": 0.3482,
-      "step": 14336
-    },
-    {
-      "epoch": 0.9369975818573949,
-      "grad_norm": 0.5028608441352844,
-      "learning_rate": 7.960692491417934e-06,
-      "loss": 0.4794,
-      "step": 14337
-    },
-    {
-      "epoch": 0.9370629370629371,
-      "grad_norm": 0.4232047200202942,
-      "learning_rate": 7.96041109424599e-06,
-      "loss": 0.3595,
-      "step": 14338
-    },
-    {
-      "epoch": 0.9371282922684792,
-      "grad_norm": 0.4544631540775299,
-      "learning_rate": 7.960129682635095e-06,
-      "loss": 0.4215,
-      "step": 14339
-    },
-    {
-      "epoch": 0.9371936474740213,
-      "grad_norm": 0.4575744569301605,
-      "learning_rate": 7.959848256586621e-06,
-      "loss": 0.4379,
-      "step": 14340
-    },
-    {
-      "epoch": 0.9372590026795634,
-      "grad_norm": 0.43337470293045044,
-      "learning_rate": 7.959566816101943e-06,
-      "loss": 0.3786,
-      "step": 14341
-    },
-    {
-      "epoch": 0.9373243578851056,
-      "grad_norm": 0.4470236897468567,
-      "learning_rate": 7.959285361182432e-06,
-      "loss": 0.4205,
-      "step": 14342
-    },
-    {
-      "epoch": 0.9373897130906477,
-      "grad_norm": 0.4310753047466278,
-      "learning_rate": 7.95900389182946e-06,
-      "loss": 0.3795,
-      "step": 14343
-    },
-    {
-      "epoch": 0.9374550682961897,
-      "grad_norm": 0.4945099651813507,
-      "learning_rate": 7.958722408044401e-06,
-      "loss": 0.4124,
-      "step": 14344
-    },
-    {
-      "epoch": 0.9375204235017319,
-      "grad_norm": 0.4588758647441864,
-      "learning_rate": 7.958440909828629e-06,
-      "loss": 0.4088,
-      "step": 14345
-    },
-    {
-      "epoch": 0.937585778707274,
-      "grad_norm": 0.440149188041687,
-      "learning_rate": 7.958159397183516e-06,
-      "loss": 0.3932,
-      "step": 14346
-    },
-    {
-      "epoch": 0.9376511339128162,
-      "grad_norm": 0.4193704426288605,
-      "learning_rate": 7.957877870110434e-06,
-      "loss": 0.3608,
-      "step": 14347
-    },
-    {
-      "epoch": 0.9377164891183583,
-      "grad_norm": 0.4371584951877594,
-      "learning_rate": 7.957596328610757e-06,
-      "loss": 0.377,
-      "step": 14348
-    },
-    {
-      "epoch": 0.9377818443239004,
-      "grad_norm": 0.44903409481048584,
-      "learning_rate": 7.957314772685858e-06,
-      "loss": 0.3684,
-      "step": 14349
-    },
-    {
-      "epoch": 0.9378471995294425,
-      "grad_norm": 0.42451244592666626,
-      "learning_rate": 7.95703320233711e-06,
-      "loss": 0.3402,
-      "step": 14350
-    },
-    {
-      "epoch": 0.9379125547349847,
-      "grad_norm": 0.444200336933136,
-      "learning_rate": 7.956751617565888e-06,
-      "loss": 0.3602,
-      "step": 14351
-    },
-    {
-      "epoch": 0.9379779099405268,
-      "grad_norm": 0.4307136833667755,
-      "learning_rate": 7.956470018373563e-06,
-      "loss": 0.3421,
-      "step": 14352
-    },
-    {
-      "epoch": 0.9380432651460688,
-      "grad_norm": 0.4414706826210022,
-      "learning_rate": 7.95618840476151e-06,
-      "loss": 0.344,
-      "step": 14353
-    },
-    {
-      "epoch": 0.938108620351611,
-      "grad_norm": 0.4512485861778259,
-      "learning_rate": 7.9559067767311e-06,
-      "loss": 0.3523,
-      "step": 14354
-    },
-    {
-      "epoch": 0.9381739755571531,
-      "grad_norm": 0.40102237462997437,
-      "learning_rate": 7.955625134283712e-06,
-      "loss": 0.3295,
-      "step": 14355
-    },
-    {
-      "epoch": 0.9382393307626953,
-      "grad_norm": 0.4707501530647278,
-      "learning_rate": 7.955343477420714e-06,
-      "loss": 0.3757,
-      "step": 14356
-    },
-    {
-      "epoch": 0.9383046859682374,
-      "grad_norm": 0.4292069971561432,
-      "learning_rate": 7.955061806143483e-06,
-      "loss": 0.3438,
-      "step": 14357
-    },
-    {
-      "epoch": 0.9383700411737795,
-      "grad_norm": 0.4546838402748108,
-      "learning_rate": 7.954780120453392e-06,
-      "loss": 0.3824,
-      "step": 14358
-    },
-    {
-      "epoch": 0.9384353963793216,
-      "grad_norm": 0.453620970249176,
-      "learning_rate": 7.954498420351814e-06,
-      "loss": 0.3681,
-      "step": 14359
-    },
-    {
-      "epoch": 0.9385007515848637,
-      "grad_norm": 0.44989854097366333,
-      "learning_rate": 7.954216705840124e-06,
-      "loss": 0.3727,
-      "step": 14360
-    },
-    {
-      "epoch": 0.9385661067904059,
-      "grad_norm": 0.4276033341884613,
-      "learning_rate": 7.953934976919695e-06,
-      "loss": 0.3781,
-      "step": 14361
-    },
-    {
-      "epoch": 0.9386314619959479,
-      "grad_norm": 0.4170686602592468,
-      "learning_rate": 7.953653233591903e-06,
-      "loss": 0.3124,
-      "step": 14362
-    },
-    {
-      "epoch": 0.9386968172014901,
-      "grad_norm": 0.45001140236854553,
-      "learning_rate": 7.953371475858119e-06,
-      "loss": 0.4009,
-      "step": 14363
-    },
-    {
-      "epoch": 0.9387621724070322,
-      "grad_norm": 0.44703391194343567,
-      "learning_rate": 7.953089703719721e-06,
-      "loss": 0.3673,
-      "step": 14364
-    },
-    {
-      "epoch": 0.9388275276125744,
-      "grad_norm": 0.4642198085784912,
-      "learning_rate": 7.95280791717808e-06,
-      "loss": 0.379,
-      "step": 14365
-    },
-    {
-      "epoch": 0.9388928828181164,
-      "grad_norm": 0.4425627291202545,
-      "learning_rate": 7.952526116234574e-06,
-      "loss": 0.3786,
-      "step": 14366
-    },
-    {
-      "epoch": 0.9389582380236586,
-      "grad_norm": 0.46043291687965393,
-      "learning_rate": 7.952244300890574e-06,
-      "loss": 0.3847,
-      "step": 14367
-    },
-    {
-      "epoch": 0.9390235932292007,
-      "grad_norm": 0.43328171968460083,
-      "learning_rate": 7.951962471147456e-06,
-      "loss": 0.3951,
-      "step": 14368
-    },
-    {
-      "epoch": 0.9390889484347428,
-      "grad_norm": 0.4538939893245697,
-      "learning_rate": 7.951680627006592e-06,
-      "loss": 0.3976,
-      "step": 14369
-    },
-    {
-      "epoch": 0.939154303640285,
-      "grad_norm": 0.46278855204582214,
-      "learning_rate": 7.951398768469363e-06,
-      "loss": 0.4246,
-      "step": 14370
-    },
-    {
-      "epoch": 0.939219658845827,
-      "grad_norm": 0.46772995591163635,
-      "learning_rate": 7.951116895537137e-06,
-      "loss": 0.416,
-      "step": 14371
-    },
-    {
-      "epoch": 0.9392850140513692,
-      "grad_norm": 0.4080042839050293,
-      "learning_rate": 7.950835008211292e-06,
-      "loss": 0.351,
-      "step": 14372
-    },
-    {
-      "epoch": 0.9393503692569113,
-      "grad_norm": 0.4318680763244629,
-      "learning_rate": 7.950553106493203e-06,
-      "loss": 0.3589,
-      "step": 14373
-    },
-    {
-      "epoch": 0.9394157244624535,
-      "grad_norm": 0.42919856309890747,
-      "learning_rate": 7.950271190384244e-06,
-      "loss": 0.3761,
-      "step": 14374
-    },
-    {
-      "epoch": 0.9394810796679955,
-      "grad_norm": 0.45179271697998047,
-      "learning_rate": 7.94998925988579e-06,
-      "loss": 0.3819,
-      "step": 14375
-    },
-    {
-      "epoch": 0.9395464348735377,
-      "grad_norm": 0.4313129186630249,
-      "learning_rate": 7.949707314999217e-06,
-      "loss": 0.3557,
-      "step": 14376
-    },
-    {
-      "epoch": 0.9396117900790798,
-      "grad_norm": 0.4538089334964752,
-      "learning_rate": 7.9494253557259e-06,
-      "loss": 0.3771,
-      "step": 14377
-    },
-    {
-      "epoch": 0.9396771452846219,
-      "grad_norm": 0.4204086363315582,
-      "learning_rate": 7.949143382067214e-06,
-      "loss": 0.3571,
-      "step": 14378
-    },
-    {
-      "epoch": 0.9397425004901641,
-      "grad_norm": 0.4639827609062195,
-      "learning_rate": 7.948861394024534e-06,
-      "loss": 0.395,
-      "step": 14379
-    },
-    {
-      "epoch": 0.9398078556957061,
-      "grad_norm": 0.4357139766216278,
-      "learning_rate": 7.948579391599233e-06,
-      "loss": 0.3651,
-      "step": 14380
-    },
-    {
-      "epoch": 0.9398732109012483,
-      "grad_norm": 0.46814483404159546,
-      "learning_rate": 7.948297374792692e-06,
-      "loss": 0.4106,
-      "step": 14381
-    },
-    {
-      "epoch": 0.9399385661067904,
-      "grad_norm": 0.4309825897216797,
-      "learning_rate": 7.948015343606281e-06,
-      "loss": 0.3979,
-      "step": 14382
-    },
-    {
-      "epoch": 0.9400039213123326,
-      "grad_norm": 0.49262019991874695,
-      "learning_rate": 7.947733298041379e-06,
-      "loss": 0.4262,
-      "step": 14383
-    },
-    {
-      "epoch": 0.9400692765178746,
-      "grad_norm": 0.45704779028892517,
-      "learning_rate": 7.94745123809936e-06,
-      "loss": 0.3921,
-      "step": 14384
-    },
-    {
-      "epoch": 0.9401346317234168,
-      "grad_norm": 0.445494145154953,
-      "learning_rate": 7.9471691637816e-06,
-      "loss": 0.4015,
-      "step": 14385
-    },
-    {
-      "epoch": 0.9401999869289589,
-      "grad_norm": 0.4356983006000519,
-      "learning_rate": 7.946887075089477e-06,
-      "loss": 0.3884,
-      "step": 14386
-    },
-    {
-      "epoch": 0.940265342134501,
-      "grad_norm": 0.4237995743751526,
-      "learning_rate": 7.946604972024363e-06,
-      "loss": 0.3447,
-      "step": 14387
-    },
-    {
-      "epoch": 0.9403306973400432,
-      "grad_norm": 0.43086937069892883,
-      "learning_rate": 7.946322854587636e-06,
-      "loss": 0.3722,
-      "step": 14388
-    },
-    {
-      "epoch": 0.9403960525455852,
-      "grad_norm": 0.4470529556274414,
-      "learning_rate": 7.946040722780672e-06,
-      "loss": 0.3997,
-      "step": 14389
-    },
-    {
-      "epoch": 0.9404614077511274,
-      "grad_norm": 0.4738561511039734,
-      "learning_rate": 7.945758576604846e-06,
-      "loss": 0.4423,
-      "step": 14390
-    },
-    {
-      "epoch": 0.9405267629566695,
-      "grad_norm": 0.45183515548706055,
-      "learning_rate": 7.945476416061535e-06,
-      "loss": 0.3991,
-      "step": 14391
-    },
-    {
-      "epoch": 0.9405921181622117,
-      "grad_norm": 0.4276072382926941,
-      "learning_rate": 7.945194241152115e-06,
-      "loss": 0.358,
-      "step": 14392
-    },
-    {
-      "epoch": 0.9406574733677537,
-      "grad_norm": 0.45037370920181274,
-      "learning_rate": 7.944912051877963e-06,
-      "loss": 0.3945,
-      "step": 14393
-    },
-    {
-      "epoch": 0.9407228285732958,
-      "grad_norm": 0.4249766767024994,
-      "learning_rate": 7.944629848240455e-06,
-      "loss": 0.393,
-      "step": 14394
-    },
-    {
-      "epoch": 0.940788183778838,
-      "grad_norm": 0.45881056785583496,
-      "learning_rate": 7.944347630240968e-06,
-      "loss": 0.3784,
-      "step": 14395
-    },
-    {
-      "epoch": 0.9408535389843801,
-      "grad_norm": 0.4441053867340088,
-      "learning_rate": 7.944065397880876e-06,
-      "loss": 0.3548,
-      "step": 14396
-    },
-    {
-      "epoch": 0.9409188941899223,
-      "grad_norm": 0.47556474804878235,
-      "learning_rate": 7.943783151161558e-06,
-      "loss": 0.4012,
-      "step": 14397
-    },
-    {
-      "epoch": 0.9409842493954643,
-      "grad_norm": 0.43414306640625,
-      "learning_rate": 7.943500890084387e-06,
-      "loss": 0.3776,
-      "step": 14398
-    },
-    {
-      "epoch": 0.9410496046010065,
-      "grad_norm": 0.4786895215511322,
-      "learning_rate": 7.943218614650745e-06,
-      "loss": 0.4717,
-      "step": 14399
-    },
-    {
-      "epoch": 0.9411149598065486,
-      "grad_norm": 0.4821021556854248,
-      "learning_rate": 7.942936324862007e-06,
-      "loss": 0.3752,
-      "step": 14400
-    },
-    {
-      "epoch": 0.9411803150120908,
-      "grad_norm": 0.4355505704879761,
-      "learning_rate": 7.942654020719548e-06,
-      "loss": 0.3903,
-      "step": 14401
-    },
-    {
-      "epoch": 0.9412456702176328,
-      "grad_norm": 0.4327200651168823,
-      "learning_rate": 7.942371702224745e-06,
-      "loss": 0.3628,
-      "step": 14402
-    },
-    {
-      "epoch": 0.9413110254231749,
-      "grad_norm": 0.4677167534828186,
-      "learning_rate": 7.942089369378977e-06,
-      "loss": 0.4297,
-      "step": 14403
-    },
-    {
-      "epoch": 0.9413763806287171,
-      "grad_norm": 0.42914190888404846,
-      "learning_rate": 7.941807022183618e-06,
-      "loss": 0.3379,
-      "step": 14404
-    },
-    {
-      "epoch": 0.9414417358342592,
-      "grad_norm": 0.3983922600746155,
-      "learning_rate": 7.94152466064005e-06,
-      "loss": 0.3071,
-      "step": 14405
-    },
-    {
-      "epoch": 0.9415070910398013,
-      "grad_norm": 0.4407617747783661,
-      "learning_rate": 7.941242284749647e-06,
-      "loss": 0.3445,
-      "step": 14406
-    },
-    {
-      "epoch": 0.9415724462453434,
-      "grad_norm": 0.42539024353027344,
-      "learning_rate": 7.940959894513784e-06,
-      "loss": 0.3476,
-      "step": 14407
-    },
-    {
-      "epoch": 0.9416378014508856,
-      "grad_norm": 0.41399627923965454,
-      "learning_rate": 7.940677489933842e-06,
-      "loss": 0.3264,
-      "step": 14408
-    },
-    {
-      "epoch": 0.9417031566564277,
-      "grad_norm": 0.4142743647098541,
-      "learning_rate": 7.940395071011196e-06,
-      "loss": 0.3233,
-      "step": 14409
-    },
-    {
-      "epoch": 0.9417685118619699,
-      "grad_norm": 0.4264567792415619,
-      "learning_rate": 7.940112637747227e-06,
-      "loss": 0.3709,
-      "step": 14410
-    },
-    {
-      "epoch": 0.9418338670675119,
-      "grad_norm": 0.43007078766822815,
-      "learning_rate": 7.939830190143308e-06,
-      "loss": 0.3273,
-      "step": 14411
-    },
-    {
-      "epoch": 0.941899222273054,
-      "grad_norm": 0.45897284150123596,
-      "learning_rate": 7.939547728200819e-06,
-      "loss": 0.4395,
-      "step": 14412
-    },
-    {
-      "epoch": 0.9419645774785962,
-      "grad_norm": 0.4416148364543915,
-      "learning_rate": 7.939265251921137e-06,
-      "loss": 0.3751,
-      "step": 14413
-    },
-    {
-      "epoch": 0.9420299326841383,
-      "grad_norm": 0.4365776479244232,
-      "learning_rate": 7.938982761305643e-06,
-      "loss": 0.3315,
-      "step": 14414
-    },
-    {
-      "epoch": 0.9420952878896804,
-      "grad_norm": 0.44567838311195374,
-      "learning_rate": 7.93870025635571e-06,
-      "loss": 0.4074,
-      "step": 14415
-    },
-    {
-      "epoch": 0.9421606430952225,
-      "grad_norm": 0.4563732147216797,
-      "learning_rate": 7.938417737072717e-06,
-      "loss": 0.4041,
-      "step": 14416
-    },
-    {
-      "epoch": 0.9422259983007647,
-      "grad_norm": 0.4569830298423767,
-      "learning_rate": 7.938135203458043e-06,
-      "loss": 0.402,
-      "step": 14417
-    },
-    {
-      "epoch": 0.9422913535063068,
-      "grad_norm": 0.47362250089645386,
-      "learning_rate": 7.937852655513065e-06,
-      "loss": 0.4038,
-      "step": 14418
-    },
-    {
-      "epoch": 0.9423567087118488,
-      "grad_norm": 0.4402032792568207,
-      "learning_rate": 7.937570093239164e-06,
-      "loss": 0.3827,
-      "step": 14419
-    },
-    {
-      "epoch": 0.942422063917391,
-      "grad_norm": 0.46891897916793823,
-      "learning_rate": 7.937287516637714e-06,
-      "loss": 0.4175,
-      "step": 14420
-    },
-    {
-      "epoch": 0.9424874191229331,
-      "grad_norm": 0.4435456693172455,
-      "learning_rate": 7.937004925710097e-06,
-      "loss": 0.3641,
-      "step": 14421
-    },
-    {
-      "epoch": 0.9425527743284753,
-      "grad_norm": 0.40931662917137146,
-      "learning_rate": 7.936722320457689e-06,
-      "loss": 0.3061,
-      "step": 14422
-    },
-    {
-      "epoch": 0.9426181295340174,
-      "grad_norm": 0.42935508489608765,
-      "learning_rate": 7.936439700881868e-06,
-      "loss": 0.3579,
-      "step": 14423
-    },
-    {
-      "epoch": 0.9426834847395595,
-      "grad_norm": 0.41177570819854736,
-      "learning_rate": 7.936157066984014e-06,
-      "loss": 0.3376,
-      "step": 14424
-    },
-    {
-      "epoch": 0.9427488399451016,
-      "grad_norm": 0.4286854565143585,
-      "learning_rate": 7.935874418765505e-06,
-      "loss": 0.346,
-      "step": 14425
-    },
-    {
-      "epoch": 0.9428141951506438,
-      "grad_norm": 0.4349355101585388,
-      "learning_rate": 7.935591756227718e-06,
-      "loss": 0.349,
-      "step": 14426
-    },
-    {
-      "epoch": 0.9428795503561859,
-      "grad_norm": 0.46722865104675293,
-      "learning_rate": 7.935309079372035e-06,
-      "loss": 0.3911,
-      "step": 14427
-    },
-    {
-      "epoch": 0.9429449055617279,
-      "grad_norm": 0.4100147485733032,
-      "learning_rate": 7.935026388199832e-06,
-      "loss": 0.3162,
-      "step": 14428
-    },
-    {
-      "epoch": 0.9430102607672701,
-      "grad_norm": 0.46007734537124634,
-      "learning_rate": 7.934743682712487e-06,
-      "loss": 0.3979,
-      "step": 14429
-    },
-    {
-      "epoch": 0.9430756159728122,
-      "grad_norm": 0.4268944263458252,
-      "learning_rate": 7.934460962911382e-06,
-      "loss": 0.3655,
-      "step": 14430
-    },
-    {
-      "epoch": 0.9431409711783544,
-      "grad_norm": 0.4173218905925751,
-      "learning_rate": 7.934178228797894e-06,
-      "loss": 0.3553,
-      "step": 14431
-    },
-    {
-      "epoch": 0.9432063263838965,
-      "grad_norm": 0.4390046000480652,
-      "learning_rate": 7.933895480373403e-06,
-      "loss": 0.3792,
-      "step": 14432
-    },
-    {
-      "epoch": 0.9432716815894386,
-      "grad_norm": 0.43881070613861084,
-      "learning_rate": 7.933612717639285e-06,
-      "loss": 0.3819,
-      "step": 14433
-    },
-    {
-      "epoch": 0.9433370367949807,
-      "grad_norm": 0.41829654574394226,
-      "learning_rate": 7.933329940596923e-06,
-      "loss": 0.3199,
-      "step": 14434
-    },
-    {
-      "epoch": 0.9434023920005229,
-      "grad_norm": 0.43282178044319153,
-      "learning_rate": 7.933047149247695e-06,
-      "loss": 0.353,
-      "step": 14435
-    },
-    {
-      "epoch": 0.943467747206065,
-      "grad_norm": 0.45350539684295654,
-      "learning_rate": 7.93276434359298e-06,
-      "loss": 0.3797,
-      "step": 14436
-    },
-    {
-      "epoch": 0.943533102411607,
-      "grad_norm": 0.43426403403282166,
-      "learning_rate": 7.932481523634157e-06,
-      "loss": 0.3329,
-      "step": 14437
-    },
-    {
-      "epoch": 0.9435984576171492,
-      "grad_norm": 0.44173911213874817,
-      "learning_rate": 7.932198689372605e-06,
-      "loss": 0.4018,
-      "step": 14438
-    },
-    {
-      "epoch": 0.9436638128226913,
-      "grad_norm": 0.43885087966918945,
-      "learning_rate": 7.931915840809705e-06,
-      "loss": 0.3791,
-      "step": 14439
-    },
-    {
-      "epoch": 0.9437291680282335,
-      "grad_norm": 0.43043452501296997,
-      "learning_rate": 7.931632977946836e-06,
-      "loss": 0.3639,
-      "step": 14440
-    },
-    {
-      "epoch": 0.9437945232337756,
-      "grad_norm": 0.41387373208999634,
-      "learning_rate": 7.931350100785377e-06,
-      "loss": 0.3323,
-      "step": 14441
-    },
-    {
-      "epoch": 0.9438598784393177,
-      "grad_norm": 0.42461729049682617,
-      "learning_rate": 7.931067209326708e-06,
-      "loss": 0.3292,
-      "step": 14442
-    },
-    {
-      "epoch": 0.9439252336448598,
-      "grad_norm": 0.4541226923465729,
-      "learning_rate": 7.930784303572207e-06,
-      "loss": 0.3808,
-      "step": 14443
-    },
-    {
-      "epoch": 0.9439905888504019,
-      "grad_norm": 0.4451179504394531,
-      "learning_rate": 7.93050138352326e-06,
-      "loss": 0.4153,
-      "step": 14444
-    },
-    {
-      "epoch": 0.9440559440559441,
-      "grad_norm": 0.4671146273612976,
-      "learning_rate": 7.93021844918124e-06,
-      "loss": 0.3824,
-      "step": 14445
-    },
-    {
-      "epoch": 0.9441212992614861,
-      "grad_norm": 0.41073986887931824,
-      "learning_rate": 7.929935500547528e-06,
-      "loss": 0.3462,
-      "step": 14446
-    },
-    {
-      "epoch": 0.9441866544670283,
-      "grad_norm": 0.4628707468509674,
-      "learning_rate": 7.929652537623507e-06,
-      "loss": 0.4038,
-      "step": 14447
-    },
-    {
-      "epoch": 0.9442520096725704,
-      "grad_norm": 0.4600439667701721,
-      "learning_rate": 7.929369560410556e-06,
-      "loss": 0.3781,
-      "step": 14448
-    },
-    {
-      "epoch": 0.9443173648781126,
-      "grad_norm": 0.4218600392341614,
-      "learning_rate": 7.929086568910053e-06,
-      "loss": 0.3257,
-      "step": 14449
-    },
-    {
-      "epoch": 0.9443827200836546,
-      "grad_norm": 0.4293709695339203,
-      "learning_rate": 7.928803563123381e-06,
-      "loss": 0.3433,
-      "step": 14450
-    },
-    {
-      "epoch": 0.9444480752891968,
-      "grad_norm": 0.4584823548793793,
-      "learning_rate": 7.928520543051921e-06,
-      "loss": 0.3536,
-      "step": 14451
-    },
-    {
-      "epoch": 0.9445134304947389,
-      "grad_norm": 0.44159606099128723,
-      "learning_rate": 7.92823750869705e-06,
-      "loss": 0.3742,
-      "step": 14452
-    },
-    {
-      "epoch": 0.944578785700281,
-      "grad_norm": 0.4415980875492096,
-      "learning_rate": 7.927954460060152e-06,
-      "loss": 0.3648,
-      "step": 14453
-    },
-    {
-      "epoch": 0.9446441409058232,
-      "grad_norm": 0.45986682176589966,
-      "learning_rate": 7.927671397142605e-06,
-      "loss": 0.4111,
-      "step": 14454
-    },
-    {
-      "epoch": 0.9447094961113652,
-      "grad_norm": 0.435773104429245,
-      "learning_rate": 7.92738831994579e-06,
-      "loss": 0.3548,
-      "step": 14455
-    },
-    {
-      "epoch": 0.9447748513169074,
-      "grad_norm": 0.4182360768318176,
-      "learning_rate": 7.927105228471086e-06,
-      "loss": 0.2893,
-      "step": 14456
-    },
-    {
-      "epoch": 0.9448402065224495,
-      "grad_norm": 0.4321918785572052,
-      "learning_rate": 7.926822122719879e-06,
-      "loss": 0.3459,
-      "step": 14457
-    },
-    {
-      "epoch": 0.9449055617279917,
-      "grad_norm": 0.4098641276359558,
-      "learning_rate": 7.926539002693546e-06,
-      "loss": 0.3425,
-      "step": 14458
-    },
-    {
-      "epoch": 0.9449709169335337,
-      "grad_norm": 0.4105234444141388,
-      "learning_rate": 7.926255868393466e-06,
-      "loss": 0.355,
-      "step": 14459
-    },
-    {
-      "epoch": 0.9450362721390759,
-      "grad_norm": 0.4359675943851471,
-      "learning_rate": 7.925972719821025e-06,
-      "loss": 0.3748,
-      "step": 14460
-    },
-    {
-      "epoch": 0.945101627344618,
-      "grad_norm": 0.47171032428741455,
-      "learning_rate": 7.9256895569776e-06,
-      "loss": 0.4521,
-      "step": 14461
-    },
-    {
-      "epoch": 0.9451669825501601,
-      "grad_norm": 0.4673002064228058,
-      "learning_rate": 7.925406379864573e-06,
-      "loss": 0.3746,
-      "step": 14462
-    },
-    {
-      "epoch": 0.9452323377557023,
-      "grad_norm": 0.44654732942581177,
-      "learning_rate": 7.925123188483328e-06,
-      "loss": 0.3558,
-      "step": 14463
-    },
-    {
-      "epoch": 0.9452976929612443,
-      "grad_norm": 0.3931218981742859,
-      "learning_rate": 7.924839982835242e-06,
-      "loss": 0.3041,
-      "step": 14464
-    },
-    {
-      "epoch": 0.9453630481667865,
-      "grad_norm": 0.40793532133102417,
-      "learning_rate": 7.9245567629217e-06,
-      "loss": 0.3247,
-      "step": 14465
-    },
-    {
-      "epoch": 0.9454284033723286,
-      "grad_norm": 0.4634629786014557,
-      "learning_rate": 7.92427352874408e-06,
-      "loss": 0.421,
-      "step": 14466
-    },
-    {
-      "epoch": 0.9454937585778708,
-      "grad_norm": 0.45527854561805725,
-      "learning_rate": 7.923990280303763e-06,
-      "loss": 0.3612,
-      "step": 14467
-    },
-    {
-      "epoch": 0.9455591137834128,
-      "grad_norm": 0.460274875164032,
-      "learning_rate": 7.923707017602135e-06,
-      "loss": 0.4056,
-      "step": 14468
-    },
-    {
-      "epoch": 0.945624468988955,
-      "grad_norm": 0.4577227830886841,
-      "learning_rate": 7.923423740640575e-06,
-      "loss": 0.3811,
-      "step": 14469
-    },
-    {
-      "epoch": 0.9456898241944971,
-      "grad_norm": 0.44590795040130615,
-      "learning_rate": 7.923140449420464e-06,
-      "loss": 0.4015,
-      "step": 14470
-    },
-    {
-      "epoch": 0.9457551794000392,
-      "grad_norm": 0.47536012530326843,
-      "learning_rate": 7.922857143943184e-06,
-      "loss": 0.4492,
-      "step": 14471
-    },
-    {
-      "epoch": 0.9458205346055814,
-      "grad_norm": 0.4406832158565521,
-      "learning_rate": 7.922573824210118e-06,
-      "loss": 0.3236,
-      "step": 14472
-    },
-    {
-      "epoch": 0.9458858898111234,
-      "grad_norm": 0.42595773935317993,
-      "learning_rate": 7.922290490222646e-06,
-      "loss": 0.3211,
-      "step": 14473
-    },
-    {
-      "epoch": 0.9459512450166656,
-      "grad_norm": 0.4478296637535095,
-      "learning_rate": 7.922007141982152e-06,
-      "loss": 0.3974,
-      "step": 14474
-    },
-    {
-      "epoch": 0.9460166002222077,
-      "grad_norm": 0.42474326491355896,
-      "learning_rate": 7.921723779490015e-06,
-      "loss": 0.3207,
-      "step": 14475
-    },
-    {
-      "epoch": 0.9460819554277499,
-      "grad_norm": 0.44154810905456543,
-      "learning_rate": 7.921440402747622e-06,
-      "loss": 0.368,
-      "step": 14476
-    },
-    {
-      "epoch": 0.9461473106332919,
-      "grad_norm": 0.40492886304855347,
-      "learning_rate": 7.92115701175635e-06,
-      "loss": 0.3183,
-      "step": 14477
-    },
-    {
-      "epoch": 0.946212665838834,
-      "grad_norm": 0.4571344554424286,
-      "learning_rate": 7.920873606517583e-06,
-      "loss": 0.3703,
-      "step": 14478
-    },
-    {
-      "epoch": 0.9462780210443762,
-      "grad_norm": 0.5302592515945435,
-      "learning_rate": 7.920590187032704e-06,
-      "loss": 0.3917,
-      "step": 14479
-    },
-    {
-      "epoch": 0.9463433762499183,
-      "grad_norm": 0.4236275255680084,
-      "learning_rate": 7.920306753303096e-06,
-      "loss": 0.3139,
-      "step": 14480
-    },
-    {
-      "epoch": 0.9464087314554605,
-      "grad_norm": 0.44597524404525757,
-      "learning_rate": 7.92002330533014e-06,
-      "loss": 0.3798,
-      "step": 14481
-    },
-    {
-      "epoch": 0.9464740866610025,
-      "grad_norm": 0.4571422338485718,
-      "learning_rate": 7.919739843115218e-06,
-      "loss": 0.4051,
-      "step": 14482
-    },
-    {
-      "epoch": 0.9465394418665447,
-      "grad_norm": 0.40712401270866394,
-      "learning_rate": 7.919456366659713e-06,
-      "loss": 0.316,
-      "step": 14483
-    },
-    {
-      "epoch": 0.9466047970720868,
-      "grad_norm": 0.45134514570236206,
-      "learning_rate": 7.919172875965008e-06,
-      "loss": 0.3486,
-      "step": 14484
-    },
-    {
-      "epoch": 0.946670152277629,
-      "grad_norm": 0.4873284697532654,
-      "learning_rate": 7.918889371032486e-06,
-      "loss": 0.4607,
-      "step": 14485
-    },
-    {
-      "epoch": 0.946735507483171,
-      "grad_norm": 0.4418301284313202,
-      "learning_rate": 7.918605851863528e-06,
-      "loss": 0.368,
-      "step": 14486
-    },
-    {
-      "epoch": 0.9468008626887131,
-      "grad_norm": 0.4351485073566437,
-      "learning_rate": 7.91832231845952e-06,
-      "loss": 0.392,
-      "step": 14487
-    },
-    {
-      "epoch": 0.9468662178942553,
-      "grad_norm": 0.4526745080947876,
-      "learning_rate": 7.918038770821844e-06,
-      "loss": 0.3803,
-      "step": 14488
-    },
-    {
-      "epoch": 0.9469315730997974,
-      "grad_norm": 0.4391677975654602,
-      "learning_rate": 7.917755208951879e-06,
-      "loss": 0.3997,
-      "step": 14489
-    },
-    {
-      "epoch": 0.9469969283053395,
-      "grad_norm": 0.43049854040145874,
-      "learning_rate": 7.917471632851013e-06,
-      "loss": 0.372,
-      "step": 14490
-    },
-    {
-      "epoch": 0.9470622835108816,
-      "grad_norm": 0.4770206809043884,
-      "learning_rate": 7.917188042520625e-06,
-      "loss": 0.3895,
-      "step": 14491
-    },
-    {
-      "epoch": 0.9471276387164238,
-      "grad_norm": 0.43693140149116516,
-      "learning_rate": 7.916904437962103e-06,
-      "loss": 0.3716,
-      "step": 14492
-    },
-    {
-      "epoch": 0.9471929939219659,
-      "grad_norm": 0.46739405393600464,
-      "learning_rate": 7.916620819176825e-06,
-      "loss": 0.4037,
-      "step": 14493
-    },
-    {
-      "epoch": 0.9472583491275081,
-      "grad_norm": 0.4732327461242676,
-      "learning_rate": 7.916337186166178e-06,
-      "loss": 0.3726,
-      "step": 14494
-    },
-    {
-      "epoch": 0.9473237043330501,
-      "grad_norm": 0.433886855840683,
-      "learning_rate": 7.916053538931545e-06,
-      "loss": 0.3584,
-      "step": 14495
-    },
-    {
-      "epoch": 0.9473890595385922,
-      "grad_norm": 0.4469623863697052,
-      "learning_rate": 7.915769877474308e-06,
-      "loss": 0.3454,
-      "step": 14496
-    },
-    {
-      "epoch": 0.9474544147441344,
-      "grad_norm": 0.4504670798778534,
-      "learning_rate": 7.915486201795853e-06,
-      "loss": 0.387,
-      "step": 14497
-    },
-    {
-      "epoch": 0.9475197699496765,
-      "grad_norm": 0.4406098425388336,
-      "learning_rate": 7.915202511897559e-06,
-      "loss": 0.3981,
-      "step": 14498
-    },
-    {
-      "epoch": 0.9475851251552186,
-      "grad_norm": 0.47060224413871765,
-      "learning_rate": 7.914918807780814e-06,
-      "loss": 0.4474,
-      "step": 14499
-    },
-    {
-      "epoch": 0.9476504803607607,
-      "grad_norm": 0.43092191219329834,
-      "learning_rate": 7.914635089447e-06,
-      "loss": 0.3676,
-      "step": 14500
-    },
-    {
-      "epoch": 0.9477158355663029,
-      "grad_norm": 0.43119972944259644,
-      "learning_rate": 7.914351356897501e-06,
-      "loss": 0.3553,
-      "step": 14501
-    },
-    {
-      "epoch": 0.947781190771845,
-      "grad_norm": 0.424753338098526,
-      "learning_rate": 7.9140676101337e-06,
-      "loss": 0.3377,
-      "step": 14502
-    },
-    {
-      "epoch": 0.947846545977387,
-      "grad_norm": 0.4283413290977478,
-      "learning_rate": 7.913783849156981e-06,
-      "loss": 0.3405,
-      "step": 14503
-    },
-    {
-      "epoch": 0.9479119011829292,
-      "grad_norm": 0.4264403283596039,
-      "learning_rate": 7.91350007396873e-06,
-      "loss": 0.3488,
-      "step": 14504
-    },
-    {
-      "epoch": 0.9479772563884713,
-      "grad_norm": 0.48066049814224243,
-      "learning_rate": 7.91321628457033e-06,
-      "loss": 0.4064,
-      "step": 14505
-    },
-    {
-      "epoch": 0.9480426115940135,
-      "grad_norm": 0.42478930950164795,
-      "learning_rate": 7.912932480963166e-06,
-      "loss": 0.377,
-      "step": 14506
-    },
-    {
-      "epoch": 0.9481079667995556,
-      "grad_norm": 0.4370587170124054,
-      "learning_rate": 7.912648663148619e-06,
-      "loss": 0.3409,
-      "step": 14507
-    },
-    {
-      "epoch": 0.9481733220050977,
-      "grad_norm": 0.47303035855293274,
-      "learning_rate": 7.912364831128076e-06,
-      "loss": 0.4258,
-      "step": 14508
-    },
-    {
-      "epoch": 0.9482386772106398,
-      "grad_norm": 0.4379528760910034,
-      "learning_rate": 7.912080984902922e-06,
-      "loss": 0.3946,
-      "step": 14509
-    },
-    {
-      "epoch": 0.948304032416182,
-      "grad_norm": 0.44799181818962097,
-      "learning_rate": 7.911797124474539e-06,
-      "loss": 0.3844,
-      "step": 14510
-    },
-    {
-      "epoch": 0.9483693876217241,
-      "grad_norm": 0.42166668176651,
-      "learning_rate": 7.911513249844314e-06,
-      "loss": 0.3829,
-      "step": 14511
-    },
-    {
-      "epoch": 0.9484347428272661,
-      "grad_norm": 0.4153478443622589,
-      "learning_rate": 7.911229361013629e-06,
-      "loss": 0.3489,
-      "step": 14512
-    },
-    {
-      "epoch": 0.9485000980328083,
-      "grad_norm": 0.4486326277256012,
-      "learning_rate": 7.91094545798387e-06,
-      "loss": 0.3955,
-      "step": 14513
-    },
-    {
-      "epoch": 0.9485654532383504,
-      "grad_norm": 0.4419936537742615,
-      "learning_rate": 7.91066154075642e-06,
-      "loss": 0.3856,
-      "step": 14514
-    },
-    {
-      "epoch": 0.9486308084438926,
-      "grad_norm": 0.42571067810058594,
-      "learning_rate": 7.910377609332666e-06,
-      "loss": 0.3343,
-      "step": 14515
-    },
-    {
-      "epoch": 0.9486961636494347,
-      "grad_norm": 0.45720112323760986,
-      "learning_rate": 7.910093663713994e-06,
-      "loss": 0.4466,
-      "step": 14516
-    },
-    {
-      "epoch": 0.9487615188549768,
-      "grad_norm": 0.45299604535102844,
-      "learning_rate": 7.909809703901787e-06,
-      "loss": 0.4152,
-      "step": 14517
-    },
-    {
-      "epoch": 0.9488268740605189,
-      "grad_norm": 0.4278375506401062,
-      "learning_rate": 7.90952572989743e-06,
-      "loss": 0.3984,
-      "step": 14518
-    },
-    {
-      "epoch": 0.9488922292660611,
-      "grad_norm": 0.4548674523830414,
-      "learning_rate": 7.909241741702306e-06,
-      "loss": 0.3994,
-      "step": 14519
-    },
-    {
-      "epoch": 0.9489575844716032,
-      "grad_norm": 0.4342322051525116,
-      "learning_rate": 7.908957739317803e-06,
-      "loss": 0.3509,
-      "step": 14520
-    },
-    {
-      "epoch": 0.9490229396771452,
-      "grad_norm": 0.455106258392334,
-      "learning_rate": 7.908673722745306e-06,
-      "loss": 0.3708,
-      "step": 14521
-    },
-    {
-      "epoch": 0.9490882948826874,
-      "grad_norm": 0.41844823956489563,
-      "learning_rate": 7.908389691986197e-06,
-      "loss": 0.3622,
-      "step": 14522
-    },
-    {
-      "epoch": 0.9491536500882295,
-      "grad_norm": 0.4724261462688446,
-      "learning_rate": 7.908105647041868e-06,
-      "loss": 0.4171,
-      "step": 14523
-    },
-    {
-      "epoch": 0.9492190052937717,
-      "grad_norm": 0.4346033036708832,
-      "learning_rate": 7.907821587913698e-06,
-      "loss": 0.3714,
-      "step": 14524
-    },
-    {
-      "epoch": 0.9492843604993138,
-      "grad_norm": 0.45755767822265625,
-      "learning_rate": 7.907537514603076e-06,
-      "loss": 0.3902,
-      "step": 14525
-    },
-    {
-      "epoch": 0.9493497157048559,
-      "grad_norm": 0.49286603927612305,
-      "learning_rate": 7.907253427111386e-06,
-      "loss": 0.4217,
-      "step": 14526
-    },
-    {
-      "epoch": 0.949415070910398,
-      "grad_norm": 0.4719918966293335,
-      "learning_rate": 7.906969325440012e-06,
-      "loss": 0.4078,
-      "step": 14527
-    },
-    {
-      "epoch": 0.9494804261159401,
-      "grad_norm": 0.4337770938873291,
-      "learning_rate": 7.906685209590343e-06,
-      "loss": 0.3701,
-      "step": 14528
-    },
-    {
-      "epoch": 0.9495457813214823,
-      "grad_norm": 0.4309212267398834,
-      "learning_rate": 7.906401079563764e-06,
-      "loss": 0.36,
-      "step": 14529
-    },
-    {
-      "epoch": 0.9496111365270243,
-      "grad_norm": 0.420102059841156,
-      "learning_rate": 7.906116935361659e-06,
-      "loss": 0.3389,
-      "step": 14530
-    },
-    {
-      "epoch": 0.9496764917325665,
-      "grad_norm": 0.43637362122535706,
-      "learning_rate": 7.905832776985415e-06,
-      "loss": 0.378,
-      "step": 14531
-    },
-    {
-      "epoch": 0.9497418469381086,
-      "grad_norm": 0.41150134801864624,
-      "learning_rate": 7.905548604436418e-06,
-      "loss": 0.3682,
-      "step": 14532
-    },
-    {
-      "epoch": 0.9498072021436508,
-      "grad_norm": 0.45195427536964417,
-      "learning_rate": 7.905264417716055e-06,
-      "loss": 0.3716,
-      "step": 14533
-    },
-    {
-      "epoch": 0.9498725573491928,
-      "grad_norm": 0.40737611055374146,
-      "learning_rate": 7.904980216825708e-06,
-      "loss": 0.3467,
-      "step": 14534
-    },
-    {
-      "epoch": 0.949937912554735,
-      "grad_norm": 0.4473322629928589,
-      "learning_rate": 7.904696001766768e-06,
-      "loss": 0.3875,
-      "step": 14535
-    },
-    {
-      "epoch": 0.9500032677602771,
-      "grad_norm": 0.4047374725341797,
-      "learning_rate": 7.90441177254062e-06,
-      "loss": 0.3549,
-      "step": 14536
-    },
-    {
-      "epoch": 0.9500686229658192,
-      "grad_norm": 0.4511867165565491,
-      "learning_rate": 7.904127529148648e-06,
-      "loss": 0.3615,
-      "step": 14537
-    },
-    {
-      "epoch": 0.9501339781713614,
-      "grad_norm": 0.4167231619358063,
-      "learning_rate": 7.903843271592242e-06,
-      "loss": 0.36,
-      "step": 14538
-    },
-    {
-      "epoch": 0.9501993333769034,
-      "grad_norm": 0.4476933479309082,
-      "learning_rate": 7.903558999872785e-06,
-      "loss": 0.3949,
-      "step": 14539
-    },
-    {
-      "epoch": 0.9502646885824456,
-      "grad_norm": 0.42178258299827576,
-      "learning_rate": 7.903274713991664e-06,
-      "loss": 0.3716,
-      "step": 14540
-    },
-    {
-      "epoch": 0.9503300437879877,
-      "grad_norm": 0.4218887388706207,
-      "learning_rate": 7.90299041395027e-06,
-      "loss": 0.3376,
-      "step": 14541
-    },
-    {
-      "epoch": 0.9503953989935299,
-      "grad_norm": 0.44128820300102234,
-      "learning_rate": 7.902706099749982e-06,
-      "loss": 0.3949,
-      "step": 14542
-    },
-    {
-      "epoch": 0.9504607541990719,
-      "grad_norm": 0.4702761471271515,
-      "learning_rate": 7.902421771392192e-06,
-      "loss": 0.3833,
-      "step": 14543
-    },
-    {
-      "epoch": 0.9505261094046141,
-      "grad_norm": 0.4297698736190796,
-      "learning_rate": 7.902137428878287e-06,
-      "loss": 0.3363,
-      "step": 14544
-    },
-    {
-      "epoch": 0.9505914646101562,
-      "grad_norm": 0.43554675579071045,
-      "learning_rate": 7.90185307220965e-06,
-      "loss": 0.3495,
-      "step": 14545
-    },
-    {
-      "epoch": 0.9506568198156983,
-      "grad_norm": 0.4012167155742645,
-      "learning_rate": 7.901568701387671e-06,
-      "loss": 0.322,
-      "step": 14546
-    },
-    {
-      "epoch": 0.9507221750212405,
-      "grad_norm": 0.40920358896255493,
-      "learning_rate": 7.901284316413738e-06,
-      "loss": 0.3448,
-      "step": 14547
-    },
-    {
-      "epoch": 0.9507875302267825,
-      "grad_norm": 0.4171324372291565,
-      "learning_rate": 7.900999917289234e-06,
-      "loss": 0.3518,
-      "step": 14548
-    },
-    {
-      "epoch": 0.9508528854323247,
-      "grad_norm": 0.43824639916419983,
-      "learning_rate": 7.90071550401555e-06,
-      "loss": 0.3964,
-      "step": 14549
-    },
-    {
-      "epoch": 0.9509182406378668,
-      "grad_norm": 0.44572529196739197,
-      "learning_rate": 7.90043107659407e-06,
-      "loss": 0.348,
-      "step": 14550
-    },
-    {
-      "epoch": 0.950983595843409,
-      "grad_norm": 0.42346009612083435,
-      "learning_rate": 7.900146635026184e-06,
-      "loss": 0.3132,
-      "step": 14551
-    },
-    {
-      "epoch": 0.951048951048951,
-      "grad_norm": 0.42569127678871155,
-      "learning_rate": 7.899862179313278e-06,
-      "loss": 0.3783,
-      "step": 14552
-    },
-    {
-      "epoch": 0.9511143062544932,
-      "grad_norm": 0.4125511646270752,
-      "learning_rate": 7.89957770945674e-06,
-      "loss": 0.3229,
-      "step": 14553
-    },
-    {
-      "epoch": 0.9511796614600353,
-      "grad_norm": 0.4504038393497467,
-      "learning_rate": 7.899293225457956e-06,
-      "loss": 0.367,
-      "step": 14554
-    },
-    {
-      "epoch": 0.9512450166655774,
-      "grad_norm": 0.4786413311958313,
-      "learning_rate": 7.899008727318315e-06,
-      "loss": 0.3977,
-      "step": 14555
-    },
-    {
-      "epoch": 0.9513103718711196,
-      "grad_norm": 0.4554285407066345,
-      "learning_rate": 7.898724215039204e-06,
-      "loss": 0.4162,
-      "step": 14556
-    },
-    {
-      "epoch": 0.9513757270766616,
-      "grad_norm": 0.4733133614063263,
-      "learning_rate": 7.898439688622011e-06,
-      "loss": 0.4213,
-      "step": 14557
-    },
-    {
-      "epoch": 0.9514410822822038,
-      "grad_norm": 0.45218655467033386,
-      "learning_rate": 7.898155148068124e-06,
-      "loss": 0.4109,
-      "step": 14558
-    },
-    {
-      "epoch": 0.9515064374877459,
-      "grad_norm": 0.436526894569397,
-      "learning_rate": 7.897870593378928e-06,
-      "loss": 0.3445,
-      "step": 14559
-    },
-    {
-      "epoch": 0.9515717926932881,
-      "grad_norm": 0.448528528213501,
-      "learning_rate": 7.897586024555816e-06,
-      "loss": 0.3581,
-      "step": 14560
-    },
-    {
-      "epoch": 0.9516371478988301,
-      "grad_norm": 0.4428890347480774,
-      "learning_rate": 7.897301441600172e-06,
-      "loss": 0.3828,
-      "step": 14561
-    },
-    {
-      "epoch": 0.9517025031043722,
-      "grad_norm": 0.46396172046661377,
-      "learning_rate": 7.897016844513386e-06,
-      "loss": 0.4034,
-      "step": 14562
-    },
-    {
-      "epoch": 0.9517678583099144,
-      "grad_norm": 0.4298264682292938,
-      "learning_rate": 7.896732233296844e-06,
-      "loss": 0.3458,
-      "step": 14563
-    },
-    {
-      "epoch": 0.9518332135154565,
-      "grad_norm": 0.4606119394302368,
-      "learning_rate": 7.896447607951936e-06,
-      "loss": 0.3958,
-      "step": 14564
-    },
-    {
-      "epoch": 0.9518985687209987,
-      "grad_norm": 0.39340078830718994,
-      "learning_rate": 7.89616296848005e-06,
-      "loss": 0.3042,
-      "step": 14565
-    },
-    {
-      "epoch": 0.9519639239265407,
-      "grad_norm": 0.45184096693992615,
-      "learning_rate": 7.895878314882572e-06,
-      "loss": 0.4126,
-      "step": 14566
-    },
-    {
-      "epoch": 0.9520292791320829,
-      "grad_norm": 0.4581175148487091,
-      "learning_rate": 7.895593647160895e-06,
-      "loss": 0.3521,
-      "step": 14567
-    },
-    {
-      "epoch": 0.952094634337625,
-      "grad_norm": 0.4212242066860199,
-      "learning_rate": 7.895308965316403e-06,
-      "loss": 0.3325,
-      "step": 14568
-    },
-    {
-      "epoch": 0.9521599895431672,
-      "grad_norm": 0.43856918811798096,
-      "learning_rate": 7.895024269350486e-06,
-      "loss": 0.3645,
-      "step": 14569
-    },
-    {
-      "epoch": 0.9522253447487092,
-      "grad_norm": 0.42181190848350525,
-      "learning_rate": 7.894739559264534e-06,
-      "loss": 0.3402,
-      "step": 14570
-    },
-    {
-      "epoch": 0.9522906999542513,
-      "grad_norm": 0.4480799436569214,
-      "learning_rate": 7.894454835059933e-06,
-      "loss": 0.4022,
-      "step": 14571
-    },
-    {
-      "epoch": 0.9523560551597935,
-      "grad_norm": 0.4167082607746124,
-      "learning_rate": 7.894170096738074e-06,
-      "loss": 0.3372,
-      "step": 14572
-    },
-    {
-      "epoch": 0.9524214103653356,
-      "grad_norm": 0.42250657081604004,
-      "learning_rate": 7.893885344300344e-06,
-      "loss": 0.3501,
-      "step": 14573
-    },
-    {
-      "epoch": 0.9524867655708777,
-      "grad_norm": 0.5498653650283813,
-      "learning_rate": 7.893600577748133e-06,
-      "loss": 0.3949,
-      "step": 14574
-    },
-    {
-      "epoch": 0.9525521207764198,
-      "grad_norm": 0.4462451934814453,
-      "learning_rate": 7.89331579708283e-06,
-      "loss": 0.3799,
-      "step": 14575
-    },
-    {
-      "epoch": 0.952617475981962,
-      "grad_norm": 0.4206028878688812,
-      "learning_rate": 7.893031002305822e-06,
-      "loss": 0.3208,
-      "step": 14576
-    },
-    {
-      "epoch": 0.9526828311875041,
-      "grad_norm": 0.418376088142395,
-      "learning_rate": 7.892746193418502e-06,
-      "loss": 0.3666,
-      "step": 14577
-    },
-    {
-      "epoch": 0.9527481863930463,
-      "grad_norm": 0.4303472340106964,
-      "learning_rate": 7.892461370422252e-06,
-      "loss": 0.3816,
-      "step": 14578
-    },
-    {
-      "epoch": 0.9528135415985883,
-      "grad_norm": 0.42829039692878723,
-      "learning_rate": 7.89217653331847e-06,
-      "loss": 0.3637,
-      "step": 14579
-    },
-    {
-      "epoch": 0.9528788968041304,
-      "grad_norm": 0.43662217259407043,
-      "learning_rate": 7.891891682108541e-06,
-      "loss": 0.3887,
-      "step": 14580
-    },
-    {
-      "epoch": 0.9529442520096726,
-      "grad_norm": 0.4312734603881836,
-      "learning_rate": 7.891606816793853e-06,
-      "loss": 0.3579,
-      "step": 14581
-    },
-    {
-      "epoch": 0.9530096072152147,
-      "grad_norm": 0.4352143406867981,
-      "learning_rate": 7.891321937375798e-06,
-      "loss": 0.3467,
-      "step": 14582
-    },
-    {
-      "epoch": 0.9530749624207568,
-      "grad_norm": 0.43026095628738403,
-      "learning_rate": 7.891037043855763e-06,
-      "loss": 0.3621,
-      "step": 14583
-    },
-    {
-      "epoch": 0.9531403176262989,
-      "grad_norm": 0.44248607754707336,
-      "learning_rate": 7.89075213623514e-06,
-      "loss": 0.3931,
-      "step": 14584
-    },
-    {
-      "epoch": 0.9532056728318411,
-      "grad_norm": 0.418418824672699,
-      "learning_rate": 7.890467214515315e-06,
-      "loss": 0.339,
-      "step": 14585
-    },
-    {
-      "epoch": 0.9532710280373832,
-      "grad_norm": 0.45571669936180115,
-      "learning_rate": 7.890182278697682e-06,
-      "loss": 0.43,
-      "step": 14586
-    },
-    {
-      "epoch": 0.9533363832429252,
-      "grad_norm": 0.8395232558250427,
-      "learning_rate": 7.889897328783628e-06,
-      "loss": 0.4226,
-      "step": 14587
-    },
-    {
-      "epoch": 0.9534017384484674,
-      "grad_norm": 0.43124720454216003,
-      "learning_rate": 7.889612364774547e-06,
-      "loss": 0.3707,
-      "step": 14588
-    },
-    {
-      "epoch": 0.9534670936540095,
-      "grad_norm": 0.48433342576026917,
-      "learning_rate": 7.88932738667182e-06,
-      "loss": 0.4657,
-      "step": 14589
-    },
-    {
-      "epoch": 0.9535324488595517,
-      "grad_norm": 0.4577804207801819,
-      "learning_rate": 7.889042394476847e-06,
-      "loss": 0.3597,
-      "step": 14590
-    },
-    {
-      "epoch": 0.9535978040650938,
-      "grad_norm": 0.43687108159065247,
-      "learning_rate": 7.88875738819101e-06,
-      "loss": 0.3499,
-      "step": 14591
-    },
-    {
-      "epoch": 0.9536631592706359,
-      "grad_norm": 0.4336129128932953,
-      "learning_rate": 7.888472367815705e-06,
-      "loss": 0.3431,
-      "step": 14592
-    },
-    {
-      "epoch": 0.953728514476178,
-      "grad_norm": 0.45300954580307007,
-      "learning_rate": 7.888187333352318e-06,
-      "loss": 0.4078,
-      "step": 14593
-    },
-    {
-      "epoch": 0.9537938696817202,
-      "grad_norm": 0.41192471981048584,
-      "learning_rate": 7.887902284802242e-06,
-      "loss": 0.3475,
-      "step": 14594
-    },
-    {
-      "epoch": 0.9538592248872623,
-      "grad_norm": 0.4481332302093506,
-      "learning_rate": 7.887617222166866e-06,
-      "loss": 0.4063,
-      "step": 14595
-    },
-    {
-      "epoch": 0.9539245800928043,
-      "grad_norm": 0.4861341416835785,
-      "learning_rate": 7.887332145447579e-06,
-      "loss": 0.3733,
-      "step": 14596
-    },
-    {
-      "epoch": 0.9539899352983465,
-      "grad_norm": 0.4583840072154999,
-      "learning_rate": 7.887047054645773e-06,
-      "loss": 0.4015,
-      "step": 14597
-    },
-    {
-      "epoch": 0.9540552905038886,
-      "grad_norm": 0.45041805505752563,
-      "learning_rate": 7.886761949762838e-06,
-      "loss": 0.3877,
-      "step": 14598
-    },
-    {
-      "epoch": 0.9541206457094308,
-      "grad_norm": 0.4253000020980835,
-      "learning_rate": 7.886476830800166e-06,
-      "loss": 0.3459,
-      "step": 14599
-    },
-    {
-      "epoch": 0.9541860009149729,
-      "grad_norm": 0.420580118894577,
-      "learning_rate": 7.886191697759146e-06,
-      "loss": 0.3515,
-      "step": 14600
-    },
-    {
-      "epoch": 0.954251356120515,
-      "grad_norm": 0.45473599433898926,
-      "learning_rate": 7.885906550641172e-06,
-      "loss": 0.3688,
-      "step": 14601
-    },
-    {
-      "epoch": 0.9543167113260571,
-      "grad_norm": 0.43718987703323364,
-      "learning_rate": 7.885621389447628e-06,
-      "loss": 0.3817,
-      "step": 14602
-    },
-    {
-      "epoch": 0.9543820665315993,
-      "grad_norm": 0.4303869307041168,
-      "learning_rate": 7.88533621417991e-06,
-      "loss": 0.3704,
-      "step": 14603
-    },
-    {
-      "epoch": 0.9544474217371414,
-      "grad_norm": 0.4161582887172699,
-      "learning_rate": 7.885051024839408e-06,
-      "loss": 0.3437,
-      "step": 14604
-    },
-    {
-      "epoch": 0.9545127769426834,
-      "grad_norm": 0.46883052587509155,
-      "learning_rate": 7.884765821427514e-06,
-      "loss": 0.3489,
-      "step": 14605
-    },
-    {
-      "epoch": 0.9545781321482256,
-      "grad_norm": 0.4682151973247528,
-      "learning_rate": 7.884480603945615e-06,
-      "loss": 0.4144,
-      "step": 14606
-    },
-    {
-      "epoch": 0.9546434873537677,
-      "grad_norm": 0.4272187054157257,
-      "learning_rate": 7.884195372395107e-06,
-      "loss": 0.3414,
-      "step": 14607
-    },
-    {
-      "epoch": 0.9547088425593099,
-      "grad_norm": 0.4370492100715637,
-      "learning_rate": 7.88391012677738e-06,
-      "loss": 0.3601,
-      "step": 14608
-    },
-    {
-      "epoch": 0.954774197764852,
-      "grad_norm": 0.4563125669956207,
-      "learning_rate": 7.88362486709382e-06,
-      "loss": 0.4149,
-      "step": 14609
-    },
-    {
-      "epoch": 0.9548395529703941,
-      "grad_norm": 0.44245320558547974,
-      "learning_rate": 7.883339593345827e-06,
-      "loss": 0.4101,
-      "step": 14610
-    },
-    {
-      "epoch": 0.9549049081759362,
-      "grad_norm": 0.4485817849636078,
-      "learning_rate": 7.883054305534786e-06,
-      "loss": 0.3841,
-      "step": 14611
-    },
-    {
-      "epoch": 0.9549702633814783,
-      "grad_norm": 0.4516976773738861,
-      "learning_rate": 7.88276900366209e-06,
-      "loss": 0.4044,
-      "step": 14612
-    },
-    {
-      "epoch": 0.9550356185870205,
-      "grad_norm": 0.5184816122055054,
-      "learning_rate": 7.882483687729131e-06,
-      "loss": 0.466,
-      "step": 14613
-    },
-    {
-      "epoch": 0.9551009737925625,
-      "grad_norm": 0.43364837765693665,
-      "learning_rate": 7.882198357737301e-06,
-      "loss": 0.3713,
-      "step": 14614
-    },
-    {
-      "epoch": 0.9551663289981047,
-      "grad_norm": 0.4156990349292755,
-      "learning_rate": 7.88191301368799e-06,
-      "loss": 0.3477,
-      "step": 14615
-    },
-    {
-      "epoch": 0.9552316842036468,
-      "grad_norm": 0.42361167073249817,
-      "learning_rate": 7.881627655582593e-06,
-      "loss": 0.3702,
-      "step": 14616
-    },
-    {
-      "epoch": 0.955297039409189,
-      "grad_norm": 0.4145409166812897,
-      "learning_rate": 7.881342283422498e-06,
-      "loss": 0.3354,
-      "step": 14617
-    },
-    {
-      "epoch": 0.955362394614731,
-      "grad_norm": 0.4254988133907318,
-      "learning_rate": 7.881056897209098e-06,
-      "loss": 0.3606,
-      "step": 14618
-    },
-    {
-      "epoch": 0.9554277498202732,
-      "grad_norm": 0.6123539805412292,
-      "learning_rate": 7.880771496943786e-06,
-      "loss": 0.3528,
-      "step": 14619
-    },
-    {
-      "epoch": 0.9554931050258153,
-      "grad_norm": 0.43436136841773987,
-      "learning_rate": 7.880486082627954e-06,
-      "loss": 0.3623,
-      "step": 14620
-    },
-    {
-      "epoch": 0.9555584602313574,
-      "grad_norm": 0.47374093532562256,
-      "learning_rate": 7.880200654262993e-06,
-      "loss": 0.4154,
-      "step": 14621
-    },
-    {
-      "epoch": 0.9556238154368996,
-      "grad_norm": 0.43287163972854614,
-      "learning_rate": 7.879915211850296e-06,
-      "loss": 0.3361,
-      "step": 14622
-    },
-    {
-      "epoch": 0.9556891706424416,
-      "grad_norm": 0.4072272479534149,
-      "learning_rate": 7.879629755391254e-06,
-      "loss": 0.3355,
-      "step": 14623
-    },
-    {
-      "epoch": 0.9557545258479838,
-      "grad_norm": 0.4534406363964081,
-      "learning_rate": 7.87934428488726e-06,
-      "loss": 0.3798,
-      "step": 14624
-    },
-    {
-      "epoch": 0.9558198810535259,
-      "grad_norm": 0.427864134311676,
-      "learning_rate": 7.879058800339708e-06,
-      "loss": 0.3236,
-      "step": 14625
-    },
-    {
-      "epoch": 0.9558852362590681,
-      "grad_norm": 0.4547727406024933,
-      "learning_rate": 7.878773301749986e-06,
-      "loss": 0.376,
-      "step": 14626
-    },
-    {
-      "epoch": 0.9559505914646101,
-      "grad_norm": 0.4406866133213043,
-      "learning_rate": 7.878487789119492e-06,
-      "loss": 0.373,
-      "step": 14627
-    },
-    {
-      "epoch": 0.9560159466701523,
-      "grad_norm": 0.45401880145072937,
-      "learning_rate": 7.878202262449615e-06,
-      "loss": 0.3786,
-      "step": 14628
-    },
-    {
-      "epoch": 0.9560813018756944,
-      "grad_norm": 0.4420032203197479,
-      "learning_rate": 7.87791672174175e-06,
-      "loss": 0.397,
-      "step": 14629
-    },
-    {
-      "epoch": 0.9561466570812365,
-      "grad_norm": 0.43429428339004517,
-      "learning_rate": 7.877631166997286e-06,
-      "loss": 0.3917,
-      "step": 14630
-    },
-    {
-      "epoch": 0.9562120122867787,
-      "grad_norm": 0.4955896735191345,
-      "learning_rate": 7.877345598217618e-06,
-      "loss": 0.4578,
-      "step": 14631
-    },
-    {
-      "epoch": 0.9562773674923207,
-      "grad_norm": 0.4426642954349518,
-      "learning_rate": 7.87706001540414e-06,
-      "loss": 0.4086,
-      "step": 14632
-    },
-    {
-      "epoch": 0.9563427226978629,
-      "grad_norm": 0.4321775734424591,
-      "learning_rate": 7.876774418558242e-06,
-      "loss": 0.3608,
-      "step": 14633
-    },
-    {
-      "epoch": 0.956408077903405,
-      "grad_norm": 0.45228639245033264,
-      "learning_rate": 7.87648880768132e-06,
-      "loss": 0.3884,
-      "step": 14634
-    },
-    {
-      "epoch": 0.9564734331089472,
-      "grad_norm": 0.4376620948314667,
-      "learning_rate": 7.876203182774764e-06,
-      "loss": 0.3385,
-      "step": 14635
-    },
-    {
-      "epoch": 0.9565387883144892,
-      "grad_norm": 0.42276784777641296,
-      "learning_rate": 7.87591754383997e-06,
-      "loss": 0.3366,
-      "step": 14636
-    },
-    {
-      "epoch": 0.9566041435200314,
-      "grad_norm": 0.4231339395046234,
-      "learning_rate": 7.87563189087833e-06,
-      "loss": 0.3407,
-      "step": 14637
-    },
-    {
-      "epoch": 0.9566694987255735,
-      "grad_norm": 0.40449103713035583,
-      "learning_rate": 7.875346223891236e-06,
-      "loss": 0.3293,
-      "step": 14638
-    },
-    {
-      "epoch": 0.9567348539311156,
-      "grad_norm": 0.4093177914619446,
-      "learning_rate": 7.875060542880083e-06,
-      "loss": 0.346,
-      "step": 14639
-    },
-    {
-      "epoch": 0.9568002091366578,
-      "grad_norm": 0.4191904067993164,
-      "learning_rate": 7.874774847846263e-06,
-      "loss": 0.3648,
-      "step": 14640
-    },
-    {
-      "epoch": 0.9568655643421998,
-      "grad_norm": 0.44178369641304016,
-      "learning_rate": 7.87448913879117e-06,
-      "loss": 0.3717,
-      "step": 14641
-    },
-    {
-      "epoch": 0.956930919547742,
-      "grad_norm": 0.4869230389595032,
-      "learning_rate": 7.8742034157162e-06,
-      "loss": 0.4358,
-      "step": 14642
-    },
-    {
-      "epoch": 0.9569962747532841,
-      "grad_norm": 0.47952571511268616,
-      "learning_rate": 7.873917678622742e-06,
-      "loss": 0.4481,
-      "step": 14643
-    },
-    {
-      "epoch": 0.9570616299588263,
-      "grad_norm": 0.4684247672557831,
-      "learning_rate": 7.873631927512192e-06,
-      "loss": 0.3839,
-      "step": 14644
-    },
-    {
-      "epoch": 0.9571269851643683,
-      "grad_norm": 0.43733447790145874,
-      "learning_rate": 7.873346162385944e-06,
-      "loss": 0.3691,
-      "step": 14645
-    },
-    {
-      "epoch": 0.9571923403699104,
-      "grad_norm": 0.4437146782875061,
-      "learning_rate": 7.873060383245391e-06,
-      "loss": 0.3883,
-      "step": 14646
-    },
-    {
-      "epoch": 0.9572576955754526,
-      "grad_norm": 0.41246497631073,
-      "learning_rate": 7.872774590091927e-06,
-      "loss": 0.3176,
-      "step": 14647
-    },
-    {
-      "epoch": 0.9573230507809947,
-      "grad_norm": 0.42654070258140564,
-      "learning_rate": 7.872488782926948e-06,
-      "loss": 0.3929,
-      "step": 14648
-    },
-    {
-      "epoch": 0.9573884059865369,
-      "grad_norm": 0.41639426350593567,
-      "learning_rate": 7.872202961751845e-06,
-      "loss": 0.3336,
-      "step": 14649
-    },
-    {
-      "epoch": 0.9574537611920789,
-      "grad_norm": 0.4656263291835785,
-      "learning_rate": 7.871917126568011e-06,
-      "loss": 0.382,
-      "step": 14650
-    },
-    {
-      "epoch": 0.9575191163976211,
-      "grad_norm": 0.4734870195388794,
-      "learning_rate": 7.871631277376845e-06,
-      "loss": 0.4118,
-      "step": 14651
-    },
-    {
-      "epoch": 0.9575844716031632,
-      "grad_norm": 0.45249998569488525,
-      "learning_rate": 7.871345414179738e-06,
-      "loss": 0.3986,
-      "step": 14652
-    },
-    {
-      "epoch": 0.9576498268087054,
-      "grad_norm": 0.4351339042186737,
-      "learning_rate": 7.871059536978085e-06,
-      "loss": 0.3862,
-      "step": 14653
-    },
-    {
-      "epoch": 0.9577151820142474,
-      "grad_norm": 0.44380736351013184,
-      "learning_rate": 7.870773645773278e-06,
-      "loss": 0.3743,
-      "step": 14654
-    },
-    {
-      "epoch": 0.9577805372197895,
-      "grad_norm": 0.42017319798469543,
-      "learning_rate": 7.870487740566714e-06,
-      "loss": 0.33,
-      "step": 14655
-    },
-    {
-      "epoch": 0.9578458924253317,
-      "grad_norm": 0.44864538311958313,
-      "learning_rate": 7.87020182135979e-06,
-      "loss": 0.368,
-      "step": 14656
-    },
-    {
-      "epoch": 0.9579112476308738,
-      "grad_norm": 0.44515731930732727,
-      "learning_rate": 7.869915888153893e-06,
-      "loss": 0.3473,
-      "step": 14657
-    },
-    {
-      "epoch": 0.957976602836416,
-      "grad_norm": 0.4643704891204834,
-      "learning_rate": 7.869629940950423e-06,
-      "loss": 0.413,
-      "step": 14658
-    },
-    {
-      "epoch": 0.958041958041958,
-      "grad_norm": 0.44642704725265503,
-      "learning_rate": 7.869343979750776e-06,
-      "loss": 0.4036,
-      "step": 14659
-    },
-    {
-      "epoch": 0.9581073132475002,
-      "grad_norm": 0.43242260813713074,
-      "learning_rate": 7.869058004556342e-06,
-      "loss": 0.3551,
-      "step": 14660
-    },
-    {
-      "epoch": 0.9581726684530423,
-      "grad_norm": 0.42438915371894836,
-      "learning_rate": 7.868772015368518e-06,
-      "loss": 0.3662,
-      "step": 14661
-    },
-    {
-      "epoch": 0.9582380236585845,
-      "grad_norm": 0.43452465534210205,
-      "learning_rate": 7.8684860121887e-06,
-      "loss": 0.3656,
-      "step": 14662
-    },
-    {
-      "epoch": 0.9583033788641265,
-      "grad_norm": 0.4757184386253357,
-      "learning_rate": 7.868199995018283e-06,
-      "loss": 0.4081,
-      "step": 14663
-    },
-    {
-      "epoch": 0.9583687340696686,
-      "grad_norm": 0.4319245517253876,
-      "learning_rate": 7.86791396385866e-06,
-      "loss": 0.3569,
-      "step": 14664
-    },
-    {
-      "epoch": 0.9584340892752108,
-      "grad_norm": 0.41545370221138,
-      "learning_rate": 7.867627918711226e-06,
-      "loss": 0.3381,
-      "step": 14665
-    },
-    {
-      "epoch": 0.9584994444807529,
-      "grad_norm": 0.40892425179481506,
-      "learning_rate": 7.86734185957738e-06,
-      "loss": 0.3569,
-      "step": 14666
-    },
-    {
-      "epoch": 0.958564799686295,
-      "grad_norm": 0.4482435882091522,
-      "learning_rate": 7.867055786458512e-06,
-      "loss": 0.3833,
-      "step": 14667
-    },
-    {
-      "epoch": 0.9586301548918371,
-      "grad_norm": 0.4535319209098816,
-      "learning_rate": 7.86676969935602e-06,
-      "loss": 0.3964,
-      "step": 14668
-    },
-    {
-      "epoch": 0.9586955100973793,
-      "grad_norm": 0.4511040449142456,
-      "learning_rate": 7.866483598271299e-06,
-      "loss": 0.3656,
-      "step": 14669
-    },
-    {
-      "epoch": 0.9587608653029214,
-      "grad_norm": 0.4494752585887909,
-      "learning_rate": 7.866197483205745e-06,
-      "loss": 0.3738,
-      "step": 14670
-    },
-    {
-      "epoch": 0.9588262205084634,
-      "grad_norm": 0.462099552154541,
-      "learning_rate": 7.865911354160754e-06,
-      "loss": 0.41,
-      "step": 14671
-    },
-    {
-      "epoch": 0.9588915757140056,
-      "grad_norm": 0.424837201833725,
-      "learning_rate": 7.865625211137717e-06,
-      "loss": 0.3404,
-      "step": 14672
-    },
-    {
-      "epoch": 0.9589569309195477,
-      "grad_norm": 0.45248329639434814,
-      "learning_rate": 7.865339054138037e-06,
-      "loss": 0.4095,
-      "step": 14673
-    },
-    {
-      "epoch": 0.9590222861250899,
-      "grad_norm": 0.4722941517829895,
-      "learning_rate": 7.865052883163104e-06,
-      "loss": 0.4126,
-      "step": 14674
-    },
-    {
-      "epoch": 0.959087641330632,
-      "grad_norm": 0.4175872802734375,
-      "learning_rate": 7.864766698214315e-06,
-      "loss": 0.3417,
-      "step": 14675
-    },
-    {
-      "epoch": 0.9591529965361741,
-      "grad_norm": 0.4347960352897644,
-      "learning_rate": 7.864480499293069e-06,
-      "loss": 0.3559,
-      "step": 14676
-    },
-    {
-      "epoch": 0.9592183517417162,
-      "grad_norm": 0.5596758127212524,
-      "learning_rate": 7.864194286400756e-06,
-      "loss": 0.3684,
-      "step": 14677
-    },
-    {
-      "epoch": 0.9592837069472584,
-      "grad_norm": 0.4326117932796478,
-      "learning_rate": 7.863908059538776e-06,
-      "loss": 0.311,
-      "step": 14678
-    },
-    {
-      "epoch": 0.9593490621528005,
-      "grad_norm": 0.4768591523170471,
-      "learning_rate": 7.863621818708526e-06,
-      "loss": 0.4144,
-      "step": 14679
-    },
-    {
-      "epoch": 0.9594144173583425,
-      "grad_norm": 0.44171446561813354,
-      "learning_rate": 7.863335563911399e-06,
-      "loss": 0.3635,
-      "step": 14680
-    },
-    {
-      "epoch": 0.9594797725638847,
-      "grad_norm": 0.43145930767059326,
-      "learning_rate": 7.863049295148793e-06,
-      "loss": 0.3436,
-      "step": 14681
-    },
-    {
-      "epoch": 0.9595451277694268,
-      "grad_norm": 0.9899649620056152,
-      "learning_rate": 7.862763012422102e-06,
-      "loss": 0.3586,
-      "step": 14682
-    },
-    {
-      "epoch": 0.959610482974969,
-      "grad_norm": 0.44094836711883545,
-      "learning_rate": 7.862476715732726e-06,
-      "loss": 0.3718,
-      "step": 14683
-    },
-    {
-      "epoch": 0.959675838180511,
-      "grad_norm": 0.5062451362609863,
-      "learning_rate": 7.862190405082057e-06,
-      "loss": 0.4498,
-      "step": 14684
-    },
-    {
-      "epoch": 0.9597411933860532,
-      "grad_norm": 0.44377270340919495,
-      "learning_rate": 7.861904080471497e-06,
-      "loss": 0.3573,
-      "step": 14685
-    },
-    {
-      "epoch": 0.9598065485915953,
-      "grad_norm": 0.5252393484115601,
-      "learning_rate": 7.861617741902437e-06,
-      "loss": 0.4324,
-      "step": 14686
-    },
-    {
-      "epoch": 0.9598719037971375,
-      "grad_norm": 0.44677838683128357,
-      "learning_rate": 7.861331389376277e-06,
-      "loss": 0.3811,
-      "step": 14687
-    },
-    {
-      "epoch": 0.9599372590026796,
-      "grad_norm": 0.4279773235321045,
-      "learning_rate": 7.861045022894414e-06,
-      "loss": 0.378,
-      "step": 14688
-    },
-    {
-      "epoch": 0.9600026142082216,
-      "grad_norm": 0.40343350172042847,
-      "learning_rate": 7.86075864245824e-06,
-      "loss": 0.3158,
-      "step": 14689
-    },
-    {
-      "epoch": 0.9600679694137638,
-      "grad_norm": 0.4713086485862732,
-      "learning_rate": 7.860472248069157e-06,
-      "loss": 0.3722,
-      "step": 14690
-    },
-    {
-      "epoch": 0.9601333246193059,
-      "grad_norm": 0.4739154577255249,
-      "learning_rate": 7.860185839728559e-06,
-      "loss": 0.3753,
-      "step": 14691
-    },
-    {
-      "epoch": 0.9601986798248481,
-      "grad_norm": 0.4589148759841919,
-      "learning_rate": 7.859899417437845e-06,
-      "loss": 0.3922,
-      "step": 14692
-    },
-    {
-      "epoch": 0.9602640350303902,
-      "grad_norm": 0.45614346861839294,
-      "learning_rate": 7.85961298119841e-06,
-      "loss": 0.3829,
-      "step": 14693
-    },
-    {
-      "epoch": 0.9603293902359323,
-      "grad_norm": 0.4335387647151947,
-      "learning_rate": 7.859326531011654e-06,
-      "loss": 0.3583,
-      "step": 14694
-    },
-    {
-      "epoch": 0.9603947454414744,
-      "grad_norm": 0.43192481994628906,
-      "learning_rate": 7.859040066878969e-06,
-      "loss": 0.3662,
-      "step": 14695
-    },
-    {
-      "epoch": 0.9604601006470166,
-      "grad_norm": 0.46029818058013916,
-      "learning_rate": 7.858753588801755e-06,
-      "loss": 0.42,
-      "step": 14696
-    },
-    {
-      "epoch": 0.9605254558525587,
-      "grad_norm": 0.48127636313438416,
-      "learning_rate": 7.858467096781411e-06,
-      "loss": 0.4115,
-      "step": 14697
-    },
-    {
-      "epoch": 0.9605908110581007,
-      "grad_norm": 0.4949004352092743,
-      "learning_rate": 7.858180590819332e-06,
-      "loss": 0.4736,
-      "step": 14698
-    },
-    {
-      "epoch": 0.9606561662636429,
-      "grad_norm": 0.44555923342704773,
-      "learning_rate": 7.857894070916915e-06,
-      "loss": 0.3926,
-      "step": 14699
-    },
-    {
-      "epoch": 0.960721521469185,
-      "grad_norm": 0.5973967909812927,
-      "learning_rate": 7.85760753707556e-06,
-      "loss": 0.3727,
-      "step": 14700
-    },
-    {
-      "epoch": 0.9607868766747272,
-      "grad_norm": 0.4441831707954407,
-      "learning_rate": 7.857320989296664e-06,
-      "loss": 0.3914,
-      "step": 14701
-    },
-    {
-      "epoch": 0.9608522318802692,
-      "grad_norm": 0.4172165095806122,
-      "learning_rate": 7.857034427581623e-06,
-      "loss": 0.3664,
-      "step": 14702
-    },
-    {
-      "epoch": 0.9609175870858114,
-      "grad_norm": 0.4362238943576813,
-      "learning_rate": 7.856747851931834e-06,
-      "loss": 0.3455,
-      "step": 14703
-    },
-    {
-      "epoch": 0.9609829422913535,
-      "grad_norm": 0.44920337200164795,
-      "learning_rate": 7.856461262348696e-06,
-      "loss": 0.3945,
-      "step": 14704
-    },
-    {
-      "epoch": 0.9610482974968956,
-      "grad_norm": 0.44160401821136475,
-      "learning_rate": 7.856174658833609e-06,
-      "loss": 0.3607,
-      "step": 14705
-    },
-    {
-      "epoch": 0.9611136527024378,
-      "grad_norm": 0.43351274728775024,
-      "learning_rate": 7.855888041387967e-06,
-      "loss": 0.3786,
-      "step": 14706
-    },
-    {
-      "epoch": 0.9611790079079798,
-      "grad_norm": 0.45133909583091736,
-      "learning_rate": 7.85560141001317e-06,
-      "loss": 0.3585,
-      "step": 14707
-    },
-    {
-      "epoch": 0.961244363113522,
-      "grad_norm": 0.4177149534225464,
-      "learning_rate": 7.855314764710616e-06,
-      "loss": 0.3664,
-      "step": 14708
-    },
-    {
-      "epoch": 0.9613097183190641,
-      "grad_norm": 0.47566458582878113,
-      "learning_rate": 7.8550281054817e-06,
-      "loss": 0.4278,
-      "step": 14709
-    },
-    {
-      "epoch": 0.9613750735246063,
-      "grad_norm": 0.39405885338783264,
-      "learning_rate": 7.854741432327827e-06,
-      "loss": 0.3067,
-      "step": 14710
-    },
-    {
-      "epoch": 0.9614404287301483,
-      "grad_norm": 0.4599893093109131,
-      "learning_rate": 7.854454745250388e-06,
-      "loss": 0.3991,
-      "step": 14711
-    },
-    {
-      "epoch": 0.9615057839356905,
-      "grad_norm": 0.4538350999355316,
-      "learning_rate": 7.854168044250788e-06,
-      "loss": 0.3745,
-      "step": 14712
-    },
-    {
-      "epoch": 0.9615711391412326,
-      "grad_norm": 0.4846935272216797,
-      "learning_rate": 7.853881329330419e-06,
-      "loss": 0.4154,
-      "step": 14713
-    },
-    {
-      "epoch": 0.9616364943467747,
-      "grad_norm": 0.4401126205921173,
-      "learning_rate": 7.853594600490681e-06,
-      "loss": 0.3682,
-      "step": 14714
-    },
-    {
-      "epoch": 0.9617018495523169,
-      "grad_norm": 0.41723647713661194,
-      "learning_rate": 7.853307857732976e-06,
-      "loss": 0.3163,
-      "step": 14715
-    },
-    {
-      "epoch": 0.9617672047578589,
-      "grad_norm": 0.4554825723171234,
-      "learning_rate": 7.8530211010587e-06,
-      "loss": 0.3879,
-      "step": 14716
-    },
-    {
-      "epoch": 0.9618325599634011,
-      "grad_norm": 0.4212765395641327,
-      "learning_rate": 7.852734330469249e-06,
-      "loss": 0.3478,
-      "step": 14717
-    },
-    {
-      "epoch": 0.9618979151689432,
-      "grad_norm": 0.4703579246997833,
-      "learning_rate": 7.852447545966026e-06,
-      "loss": 0.3866,
-      "step": 14718
-    },
-    {
-      "epoch": 0.9619632703744854,
-      "grad_norm": 0.4920124411582947,
-      "learning_rate": 7.85216074755043e-06,
-      "loss": 0.4278,
-      "step": 14719
-    },
-    {
-      "epoch": 0.9620286255800274,
-      "grad_norm": 0.41396501660346985,
-      "learning_rate": 7.851873935223856e-06,
-      "loss": 0.3861,
-      "step": 14720
-    },
-    {
-      "epoch": 0.9620939807855696,
-      "grad_norm": 0.43075990676879883,
-      "learning_rate": 7.851587108987705e-06,
-      "loss": 0.3568,
-      "step": 14721
-    },
-    {
-      "epoch": 0.9621593359911117,
-      "grad_norm": 0.42916256189346313,
-      "learning_rate": 7.851300268843376e-06,
-      "loss": 0.3821,
-      "step": 14722
-    },
-    {
-      "epoch": 0.9622246911966538,
-      "grad_norm": 0.42286500334739685,
-      "learning_rate": 7.851013414792267e-06,
-      "loss": 0.3592,
-      "step": 14723
-    },
-    {
-      "epoch": 0.962290046402196,
-      "grad_norm": 0.4404940605163574,
-      "learning_rate": 7.850726546835779e-06,
-      "loss": 0.3643,
-      "step": 14724
-    },
-    {
-      "epoch": 0.962355401607738,
-      "grad_norm": 0.41407260298728943,
-      "learning_rate": 7.85043966497531e-06,
-      "loss": 0.3598,
-      "step": 14725
-    },
-    {
-      "epoch": 0.9624207568132802,
-      "grad_norm": 0.4436608552932739,
-      "learning_rate": 7.850152769212258e-06,
-      "loss": 0.3768,
-      "step": 14726
-    },
-    {
-      "epoch": 0.9624861120188223,
-      "grad_norm": 0.4768117666244507,
-      "learning_rate": 7.849865859548025e-06,
-      "loss": 0.4293,
-      "step": 14727
-    },
-    {
-      "epoch": 0.9625514672243645,
-      "grad_norm": 0.42851582169532776,
-      "learning_rate": 7.849578935984007e-06,
-      "loss": 0.3621,
-      "step": 14728
-    },
-    {
-      "epoch": 0.9626168224299065,
-      "grad_norm": 0.46475011110305786,
-      "learning_rate": 7.849291998521608e-06,
-      "loss": 0.3942,
-      "step": 14729
-    },
-    {
-      "epoch": 0.9626821776354486,
-      "grad_norm": 0.420767217874527,
-      "learning_rate": 7.849005047162223e-06,
-      "loss": 0.3538,
-      "step": 14730
-    },
-    {
-      "epoch": 0.9627475328409908,
-      "grad_norm": 0.4748491048812866,
-      "learning_rate": 7.848718081907253e-06,
-      "loss": 0.3727,
-      "step": 14731
-    },
-    {
-      "epoch": 0.9628128880465329,
-      "grad_norm": 0.43817439675331116,
-      "learning_rate": 7.848431102758101e-06,
-      "loss": 0.3554,
-      "step": 14732
-    },
-    {
-      "epoch": 0.962878243252075,
-      "grad_norm": 0.4493640959262848,
-      "learning_rate": 7.84814410971616e-06,
-      "loss": 0.3965,
-      "step": 14733
-    },
-    {
-      "epoch": 0.9629435984576171,
-      "grad_norm": 0.4303871691226959,
-      "learning_rate": 7.847857102782836e-06,
-      "loss": 0.3788,
-      "step": 14734
-    },
-    {
-      "epoch": 0.9630089536631593,
-      "grad_norm": 0.40862104296684265,
-      "learning_rate": 7.847570081959525e-06,
-      "loss": 0.312,
-      "step": 14735
-    },
-    {
-      "epoch": 0.9630743088687014,
-      "grad_norm": 0.44714468717575073,
-      "learning_rate": 7.847283047247629e-06,
-      "loss": 0.3861,
-      "step": 14736
-    },
-    {
-      "epoch": 0.9631396640742436,
-      "grad_norm": 0.4178265333175659,
-      "learning_rate": 7.846995998648547e-06,
-      "loss": 0.3532,
-      "step": 14737
-    },
-    {
-      "epoch": 0.9632050192797856,
-      "grad_norm": 0.4428759813308716,
-      "learning_rate": 7.846708936163679e-06,
-      "loss": 0.3965,
-      "step": 14738
-    },
-    {
-      "epoch": 0.9632703744853277,
-      "grad_norm": 0.41116204857826233,
-      "learning_rate": 7.846421859794426e-06,
-      "loss": 0.3523,
-      "step": 14739
-    },
-    {
-      "epoch": 0.9633357296908699,
-      "grad_norm": 0.39715877175331116,
-      "learning_rate": 7.846134769542186e-06,
-      "loss": 0.3151,
-      "step": 14740
-    },
-    {
-      "epoch": 0.963401084896412,
-      "grad_norm": 0.4801284372806549,
-      "learning_rate": 7.845847665408362e-06,
-      "loss": 0.3931,
-      "step": 14741
-    },
-    {
-      "epoch": 0.9634664401019541,
-      "grad_norm": 0.43569839000701904,
-      "learning_rate": 7.845560547394353e-06,
-      "loss": 0.3778,
-      "step": 14742
-    },
-    {
-      "epoch": 0.9635317953074962,
-      "grad_norm": 0.4289999008178711,
-      "learning_rate": 7.84527341550156e-06,
-      "loss": 0.395,
-      "step": 14743
-    },
-    {
-      "epoch": 0.9635971505130384,
-      "grad_norm": 0.4477882981300354,
-      "learning_rate": 7.844986269731381e-06,
-      "loss": 0.3788,
-      "step": 14744
-    },
-    {
-      "epoch": 0.9636625057185805,
-      "grad_norm": 0.4254097044467926,
-      "learning_rate": 7.844699110085218e-06,
-      "loss": 0.3743,
-      "step": 14745
-    },
-    {
-      "epoch": 0.9637278609241227,
-      "grad_norm": 0.45810019969940186,
-      "learning_rate": 7.844411936564475e-06,
-      "loss": 0.385,
-      "step": 14746
-    },
-    {
-      "epoch": 0.9637932161296647,
-      "grad_norm": 0.42661023139953613,
-      "learning_rate": 7.844124749170547e-06,
-      "loss": 0.3897,
-      "step": 14747
-    },
-    {
-      "epoch": 0.9638585713352068,
-      "grad_norm": 0.45324674248695374,
-      "learning_rate": 7.843837547904838e-06,
-      "loss": 0.4,
-      "step": 14748
-    },
-    {
-      "epoch": 0.963923926540749,
-      "grad_norm": 0.4282146394252777,
-      "learning_rate": 7.843550332768747e-06,
-      "loss": 0.3479,
-      "step": 14749
-    },
-    {
-      "epoch": 0.9639892817462911,
-      "grad_norm": 0.47269314527511597,
-      "learning_rate": 7.84326310376368e-06,
-      "loss": 0.4153,
-      "step": 14750
-    },
-    {
-      "epoch": 0.9640546369518332,
-      "grad_norm": 0.431690514087677,
-      "learning_rate": 7.842975860891029e-06,
-      "loss": 0.3873,
-      "step": 14751
-    },
-    {
-      "epoch": 0.9641199921573753,
-      "grad_norm": 0.5653694272041321,
-      "learning_rate": 7.842688604152202e-06,
-      "loss": 0.3715,
-      "step": 14752
-    },
-    {
-      "epoch": 0.9641853473629175,
-      "grad_norm": 0.44812360405921936,
-      "learning_rate": 7.842401333548599e-06,
-      "loss": 0.4027,
-      "step": 14753
-    },
-    {
-      "epoch": 0.9642507025684596,
-      "grad_norm": 0.46026045083999634,
-      "learning_rate": 7.842114049081618e-06,
-      "loss": 0.4292,
-      "step": 14754
-    },
-    {
-      "epoch": 0.9643160577740016,
-      "grad_norm": 0.44783565402030945,
-      "learning_rate": 7.841826750752663e-06,
-      "loss": 0.3687,
-      "step": 14755
-    },
-    {
-      "epoch": 0.9643814129795438,
-      "grad_norm": 0.4119025468826294,
-      "learning_rate": 7.841539438563134e-06,
-      "loss": 0.3316,
-      "step": 14756
-    },
-    {
-      "epoch": 0.9644467681850859,
-      "grad_norm": 0.43547311425209045,
-      "learning_rate": 7.841252112514433e-06,
-      "loss": 0.3565,
-      "step": 14757
-    },
-    {
-      "epoch": 0.9645121233906281,
-      "grad_norm": 0.4158417284488678,
-      "learning_rate": 7.840964772607962e-06,
-      "loss": 0.3285,
-      "step": 14758
-    },
-    {
-      "epoch": 0.9645774785961702,
-      "grad_norm": 0.4588179886341095,
-      "learning_rate": 7.840677418845119e-06,
-      "loss": 0.3837,
-      "step": 14759
-    },
-    {
-      "epoch": 0.9646428338017123,
-      "grad_norm": 0.4297809898853302,
-      "learning_rate": 7.84039005122731e-06,
-      "loss": 0.3471,
-      "step": 14760
-    },
-    {
-      "epoch": 0.9647081890072544,
-      "grad_norm": 0.4243960678577423,
-      "learning_rate": 7.840102669755936e-06,
-      "loss": 0.3413,
-      "step": 14761
-    },
-    {
-      "epoch": 0.9647735442127966,
-      "grad_norm": 0.5048423409461975,
-      "learning_rate": 7.839815274432397e-06,
-      "loss": 0.3964,
-      "step": 14762
-    },
-    {
-      "epoch": 0.9648388994183387,
-      "grad_norm": 0.45856770873069763,
-      "learning_rate": 7.839527865258093e-06,
-      "loss": 0.4471,
-      "step": 14763
-    },
-    {
-      "epoch": 0.9649042546238807,
-      "grad_norm": 0.4537294805049896,
-      "learning_rate": 7.839240442234428e-06,
-      "loss": 0.3744,
-      "step": 14764
-    },
-    {
-      "epoch": 0.9649696098294229,
-      "grad_norm": 0.4455563426017761,
-      "learning_rate": 7.838953005362807e-06,
-      "loss": 0.3939,
-      "step": 14765
-    },
-    {
-      "epoch": 0.965034965034965,
-      "grad_norm": 0.4145625829696655,
-      "learning_rate": 7.838665554644624e-06,
-      "loss": 0.3369,
-      "step": 14766
-    },
-    {
-      "epoch": 0.9651003202405072,
-      "grad_norm": 0.4203234016895294,
-      "learning_rate": 7.838378090081289e-06,
-      "loss": 0.3372,
-      "step": 14767
-    },
-    {
-      "epoch": 0.9651656754460493,
-      "grad_norm": 0.48393452167510986,
-      "learning_rate": 7.838090611674199e-06,
-      "loss": 0.4139,
-      "step": 14768
-    },
-    {
-      "epoch": 0.9652310306515914,
-      "grad_norm": 0.48538827896118164,
-      "learning_rate": 7.837803119424759e-06,
-      "loss": 0.4358,
-      "step": 14769
-    },
-    {
-      "epoch": 0.9652963858571335,
-      "grad_norm": 0.46151402592658997,
-      "learning_rate": 7.83751561333437e-06,
-      "loss": 0.4115,
-      "step": 14770
-    },
-    {
-      "epoch": 0.9653617410626757,
-      "grad_norm": 0.4766049385070801,
-      "learning_rate": 7.837228093404434e-06,
-      "loss": 0.3971,
-      "step": 14771
-    },
-    {
-      "epoch": 0.9654270962682178,
-      "grad_norm": 0.46706485748291016,
-      "learning_rate": 7.836940559636354e-06,
-      "loss": 0.4104,
-      "step": 14772
-    },
-    {
-      "epoch": 0.9654924514737598,
-      "grad_norm": 0.45811277627944946,
-      "learning_rate": 7.836653012031533e-06,
-      "loss": 0.4028,
-      "step": 14773
-    },
-    {
-      "epoch": 0.965557806679302,
-      "grad_norm": 0.4126622676849365,
-      "learning_rate": 7.83636545059137e-06,
-      "loss": 0.336,
-      "step": 14774
-    },
-    {
-      "epoch": 0.9656231618848441,
-      "grad_norm": 0.4496912956237793,
-      "learning_rate": 7.836077875317271e-06,
-      "loss": 0.3969,
-      "step": 14775
-    },
-    {
-      "epoch": 0.9656885170903863,
-      "grad_norm": 0.44422343373298645,
-      "learning_rate": 7.835790286210639e-06,
-      "loss": 0.3789,
-      "step": 14776
-    },
-    {
-      "epoch": 0.9657538722959284,
-      "grad_norm": 0.4231109023094177,
-      "learning_rate": 7.835502683272874e-06,
-      "loss": 0.3419,
-      "step": 14777
-    },
-    {
-      "epoch": 0.9658192275014705,
-      "grad_norm": 0.4425690770149231,
-      "learning_rate": 7.835215066505382e-06,
-      "loss": 0.3368,
-      "step": 14778
-    },
-    {
-      "epoch": 0.9658845827070126,
-      "grad_norm": 0.4472150206565857,
-      "learning_rate": 7.834927435909562e-06,
-      "loss": 0.422,
-      "step": 14779
-    },
-    {
-      "epoch": 0.9659499379125548,
-      "grad_norm": 0.48612692952156067,
-      "learning_rate": 7.83463979148682e-06,
-      "loss": 0.4519,
-      "step": 14780
-    },
-    {
-      "epoch": 0.9660152931180969,
-      "grad_norm": 0.4408412277698517,
-      "learning_rate": 7.834352133238558e-06,
-      "loss": 0.3827,
-      "step": 14781
-    },
-    {
-      "epoch": 0.9660806483236389,
-      "grad_norm": 0.43729573488235474,
-      "learning_rate": 7.834064461166178e-06,
-      "loss": 0.3933,
-      "step": 14782
-    },
-    {
-      "epoch": 0.9661460035291811,
-      "grad_norm": 0.41399234533309937,
-      "learning_rate": 7.833776775271083e-06,
-      "loss": 0.3313,
-      "step": 14783
-    },
-    {
-      "epoch": 0.9662113587347232,
-      "grad_norm": 0.4549391269683838,
-      "learning_rate": 7.833489075554679e-06,
-      "loss": 0.3768,
-      "step": 14784
-    },
-    {
-      "epoch": 0.9662767139402654,
-      "grad_norm": 0.4329865276813507,
-      "learning_rate": 7.833201362018364e-06,
-      "loss": 0.3603,
-      "step": 14785
-    },
-    {
-      "epoch": 0.9663420691458074,
-      "grad_norm": 0.42552855610847473,
-      "learning_rate": 7.832913634663545e-06,
-      "loss": 0.3512,
-      "step": 14786
-    },
-    {
-      "epoch": 0.9664074243513496,
-      "grad_norm": 0.4280606210231781,
-      "learning_rate": 7.832625893491627e-06,
-      "loss": 0.3397,
-      "step": 14787
-    },
-    {
-      "epoch": 0.9664727795568917,
-      "grad_norm": 0.47604286670684814,
-      "learning_rate": 7.83233813850401e-06,
-      "loss": 0.4267,
-      "step": 14788
-    },
-    {
-      "epoch": 0.9665381347624338,
-      "grad_norm": 0.4646662771701813,
-      "learning_rate": 7.8320503697021e-06,
-      "loss": 0.383,
-      "step": 14789
-    },
-    {
-      "epoch": 0.966603489967976,
-      "grad_norm": 0.4610874652862549,
-      "learning_rate": 7.831762587087297e-06,
-      "loss": 0.4114,
-      "step": 14790
-    },
-    {
-      "epoch": 0.966668845173518,
-      "grad_norm": 0.45965999364852905,
-      "learning_rate": 7.831474790661009e-06,
-      "loss": 0.3919,
-      "step": 14791
-    },
-    {
-      "epoch": 0.9667342003790602,
-      "grad_norm": 0.43267151713371277,
-      "learning_rate": 7.831186980424637e-06,
-      "loss": 0.3704,
-      "step": 14792
-    },
-    {
-      "epoch": 0.9667995555846023,
-      "grad_norm": 0.4539198875427246,
-      "learning_rate": 7.830899156379584e-06,
-      "loss": 0.3653,
-      "step": 14793
-    },
-    {
-      "epoch": 0.9668649107901445,
-      "grad_norm": 0.4215865135192871,
-      "learning_rate": 7.830611318527254e-06,
-      "loss": 0.3673,
-      "step": 14794
-    },
-    {
-      "epoch": 0.9669302659956865,
-      "grad_norm": 0.4116964638233185,
-      "learning_rate": 7.830323466869055e-06,
-      "loss": 0.3611,
-      "step": 14795
-    },
-    {
-      "epoch": 0.9669956212012287,
-      "grad_norm": 0.5104700326919556,
-      "learning_rate": 7.830035601406386e-06,
-      "loss": 0.3595,
-      "step": 14796
-    },
-    {
-      "epoch": 0.9670609764067708,
-      "grad_norm": 0.4243166446685791,
-      "learning_rate": 7.829747722140656e-06,
-      "loss": 0.3695,
-      "step": 14797
-    },
-    {
-      "epoch": 0.9671263316123129,
-      "grad_norm": 0.42462068796157837,
-      "learning_rate": 7.829459829073263e-06,
-      "loss": 0.3653,
-      "step": 14798
-    },
-    {
-      "epoch": 0.9671916868178551,
-      "grad_norm": 0.4319928288459778,
-      "learning_rate": 7.829171922205615e-06,
-      "loss": 0.3521,
-      "step": 14799
-    },
-    {
-      "epoch": 0.9672570420233971,
-      "grad_norm": 0.43317168951034546,
-      "learning_rate": 7.828884001539117e-06,
-      "loss": 0.3671,
-      "step": 14800
-    },
-    {
-      "epoch": 0.9673223972289393,
-      "grad_norm": 0.4600657522678375,
-      "learning_rate": 7.82859606707517e-06,
-      "loss": 0.3758,
-      "step": 14801
-    },
-    {
-      "epoch": 0.9673877524344814,
-      "grad_norm": 0.4432556629180908,
-      "learning_rate": 7.82830811881518e-06,
-      "loss": 0.3742,
-      "step": 14802
-    },
-    {
-      "epoch": 0.9674531076400236,
-      "grad_norm": 0.4621134102344513,
-      "learning_rate": 7.82802015676055e-06,
-      "loss": 0.365,
-      "step": 14803
-    },
-    {
-      "epoch": 0.9675184628455656,
-      "grad_norm": 0.3997276723384857,
-      "learning_rate": 7.827732180912689e-06,
-      "loss": 0.3464,
-      "step": 14804
-    },
-    {
-      "epoch": 0.9675838180511078,
-      "grad_norm": 0.46412381529808044,
-      "learning_rate": 7.827444191272997e-06,
-      "loss": 0.4052,
-      "step": 14805
-    },
-    {
-      "epoch": 0.9676491732566499,
-      "grad_norm": 0.4773043990135193,
-      "learning_rate": 7.82715618784288e-06,
-      "loss": 0.4325,
-      "step": 14806
-    },
-    {
-      "epoch": 0.967714528462192,
-      "grad_norm": 0.4289936423301697,
-      "learning_rate": 7.826868170623742e-06,
-      "loss": 0.3608,
-      "step": 14807
-    },
-    {
-      "epoch": 0.9677798836677342,
-      "grad_norm": 0.4527800381183624,
-      "learning_rate": 7.82658013961699e-06,
-      "loss": 0.3931,
-      "step": 14808
-    },
-    {
-      "epoch": 0.9678452388732762,
-      "grad_norm": 0.4790951907634735,
-      "learning_rate": 7.826292094824029e-06,
-      "loss": 0.4221,
-      "step": 14809
-    },
-    {
-      "epoch": 0.9679105940788184,
-      "grad_norm": 0.4474199414253235,
-      "learning_rate": 7.82600403624626e-06,
-      "loss": 0.3521,
-      "step": 14810
-    },
-    {
-      "epoch": 0.9679759492843605,
-      "grad_norm": 0.4509504735469818,
-      "learning_rate": 7.82571596388509e-06,
-      "loss": 0.3564,
-      "step": 14811
-    },
-    {
-      "epoch": 0.9680413044899027,
-      "grad_norm": 0.44744789600372314,
-      "learning_rate": 7.825427877741925e-06,
-      "loss": 0.3412,
-      "step": 14812
-    },
-    {
-      "epoch": 0.9681066596954447,
-      "grad_norm": 0.4231618344783783,
-      "learning_rate": 7.825139777818169e-06,
-      "loss": 0.3509,
-      "step": 14813
-    },
-    {
-      "epoch": 0.9681720149009868,
-      "grad_norm": 0.4987364709377289,
-      "learning_rate": 7.82485166411523e-06,
-      "loss": 0.3934,
-      "step": 14814
-    },
-    {
-      "epoch": 0.968237370106529,
-      "grad_norm": 0.45702850818634033,
-      "learning_rate": 7.824563536634507e-06,
-      "loss": 0.3632,
-      "step": 14815
-    },
-    {
-      "epoch": 0.9683027253120711,
-      "grad_norm": 0.40645474195480347,
-      "learning_rate": 7.824275395377411e-06,
-      "loss": 0.3432,
-      "step": 14816
-    },
-    {
-      "epoch": 0.9683680805176133,
-      "grad_norm": 0.4501011371612549,
-      "learning_rate": 7.823987240345346e-06,
-      "loss": 0.3772,
-      "step": 14817
-    },
-    {
-      "epoch": 0.9684334357231553,
-      "grad_norm": 0.4299066662788391,
-      "learning_rate": 7.823699071539717e-06,
-      "loss": 0.354,
-      "step": 14818
-    },
-    {
-      "epoch": 0.9684987909286975,
-      "grad_norm": 0.43744340538978577,
-      "learning_rate": 7.823410888961928e-06,
-      "loss": 0.3451,
-      "step": 14819
-    },
-    {
-      "epoch": 0.9685641461342396,
-      "grad_norm": 0.44244512915611267,
-      "learning_rate": 7.82312269261339e-06,
-      "loss": 0.3441,
-      "step": 14820
-    },
-    {
-      "epoch": 0.9686295013397818,
-      "grad_norm": 0.45788103342056274,
-      "learning_rate": 7.8228344824955e-06,
-      "loss": 0.3931,
-      "step": 14821
-    },
-    {
-      "epoch": 0.9686948565453238,
-      "grad_norm": 0.45720037817955017,
-      "learning_rate": 7.82254625860967e-06,
-      "loss": 0.4013,
-      "step": 14822
-    },
-    {
-      "epoch": 0.9687602117508659,
-      "grad_norm": 0.4614906311035156,
-      "learning_rate": 7.822258020957302e-06,
-      "loss": 0.4173,
-      "step": 14823
-    },
-    {
-      "epoch": 0.9688255669564081,
-      "grad_norm": 0.4436750113964081,
-      "learning_rate": 7.821969769539806e-06,
-      "loss": 0.3766,
-      "step": 14824
-    },
-    {
-      "epoch": 0.9688909221619502,
-      "grad_norm": 0.3952990472316742,
-      "learning_rate": 7.821681504358587e-06,
-      "loss": 0.2851,
-      "step": 14825
-    },
-    {
-      "epoch": 0.9689562773674923,
-      "grad_norm": 0.4649125933647156,
-      "learning_rate": 7.821393225415047e-06,
-      "loss": 0.3955,
-      "step": 14826
-    },
-    {
-      "epoch": 0.9690216325730344,
-      "grad_norm": 0.42980360984802246,
-      "learning_rate": 7.821104932710597e-06,
-      "loss": 0.3687,
-      "step": 14827
-    },
-    {
-      "epoch": 0.9690869877785766,
-      "grad_norm": 0.4904744327068329,
-      "learning_rate": 7.82081662624664e-06,
-      "loss": 0.455,
-      "step": 14828
-    },
-    {
-      "epoch": 0.9691523429841187,
-      "grad_norm": 0.4130229651927948,
-      "learning_rate": 7.820528306024583e-06,
-      "loss": 0.3605,
-      "step": 14829
-    },
-    {
-      "epoch": 0.9692176981896609,
-      "grad_norm": 0.5697034597396851,
-      "learning_rate": 7.820239972045832e-06,
-      "loss": 0.4163,
-      "step": 14830
-    },
-    {
-      "epoch": 0.9692830533952029,
-      "grad_norm": 0.44115638732910156,
-      "learning_rate": 7.819951624311794e-06,
-      "loss": 0.3685,
-      "step": 14831
-    },
-    {
-      "epoch": 0.969348408600745,
-      "grad_norm": 0.4227757155895233,
-      "learning_rate": 7.819663262823876e-06,
-      "loss": 0.323,
-      "step": 14832
-    },
-    {
-      "epoch": 0.9694137638062872,
-      "grad_norm": 0.47103533148765564,
-      "learning_rate": 7.819374887583481e-06,
-      "loss": 0.3533,
-      "step": 14833
-    },
-    {
-      "epoch": 0.9694791190118293,
-      "grad_norm": 0.4258157014846802,
-      "learning_rate": 7.81908649859202e-06,
-      "loss": 0.333,
-      "step": 14834
-    },
-    {
-      "epoch": 0.9695444742173714,
-      "grad_norm": 0.46860530972480774,
-      "learning_rate": 7.818798095850897e-06,
-      "loss": 0.3989,
-      "step": 14835
-    },
-    {
-      "epoch": 0.9696098294229135,
-      "grad_norm": 0.49100691080093384,
-      "learning_rate": 7.81850967936152e-06,
-      "loss": 0.4704,
-      "step": 14836
-    },
-    {
-      "epoch": 0.9696751846284557,
-      "grad_norm": 0.44359099864959717,
-      "learning_rate": 7.818221249125293e-06,
-      "loss": 0.4016,
-      "step": 14837
-    },
-    {
-      "epoch": 0.9697405398339978,
-      "grad_norm": 0.44306522607803345,
-      "learning_rate": 7.817932805143627e-06,
-      "loss": 0.3773,
-      "step": 14838
-    },
-    {
-      "epoch": 0.9698058950395398,
-      "grad_norm": 0.4392610490322113,
-      "learning_rate": 7.817644347417924e-06,
-      "loss": 0.3406,
-      "step": 14839
-    },
-    {
-      "epoch": 0.969871250245082,
-      "grad_norm": 0.42378175258636475,
-      "learning_rate": 7.817355875949594e-06,
-      "loss": 0.3291,
-      "step": 14840
-    },
-    {
-      "epoch": 0.9699366054506241,
-      "grad_norm": 0.4888037145137787,
-      "learning_rate": 7.817067390740046e-06,
-      "loss": 0.4358,
-      "step": 14841
-    },
-    {
-      "epoch": 0.9700019606561663,
-      "grad_norm": 0.45393961668014526,
-      "learning_rate": 7.816778891790682e-06,
-      "loss": 0.3731,
-      "step": 14842
-    },
-    {
-      "epoch": 0.9700673158617084,
-      "grad_norm": 0.43428176641464233,
-      "learning_rate": 7.816490379102912e-06,
-      "loss": 0.3791,
-      "step": 14843
-    },
-    {
-      "epoch": 0.9701326710672505,
-      "grad_norm": 0.46117889881134033,
-      "learning_rate": 7.816201852678143e-06,
-      "loss": 0.4169,
-      "step": 14844
-    },
-    {
-      "epoch": 0.9701980262727926,
-      "grad_norm": 0.43631690740585327,
-      "learning_rate": 7.81591331251778e-06,
-      "loss": 0.3376,
-      "step": 14845
-    },
-    {
-      "epoch": 0.9702633814783348,
-      "grad_norm": 0.43752554059028625,
-      "learning_rate": 7.815624758623235e-06,
-      "loss": 0.3766,
-      "step": 14846
-    },
-    {
-      "epoch": 0.9703287366838769,
-      "grad_norm": 0.40561220049858093,
-      "learning_rate": 7.81533619099591e-06,
-      "loss": 0.3261,
-      "step": 14847
-    },
-    {
-      "epoch": 0.9703940918894189,
-      "grad_norm": 0.42520958185195923,
-      "learning_rate": 7.815047609637216e-06,
-      "loss": 0.363,
-      "step": 14848
-    },
-    {
-      "epoch": 0.9704594470949611,
-      "grad_norm": 0.44699394702911377,
-      "learning_rate": 7.814759014548561e-06,
-      "loss": 0.3687,
-      "step": 14849
-    },
-    {
-      "epoch": 0.9705248023005032,
-      "grad_norm": 0.4572710394859314,
-      "learning_rate": 7.81447040573135e-06,
-      "loss": 0.3858,
-      "step": 14850
-    },
-    {
-      "epoch": 0.9705901575060454,
-      "grad_norm": 0.41853249073028564,
-      "learning_rate": 7.814181783186992e-06,
-      "loss": 0.3659,
-      "step": 14851
-    },
-    {
-      "epoch": 0.9706555127115875,
-      "grad_norm": 0.4620044231414795,
-      "learning_rate": 7.813893146916895e-06,
-      "loss": 0.4128,
-      "step": 14852
-    },
-    {
-      "epoch": 0.9707208679171296,
-      "grad_norm": 0.4533671736717224,
-      "learning_rate": 7.813604496922465e-06,
-      "loss": 0.3396,
-      "step": 14853
-    },
-    {
-      "epoch": 0.9707862231226717,
-      "grad_norm": 0.5005420446395874,
-      "learning_rate": 7.813315833205114e-06,
-      "loss": 0.4176,
-      "step": 14854
-    },
-    {
-      "epoch": 0.9708515783282139,
-      "grad_norm": 0.46725791692733765,
-      "learning_rate": 7.813027155766244e-06,
-      "loss": 0.4292,
-      "step": 14855
-    },
-    {
-      "epoch": 0.970916933533756,
-      "grad_norm": 0.4560980200767517,
-      "learning_rate": 7.812738464607266e-06,
-      "loss": 0.4013,
-      "step": 14856
-    },
-    {
-      "epoch": 0.970982288739298,
-      "grad_norm": 0.4414271414279938,
-      "learning_rate": 7.81244975972959e-06,
-      "loss": 0.3898,
-      "step": 14857
-    },
-    {
-      "epoch": 0.9710476439448402,
-      "grad_norm": 0.5079962015151978,
-      "learning_rate": 7.81216104113462e-06,
-      "loss": 0.4037,
-      "step": 14858
-    },
-    {
-      "epoch": 0.9711129991503823,
-      "grad_norm": 0.46962466835975647,
-      "learning_rate": 7.811872308823768e-06,
-      "loss": 0.4226,
-      "step": 14859
-    },
-    {
-      "epoch": 0.9711783543559245,
-      "grad_norm": 0.4703862965106964,
-      "learning_rate": 7.81158356279844e-06,
-      "loss": 0.4363,
-      "step": 14860
-    },
-    {
-      "epoch": 0.9712437095614666,
-      "grad_norm": 0.4480913281440735,
-      "learning_rate": 7.811294803060046e-06,
-      "loss": 0.3877,
-      "step": 14861
-    },
-    {
-      "epoch": 0.9713090647670087,
-      "grad_norm": 0.4308094084262848,
-      "learning_rate": 7.811006029609993e-06,
-      "loss": 0.3454,
-      "step": 14862
-    },
-    {
-      "epoch": 0.9713744199725508,
-      "grad_norm": 0.42413821816444397,
-      "learning_rate": 7.810717242449689e-06,
-      "loss": 0.3517,
-      "step": 14863
-    },
-    {
-      "epoch": 0.971439775178093,
-      "grad_norm": 0.43644389510154724,
-      "learning_rate": 7.810428441580544e-06,
-      "loss": 0.3773,
-      "step": 14864
-    },
-    {
-      "epoch": 0.9715051303836351,
-      "grad_norm": 0.4299981892108917,
-      "learning_rate": 7.810139627003966e-06,
-      "loss": 0.3818,
-      "step": 14865
-    },
-    {
-      "epoch": 0.9715704855891771,
-      "grad_norm": 0.43086904287338257,
-      "learning_rate": 7.809850798721362e-06,
-      "loss": 0.3508,
-      "step": 14866
-    },
-    {
-      "epoch": 0.9716358407947193,
-      "grad_norm": 0.44815102219581604,
-      "learning_rate": 7.809561956734145e-06,
-      "loss": 0.3749,
-      "step": 14867
-    },
-    {
-      "epoch": 0.9717011960002614,
-      "grad_norm": 0.4685414731502533,
-      "learning_rate": 7.809273101043717e-06,
-      "loss": 0.4099,
-      "step": 14868
-    },
-    {
-      "epoch": 0.9717665512058036,
-      "grad_norm": 0.4301880896091461,
-      "learning_rate": 7.808984231651492e-06,
-      "loss": 0.3639,
-      "step": 14869
-    },
-    {
-      "epoch": 0.9718319064113456,
-      "grad_norm": 0.43557488918304443,
-      "learning_rate": 7.80869534855888e-06,
-      "loss": 0.3494,
-      "step": 14870
-    },
-    {
-      "epoch": 0.9718972616168878,
-      "grad_norm": 0.4650508463382721,
-      "learning_rate": 7.808406451767287e-06,
-      "loss": 0.4192,
-      "step": 14871
-    },
-    {
-      "epoch": 0.9719626168224299,
-      "grad_norm": 0.44013047218322754,
-      "learning_rate": 7.808117541278121e-06,
-      "loss": 0.374,
-      "step": 14872
-    },
-    {
-      "epoch": 0.972027972027972,
-      "grad_norm": 0.47921615839004517,
-      "learning_rate": 7.807828617092796e-06,
-      "loss": 0.4231,
-      "step": 14873
-    },
-    {
-      "epoch": 0.9720933272335142,
-      "grad_norm": 0.4601583778858185,
-      "learning_rate": 7.807539679212716e-06,
-      "loss": 0.4013,
-      "step": 14874
-    },
-    {
-      "epoch": 0.9721586824390562,
-      "grad_norm": 0.4489063620567322,
-      "learning_rate": 7.807250727639293e-06,
-      "loss": 0.3465,
-      "step": 14875
-    },
-    {
-      "epoch": 0.9722240376445984,
-      "grad_norm": 0.45170992612838745,
-      "learning_rate": 7.806961762373935e-06,
-      "loss": 0.3802,
-      "step": 14876
-    },
-    {
-      "epoch": 0.9722893928501405,
-      "grad_norm": 0.44086071848869324,
-      "learning_rate": 7.806672783418053e-06,
-      "loss": 0.3723,
-      "step": 14877
-    },
-    {
-      "epoch": 0.9723547480556827,
-      "grad_norm": 0.4645160734653473,
-      "learning_rate": 7.806383790773055e-06,
-      "loss": 0.3882,
-      "step": 14878
-    },
-    {
-      "epoch": 0.9724201032612247,
-      "grad_norm": 0.44723886251449585,
-      "learning_rate": 7.806094784440351e-06,
-      "loss": 0.396,
-      "step": 14879
-    },
-    {
-      "epoch": 0.9724854584667669,
-      "grad_norm": 0.4568825960159302,
-      "learning_rate": 7.80580576442135e-06,
-      "loss": 0.4182,
-      "step": 14880
-    },
-    {
-      "epoch": 0.972550813672309,
-      "grad_norm": 0.44152265787124634,
-      "learning_rate": 7.805516730717464e-06,
-      "loss": 0.3922,
-      "step": 14881
-    },
-    {
-      "epoch": 0.9726161688778511,
-      "grad_norm": 0.4171949028968811,
-      "learning_rate": 7.8052276833301e-06,
-      "loss": 0.349,
-      "step": 14882
-    },
-    {
-      "epoch": 0.9726815240833933,
-      "grad_norm": 0.4647041857242584,
-      "learning_rate": 7.804938622260669e-06,
-      "loss": 0.4251,
-      "step": 14883
-    },
-    {
-      "epoch": 0.9727468792889353,
-      "grad_norm": 0.4064752459526062,
-      "learning_rate": 7.804649547510581e-06,
-      "loss": 0.3155,
-      "step": 14884
-    },
-    {
-      "epoch": 0.9728122344944775,
-      "grad_norm": 0.40594616532325745,
-      "learning_rate": 7.804360459081244e-06,
-      "loss": 0.3373,
-      "step": 14885
-    },
-    {
-      "epoch": 0.9728775897000196,
-      "grad_norm": 0.4173528254032135,
-      "learning_rate": 7.804071356974071e-06,
-      "loss": 0.3313,
-      "step": 14886
-    },
-    {
-      "epoch": 0.9729429449055618,
-      "grad_norm": 0.44771790504455566,
-      "learning_rate": 7.803782241190469e-06,
-      "loss": 0.3663,
-      "step": 14887
-    },
-    {
-      "epoch": 0.9730083001111038,
-      "grad_norm": 0.4582705497741699,
-      "learning_rate": 7.803493111731852e-06,
-      "loss": 0.3892,
-      "step": 14888
-    },
-    {
-      "epoch": 0.973073655316646,
-      "grad_norm": 0.4479402005672455,
-      "learning_rate": 7.803203968599626e-06,
-      "loss": 0.4031,
-      "step": 14889
-    },
-    {
-      "epoch": 0.9731390105221881,
-      "grad_norm": 0.4287157356739044,
-      "learning_rate": 7.802914811795204e-06,
-      "loss": 0.3344,
-      "step": 14890
-    },
-    {
-      "epoch": 0.9732043657277302,
-      "grad_norm": 0.45062702894210815,
-      "learning_rate": 7.802625641319994e-06,
-      "loss": 0.3422,
-      "step": 14891
-    },
-    {
-      "epoch": 0.9732697209332724,
-      "grad_norm": 0.4496932029724121,
-      "learning_rate": 7.80233645717541e-06,
-      "loss": 0.3813,
-      "step": 14892
-    },
-    {
-      "epoch": 0.9733350761388144,
-      "grad_norm": 0.4372115433216095,
-      "learning_rate": 7.80204725936286e-06,
-      "loss": 0.3871,
-      "step": 14893
-    },
-    {
-      "epoch": 0.9734004313443566,
-      "grad_norm": 0.5093250274658203,
-      "learning_rate": 7.801758047883752e-06,
-      "loss": 0.4199,
-      "step": 14894
-    },
-    {
-      "epoch": 0.9734657865498987,
-      "grad_norm": 0.41162726283073425,
-      "learning_rate": 7.801468822739502e-06,
-      "loss": 0.3288,
-      "step": 14895
-    },
-    {
-      "epoch": 0.9735311417554409,
-      "grad_norm": 0.4383789598941803,
-      "learning_rate": 7.801179583931517e-06,
-      "loss": 0.3474,
-      "step": 14896
-    },
-    {
-      "epoch": 0.9735964969609829,
-      "grad_norm": 0.441730797290802,
-      "learning_rate": 7.80089033146121e-06,
-      "loss": 0.3567,
-      "step": 14897
-    },
-    {
-      "epoch": 0.973661852166525,
-      "grad_norm": 0.44002655148506165,
-      "learning_rate": 7.80060106532999e-06,
-      "loss": 0.4038,
-      "step": 14898
-    },
-    {
-      "epoch": 0.9737272073720672,
-      "grad_norm": 0.42360547184944153,
-      "learning_rate": 7.800311785539267e-06,
-      "loss": 0.3493,
-      "step": 14899
-    },
-    {
-      "epoch": 0.9737925625776093,
-      "grad_norm": 0.46605971455574036,
-      "learning_rate": 7.800022492090455e-06,
-      "loss": 0.4651,
-      "step": 14900
-    },
-    {
-      "epoch": 0.9738579177831514,
-      "grad_norm": 0.44473716616630554,
-      "learning_rate": 7.799733184984961e-06,
-      "loss": 0.3775,
-      "step": 14901
-    },
-    {
-      "epoch": 0.9739232729886935,
-      "grad_norm": 0.49451306462287903,
-      "learning_rate": 7.7994438642242e-06,
-      "loss": 0.3583,
-      "step": 14902
-    },
-    {
-      "epoch": 0.9739886281942357,
-      "grad_norm": 0.4529349207878113,
-      "learning_rate": 7.799154529809583e-06,
-      "loss": 0.3843,
-      "step": 14903
-    },
-    {
-      "epoch": 0.9740539833997778,
-      "grad_norm": 0.42561668157577515,
-      "learning_rate": 7.798865181742518e-06,
-      "loss": 0.3504,
-      "step": 14904
-    },
-    {
-      "epoch": 0.97411933860532,
-      "grad_norm": 0.4208686053752899,
-      "learning_rate": 7.798575820024418e-06,
-      "loss": 0.3495,
-      "step": 14905
-    },
-    {
-      "epoch": 0.974184693810862,
-      "grad_norm": 0.44645193219184875,
-      "learning_rate": 7.798286444656694e-06,
-      "loss": 0.3732,
-      "step": 14906
-    },
-    {
-      "epoch": 0.9742500490164041,
-      "grad_norm": 0.4234239459037781,
-      "learning_rate": 7.797997055640758e-06,
-      "loss": 0.3313,
-      "step": 14907
-    },
-    {
-      "epoch": 0.9743154042219463,
-      "grad_norm": 0.42849862575531006,
-      "learning_rate": 7.79770765297802e-06,
-      "loss": 0.3701,
-      "step": 14908
-    },
-    {
-      "epoch": 0.9743807594274884,
-      "grad_norm": 0.4528283476829529,
-      "learning_rate": 7.797418236669894e-06,
-      "loss": 0.4159,
-      "step": 14909
-    },
-    {
-      "epoch": 0.9744461146330305,
-      "grad_norm": 0.4759887158870697,
-      "learning_rate": 7.79712880671779e-06,
-      "loss": 0.4212,
-      "step": 14910
-    },
-    {
-      "epoch": 0.9745114698385726,
-      "grad_norm": 0.460155725479126,
-      "learning_rate": 7.796839363123119e-06,
-      "loss": 0.3844,
-      "step": 14911
-    },
-    {
-      "epoch": 0.9745768250441148,
-      "grad_norm": 0.4385068118572235,
-      "learning_rate": 7.796549905887293e-06,
-      "loss": 0.3679,
-      "step": 14912
-    },
-    {
-      "epoch": 0.9746421802496569,
-      "grad_norm": 0.45536938309669495,
-      "learning_rate": 7.796260435011726e-06,
-      "loss": 0.3861,
-      "step": 14913
-    },
-    {
-      "epoch": 0.9747075354551991,
-      "grad_norm": 0.4241807162761688,
-      "learning_rate": 7.795970950497826e-06,
-      "loss": 0.3312,
-      "step": 14914
-    },
-    {
-      "epoch": 0.9747728906607411,
-      "grad_norm": 0.45421335101127625,
-      "learning_rate": 7.795681452347009e-06,
-      "loss": 0.4028,
-      "step": 14915
-    },
-    {
-      "epoch": 0.9748382458662832,
-      "grad_norm": 0.40803903341293335,
-      "learning_rate": 7.795391940560684e-06,
-      "loss": 0.3306,
-      "step": 14916
-    },
-    {
-      "epoch": 0.9749036010718254,
-      "grad_norm": 0.43891775608062744,
-      "learning_rate": 7.795102415140265e-06,
-      "loss": 0.3945,
-      "step": 14917
-    },
-    {
-      "epoch": 0.9749689562773675,
-      "grad_norm": 0.41384536027908325,
-      "learning_rate": 7.794812876087161e-06,
-      "loss": 0.3119,
-      "step": 14918
-    },
-    {
-      "epoch": 0.9750343114829096,
-      "grad_norm": 0.42975956201553345,
-      "learning_rate": 7.794523323402788e-06,
-      "loss": 0.3619,
-      "step": 14919
-    },
-    {
-      "epoch": 0.9750996666884517,
-      "grad_norm": 0.4349423050880432,
-      "learning_rate": 7.794233757088558e-06,
-      "loss": 0.3588,
-      "step": 14920
-    },
-    {
-      "epoch": 0.9751650218939939,
-      "grad_norm": 0.43726763129234314,
-      "learning_rate": 7.79394417714588e-06,
-      "loss": 0.365,
-      "step": 14921
-    },
-    {
-      "epoch": 0.975230377099536,
-      "grad_norm": 0.45229989290237427,
-      "learning_rate": 7.793654583576168e-06,
-      "loss": 0.3725,
-      "step": 14922
-    },
-    {
-      "epoch": 0.975295732305078,
-      "grad_norm": 0.4289722144603729,
-      "learning_rate": 7.793364976380837e-06,
-      "loss": 0.3946,
-      "step": 14923
-    },
-    {
-      "epoch": 0.9753610875106202,
-      "grad_norm": 0.4572765827178955,
-      "learning_rate": 7.793075355561295e-06,
-      "loss": 0.3567,
-      "step": 14924
-    },
-    {
-      "epoch": 0.9754264427161623,
-      "grad_norm": 0.42398735880851746,
-      "learning_rate": 7.792785721118959e-06,
-      "loss": 0.3585,
-      "step": 14925
-    },
-    {
-      "epoch": 0.9754917979217045,
-      "grad_norm": 0.43802231550216675,
-      "learning_rate": 7.792496073055238e-06,
-      "loss": 0.3532,
-      "step": 14926
-    },
-    {
-      "epoch": 0.9755571531272466,
-      "grad_norm": 0.4520270824432373,
-      "learning_rate": 7.792206411371547e-06,
-      "loss": 0.4067,
-      "step": 14927
-    },
-    {
-      "epoch": 0.9756225083327887,
-      "grad_norm": 0.4393850266933441,
-      "learning_rate": 7.791916736069298e-06,
-      "loss": 0.3951,
-      "step": 14928
-    },
-    {
-      "epoch": 0.9756878635383308,
-      "grad_norm": 0.4602534770965576,
-      "learning_rate": 7.791627047149903e-06,
-      "loss": 0.405,
-      "step": 14929
-    },
-    {
-      "epoch": 0.975753218743873,
-      "grad_norm": 0.4499501585960388,
-      "learning_rate": 7.791337344614776e-06,
-      "loss": 0.3758,
-      "step": 14930
-    },
-    {
-      "epoch": 0.9758185739494151,
-      "grad_norm": 0.48889756202697754,
-      "learning_rate": 7.79104762846533e-06,
-      "loss": 0.3803,
-      "step": 14931
-    },
-    {
-      "epoch": 0.9758839291549571,
-      "grad_norm": 0.4424671530723572,
-      "learning_rate": 7.79075789870298e-06,
-      "loss": 0.3588,
-      "step": 14932
-    },
-    {
-      "epoch": 0.9759492843604993,
-      "grad_norm": 0.4272640347480774,
-      "learning_rate": 7.790468155329132e-06,
-      "loss": 0.3631,
-      "step": 14933
-    },
-    {
-      "epoch": 0.9760146395660414,
-      "grad_norm": 0.4260696470737457,
-      "learning_rate": 7.790178398345208e-06,
-      "loss": 0.3453,
-      "step": 14934
-    },
-    {
-      "epoch": 0.9760799947715836,
-      "grad_norm": 0.4083070456981659,
-      "learning_rate": 7.789888627752616e-06,
-      "loss": 0.3532,
-      "step": 14935
-    },
-    {
-      "epoch": 0.9761453499771257,
-      "grad_norm": 0.4304865002632141,
-      "learning_rate": 7.78959884355277e-06,
-      "loss": 0.333,
-      "step": 14936
-    },
-    {
-      "epoch": 0.9762107051826678,
-      "grad_norm": 0.4329855740070343,
-      "learning_rate": 7.789309045747085e-06,
-      "loss": 0.3387,
-      "step": 14937
-    },
-    {
-      "epoch": 0.9762760603882099,
-      "grad_norm": 0.48346036672592163,
-      "learning_rate": 7.789019234336974e-06,
-      "loss": 0.3844,
-      "step": 14938
-    },
-    {
-      "epoch": 0.9763414155937521,
-      "grad_norm": 0.4386554956436157,
-      "learning_rate": 7.78872940932385e-06,
-      "loss": 0.3475,
-      "step": 14939
-    },
-    {
-      "epoch": 0.9764067707992942,
-      "grad_norm": 0.4556921422481537,
-      "learning_rate": 7.788439570709126e-06,
-      "loss": 0.3973,
-      "step": 14940
-    },
-    {
-      "epoch": 0.9764721260048362,
-      "grad_norm": 0.4295085072517395,
-      "learning_rate": 7.788149718494215e-06,
-      "loss": 0.3681,
-      "step": 14941
-    },
-    {
-      "epoch": 0.9765374812103784,
-      "grad_norm": 0.42635348439216614,
-      "learning_rate": 7.787859852680533e-06,
-      "loss": 0.3908,
-      "step": 14942
-    },
-    {
-      "epoch": 0.9766028364159205,
-      "grad_norm": 0.42449191212654114,
-      "learning_rate": 7.787569973269493e-06,
-      "loss": 0.3439,
-      "step": 14943
-    },
-    {
-      "epoch": 0.9766681916214627,
-      "grad_norm": 0.4358311593532562,
-      "learning_rate": 7.787280080262509e-06,
-      "loss": 0.3678,
-      "step": 14944
-    },
-    {
-      "epoch": 0.9767335468270048,
-      "grad_norm": 0.4210241436958313,
-      "learning_rate": 7.786990173660993e-06,
-      "loss": 0.3371,
-      "step": 14945
-    },
-    {
-      "epoch": 0.9767989020325469,
-      "grad_norm": 0.47323113679885864,
-      "learning_rate": 7.786700253466362e-06,
-      "loss": 0.4061,
-      "step": 14946
-    },
-    {
-      "epoch": 0.976864257238089,
-      "grad_norm": 0.41978371143341064,
-      "learning_rate": 7.786410319680027e-06,
-      "loss": 0.3564,
-      "step": 14947
-    },
-    {
-      "epoch": 0.9769296124436312,
-      "grad_norm": 0.4457114338874817,
-      "learning_rate": 7.786120372303404e-06,
-      "loss": 0.3786,
-      "step": 14948
-    },
-    {
-      "epoch": 0.9769949676491733,
-      "grad_norm": 0.45904749631881714,
-      "learning_rate": 7.785830411337906e-06,
-      "loss": 0.3654,
-      "step": 14949
-    },
-    {
-      "epoch": 0.9770603228547153,
-      "grad_norm": 0.45155349373817444,
-      "learning_rate": 7.785540436784951e-06,
-      "loss": 0.3628,
-      "step": 14950
-    },
-    {
-      "epoch": 0.9771256780602575,
-      "grad_norm": 0.457391619682312,
-      "learning_rate": 7.785250448645947e-06,
-      "loss": 0.4203,
-      "step": 14951
-    },
-    {
-      "epoch": 0.9771910332657996,
-      "grad_norm": 0.4627765119075775,
-      "learning_rate": 7.784960446922313e-06,
-      "loss": 0.4565,
-      "step": 14952
-    },
-    {
-      "epoch": 0.9772563884713418,
-      "grad_norm": 0.4499048888683319,
-      "learning_rate": 7.784670431615462e-06,
-      "loss": 0.381,
-      "step": 14953
-    },
-    {
-      "epoch": 0.9773217436768838,
-      "grad_norm": 0.43460148572921753,
-      "learning_rate": 7.784380402726807e-06,
-      "loss": 0.3833,
-      "step": 14954
-    },
-    {
-      "epoch": 0.977387098882426,
-      "grad_norm": 0.42551854252815247,
-      "learning_rate": 7.784090360257766e-06,
-      "loss": 0.3546,
-      "step": 14955
-    },
-    {
-      "epoch": 0.9774524540879681,
-      "grad_norm": 0.4280862808227539,
-      "learning_rate": 7.783800304209752e-06,
-      "loss": 0.3759,
-      "step": 14956
-    },
-    {
-      "epoch": 0.9775178092935102,
-      "grad_norm": 0.42694562673568726,
-      "learning_rate": 7.783510234584179e-06,
-      "loss": 0.3735,
-      "step": 14957
-    },
-    {
-      "epoch": 0.9775831644990524,
-      "grad_norm": 0.4526824653148651,
-      "learning_rate": 7.783220151382462e-06,
-      "loss": 0.4044,
-      "step": 14958
-    },
-    {
-      "epoch": 0.9776485197045944,
-      "grad_norm": 0.4598309397697449,
-      "learning_rate": 7.782930054606017e-06,
-      "loss": 0.4073,
-      "step": 14959
-    },
-    {
-      "epoch": 0.9777138749101366,
-      "grad_norm": 0.4279153645038605,
-      "learning_rate": 7.782639944256257e-06,
-      "loss": 0.336,
-      "step": 14960
-    },
-    {
-      "epoch": 0.9777792301156787,
-      "grad_norm": 0.4534274637699127,
-      "learning_rate": 7.782349820334598e-06,
-      "loss": 0.3593,
-      "step": 14961
-    },
-    {
-      "epoch": 0.9778445853212209,
-      "grad_norm": 0.4329860210418701,
-      "learning_rate": 7.782059682842455e-06,
-      "loss": 0.3679,
-      "step": 14962
-    },
-    {
-      "epoch": 0.9779099405267629,
-      "grad_norm": 0.4311714470386505,
-      "learning_rate": 7.781769531781244e-06,
-      "loss": 0.3706,
-      "step": 14963
-    },
-    {
-      "epoch": 0.9779752957323051,
-      "grad_norm": 0.44840192794799805,
-      "learning_rate": 7.78147936715238e-06,
-      "loss": 0.377,
-      "step": 14964
-    },
-    {
-      "epoch": 0.9780406509378472,
-      "grad_norm": 0.4737732410430908,
-      "learning_rate": 7.781189188957276e-06,
-      "loss": 0.4512,
-      "step": 14965
-    },
-    {
-      "epoch": 0.9781060061433893,
-      "grad_norm": 0.4669395983219147,
-      "learning_rate": 7.780898997197348e-06,
-      "loss": 0.3887,
-      "step": 14966
-    },
-    {
-      "epoch": 0.9781713613489315,
-      "grad_norm": 0.4276529550552368,
-      "learning_rate": 7.780608791874014e-06,
-      "loss": 0.3566,
-      "step": 14967
-    },
-    {
-      "epoch": 0.9782367165544735,
-      "grad_norm": 0.4112605154514313,
-      "learning_rate": 7.780318572988688e-06,
-      "loss": 0.3341,
-      "step": 14968
-    },
-    {
-      "epoch": 0.9783020717600157,
-      "grad_norm": 0.5153815150260925,
-      "learning_rate": 7.780028340542785e-06,
-      "loss": 0.419,
-      "step": 14969
-    },
-    {
-      "epoch": 0.9783674269655578,
-      "grad_norm": 0.4861021935939789,
-      "learning_rate": 7.779738094537718e-06,
-      "loss": 0.4132,
-      "step": 14970
-    },
-    {
-      "epoch": 0.9784327821711,
-      "grad_norm": 0.4335067868232727,
-      "learning_rate": 7.779447834974909e-06,
-      "loss": 0.3842,
-      "step": 14971
-    },
-    {
-      "epoch": 0.978498137376642,
-      "grad_norm": 0.4382305443286896,
-      "learning_rate": 7.779157561855767e-06,
-      "loss": 0.3874,
-      "step": 14972
-    },
-    {
-      "epoch": 0.9785634925821842,
-      "grad_norm": 0.4398297965526581,
-      "learning_rate": 7.778867275181712e-06,
-      "loss": 0.3877,
-      "step": 14973
-    },
-    {
-      "epoch": 0.9786288477877263,
-      "grad_norm": 0.43479427695274353,
-      "learning_rate": 7.77857697495416e-06,
-      "loss": 0.3797,
-      "step": 14974
-    },
-    {
-      "epoch": 0.9786942029932684,
-      "grad_norm": 0.41617661714553833,
-      "learning_rate": 7.778286661174523e-06,
-      "loss": 0.3377,
-      "step": 14975
-    },
-    {
-      "epoch": 0.9787595581988106,
-      "grad_norm": 0.468075692653656,
-      "learning_rate": 7.777996333844219e-06,
-      "loss": 0.4461,
-      "step": 14976
-    },
-    {
-      "epoch": 0.9788249134043526,
-      "grad_norm": 0.47167110443115234,
-      "learning_rate": 7.777705992964668e-06,
-      "loss": 0.388,
-      "step": 14977
-    },
-    {
-      "epoch": 0.9788902686098948,
-      "grad_norm": 0.44846072793006897,
-      "learning_rate": 7.77741563853728e-06,
-      "loss": 0.4309,
-      "step": 14978
-    },
-    {
-      "epoch": 0.9789556238154369,
-      "grad_norm": 0.41806384921073914,
-      "learning_rate": 7.777125270563474e-06,
-      "loss": 0.3486,
-      "step": 14979
-    },
-    {
-      "epoch": 0.9790209790209791,
-      "grad_norm": 0.418844997882843,
-      "learning_rate": 7.776834889044666e-06,
-      "loss": 0.3492,
-      "step": 14980
-    },
-    {
-      "epoch": 0.9790863342265211,
-      "grad_norm": 0.43911558389663696,
-      "learning_rate": 7.776544493982274e-06,
-      "loss": 0.3571,
-      "step": 14981
-    },
-    {
-      "epoch": 0.9791516894320632,
-      "grad_norm": 0.4887353777885437,
-      "learning_rate": 7.77625408537771e-06,
-      "loss": 0.4171,
-      "step": 14982
-    },
-    {
-      "epoch": 0.9792170446376054,
-      "grad_norm": 0.3884548246860504,
-      "learning_rate": 7.775963663232395e-06,
-      "loss": 0.2972,
-      "step": 14983
-    },
-    {
-      "epoch": 0.9792823998431475,
-      "grad_norm": 0.44042345881462097,
-      "learning_rate": 7.775673227547743e-06,
-      "loss": 0.3984,
-      "step": 14984
-    },
-    {
-      "epoch": 0.9793477550486896,
-      "grad_norm": 0.43387743830680847,
-      "learning_rate": 7.775382778325171e-06,
-      "loss": 0.3768,
-      "step": 14985
-    },
-    {
-      "epoch": 0.9794131102542317,
-      "grad_norm": 0.4563215672969818,
-      "learning_rate": 7.775092315566095e-06,
-      "loss": 0.3593,
-      "step": 14986
-    },
-    {
-      "epoch": 0.9794784654597739,
-      "grad_norm": 0.4310094118118286,
-      "learning_rate": 7.774801839271933e-06,
-      "loss": 0.3451,
-      "step": 14987
-    },
-    {
-      "epoch": 0.979543820665316,
-      "grad_norm": 0.4018457233905792,
-      "learning_rate": 7.774511349444103e-06,
-      "loss": 0.3125,
-      "step": 14988
-    },
-    {
-      "epoch": 0.9796091758708582,
-      "grad_norm": 0.4388629198074341,
-      "learning_rate": 7.774220846084017e-06,
-      "loss": 0.3657,
-      "step": 14989
-    },
-    {
-      "epoch": 0.9796745310764002,
-      "grad_norm": 0.3832870125770569,
-      "learning_rate": 7.773930329193096e-06,
-      "loss": 0.2841,
-      "step": 14990
-    },
-    {
-      "epoch": 0.9797398862819423,
-      "grad_norm": 0.4359722435474396,
-      "learning_rate": 7.773639798772755e-06,
-      "loss": 0.3823,
-      "step": 14991
-    },
-    {
-      "epoch": 0.9798052414874845,
-      "grad_norm": 0.4525127410888672,
-      "learning_rate": 7.773349254824412e-06,
-      "loss": 0.4331,
-      "step": 14992
-    },
-    {
-      "epoch": 0.9798705966930266,
-      "grad_norm": 0.44532421231269836,
-      "learning_rate": 7.773058697349485e-06,
-      "loss": 0.3725,
-      "step": 14993
-    },
-    {
-      "epoch": 0.9799359518985687,
-      "grad_norm": 0.4242517948150635,
-      "learning_rate": 7.772768126349391e-06,
-      "loss": 0.3564,
-      "step": 14994
-    },
-    {
-      "epoch": 0.9800013071041108,
-      "grad_norm": 0.41339805722236633,
-      "learning_rate": 7.772477541825542e-06,
-      "loss": 0.3018,
-      "step": 14995
-    },
-    {
-      "epoch": 0.980066662309653,
-      "grad_norm": 0.40947771072387695,
-      "learning_rate": 7.772186943779365e-06,
-      "loss": 0.3315,
-      "step": 14996
-    },
-    {
-      "epoch": 0.9801320175151951,
-      "grad_norm": 0.43682199716567993,
-      "learning_rate": 7.771896332212268e-06,
-      "loss": 0.3525,
-      "step": 14997
-    },
-    {
-      "epoch": 0.9801973727207373,
-      "grad_norm": 0.4362927973270416,
-      "learning_rate": 7.771605707125673e-06,
-      "loss": 0.3632,
-      "step": 14998
-    },
-    {
-      "epoch": 0.9802627279262793,
-      "grad_norm": 0.4838549792766571,
-      "learning_rate": 7.771315068520997e-06,
-      "loss": 0.4175,
-      "step": 14999
-    },
-    {
-      "epoch": 0.9803280831318214,
-      "grad_norm": 0.4268008768558502,
-      "learning_rate": 7.771024416399658e-06,
-      "loss": 0.3336,
-      "step": 15000
-    },
-    {
-      "epoch": 0.9803934383373636,
-      "grad_norm": 0.4409523904323578,
-      "learning_rate": 7.770733750763072e-06,
-      "loss": 0.3742,
-      "step": 15001
-    },
-    {
-      "epoch": 0.9804587935429057,
-      "grad_norm": 0.4567115008831024,
-      "learning_rate": 7.770443071612658e-06,
-      "loss": 0.4437,
-      "step": 15002
-    },
-    {
-      "epoch": 0.9805241487484478,
-      "grad_norm": 0.42176324129104614,
-      "learning_rate": 7.770152378949834e-06,
-      "loss": 0.3487,
-      "step": 15003
-    },
-    {
-      "epoch": 0.9805895039539899,
-      "grad_norm": 0.4577464163303375,
-      "learning_rate": 7.769861672776018e-06,
-      "loss": 0.4139,
-      "step": 15004
-    },
-    {
-      "epoch": 0.9806548591595321,
-      "grad_norm": 0.5067284107208252,
-      "learning_rate": 7.769570953092625e-06,
-      "loss": 0.4526,
-      "step": 15005
-    },
-    {
-      "epoch": 0.9807202143650742,
-      "grad_norm": 0.45248618721961975,
-      "learning_rate": 7.769280219901077e-06,
-      "loss": 0.3726,
-      "step": 15006
-    },
-    {
-      "epoch": 0.9807855695706162,
-      "grad_norm": 0.47500354051589966,
-      "learning_rate": 7.768989473202789e-06,
-      "loss": 0.4138,
-      "step": 15007
-    },
-    {
-      "epoch": 0.9808509247761584,
-      "grad_norm": 0.4679728150367737,
-      "learning_rate": 7.768698712999179e-06,
-      "loss": 0.4271,
-      "step": 15008
-    },
-    {
-      "epoch": 0.9809162799817005,
-      "grad_norm": 0.5358428955078125,
-      "learning_rate": 7.768407939291667e-06,
-      "loss": 0.3613,
-      "step": 15009
-    },
-    {
-      "epoch": 0.9809816351872427,
-      "grad_norm": 0.42767661809921265,
-      "learning_rate": 7.768117152081672e-06,
-      "loss": 0.3778,
-      "step": 15010
-    },
-    {
-      "epoch": 0.9810469903927848,
-      "grad_norm": 0.5026191473007202,
-      "learning_rate": 7.767826351370608e-06,
-      "loss": 0.4399,
-      "step": 15011
-    },
-    {
-      "epoch": 0.9811123455983269,
-      "grad_norm": 0.5366755127906799,
-      "learning_rate": 7.767535537159896e-06,
-      "loss": 0.3971,
-      "step": 15012
-    },
-    {
-      "epoch": 0.981177700803869,
-      "grad_norm": 0.443352073431015,
-      "learning_rate": 7.767244709450955e-06,
-      "loss": 0.3624,
-      "step": 15013
-    },
-    {
-      "epoch": 0.9812430560094112,
-      "grad_norm": 0.4398881196975708,
-      "learning_rate": 7.766953868245205e-06,
-      "loss": 0.3638,
-      "step": 15014
-    },
-    {
-      "epoch": 0.9813084112149533,
-      "grad_norm": 0.4046792984008789,
-      "learning_rate": 7.766663013544062e-06,
-      "loss": 0.2867,
-      "step": 15015
-    },
-    {
-      "epoch": 0.9813737664204953,
-      "grad_norm": 0.41330233216285706,
-      "learning_rate": 7.766372145348944e-06,
-      "loss": 0.3038,
-      "step": 15016
-    },
-    {
-      "epoch": 0.9814391216260375,
-      "grad_norm": 0.4522208273410797,
-      "learning_rate": 7.76608126366127e-06,
-      "loss": 0.4102,
-      "step": 15017
-    },
-    {
-      "epoch": 0.9815044768315796,
-      "grad_norm": 0.43401139974594116,
-      "learning_rate": 7.76579036848246e-06,
-      "loss": 0.3851,
-      "step": 15018
-    },
-    {
-      "epoch": 0.9815698320371218,
-      "grad_norm": 0.44505298137664795,
-      "learning_rate": 7.765499459813932e-06,
-      "loss": 0.3695,
-      "step": 15019
-    },
-    {
-      "epoch": 0.9816351872426639,
-      "grad_norm": 0.45697569847106934,
-      "learning_rate": 7.765208537657106e-06,
-      "loss": 0.3891,
-      "step": 15020
-    },
-    {
-      "epoch": 0.981700542448206,
-      "grad_norm": 0.49235251545906067,
-      "learning_rate": 7.764917602013398e-06,
-      "loss": 0.4273,
-      "step": 15021
-    },
-    {
-      "epoch": 0.9817658976537481,
-      "grad_norm": 0.46051719784736633,
-      "learning_rate": 7.76462665288423e-06,
-      "loss": 0.3993,
-      "step": 15022
-    },
-    {
-      "epoch": 0.9818312528592903,
-      "grad_norm": 0.4594170153141022,
-      "learning_rate": 7.764335690271022e-06,
-      "loss": 0.3817,
-      "step": 15023
-    },
-    {
-      "epoch": 0.9818966080648324,
-      "grad_norm": 0.4451127350330353,
-      "learning_rate": 7.764044714175188e-06,
-      "loss": 0.3676,
-      "step": 15024
-    },
-    {
-      "epoch": 0.9819619632703744,
-      "grad_norm": 0.4703153371810913,
-      "learning_rate": 7.763753724598153e-06,
-      "loss": 0.4086,
-      "step": 15025
-    },
-    {
-      "epoch": 0.9820273184759166,
-      "grad_norm": 0.42368975281715393,
-      "learning_rate": 7.763462721541332e-06,
-      "loss": 0.3568,
-      "step": 15026
-    },
-    {
-      "epoch": 0.9820926736814587,
-      "grad_norm": 0.4411110281944275,
-      "learning_rate": 7.763171705006147e-06,
-      "loss": 0.3804,
-      "step": 15027
-    },
-    {
-      "epoch": 0.9821580288870009,
-      "grad_norm": 0.4607815742492676,
-      "learning_rate": 7.762880674994015e-06,
-      "loss": 0.3616,
-      "step": 15028
-    },
-    {
-      "epoch": 0.982223384092543,
-      "grad_norm": 0.48068633675575256,
-      "learning_rate": 7.762589631506358e-06,
-      "loss": 0.4028,
-      "step": 15029
-    },
-    {
-      "epoch": 0.9822887392980851,
-      "grad_norm": 0.4616956114768982,
-      "learning_rate": 7.762298574544594e-06,
-      "loss": 0.3516,
-      "step": 15030
-    },
-    {
-      "epoch": 0.9823540945036272,
-      "grad_norm": 0.42269325256347656,
-      "learning_rate": 7.762007504110143e-06,
-      "loss": 0.3247,
-      "step": 15031
-    },
-    {
-      "epoch": 0.9824194497091694,
-      "grad_norm": 0.4312247037887573,
-      "learning_rate": 7.761716420204423e-06,
-      "loss": 0.3971,
-      "step": 15032
-    },
-    {
-      "epoch": 0.9824848049147115,
-      "grad_norm": 0.4473051428794861,
-      "learning_rate": 7.761425322828859e-06,
-      "loss": 0.3676,
-      "step": 15033
-    },
-    {
-      "epoch": 0.9825501601202535,
-      "grad_norm": 0.4145049750804901,
-      "learning_rate": 7.761134211984864e-06,
-      "loss": 0.3693,
-      "step": 15034
-    },
-    {
-      "epoch": 0.9826155153257957,
-      "grad_norm": 0.4665650427341461,
-      "learning_rate": 7.760843087673861e-06,
-      "loss": 0.397,
-      "step": 15035
-    },
-    {
-      "epoch": 0.9826808705313378,
-      "grad_norm": 0.41736987233161926,
-      "learning_rate": 7.76055194989727e-06,
-      "loss": 0.3444,
-      "step": 15036
-    },
-    {
-      "epoch": 0.98274622573688,
-      "grad_norm": 0.4125063121318817,
-      "learning_rate": 7.760260798656512e-06,
-      "loss": 0.3401,
-      "step": 15037
-    },
-    {
-      "epoch": 0.982811580942422,
-      "grad_norm": 0.43488919734954834,
-      "learning_rate": 7.759969633953006e-06,
-      "loss": 0.4122,
-      "step": 15038
-    },
-    {
-      "epoch": 0.9828769361479642,
-      "grad_norm": 0.43481025099754333,
-      "learning_rate": 7.759678455788169e-06,
-      "loss": 0.3605,
-      "step": 15039
-    },
-    {
-      "epoch": 0.9829422913535063,
-      "grad_norm": 0.44786256551742554,
-      "learning_rate": 7.759387264163427e-06,
-      "loss": 0.4014,
-      "step": 15040
-    },
-    {
-      "epoch": 0.9830076465590484,
-      "grad_norm": 0.44559410214424133,
-      "learning_rate": 7.759096059080196e-06,
-      "loss": 0.4099,
-      "step": 15041
-    },
-    {
-      "epoch": 0.9830730017645906,
-      "grad_norm": 0.4063272476196289,
-      "learning_rate": 7.7588048405399e-06,
-      "loss": 0.3129,
-      "step": 15042
-    },
-    {
-      "epoch": 0.9831383569701326,
-      "grad_norm": 0.44426804780960083,
-      "learning_rate": 7.758513608543954e-06,
-      "loss": 0.3615,
-      "step": 15043
-    },
-    {
-      "epoch": 0.9832037121756748,
-      "grad_norm": 0.4609695076942444,
-      "learning_rate": 7.758222363093783e-06,
-      "loss": 0.4033,
-      "step": 15044
-    },
-    {
-      "epoch": 0.9832690673812169,
-      "grad_norm": 0.4464883804321289,
-      "learning_rate": 7.757931104190806e-06,
-      "loss": 0.38,
-      "step": 15045
-    },
-    {
-      "epoch": 0.9833344225867591,
-      "grad_norm": 0.4010449945926666,
-      "learning_rate": 7.757639831836443e-06,
-      "loss": 0.3287,
-      "step": 15046
-    },
-    {
-      "epoch": 0.9833997777923011,
-      "grad_norm": 0.4331609904766083,
-      "learning_rate": 7.757348546032114e-06,
-      "loss": 0.4166,
-      "step": 15047
-    },
-    {
-      "epoch": 0.9834651329978433,
-      "grad_norm": 0.4953961670398712,
-      "learning_rate": 7.757057246779242e-06,
-      "loss": 0.4914,
-      "step": 15048
-    },
-    {
-      "epoch": 0.9835304882033854,
-      "grad_norm": 0.44227883219718933,
-      "learning_rate": 7.75676593407925e-06,
-      "loss": 0.4118,
-      "step": 15049
-    },
-    {
-      "epoch": 0.9835958434089275,
-      "grad_norm": 0.4692818820476532,
-      "learning_rate": 7.756474607933552e-06,
-      "loss": 0.4065,
-      "step": 15050
-    },
-    {
-      "epoch": 0.9836611986144697,
-      "grad_norm": 0.4538833200931549,
-      "learning_rate": 7.756183268343574e-06,
-      "loss": 0.3836,
-      "step": 15051
-    },
-    {
-      "epoch": 0.9837265538200117,
-      "grad_norm": 0.45548778772354126,
-      "learning_rate": 7.755891915310733e-06,
-      "loss": 0.3752,
-      "step": 15052
-    },
-    {
-      "epoch": 0.9837919090255539,
-      "grad_norm": 0.44198480248451233,
-      "learning_rate": 7.755600548836454e-06,
-      "loss": 0.3788,
-      "step": 15053
-    },
-    {
-      "epoch": 0.983857264231096,
-      "grad_norm": 0.45371824502944946,
-      "learning_rate": 7.755309168922156e-06,
-      "loss": 0.4105,
-      "step": 15054
-    },
-    {
-      "epoch": 0.9839226194366382,
-      "grad_norm": 0.47075727581977844,
-      "learning_rate": 7.75501777556926e-06,
-      "loss": 0.4289,
-      "step": 15055
-    },
-    {
-      "epoch": 0.9839879746421802,
-      "grad_norm": 0.4131263792514801,
-      "learning_rate": 7.75472636877919e-06,
-      "loss": 0.3411,
-      "step": 15056
-    },
-    {
-      "epoch": 0.9840533298477224,
-      "grad_norm": 0.4250119626522064,
-      "learning_rate": 7.754434948553364e-06,
-      "loss": 0.3678,
-      "step": 15057
-    },
-    {
-      "epoch": 0.9841186850532645,
-      "grad_norm": 0.6285251975059509,
-      "learning_rate": 7.754143514893204e-06,
-      "loss": 0.4136,
-      "step": 15058
-    },
-    {
-      "epoch": 0.9841840402588066,
-      "grad_norm": 0.424927294254303,
-      "learning_rate": 7.753852067800131e-06,
-      "loss": 0.3615,
-      "step": 15059
-    },
-    {
-      "epoch": 0.9842493954643488,
-      "grad_norm": 0.4634118378162384,
-      "learning_rate": 7.75356060727557e-06,
-      "loss": 0.3876,
-      "step": 15060
-    },
-    {
-      "epoch": 0.9843147506698908,
-      "grad_norm": 0.4112243950366974,
-      "learning_rate": 7.75326913332094e-06,
-      "loss": 0.3078,
-      "step": 15061
-    },
-    {
-      "epoch": 0.984380105875433,
-      "grad_norm": 0.4566067159175873,
-      "learning_rate": 7.75297764593766e-06,
-      "loss": 0.3923,
-      "step": 15062
-    },
-    {
-      "epoch": 0.9844454610809751,
-      "grad_norm": 0.45016443729400635,
-      "learning_rate": 7.752686145127157e-06,
-      "loss": 0.3537,
-      "step": 15063
-    },
-    {
-      "epoch": 0.9845108162865173,
-      "grad_norm": 0.4299840033054352,
-      "learning_rate": 7.752394630890846e-06,
-      "loss": 0.3854,
-      "step": 15064
-    },
-    {
-      "epoch": 0.9845761714920593,
-      "grad_norm": 0.4266165792942047,
-      "learning_rate": 7.752103103230158e-06,
-      "loss": 0.3752,
-      "step": 15065
-    },
-    {
-      "epoch": 0.9846415266976014,
-      "grad_norm": 0.4175054728984833,
-      "learning_rate": 7.751811562146506e-06,
-      "loss": 0.3811,
-      "step": 15066
-    },
-    {
-      "epoch": 0.9847068819031436,
-      "grad_norm": 0.4678158164024353,
-      "learning_rate": 7.751520007641318e-06,
-      "loss": 0.3762,
-      "step": 15067
-    },
-    {
-      "epoch": 0.9847722371086857,
-      "grad_norm": 0.4228020906448364,
-      "learning_rate": 7.751228439716012e-06,
-      "loss": 0.3345,
-      "step": 15068
-    },
-    {
-      "epoch": 0.9848375923142278,
-      "grad_norm": 0.4379521906375885,
-      "learning_rate": 7.750936858372014e-06,
-      "loss": 0.3681,
-      "step": 15069
-    },
-    {
-      "epoch": 0.9849029475197699,
-      "grad_norm": 0.45895805954933167,
-      "learning_rate": 7.750645263610742e-06,
-      "loss": 0.3839,
-      "step": 15070
-    },
-    {
-      "epoch": 0.9849683027253121,
-      "grad_norm": 0.4747651219367981,
-      "learning_rate": 7.750353655433621e-06,
-      "loss": 0.3837,
-      "step": 15071
-    },
-    {
-      "epoch": 0.9850336579308542,
-      "grad_norm": 0.4802089035511017,
-      "learning_rate": 7.750062033842071e-06,
-      "loss": 0.3606,
-      "step": 15072
-    },
-    {
-      "epoch": 0.9850990131363964,
-      "grad_norm": 0.4471137225627899,
-      "learning_rate": 7.749770398837518e-06,
-      "loss": 0.3585,
-      "step": 15073
-    },
-    {
-      "epoch": 0.9851643683419384,
-      "grad_norm": 0.4187166094779968,
-      "learning_rate": 7.749478750421381e-06,
-      "loss": 0.3396,
-      "step": 15074
-    },
-    {
-      "epoch": 0.9852297235474805,
-      "grad_norm": 0.4761616587638855,
-      "learning_rate": 7.749187088595084e-06,
-      "loss": 0.4656,
-      "step": 15075
-    },
-    {
-      "epoch": 0.9852950787530227,
-      "grad_norm": 0.45303037762641907,
-      "learning_rate": 7.748895413360048e-06,
-      "loss": 0.4055,
-      "step": 15076
-    },
-    {
-      "epoch": 0.9853604339585648,
-      "grad_norm": 0.47813448309898376,
-      "learning_rate": 7.748603724717699e-06,
-      "loss": 0.4264,
-      "step": 15077
-    },
-    {
-      "epoch": 0.985425789164107,
-      "grad_norm": 0.4306236803531647,
-      "learning_rate": 7.748312022669454e-06,
-      "loss": 0.3509,
-      "step": 15078
-    },
-    {
-      "epoch": 0.985491144369649,
-      "grad_norm": 0.4116508960723877,
-      "learning_rate": 7.748020307216742e-06,
-      "loss": 0.3041,
-      "step": 15079
-    },
-    {
-      "epoch": 0.9855564995751912,
-      "grad_norm": 0.44778963923454285,
-      "learning_rate": 7.747728578360981e-06,
-      "loss": 0.38,
-      "step": 15080
-    },
-    {
-      "epoch": 0.9856218547807333,
-      "grad_norm": 0.44965118169784546,
-      "learning_rate": 7.747436836103598e-06,
-      "loss": 0.3644,
-      "step": 15081
-    },
-    {
-      "epoch": 0.9856872099862755,
-      "grad_norm": 0.43906593322753906,
-      "learning_rate": 7.747145080446013e-06,
-      "loss": 0.3662,
-      "step": 15082
-    },
-    {
-      "epoch": 0.9857525651918175,
-      "grad_norm": 0.4133242070674896,
-      "learning_rate": 7.746853311389649e-06,
-      "loss": 0.3374,
-      "step": 15083
-    },
-    {
-      "epoch": 0.9858179203973596,
-      "grad_norm": 0.4331609904766083,
-      "learning_rate": 7.746561528935929e-06,
-      "loss": 0.3654,
-      "step": 15084
-    },
-    {
-      "epoch": 0.9858832756029018,
-      "grad_norm": 0.4636089503765106,
-      "learning_rate": 7.746269733086278e-06,
-      "loss": 0.4293,
-      "step": 15085
-    },
-    {
-      "epoch": 0.9859486308084439,
-      "grad_norm": 0.4301362931728363,
-      "learning_rate": 7.745977923842119e-06,
-      "loss": 0.3659,
-      "step": 15086
-    },
-    {
-      "epoch": 0.986013986013986,
-      "grad_norm": 0.43521547317504883,
-      "learning_rate": 7.745686101204872e-06,
-      "loss": 0.3904,
-      "step": 15087
-    },
-    {
-      "epoch": 0.9860793412195281,
-      "grad_norm": 0.4536517560482025,
-      "learning_rate": 7.745394265175965e-06,
-      "loss": 0.3839,
-      "step": 15088
-    },
-    {
-      "epoch": 0.9861446964250703,
-      "grad_norm": 0.43011221289634705,
-      "learning_rate": 7.745102415756819e-06,
-      "loss": 0.3868,
-      "step": 15089
-    },
-    {
-      "epoch": 0.9862100516306124,
-      "grad_norm": 0.41555312275886536,
-      "learning_rate": 7.744810552948856e-06,
-      "loss": 0.3491,
-      "step": 15090
-    },
-    {
-      "epoch": 0.9862754068361544,
-      "grad_norm": 0.5033795833587646,
-      "learning_rate": 7.744518676753503e-06,
-      "loss": 0.4098,
-      "step": 15091
-    },
-    {
-      "epoch": 0.9863407620416966,
-      "grad_norm": 0.43643197417259216,
-      "learning_rate": 7.744226787172179e-06,
-      "loss": 0.3489,
-      "step": 15092
-    },
-    {
-      "epoch": 0.9864061172472387,
-      "grad_norm": 0.46359705924987793,
-      "learning_rate": 7.743934884206313e-06,
-      "loss": 0.4206,
-      "step": 15093
-    },
-    {
-      "epoch": 0.9864714724527809,
-      "grad_norm": 0.4805338680744171,
-      "learning_rate": 7.743642967857325e-06,
-      "loss": 0.4336,
-      "step": 15094
-    },
-    {
-      "epoch": 0.986536827658323,
-      "grad_norm": 0.47086626291275024,
-      "learning_rate": 7.743351038126639e-06,
-      "loss": 0.4097,
-      "step": 15095
-    },
-    {
-      "epoch": 0.9866021828638651,
-      "grad_norm": 0.4841609597206116,
-      "learning_rate": 7.74305909501568e-06,
-      "loss": 0.4055,
-      "step": 15096
-    },
-    {
-      "epoch": 0.9866675380694072,
-      "grad_norm": 0.4449782073497772,
-      "learning_rate": 7.742767138525872e-06,
-      "loss": 0.3974,
-      "step": 15097
-    },
-    {
-      "epoch": 0.9867328932749494,
-      "grad_norm": 0.4355780780315399,
-      "learning_rate": 7.742475168658638e-06,
-      "loss": 0.3612,
-      "step": 15098
-    },
-    {
-      "epoch": 0.9867982484804915,
-      "grad_norm": 0.4568578898906708,
-      "learning_rate": 7.742183185415402e-06,
-      "loss": 0.3879,
-      "step": 15099
-    },
-    {
-      "epoch": 0.9868636036860335,
-      "grad_norm": 0.4183160960674286,
-      "learning_rate": 7.74189118879759e-06,
-      "loss": 0.3242,
-      "step": 15100
-    },
-    {
-      "epoch": 0.9869289588915757,
-      "grad_norm": 0.4333636164665222,
-      "learning_rate": 7.741599178806625e-06,
-      "loss": 0.3584,
-      "step": 15101
-    },
-    {
-      "epoch": 0.9869943140971178,
-      "grad_norm": 0.46936675906181335,
-      "learning_rate": 7.74130715544393e-06,
-      "loss": 0.4096,
-      "step": 15102
-    },
-    {
-      "epoch": 0.98705966930266,
-      "grad_norm": 0.4286305606365204,
-      "learning_rate": 7.74101511871093e-06,
-      "loss": 0.3366,
-      "step": 15103
-    },
-    {
-      "epoch": 0.987125024508202,
-      "grad_norm": 0.4025880992412567,
-      "learning_rate": 7.740723068609049e-06,
-      "loss": 0.3235,
-      "step": 15104
-    },
-    {
-      "epoch": 0.9871903797137442,
-      "grad_norm": 0.4574611186981201,
-      "learning_rate": 7.740431005139712e-06,
-      "loss": 0.3791,
-      "step": 15105
-    },
-    {
-      "epoch": 0.9872557349192863,
-      "grad_norm": 0.44784244894981384,
-      "learning_rate": 7.740138928304345e-06,
-      "loss": 0.3748,
-      "step": 15106
-    },
-    {
-      "epoch": 0.9873210901248285,
-      "grad_norm": 0.5053922533988953,
-      "learning_rate": 7.739846838104372e-06,
-      "loss": 0.4492,
-      "step": 15107
-    },
-    {
-      "epoch": 0.9873864453303706,
-      "grad_norm": 0.4396847188472748,
-      "learning_rate": 7.739554734541216e-06,
-      "loss": 0.3866,
-      "step": 15108
-    },
-    {
-      "epoch": 0.9874518005359126,
-      "grad_norm": 0.4559018909931183,
-      "learning_rate": 7.739262617616303e-06,
-      "loss": 0.4325,
-      "step": 15109
-    },
-    {
-      "epoch": 0.9875171557414548,
-      "grad_norm": 0.41015422344207764,
-      "learning_rate": 7.738970487331056e-06,
-      "loss": 0.3572,
-      "step": 15110
-    },
-    {
-      "epoch": 0.9875825109469969,
-      "grad_norm": 0.43143847584724426,
-      "learning_rate": 7.7386783436869e-06,
-      "loss": 0.364,
-      "step": 15111
-    },
-    {
-      "epoch": 0.9876478661525391,
-      "grad_norm": 0.41663238406181335,
-      "learning_rate": 7.738386186685262e-06,
-      "loss": 0.3503,
-      "step": 15112
-    },
-    {
-      "epoch": 0.9877132213580812,
-      "grad_norm": 0.4315636157989502,
-      "learning_rate": 7.738094016327568e-06,
-      "loss": 0.3623,
-      "step": 15113
-    },
-    {
-      "epoch": 0.9877785765636233,
-      "grad_norm": 0.43158262968063354,
-      "learning_rate": 7.737801832615239e-06,
-      "loss": 0.3761,
-      "step": 15114
-    },
-    {
-      "epoch": 0.9878439317691654,
-      "grad_norm": 0.43610435724258423,
-      "learning_rate": 7.737509635549703e-06,
-      "loss": 0.369,
-      "step": 15115
-    },
-    {
-      "epoch": 0.9879092869747076,
-      "grad_norm": 0.42928194999694824,
-      "learning_rate": 7.737217425132385e-06,
-      "loss": 0.3565,
-      "step": 15116
-    },
-    {
-      "epoch": 0.9879746421802497,
-      "grad_norm": 0.42556947469711304,
-      "learning_rate": 7.736925201364706e-06,
-      "loss": 0.3466,
-      "step": 15117
-    },
-    {
-      "epoch": 0.9880399973857917,
-      "grad_norm": 0.42862996459007263,
-      "learning_rate": 7.736632964248096e-06,
-      "loss": 0.3509,
-      "step": 15118
-    },
-    {
-      "epoch": 0.9881053525913339,
-      "grad_norm": 0.4334465265274048,
-      "learning_rate": 7.73634071378398e-06,
-      "loss": 0.3704,
-      "step": 15119
-    },
-    {
-      "epoch": 0.988170707796876,
-      "grad_norm": 0.45901745557785034,
-      "learning_rate": 7.736048449973781e-06,
-      "loss": 0.389,
-      "step": 15120
-    },
-    {
-      "epoch": 0.9882360630024182,
-      "grad_norm": 0.44808298349380493,
-      "learning_rate": 7.735756172818927e-06,
-      "loss": 0.4065,
-      "step": 15121
-    },
-    {
-      "epoch": 0.9883014182079602,
-      "grad_norm": 0.48540815711021423,
-      "learning_rate": 7.735463882320842e-06,
-      "loss": 0.4113,
-      "step": 15122
-    },
-    {
-      "epoch": 0.9883667734135024,
-      "grad_norm": 0.45461204648017883,
-      "learning_rate": 7.735171578480952e-06,
-      "loss": 0.4028,
-      "step": 15123
-    },
-    {
-      "epoch": 0.9884321286190445,
-      "grad_norm": 0.42914292216300964,
-      "learning_rate": 7.734879261300683e-06,
-      "loss": 0.3713,
-      "step": 15124
-    },
-    {
-      "epoch": 0.9884974838245866,
-      "grad_norm": 0.48916131258010864,
-      "learning_rate": 7.73458693078146e-06,
-      "loss": 0.4628,
-      "step": 15125
-    },
-    {
-      "epoch": 0.9885628390301288,
-      "grad_norm": 0.4181376099586487,
-      "learning_rate": 7.73429458692471e-06,
-      "loss": 0.3482,
-      "step": 15126
-    },
-    {
-      "epoch": 0.9886281942356708,
-      "grad_norm": 0.4241492450237274,
-      "learning_rate": 7.734002229731855e-06,
-      "loss": 0.3357,
-      "step": 15127
-    },
-    {
-      "epoch": 0.988693549441213,
-      "grad_norm": 0.4653661847114563,
-      "learning_rate": 7.733709859204328e-06,
-      "loss": 0.4316,
-      "step": 15128
-    },
-    {
-      "epoch": 0.9887589046467551,
-      "grad_norm": 0.4530976414680481,
-      "learning_rate": 7.73341747534355e-06,
-      "loss": 0.4004,
-      "step": 15129
-    },
-    {
-      "epoch": 0.9888242598522973,
-      "grad_norm": 0.44103583693504333,
-      "learning_rate": 7.733125078150947e-06,
-      "loss": 0.3703,
-      "step": 15130
-    },
-    {
-      "epoch": 0.9888896150578393,
-      "grad_norm": 0.3934401273727417,
-      "learning_rate": 7.732832667627946e-06,
-      "loss": 0.2976,
-      "step": 15131
-    },
-    {
-      "epoch": 0.9889549702633815,
-      "grad_norm": 0.4212241470813751,
-      "learning_rate": 7.732540243775972e-06,
-      "loss": 0.3617,
-      "step": 15132
-    },
-    {
-      "epoch": 0.9890203254689236,
-      "grad_norm": 0.4378197193145752,
-      "learning_rate": 7.732247806596455e-06,
-      "loss": 0.3672,
-      "step": 15133
-    },
-    {
-      "epoch": 0.9890856806744657,
-      "grad_norm": 0.45740073919296265,
-      "learning_rate": 7.731955356090818e-06,
-      "loss": 0.36,
-      "step": 15134
-    },
-    {
-      "epoch": 0.9891510358800079,
-      "grad_norm": 0.4653393030166626,
-      "learning_rate": 7.73166289226049e-06,
-      "loss": 0.3825,
-      "step": 15135
-    },
-    {
-      "epoch": 0.9892163910855499,
-      "grad_norm": 0.42049020528793335,
-      "learning_rate": 7.731370415106893e-06,
-      "loss": 0.3627,
-      "step": 15136
-    },
-    {
-      "epoch": 0.9892817462910921,
-      "grad_norm": 0.4357064664363861,
-      "learning_rate": 7.731077924631458e-06,
-      "loss": 0.3699,
-      "step": 15137
-    },
-    {
-      "epoch": 0.9893471014966342,
-      "grad_norm": 0.4246349334716797,
-      "learning_rate": 7.73078542083561e-06,
-      "loss": 0.3611,
-      "step": 15138
-    },
-    {
-      "epoch": 0.9894124567021764,
-      "grad_norm": 0.44023597240448,
-      "learning_rate": 7.730492903720774e-06,
-      "loss": 0.3727,
-      "step": 15139
-    },
-    {
-      "epoch": 0.9894778119077184,
-      "grad_norm": 0.43610572814941406,
-      "learning_rate": 7.73020037328838e-06,
-      "loss": 0.3794,
-      "step": 15140
-    },
-    {
-      "epoch": 0.9895431671132606,
-      "grad_norm": 0.46135908365249634,
-      "learning_rate": 7.729907829539851e-06,
-      "loss": 0.4275,
-      "step": 15141
-    },
-    {
-      "epoch": 0.9896085223188027,
-      "grad_norm": 0.4838557243347168,
-      "learning_rate": 7.729615272476617e-06,
-      "loss": 0.4404,
-      "step": 15142
-    },
-    {
-      "epoch": 0.9896738775243448,
-      "grad_norm": 0.4211517870426178,
-      "learning_rate": 7.729322702100103e-06,
-      "loss": 0.3532,
-      "step": 15143
-    },
-    {
-      "epoch": 0.989739232729887,
-      "grad_norm": 0.4415057897567749,
-      "learning_rate": 7.729030118411737e-06,
-      "loss": 0.3632,
-      "step": 15144
-    },
-    {
-      "epoch": 0.989804587935429,
-      "grad_norm": 0.43831971287727356,
-      "learning_rate": 7.728737521412946e-06,
-      "loss": 0.3539,
-      "step": 15145
-    },
-    {
-      "epoch": 0.9898699431409712,
-      "grad_norm": 0.4630366265773773,
-      "learning_rate": 7.728444911105158e-06,
-      "loss": 0.433,
-      "step": 15146
-    },
-    {
-      "epoch": 0.9899352983465133,
-      "grad_norm": 0.44352343678474426,
-      "learning_rate": 7.728152287489796e-06,
-      "loss": 0.3744,
-      "step": 15147
-    },
-    {
-      "epoch": 0.9900006535520555,
-      "grad_norm": 0.4524170160293579,
-      "learning_rate": 7.727859650568292e-06,
-      "loss": 0.3961,
-      "step": 15148
-    },
-    {
-      "epoch": 0.9900660087575975,
-      "grad_norm": 0.43163421750068665,
-      "learning_rate": 7.727567000342071e-06,
-      "loss": 0.3605,
-      "step": 15149
-    },
-    {
-      "epoch": 0.9901313639631396,
-      "grad_norm": 0.45648854970932007,
-      "learning_rate": 7.72727433681256e-06,
-      "loss": 0.4121,
-      "step": 15150
-    },
-    {
-      "epoch": 0.9901967191686818,
-      "grad_norm": 0.46180275082588196,
-      "learning_rate": 7.726981659981188e-06,
-      "loss": 0.4064,
-      "step": 15151
-    },
-    {
-      "epoch": 0.9902620743742239,
-      "grad_norm": 0.40367501974105835,
-      "learning_rate": 7.726688969849383e-06,
-      "loss": 0.3111,
-      "step": 15152
-    },
-    {
-      "epoch": 0.990327429579766,
-      "grad_norm": 0.46861517429351807,
-      "learning_rate": 7.72639626641857e-06,
-      "loss": 0.3952,
-      "step": 15153
-    },
-    {
-      "epoch": 0.9903927847853081,
-      "grad_norm": 0.4397759437561035,
-      "learning_rate": 7.726103549690178e-06,
-      "loss": 0.4075,
-      "step": 15154
-    },
-    {
-      "epoch": 0.9904581399908503,
-      "grad_norm": 0.4731493890285492,
-      "learning_rate": 7.725810819665635e-06,
-      "loss": 0.4191,
-      "step": 15155
-    },
-    {
-      "epoch": 0.9905234951963924,
-      "grad_norm": 0.43350693583488464,
-      "learning_rate": 7.725518076346368e-06,
-      "loss": 0.3566,
-      "step": 15156
-    },
-    {
-      "epoch": 0.9905888504019346,
-      "grad_norm": 0.45391082763671875,
-      "learning_rate": 7.725225319733806e-06,
-      "loss": 0.3863,
-      "step": 15157
-    },
-    {
-      "epoch": 0.9906542056074766,
-      "grad_norm": 0.4475744664669037,
-      "learning_rate": 7.724932549829373e-06,
-      "loss": 0.3566,
-      "step": 15158
-    },
-    {
-      "epoch": 0.9907195608130187,
-      "grad_norm": 0.4379841983318329,
-      "learning_rate": 7.724639766634503e-06,
-      "loss": 0.3772,
-      "step": 15159
-    },
-    {
-      "epoch": 0.9907849160185609,
-      "grad_norm": 0.4198358952999115,
-      "learning_rate": 7.72434697015062e-06,
-      "loss": 0.3373,
-      "step": 15160
-    },
-    {
-      "epoch": 0.990850271224103,
-      "grad_norm": 0.4721154272556305,
-      "learning_rate": 7.724054160379153e-06,
-      "loss": 0.3497,
-      "step": 15161
-    },
-    {
-      "epoch": 0.9909156264296451,
-      "grad_norm": 0.5004555583000183,
-      "learning_rate": 7.72376133732153e-06,
-      "loss": 0.3997,
-      "step": 15162
-    },
-    {
-      "epoch": 0.9909809816351872,
-      "grad_norm": 0.42507079243659973,
-      "learning_rate": 7.72346850097918e-06,
-      "loss": 0.3492,
-      "step": 15163
-    },
-    {
-      "epoch": 0.9910463368407294,
-      "grad_norm": 0.4355694353580475,
-      "learning_rate": 7.72317565135353e-06,
-      "loss": 0.3697,
-      "step": 15164
-    },
-    {
-      "epoch": 0.9911116920462715,
-      "grad_norm": 0.4397289752960205,
-      "learning_rate": 7.722882788446009e-06,
-      "loss": 0.3895,
-      "step": 15165
-    },
-    {
-      "epoch": 0.9911770472518137,
-      "grad_norm": 0.4112222492694855,
-      "learning_rate": 7.722589912258045e-06,
-      "loss": 0.324,
-      "step": 15166
-    },
-    {
-      "epoch": 0.9912424024573557,
-      "grad_norm": 0.4551739990711212,
-      "learning_rate": 7.722297022791067e-06,
-      "loss": 0.4271,
-      "step": 15167
-    },
-    {
-      "epoch": 0.9913077576628978,
-      "grad_norm": 0.42867451906204224,
-      "learning_rate": 7.722004120046504e-06,
-      "loss": 0.3599,
-      "step": 15168
-    },
-    {
-      "epoch": 0.99137311286844,
-      "grad_norm": 0.4233817160129547,
-      "learning_rate": 7.721711204025784e-06,
-      "loss": 0.3715,
-      "step": 15169
-    },
-    {
-      "epoch": 0.9914384680739821,
-      "grad_norm": 0.4558604061603546,
-      "learning_rate": 7.721418274730335e-06,
-      "loss": 0.3593,
-      "step": 15170
-    },
-    {
-      "epoch": 0.9915038232795242,
-      "grad_norm": 0.455513596534729,
-      "learning_rate": 7.721125332161585e-06,
-      "loss": 0.4041,
-      "step": 15171
-    },
-    {
-      "epoch": 0.9915691784850663,
-      "grad_norm": 0.4515846371650696,
-      "learning_rate": 7.720832376320967e-06,
-      "loss": 0.4205,
-      "step": 15172
-    },
-    {
-      "epoch": 0.9916345336906085,
-      "grad_norm": 0.43746036291122437,
-      "learning_rate": 7.720539407209905e-06,
-      "loss": 0.3791,
-      "step": 15173
-    },
-    {
-      "epoch": 0.9916998888961506,
-      "grad_norm": 0.4344053268432617,
-      "learning_rate": 7.720246424829829e-06,
-      "loss": 0.3499,
-      "step": 15174
-    },
-    {
-      "epoch": 0.9917652441016926,
-      "grad_norm": 0.4476267397403717,
-      "learning_rate": 7.71995342918217e-06,
-      "loss": 0.4105,
-      "step": 15175
-    },
-    {
-      "epoch": 0.9918305993072348,
-      "grad_norm": 0.41979485750198364,
-      "learning_rate": 7.719660420268359e-06,
-      "loss": 0.3365,
-      "step": 15176
-    },
-    {
-      "epoch": 0.9918959545127769,
-      "grad_norm": 0.46881213784217834,
-      "learning_rate": 7.719367398089816e-06,
-      "loss": 0.4182,
-      "step": 15177
-    },
-    {
-      "epoch": 0.9919613097183191,
-      "grad_norm": 0.4455852210521698,
-      "learning_rate": 7.71907436264798e-06,
-      "loss": 0.3764,
-      "step": 15178
-    },
-    {
-      "epoch": 0.9920266649238612,
-      "grad_norm": 0.43420225381851196,
-      "learning_rate": 7.718781313944274e-06,
-      "loss": 0.389,
-      "step": 15179
-    },
-    {
-      "epoch": 0.9920920201294033,
-      "grad_norm": 0.4379095435142517,
-      "learning_rate": 7.718488251980131e-06,
-      "loss": 0.3902,
-      "step": 15180
-    },
-    {
-      "epoch": 0.9921573753349454,
-      "grad_norm": 0.40903136134147644,
-      "learning_rate": 7.71819517675698e-06,
-      "loss": 0.2969,
-      "step": 15181
-    },
-    {
-      "epoch": 0.9922227305404876,
-      "grad_norm": 0.4408978819847107,
-      "learning_rate": 7.717902088276247e-06,
-      "loss": 0.395,
-      "step": 15182
-    },
-    {
-      "epoch": 0.9922880857460297,
-      "grad_norm": 0.4464319944381714,
-      "learning_rate": 7.717608986539366e-06,
-      "loss": 0.4024,
-      "step": 15183
-    },
-    {
-      "epoch": 0.9923534409515717,
-      "grad_norm": 0.48024967312812805,
-      "learning_rate": 7.717315871547764e-06,
-      "loss": 0.4134,
-      "step": 15184
-    },
-    {
-      "epoch": 0.9924187961571139,
-      "grad_norm": 0.43580445647239685,
-      "learning_rate": 7.717022743302871e-06,
-      "loss": 0.372,
-      "step": 15185
-    },
-    {
-      "epoch": 0.992484151362656,
-      "grad_norm": 0.46773090958595276,
-      "learning_rate": 7.716729601806117e-06,
-      "loss": 0.3923,
-      "step": 15186
-    },
-    {
-      "epoch": 0.9925495065681982,
-      "grad_norm": 0.4395577609539032,
-      "learning_rate": 7.71643644705893e-06,
-      "loss": 0.388,
-      "step": 15187
-    },
-    {
-      "epoch": 0.9926148617737403,
-      "grad_norm": 0.4634782671928406,
-      "learning_rate": 7.716143279062743e-06,
-      "loss": 0.4041,
-      "step": 15188
-    },
-    {
-      "epoch": 0.9926802169792824,
-      "grad_norm": 0.45354804396629333,
-      "learning_rate": 7.715850097818985e-06,
-      "loss": 0.3718,
-      "step": 15189
-    },
-    {
-      "epoch": 0.9927455721848245,
-      "grad_norm": 0.41226768493652344,
-      "learning_rate": 7.715556903329084e-06,
-      "loss": 0.3385,
-      "step": 15190
-    },
-    {
-      "epoch": 0.9928109273903667,
-      "grad_norm": 0.44063106179237366,
-      "learning_rate": 7.71526369559447e-06,
-      "loss": 0.3482,
-      "step": 15191
-    },
-    {
-      "epoch": 0.9928762825959088,
-      "grad_norm": 0.48568663001060486,
-      "learning_rate": 7.714970474616577e-06,
-      "loss": 0.4381,
-      "step": 15192
-    },
-    {
-      "epoch": 0.9929416378014508,
-      "grad_norm": 0.4456183612346649,
-      "learning_rate": 7.71467724039683e-06,
-      "loss": 0.372,
-      "step": 15193
-    },
-    {
-      "epoch": 0.993006993006993,
-      "grad_norm": 0.470907598733902,
-      "learning_rate": 7.714383992936663e-06,
-      "loss": 0.3964,
-      "step": 15194
-    },
-    {
-      "epoch": 0.9930723482125351,
-      "grad_norm": 0.4126276671886444,
-      "learning_rate": 7.714090732237505e-06,
-      "loss": 0.3026,
-      "step": 15195
-    },
-    {
-      "epoch": 0.9931377034180773,
-      "grad_norm": 0.42819276452064514,
-      "learning_rate": 7.713797458300785e-06,
-      "loss": 0.3508,
-      "step": 15196
-    },
-    {
-      "epoch": 0.9932030586236194,
-      "grad_norm": 0.43126991391181946,
-      "learning_rate": 7.713504171127937e-06,
-      "loss": 0.3684,
-      "step": 15197
-    },
-    {
-      "epoch": 0.9932684138291615,
-      "grad_norm": 0.43060463666915894,
-      "learning_rate": 7.713210870720388e-06,
-      "loss": 0.3688,
-      "step": 15198
-    },
-    {
-      "epoch": 0.9933337690347036,
-      "grad_norm": 0.43413522839546204,
-      "learning_rate": 7.712917557079568e-06,
-      "loss": 0.3755,
-      "step": 15199
-    },
-    {
-      "epoch": 0.9933991242402458,
-      "grad_norm": 0.4354079067707062,
-      "learning_rate": 7.712624230206911e-06,
-      "loss": 0.3722,
-      "step": 15200
-    },
-    {
-      "epoch": 0.9934644794457879,
-      "grad_norm": 0.47235366702079773,
-      "learning_rate": 7.712330890103845e-06,
-      "loss": 0.482,
-      "step": 15201
-    },
-    {
-      "epoch": 0.9935298346513299,
-      "grad_norm": 0.38225215673446655,
-      "learning_rate": 7.712037536771802e-06,
-      "loss": 0.3072,
-      "step": 15202
-    },
-    {
-      "epoch": 0.9935951898568721,
-      "grad_norm": 0.4030027687549591,
-      "learning_rate": 7.711744170212214e-06,
-      "loss": 0.3259,
-      "step": 15203
-    },
-    {
-      "epoch": 0.9936605450624142,
-      "grad_norm": 0.42075735330581665,
-      "learning_rate": 7.711450790426508e-06,
-      "loss": 0.3567,
-      "step": 15204
-    },
-    {
-      "epoch": 0.9937259002679564,
-      "grad_norm": 0.37558045983314514,
-      "learning_rate": 7.711157397416118e-06,
-      "loss": 0.2928,
-      "step": 15205
-    },
-    {
-      "epoch": 0.9937912554734984,
-      "grad_norm": 0.4586002826690674,
-      "learning_rate": 7.710863991182473e-06,
-      "loss": 0.4227,
-      "step": 15206
-    },
-    {
-      "epoch": 0.9938566106790406,
-      "grad_norm": 0.4154171049594879,
-      "learning_rate": 7.710570571727007e-06,
-      "loss": 0.3406,
-      "step": 15207
-    },
-    {
-      "epoch": 0.9939219658845827,
-      "grad_norm": 0.4015105068683624,
-      "learning_rate": 7.710277139051148e-06,
-      "loss": 0.333,
-      "step": 15208
-    },
-    {
-      "epoch": 0.9939873210901248,
-      "grad_norm": 0.4767809510231018,
-      "learning_rate": 7.709983693156328e-06,
-      "loss": 0.3896,
-      "step": 15209
-    },
-    {
-      "epoch": 0.994052676295667,
-      "grad_norm": 0.4133701026439667,
-      "learning_rate": 7.709690234043981e-06,
-      "loss": 0.3661,
-      "step": 15210
-    },
-    {
-      "epoch": 0.994118031501209,
-      "grad_norm": 0.45379364490509033,
-      "learning_rate": 7.709396761715535e-06,
-      "loss": 0.3688,
-      "step": 15211
-    },
-    {
-      "epoch": 0.9941833867067512,
-      "grad_norm": 0.44322243332862854,
-      "learning_rate": 7.709103276172421e-06,
-      "loss": 0.387,
-      "step": 15212
-    },
-    {
-      "epoch": 0.9942487419122933,
-      "grad_norm": 0.416967511177063,
-      "learning_rate": 7.708809777416073e-06,
-      "loss": 0.3111,
-      "step": 15213
-    },
-    {
-      "epoch": 0.9943140971178355,
-      "grad_norm": 0.4574109613895416,
-      "learning_rate": 7.708516265447921e-06,
-      "loss": 0.4361,
-      "step": 15214
-    },
-    {
-      "epoch": 0.9943794523233775,
-      "grad_norm": 0.46277549862861633,
-      "learning_rate": 7.708222740269396e-06,
-      "loss": 0.4233,
-      "step": 15215
-    },
-    {
-      "epoch": 0.9944448075289197,
-      "grad_norm": 0.44610607624053955,
-      "learning_rate": 7.707929201881931e-06,
-      "loss": 0.4095,
-      "step": 15216
-    },
-    {
-      "epoch": 0.9945101627344618,
-      "grad_norm": 0.3953724801540375,
-      "learning_rate": 7.707635650286958e-06,
-      "loss": 0.3288,
-      "step": 15217
-    },
-    {
-      "epoch": 0.9945755179400039,
-      "grad_norm": 0.3972908556461334,
-      "learning_rate": 7.707342085485909e-06,
-      "loss": 0.3129,
-      "step": 15218
-    },
-    {
-      "epoch": 0.9946408731455461,
-      "grad_norm": 0.4491443336009979,
-      "learning_rate": 7.707048507480213e-06,
-      "loss": 0.406,
-      "step": 15219
-    },
-    {
-      "epoch": 0.9947062283510881,
-      "grad_norm": 0.43897825479507446,
-      "learning_rate": 7.706754916271304e-06,
-      "loss": 0.3719,
-      "step": 15220
-    },
-    {
-      "epoch": 0.9947715835566303,
-      "grad_norm": 0.4349232017993927,
-      "learning_rate": 7.706461311860614e-06,
-      "loss": 0.3875,
-      "step": 15221
-    },
-    {
-      "epoch": 0.9948369387621724,
-      "grad_norm": 0.4542638957500458,
-      "learning_rate": 7.706167694249573e-06,
-      "loss": 0.4226,
-      "step": 15222
-    },
-    {
-      "epoch": 0.9949022939677146,
-      "grad_norm": 0.43496763706207275,
-      "learning_rate": 7.705874063439616e-06,
-      "loss": 0.3792,
-      "step": 15223
-    },
-    {
-      "epoch": 0.9949676491732566,
-      "grad_norm": 0.4258120357990265,
-      "learning_rate": 7.705580419432172e-06,
-      "loss": 0.3533,
-      "step": 15224
-    },
-    {
-      "epoch": 0.9950330043787988,
-      "grad_norm": 0.45798373222351074,
-      "learning_rate": 7.705286762228678e-06,
-      "loss": 0.4284,
-      "step": 15225
-    },
-    {
-      "epoch": 0.9950983595843409,
-      "grad_norm": 0.4548629820346832,
-      "learning_rate": 7.704993091830561e-06,
-      "loss": 0.3759,
-      "step": 15226
-    },
-    {
-      "epoch": 0.995163714789883,
-      "grad_norm": 0.44336462020874023,
-      "learning_rate": 7.704699408239255e-06,
-      "loss": 0.3659,
-      "step": 15227
-    },
-    {
-      "epoch": 0.9952290699954252,
-      "grad_norm": 0.4453061819076538,
-      "learning_rate": 7.704405711456195e-06,
-      "loss": 0.3722,
-      "step": 15228
-    },
-    {
-      "epoch": 0.9952944252009672,
-      "grad_norm": 0.42155423760414124,
-      "learning_rate": 7.704112001482812e-06,
-      "loss": 0.373,
-      "step": 15229
-    },
-    {
-      "epoch": 0.9953597804065094,
-      "grad_norm": 0.45993250608444214,
-      "learning_rate": 7.703818278320537e-06,
-      "loss": 0.37,
-      "step": 15230
-    },
-    {
-      "epoch": 0.9954251356120515,
-      "grad_norm": 0.4333488643169403,
-      "learning_rate": 7.703524541970801e-06,
-      "loss": 0.3474,
-      "step": 15231
-    },
-    {
-      "epoch": 0.9954904908175937,
-      "grad_norm": 0.4314553141593933,
-      "learning_rate": 7.703230792435043e-06,
-      "loss": 0.3741,
-      "step": 15232
-    },
-    {
-      "epoch": 0.9955558460231357,
-      "grad_norm": 0.4804209768772125,
-      "learning_rate": 7.70293702971469e-06,
-      "loss": 0.436,
-      "step": 15233
-    },
-    {
-      "epoch": 0.9956212012286778,
-      "grad_norm": 0.4595136046409607,
-      "learning_rate": 7.702643253811177e-06,
-      "loss": 0.4032,
-      "step": 15234
-    },
-    {
-      "epoch": 0.99568655643422,
-      "grad_norm": 0.4507693946361542,
-      "learning_rate": 7.702349464725936e-06,
-      "loss": 0.3555,
-      "step": 15235
-    },
-    {
-      "epoch": 0.9957519116397621,
-      "grad_norm": 0.46738341450691223,
-      "learning_rate": 7.702055662460401e-06,
-      "loss": 0.4038,
-      "step": 15236
-    },
-    {
-      "epoch": 0.9958172668453042,
-      "grad_norm": 0.44166356325149536,
-      "learning_rate": 7.701761847016005e-06,
-      "loss": 0.4056,
-      "step": 15237
-    },
-    {
-      "epoch": 0.9958826220508463,
-      "grad_norm": 0.44396546483039856,
-      "learning_rate": 7.701468018394181e-06,
-      "loss": 0.3467,
-      "step": 15238
-    },
-    {
-      "epoch": 0.9959479772563885,
-      "grad_norm": 0.4619816541671753,
-      "learning_rate": 7.701174176596362e-06,
-      "loss": 0.4245,
-      "step": 15239
-    },
-    {
-      "epoch": 0.9960133324619306,
-      "grad_norm": 0.42113983631134033,
-      "learning_rate": 7.700880321623978e-06,
-      "loss": 0.3423,
-      "step": 15240
-    },
-    {
-      "epoch": 0.9960786876674728,
-      "grad_norm": 0.4135285019874573,
-      "learning_rate": 7.700586453478467e-06,
-      "loss": 0.3514,
-      "step": 15241
-    },
-    {
-      "epoch": 0.9961440428730148,
-      "grad_norm": 0.4091458022594452,
-      "learning_rate": 7.700292572161258e-06,
-      "loss": 0.3323,
-      "step": 15242
-    },
-    {
-      "epoch": 0.9962093980785569,
-      "grad_norm": 0.43799135088920593,
-      "learning_rate": 7.69999867767379e-06,
-      "loss": 0.3518,
-      "step": 15243
-    },
-    {
-      "epoch": 0.9962747532840991,
-      "grad_norm": 0.44951286911964417,
-      "learning_rate": 7.69970477001749e-06,
-      "loss": 0.4021,
-      "step": 15244
-    },
-    {
-      "epoch": 0.9963401084896412,
-      "grad_norm": 0.4571133852005005,
-      "learning_rate": 7.699410849193796e-06,
-      "loss": 0.3498,
-      "step": 15245
-    },
-    {
-      "epoch": 0.9964054636951833,
-      "grad_norm": 0.4132849872112274,
-      "learning_rate": 7.699116915204142e-06,
-      "loss": 0.3502,
-      "step": 15246
-    },
-    {
-      "epoch": 0.9964708189007254,
-      "grad_norm": 0.452383428812027,
-      "learning_rate": 7.698822968049957e-06,
-      "loss": 0.4164,
-      "step": 15247
-    },
-    {
-      "epoch": 0.9965361741062676,
-      "grad_norm": 0.42526212334632874,
-      "learning_rate": 7.698529007732678e-06,
-      "loss": 0.3573,
-      "step": 15248
-    },
-    {
-      "epoch": 0.9966015293118097,
-      "grad_norm": 0.4335181415081024,
-      "learning_rate": 7.698235034253739e-06,
-      "loss": 0.3501,
-      "step": 15249
-    },
-    {
-      "epoch": 0.9966668845173519,
-      "grad_norm": 0.4348534941673279,
-      "learning_rate": 7.697941047614571e-06,
-      "loss": 0.3519,
-      "step": 15250
-    },
-    {
-      "epoch": 0.9967322397228939,
-      "grad_norm": 0.44404658675193787,
-      "learning_rate": 7.697647047816612e-06,
-      "loss": 0.3901,
-      "step": 15251
-    },
-    {
-      "epoch": 0.996797594928436,
-      "grad_norm": 0.46016421914100647,
-      "learning_rate": 7.697353034861294e-06,
-      "loss": 0.3907,
-      "step": 15252
-    },
-    {
-      "epoch": 0.9968629501339782,
-      "grad_norm": 0.4186476171016693,
-      "learning_rate": 7.69705900875005e-06,
-      "loss": 0.3281,
-      "step": 15253
-    },
-    {
-      "epoch": 0.9969283053395203,
-      "grad_norm": 0.45298847556114197,
-      "learning_rate": 7.696764969484313e-06,
-      "loss": 0.3973,
-      "step": 15254
-    },
-    {
-      "epoch": 0.9969936605450624,
-      "grad_norm": 0.5015976428985596,
-      "learning_rate": 7.696470917065522e-06,
-      "loss": 0.4053,
-      "step": 15255
-    },
-    {
-      "epoch": 0.9970590157506045,
-      "grad_norm": 0.4459259808063507,
-      "learning_rate": 7.696176851495107e-06,
-      "loss": 0.3998,
-      "step": 15256
-    },
-    {
-      "epoch": 0.9971243709561467,
-      "grad_norm": 0.4428805708885193,
-      "learning_rate": 7.695882772774503e-06,
-      "loss": 0.406,
-      "step": 15257
-    },
-    {
-      "epoch": 0.9971897261616888,
-      "grad_norm": 0.4324735403060913,
-      "learning_rate": 7.695588680905145e-06,
-      "loss": 0.3621,
-      "step": 15258
-    },
-    {
-      "epoch": 0.9972550813672308,
-      "grad_norm": 0.46941500902175903,
-      "learning_rate": 7.695294575888467e-06,
-      "loss": 0.3866,
-      "step": 15259
-    },
-    {
-      "epoch": 0.997320436572773,
-      "grad_norm": 0.41056957840919495,
-      "learning_rate": 7.695000457725905e-06,
-      "loss": 0.3619,
-      "step": 15260
-    },
-    {
-      "epoch": 0.9973857917783151,
-      "grad_norm": 0.4279012084007263,
-      "learning_rate": 7.694706326418892e-06,
-      "loss": 0.3594,
-      "step": 15261
-    },
-    {
-      "epoch": 0.9974511469838573,
-      "grad_norm": 0.468083918094635,
-      "learning_rate": 7.69441218196886e-06,
-      "loss": 0.4272,
-      "step": 15262
-    },
-    {
-      "epoch": 0.9975165021893994,
-      "grad_norm": 0.42653143405914307,
-      "learning_rate": 7.69411802437725e-06,
-      "loss": 0.362,
-      "step": 15263
-    },
-    {
-      "epoch": 0.9975818573949415,
-      "grad_norm": 0.4228259027004242,
-      "learning_rate": 7.69382385364549e-06,
-      "loss": 0.3308,
-      "step": 15264
-    },
-    {
-      "epoch": 0.9976472126004836,
-      "grad_norm": 0.45300742983818054,
-      "learning_rate": 7.693529669775019e-06,
-      "loss": 0.415,
-      "step": 15265
-    },
-    {
-      "epoch": 0.9977125678060258,
-      "grad_norm": 0.4374513328075409,
-      "learning_rate": 7.69323547276727e-06,
-      "loss": 0.4007,
-      "step": 15266
-    },
-    {
-      "epoch": 0.9977779230115679,
-      "grad_norm": 0.4085589647293091,
-      "learning_rate": 7.692941262623681e-06,
-      "loss": 0.3631,
-      "step": 15267
-    },
-    {
-      "epoch": 0.9978432782171099,
-      "grad_norm": 0.4044470191001892,
-      "learning_rate": 7.692647039345682e-06,
-      "loss": 0.3021,
-      "step": 15268
-    },
-    {
-      "epoch": 0.9979086334226521,
-      "grad_norm": 0.4228689968585968,
-      "learning_rate": 7.692352802934711e-06,
-      "loss": 0.3354,
-      "step": 15269
-    },
-    {
-      "epoch": 0.9979739886281942,
-      "grad_norm": 0.43912720680236816,
-      "learning_rate": 7.692058553392204e-06,
-      "loss": 0.3761,
-      "step": 15270
-    },
-    {
-      "epoch": 0.9980393438337364,
-      "grad_norm": 0.4621241092681885,
-      "learning_rate": 7.691764290719593e-06,
-      "loss": 0.3938,
-      "step": 15271
-    },
-    {
-      "epoch": 0.9981046990392785,
-      "grad_norm": 0.41059187054634094,
-      "learning_rate": 7.691470014918316e-06,
-      "loss": 0.3337,
-      "step": 15272
-    },
-    {
-      "epoch": 0.9981700542448206,
-      "grad_norm": 0.4118594229221344,
-      "learning_rate": 7.691175725989808e-06,
-      "loss": 0.3332,
-      "step": 15273
-    },
-    {
-      "epoch": 0.9982354094503627,
-      "grad_norm": 0.45159950852394104,
-      "learning_rate": 7.690881423935502e-06,
-      "loss": 0.3743,
-      "step": 15274
-    },
-    {
-      "epoch": 0.9983007646559049,
-      "grad_norm": 0.45238029956817627,
-      "learning_rate": 7.690587108756837e-06,
-      "loss": 0.389,
-      "step": 15275
-    },
-    {
-      "epoch": 0.998366119861447,
-      "grad_norm": 0.41418617963790894,
-      "learning_rate": 7.690292780455244e-06,
-      "loss": 0.3242,
-      "step": 15276
-    },
-    {
-      "epoch": 0.998431475066989,
-      "grad_norm": 0.45793813467025757,
-      "learning_rate": 7.689998439032164e-06,
-      "loss": 0.3962,
-      "step": 15277
-    },
-    {
-      "epoch": 0.9984968302725312,
-      "grad_norm": 0.4461209774017334,
-      "learning_rate": 7.689704084489027e-06,
-      "loss": 0.3709,
-      "step": 15278
-    },
-    {
-      "epoch": 0.9985621854780733,
-      "grad_norm": 0.4224388897418976,
-      "learning_rate": 7.689409716827274e-06,
-      "loss": 0.3789,
-      "step": 15279
-    },
-    {
-      "epoch": 0.9986275406836155,
-      "grad_norm": 0.45713791251182556,
-      "learning_rate": 7.689115336048338e-06,
-      "loss": 0.3699,
-      "step": 15280
-    },
-    {
-      "epoch": 0.9986928958891576,
-      "grad_norm": 0.5281928181648254,
-      "learning_rate": 7.688820942153653e-06,
-      "loss": 0.497,
-      "step": 15281
-    },
-    {
-      "epoch": 0.9987582510946997,
-      "grad_norm": 0.4724256992340088,
-      "learning_rate": 7.688526535144658e-06,
-      "loss": 0.3828,
-      "step": 15282
-    },
-    {
-      "epoch": 0.9988236063002418,
-      "grad_norm": 0.47158998250961304,
-      "learning_rate": 7.688232115022786e-06,
-      "loss": 0.4414,
-      "step": 15283
-    },
-    {
-      "epoch": 0.998888961505784,
-      "grad_norm": 0.47896477580070496,
-      "learning_rate": 7.687937681789477e-06,
-      "loss": 0.4437,
-      "step": 15284
-    },
-    {
-      "epoch": 0.9989543167113261,
-      "grad_norm": 0.4669880270957947,
-      "learning_rate": 7.687643235446162e-06,
-      "loss": 0.378,
-      "step": 15285
-    },
-    {
-      "epoch": 0.9990196719168681,
-      "grad_norm": 0.4483124613761902,
-      "learning_rate": 7.687348775994283e-06,
-      "loss": 0.3631,
-      "step": 15286
-    },
-    {
-      "epoch": 0.9990850271224103,
-      "grad_norm": 0.4364282190799713,
-      "learning_rate": 7.687054303435271e-06,
-      "loss": 0.37,
-      "step": 15287
-    },
-    {
-      "epoch": 0.9991503823279524,
-      "grad_norm": 0.4038535952568054,
-      "learning_rate": 7.686759817770565e-06,
-      "loss": 0.3372,
-      "step": 15288
-    },
-    {
-      "epoch": 0.9992157375334946,
-      "grad_norm": 0.44478845596313477,
-      "learning_rate": 7.6864653190016e-06,
-      "loss": 0.3814,
-      "step": 15289
-    },
-    {
-      "epoch": 0.9992810927390366,
-      "grad_norm": 0.412781685590744,
-      "learning_rate": 7.686170807129814e-06,
-      "loss": 0.3322,
-      "step": 15290
-    },
-    {
-      "epoch": 0.9993464479445788,
-      "grad_norm": 0.4233294129371643,
-      "learning_rate": 7.68587628215664e-06,
-      "loss": 0.322,
-      "step": 15291
-    },
-    {
-      "epoch": 0.9994118031501209,
-      "grad_norm": 0.4360736906528473,
-      "learning_rate": 7.68558174408352e-06,
-      "loss": 0.3529,
-      "step": 15292
-    },
-    {
-      "epoch": 0.999477158355663,
-      "grad_norm": 0.4752248227596283,
-      "learning_rate": 7.685287192911886e-06,
-      "loss": 0.3983,
-      "step": 15293
-    },
-    {
-      "epoch": 0.9995425135612052,
-      "grad_norm": 0.42139577865600586,
-      "learning_rate": 7.684992628643176e-06,
-      "loss": 0.3448,
-      "step": 15294
-    },
-    {
-      "epoch": 0.9996078687667472,
-      "grad_norm": 0.4572370648384094,
-      "learning_rate": 7.684698051278826e-06,
-      "loss": 0.3902,
-      "step": 15295
-    },
-    {
-      "epoch": 0.9996732239722894,
-      "grad_norm": 0.4292619228363037,
-      "learning_rate": 7.684403460820276e-06,
-      "loss": 0.3385,
-      "step": 15296
-    },
-    {
-      "epoch": 0.9997385791778315,
-      "grad_norm": 0.44764095544815063,
-      "learning_rate": 7.68410885726896e-06,
-      "loss": 0.3863,
-      "step": 15297
-    },
-    {
-      "epoch": 0.9998039343833737,
-      "grad_norm": 0.4193493127822876,
-      "learning_rate": 7.683814240626313e-06,
-      "loss": 0.3307,
-      "step": 15298
-    },
-    {
-      "epoch": 0.9998692895889157,
-      "grad_norm": 0.47969362139701843,
-      "learning_rate": 7.683519610893776e-06,
-      "loss": 0.4549,
-      "step": 15299
-    },
-    {
-      "epoch": 0.9999346447944579,
-      "grad_norm": 0.4604553282260895,
-      "learning_rate": 7.683224968072782e-06,
-      "loss": 0.4099,
-      "step": 15300
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 0.4571715295314789,
-      "learning_rate": 7.682930312164771e-06,
-      "loss": 0.4088,
-      "step": 15301
-    },
-    {
-      "epoch": 1.000065355205542,
-      "grad_norm": 0.45919936895370483,
-      "learning_rate": 7.682635643171181e-06,
-      "loss": 0.3865,
-      "step": 15302
-    },
-    {
-      "epoch": 1.0001307104110841,
-      "grad_norm": 0.45753178000450134,
-      "learning_rate": 7.682340961093447e-06,
-      "loss": 0.3374,
-      "step": 15303
-    },
-    {
-      "epoch": 1.0001960656166264,
-      "grad_norm": 0.4426369369029999,
-      "learning_rate": 7.682046265933007e-06,
-      "loss": 0.3333,
-      "step": 15304
-    },
-    {
-      "epoch": 1.0002614208221685,
-      "grad_norm": 0.4452736973762512,
-      "learning_rate": 7.681751557691298e-06,
-      "loss": 0.3645,
-      "step": 15305
-    },
-    {
-      "epoch": 1.0003267760277106,
-      "grad_norm": 0.43928152322769165,
-      "learning_rate": 7.681456836369758e-06,
-      "loss": 0.3344,
-      "step": 15306
-    },
-    {
-      "epoch": 1.0003921312332527,
-      "grad_norm": 0.45254141092300415,
-      "learning_rate": 7.681162101969822e-06,
-      "loss": 0.3561,
-      "step": 15307
-    },
-    {
-      "epoch": 1.000457486438795,
-      "grad_norm": 0.471185564994812,
-      "learning_rate": 7.680867354492932e-06,
-      "loss": 0.4038,
-      "step": 15308
-    },
-    {
-      "epoch": 1.000522841644337,
-      "grad_norm": 0.42634573578834534,
-      "learning_rate": 7.680572593940521e-06,
-      "loss": 0.3478,
-      "step": 15309
-    },
-    {
-      "epoch": 1.000588196849879,
-      "grad_norm": 0.45847830176353455,
-      "learning_rate": 7.680277820314032e-06,
-      "loss": 0.3783,
-      "step": 15310
-    },
-    {
-      "epoch": 1.0006535520554212,
-      "grad_norm": 0.49452534317970276,
-      "learning_rate": 7.679983033614897e-06,
-      "loss": 0.3404,
-      "step": 15311
-    },
-    {
-      "epoch": 1.0007189072609632,
-      "grad_norm": 0.4476128816604614,
-      "learning_rate": 7.679688233844557e-06,
-      "loss": 0.3377,
-      "step": 15312
-    },
-    {
-      "epoch": 1.0007842624665055,
-      "grad_norm": 0.4552595317363739,
-      "learning_rate": 7.679393421004449e-06,
-      "loss": 0.317,
-      "step": 15313
-    },
-    {
-      "epoch": 1.0008496176720476,
-      "grad_norm": 0.4845251739025116,
-      "learning_rate": 7.679098595096011e-06,
-      "loss": 0.3454,
-      "step": 15314
-    },
-    {
-      "epoch": 1.0009149728775897,
-      "grad_norm": 0.4067078232765198,
-      "learning_rate": 7.67880375612068e-06,
-      "loss": 0.297,
-      "step": 15315
-    },
-    {
-      "epoch": 1.0009803280831318,
-      "grad_norm": 0.4564893841743469,
-      "learning_rate": 7.678508904079898e-06,
-      "loss": 0.3575,
-      "step": 15316
-    },
-    {
-      "epoch": 1.001045683288674,
-      "grad_norm": 0.43375900387763977,
-      "learning_rate": 7.678214038975098e-06,
-      "loss": 0.3163,
-      "step": 15317
-    },
-    {
-      "epoch": 1.0011110384942161,
-      "grad_norm": 0.4573245048522949,
-      "learning_rate": 7.67791916080772e-06,
-      "loss": 0.3447,
-      "step": 15318
-    },
-    {
-      "epoch": 1.0011763936997582,
-      "grad_norm": 0.4501189887523651,
-      "learning_rate": 7.677624269579204e-06,
-      "loss": 0.3953,
-      "step": 15319
-    },
-    {
-      "epoch": 1.0012417489053003,
-      "grad_norm": 0.45240911841392517,
-      "learning_rate": 7.677329365290986e-06,
-      "loss": 0.3691,
-      "step": 15320
-    },
-    {
-      "epoch": 1.0013071041108423,
-      "grad_norm": 0.452606201171875,
-      "learning_rate": 7.677034447944505e-06,
-      "loss": 0.3414,
-      "step": 15321
-    },
-    {
-      "epoch": 1.0013724593163846,
-      "grad_norm": 0.40821969509124756,
-      "learning_rate": 7.676739517541199e-06,
-      "loss": 0.3018,
-      "step": 15322
-    },
-    {
-      "epoch": 1.0014378145219267,
-      "grad_norm": 0.43178510665893555,
-      "learning_rate": 7.67644457408251e-06,
-      "loss": 0.3211,
-      "step": 15323
-    },
-    {
-      "epoch": 1.0015031697274688,
-      "grad_norm": 0.48483988642692566,
-      "learning_rate": 7.67614961756987e-06,
-      "loss": 0.3888,
-      "step": 15324
-    },
-    {
-      "epoch": 1.0015685249330109,
-      "grad_norm": 0.45330503582954407,
-      "learning_rate": 7.675854648004724e-06,
-      "loss": 0.3478,
-      "step": 15325
-    },
-    {
-      "epoch": 1.0016338801385531,
-      "grad_norm": 0.448544442653656,
-      "learning_rate": 7.675559665388506e-06,
-      "loss": 0.3452,
-      "step": 15326
-    },
-    {
-      "epoch": 1.0016992353440952,
-      "grad_norm": 0.4484884440898895,
-      "learning_rate": 7.675264669722657e-06,
-      "loss": 0.3512,
-      "step": 15327
-    },
-    {
-      "epoch": 1.0017645905496373,
-      "grad_norm": 0.46183279156684875,
-      "learning_rate": 7.674969661008615e-06,
-      "loss": 0.3651,
-      "step": 15328
-    },
-    {
-      "epoch": 1.0018299457551794,
-      "grad_norm": 0.4510844349861145,
-      "learning_rate": 7.67467463924782e-06,
-      "loss": 0.3529,
-      "step": 15329
-    },
-    {
-      "epoch": 1.0018953009607214,
-      "grad_norm": 0.4594692289829254,
-      "learning_rate": 7.674379604441711e-06,
-      "loss": 0.353,
-      "step": 15330
-    },
-    {
-      "epoch": 1.0019606561662637,
-      "grad_norm": 0.3801827132701874,
-      "learning_rate": 7.674084556591726e-06,
-      "loss": 0.2522,
-      "step": 15331
-    },
-    {
-      "epoch": 1.0020260113718058,
-      "grad_norm": 0.4472469985485077,
-      "learning_rate": 7.673789495699303e-06,
-      "loss": 0.3603,
-      "step": 15332
-    },
-    {
-      "epoch": 1.0020913665773479,
-      "grad_norm": 0.43683260679244995,
-      "learning_rate": 7.67349442176588e-06,
-      "loss": 0.318,
-      "step": 15333
-    },
-    {
-      "epoch": 1.00215672178289,
-      "grad_norm": 0.4052855968475342,
-      "learning_rate": 7.673199334792902e-06,
-      "loss": 0.309,
-      "step": 15334
-    },
-    {
-      "epoch": 1.002222076988432,
-      "grad_norm": 0.4850192666053772,
-      "learning_rate": 7.672904234781805e-06,
-      "loss": 0.4052,
-      "step": 15335
-    },
-    {
-      "epoch": 1.0022874321939743,
-      "grad_norm": 0.45603764057159424,
-      "learning_rate": 7.672609121734026e-06,
-      "loss": 0.3831,
-      "step": 15336
-    },
-    {
-      "epoch": 1.0023527873995164,
-      "grad_norm": 0.4760010540485382,
-      "learning_rate": 7.672313995651009e-06,
-      "loss": 0.3299,
-      "step": 15337
-    },
-    {
-      "epoch": 1.0024181426050585,
-      "grad_norm": 0.44374123215675354,
-      "learning_rate": 7.672018856534188e-06,
-      "loss": 0.3344,
-      "step": 15338
-    },
-    {
-      "epoch": 1.0024834978106005,
-      "grad_norm": 0.4514628052711487,
-      "learning_rate": 7.671723704385007e-06,
-      "loss": 0.3269,
-      "step": 15339
-    },
-    {
-      "epoch": 1.0025488530161428,
-      "grad_norm": 0.45178747177124023,
-      "learning_rate": 7.671428539204903e-06,
-      "loss": 0.3727,
-      "step": 15340
-    },
-    {
-      "epoch": 1.002614208221685,
-      "grad_norm": 0.4662570059299469,
-      "learning_rate": 7.671133360995317e-06,
-      "loss": 0.3559,
-      "step": 15341
-    },
-    {
-      "epoch": 1.002679563427227,
-      "grad_norm": 0.4465523362159729,
-      "learning_rate": 7.67083816975769e-06,
-      "loss": 0.3418,
-      "step": 15342
-    },
-    {
-      "epoch": 1.002744918632769,
-      "grad_norm": 0.43686559796333313,
-      "learning_rate": 7.670542965493456e-06,
-      "loss": 0.3217,
-      "step": 15343
-    },
-    {
-      "epoch": 1.0028102738383111,
-      "grad_norm": 0.45063772797584534,
-      "learning_rate": 7.670247748204065e-06,
-      "loss": 0.3198,
-      "step": 15344
-    },
-    {
-      "epoch": 1.0028756290438534,
-      "grad_norm": 0.45118752121925354,
-      "learning_rate": 7.669952517890946e-06,
-      "loss": 0.3677,
-      "step": 15345
-    },
-    {
-      "epoch": 1.0029409842493955,
-      "grad_norm": 0.4444842040538788,
-      "learning_rate": 7.669657274555545e-06,
-      "loss": 0.3484,
-      "step": 15346
-    },
-    {
-      "epoch": 1.0030063394549376,
-      "grad_norm": 0.43761003017425537,
-      "learning_rate": 7.669362018199298e-06,
-      "loss": 0.3364,
-      "step": 15347
-    },
-    {
-      "epoch": 1.0030716946604796,
-      "grad_norm": 0.48077958822250366,
-      "learning_rate": 7.66906674882365e-06,
-      "loss": 0.3413,
-      "step": 15348
-    },
-    {
-      "epoch": 1.003137049866022,
-      "grad_norm": 0.46071937680244446,
-      "learning_rate": 7.66877146643004e-06,
-      "loss": 0.33,
-      "step": 15349
-    },
-    {
-      "epoch": 1.003202405071564,
-      "grad_norm": 0.4771740138530731,
-      "learning_rate": 7.668476171019905e-06,
-      "loss": 0.3935,
-      "step": 15350
-    },
-    {
-      "epoch": 1.003267760277106,
-      "grad_norm": 0.4689539074897766,
-      "learning_rate": 7.668180862594688e-06,
-      "loss": 0.3358,
-      "step": 15351
-    },
-    {
-      "epoch": 1.0033331154826481,
-      "grad_norm": 0.46587684750556946,
-      "learning_rate": 7.667885541155828e-06,
-      "loss": 0.3521,
-      "step": 15352
-    },
-    {
-      "epoch": 1.0033984706881902,
-      "grad_norm": 0.4376724362373352,
-      "learning_rate": 7.667590206704767e-06,
-      "loss": 0.3183,
-      "step": 15353
-    },
-    {
-      "epoch": 1.0034638258937325,
-      "grad_norm": 0.4335780441761017,
-      "learning_rate": 7.667294859242945e-06,
-      "loss": 0.351,
-      "step": 15354
-    },
-    {
-      "epoch": 1.0035291810992746,
-      "grad_norm": 0.42169657349586487,
-      "learning_rate": 7.666999498771799e-06,
-      "loss": 0.3094,
-      "step": 15355
-    },
-    {
-      "epoch": 1.0035945363048167,
-      "grad_norm": 0.44098904728889465,
-      "learning_rate": 7.666704125292775e-06,
-      "loss": 0.3119,
-      "step": 15356
-    },
-    {
-      "epoch": 1.0036598915103587,
-      "grad_norm": 0.45945441722869873,
-      "learning_rate": 7.666408738807309e-06,
-      "loss": 0.3185,
-      "step": 15357
-    },
-    {
-      "epoch": 1.003725246715901,
-      "grad_norm": 0.460377961397171,
-      "learning_rate": 7.666113339316847e-06,
-      "loss": 0.3729,
-      "step": 15358
-    },
-    {
-      "epoch": 1.003790601921443,
-      "grad_norm": 0.43052223324775696,
-      "learning_rate": 7.665817926822824e-06,
-      "loss": 0.3301,
-      "step": 15359
-    },
-    {
-      "epoch": 1.0038559571269852,
-      "grad_norm": 0.522735595703125,
-      "learning_rate": 7.665522501326685e-06,
-      "loss": 0.3339,
-      "step": 15360
-    },
-    {
-      "epoch": 1.0039213123325272,
-      "grad_norm": 0.46012312173843384,
-      "learning_rate": 7.665227062829868e-06,
-      "loss": 0.3682,
-      "step": 15361
-    },
-    {
-      "epoch": 1.0039866675380693,
-      "grad_norm": 0.4636308550834656,
-      "learning_rate": 7.664931611333817e-06,
-      "loss": 0.343,
-      "step": 15362
-    },
-    {
-      "epoch": 1.0040520227436116,
-      "grad_norm": 0.4381164312362671,
-      "learning_rate": 7.66463614683997e-06,
-      "loss": 0.321,
-      "step": 15363
-    },
-    {
-      "epoch": 1.0041173779491537,
-      "grad_norm": 0.43972429633140564,
-      "learning_rate": 7.664340669349768e-06,
-      "loss": 0.3149,
-      "step": 15364
-    },
-    {
-      "epoch": 1.0041827331546958,
-      "grad_norm": 0.48464396595954895,
-      "learning_rate": 7.664045178864657e-06,
-      "loss": 0.348,
-      "step": 15365
-    },
-    {
-      "epoch": 1.0042480883602378,
-      "grad_norm": 0.5554192662239075,
-      "learning_rate": 7.663749675386072e-06,
-      "loss": 0.3856,
-      "step": 15366
-    },
-    {
-      "epoch": 1.0043134435657801,
-      "grad_norm": 0.420454204082489,
-      "learning_rate": 7.663454158915459e-06,
-      "loss": 0.3005,
-      "step": 15367
-    },
-    {
-      "epoch": 1.0043787987713222,
-      "grad_norm": 0.4903525114059448,
-      "learning_rate": 7.663158629454256e-06,
-      "loss": 0.3939,
-      "step": 15368
-    },
-    {
-      "epoch": 1.0044441539768643,
-      "grad_norm": 0.4578380286693573,
-      "learning_rate": 7.662863087003907e-06,
-      "loss": 0.3386,
-      "step": 15369
-    },
-    {
-      "epoch": 1.0045095091824063,
-      "grad_norm": 0.47643154859542847,
-      "learning_rate": 7.66256753156585e-06,
-      "loss": 0.3828,
-      "step": 15370
-    },
-    {
-      "epoch": 1.0045748643879484,
-      "grad_norm": 0.4753783643245697,
-      "learning_rate": 7.66227196314153e-06,
-      "loss": 0.3813,
-      "step": 15371
-    },
-    {
-      "epoch": 1.0046402195934907,
-      "grad_norm": 0.469293475151062,
-      "learning_rate": 7.661976381732387e-06,
-      "loss": 0.3765,
-      "step": 15372
-    },
-    {
-      "epoch": 1.0047055747990328,
-      "grad_norm": 0.4429028034210205,
-      "learning_rate": 7.661680787339864e-06,
-      "loss": 0.317,
-      "step": 15373
-    },
-    {
-      "epoch": 1.0047709300045748,
-      "grad_norm": 0.476163774728775,
-      "learning_rate": 7.661385179965402e-06,
-      "loss": 0.3822,
-      "step": 15374
-    },
-    {
-      "epoch": 1.004836285210117,
-      "grad_norm": 0.4372885525226593,
-      "learning_rate": 7.661089559610442e-06,
-      "loss": 0.3428,
-      "step": 15375
-    },
-    {
-      "epoch": 1.0049016404156592,
-      "grad_norm": 0.43586692214012146,
-      "learning_rate": 7.660793926276426e-06,
-      "loss": 0.3577,
-      "step": 15376
-    },
-    {
-      "epoch": 1.0049669956212013,
-      "grad_norm": 0.47613614797592163,
-      "learning_rate": 7.660498279964794e-06,
-      "loss": 0.3437,
-      "step": 15377
-    },
-    {
-      "epoch": 1.0050323508267434,
-      "grad_norm": 0.4132579565048218,
-      "learning_rate": 7.660202620676992e-06,
-      "loss": 0.303,
-      "step": 15378
-    },
-    {
-      "epoch": 1.0050977060322854,
-      "grad_norm": 0.43297308683395386,
-      "learning_rate": 7.659906948414462e-06,
-      "loss": 0.3349,
-      "step": 15379
-    },
-    {
-      "epoch": 1.0051630612378275,
-      "grad_norm": 0.45414796471595764,
-      "learning_rate": 7.659611263178643e-06,
-      "loss": 0.3343,
-      "step": 15380
-    },
-    {
-      "epoch": 1.0052284164433698,
-      "grad_norm": 0.45960375666618347,
-      "learning_rate": 7.659315564970977e-06,
-      "loss": 0.3439,
-      "step": 15381
-    },
-    {
-      "epoch": 1.0052937716489119,
-      "grad_norm": 0.4501376152038574,
-      "learning_rate": 7.65901985379291e-06,
-      "loss": 0.3565,
-      "step": 15382
-    },
-    {
-      "epoch": 1.005359126854454,
-      "grad_norm": 0.48546937108039856,
-      "learning_rate": 7.658724129645881e-06,
-      "loss": 0.3952,
-      "step": 15383
-    },
-    {
-      "epoch": 1.005424482059996,
-      "grad_norm": 0.4542507827281952,
-      "learning_rate": 7.658428392531333e-06,
-      "loss": 0.3589,
-      "step": 15384
-    },
-    {
-      "epoch": 1.005489837265538,
-      "grad_norm": 0.4576326906681061,
-      "learning_rate": 7.65813264245071e-06,
-      "loss": 0.3754,
-      "step": 15385
-    },
-    {
-      "epoch": 1.0055551924710804,
-      "grad_norm": 0.4136732518672943,
-      "learning_rate": 7.657836879405454e-06,
-      "loss": 0.3026,
-      "step": 15386
-    },
-    {
-      "epoch": 1.0056205476766225,
-      "grad_norm": 0.44388332962989807,
-      "learning_rate": 7.657541103397005e-06,
-      "loss": 0.3383,
-      "step": 15387
-    },
-    {
-      "epoch": 1.0056859028821645,
-      "grad_norm": 0.45588627457618713,
-      "learning_rate": 7.657245314426807e-06,
-      "loss": 0.3403,
-      "step": 15388
-    },
-    {
-      "epoch": 1.0057512580877066,
-      "grad_norm": 0.4652750790119171,
-      "learning_rate": 7.656949512496304e-06,
-      "loss": 0.3716,
-      "step": 15389
-    },
-    {
-      "epoch": 1.005816613293249,
-      "grad_norm": 0.4347158372402191,
-      "learning_rate": 7.656653697606937e-06,
-      "loss": 0.3183,
-      "step": 15390
-    },
-    {
-      "epoch": 1.005881968498791,
-      "grad_norm": 0.4857756793498993,
-      "learning_rate": 7.65635786976015e-06,
-      "loss": 0.3689,
-      "step": 15391
-    },
-    {
-      "epoch": 1.005947323704333,
-      "grad_norm": 0.4873138964176178,
-      "learning_rate": 7.656062028957385e-06,
-      "loss": 0.3659,
-      "step": 15392
-    },
-    {
-      "epoch": 1.0060126789098751,
-      "grad_norm": 0.4715665876865387,
-      "learning_rate": 7.655766175200086e-06,
-      "loss": 0.3644,
-      "step": 15393
-    },
-    {
-      "epoch": 1.0060780341154172,
-      "grad_norm": 0.44285252690315247,
-      "learning_rate": 7.655470308489694e-06,
-      "loss": 0.3466,
-      "step": 15394
-    },
-    {
-      "epoch": 1.0061433893209595,
-      "grad_norm": 0.4487248659133911,
-      "learning_rate": 7.655174428827655e-06,
-      "loss": 0.3311,
-      "step": 15395
-    },
-    {
-      "epoch": 1.0062087445265016,
-      "grad_norm": 0.4194047749042511,
-      "learning_rate": 7.654878536215412e-06,
-      "loss": 0.3137,
-      "step": 15396
-    },
-    {
-      "epoch": 1.0062740997320436,
-      "grad_norm": 0.4552207291126251,
-      "learning_rate": 7.654582630654404e-06,
-      "loss": 0.354,
-      "step": 15397
-    },
-    {
-      "epoch": 1.0063394549375857,
-      "grad_norm": 0.4308977425098419,
-      "learning_rate": 7.654286712146077e-06,
-      "loss": 0.3331,
-      "step": 15398
-    },
-    {
-      "epoch": 1.006404810143128,
-      "grad_norm": 0.45188385248184204,
-      "learning_rate": 7.653990780691874e-06,
-      "loss": 0.3712,
-      "step": 15399
-    },
-    {
-      "epoch": 1.00647016534867,
-      "grad_norm": 0.45716938376426697,
-      "learning_rate": 7.65369483629324e-06,
-      "loss": 0.3443,
-      "step": 15400
-    },
-    {
-      "epoch": 1.0065355205542121,
-      "grad_norm": 0.42230695486068726,
-      "learning_rate": 7.653398878951615e-06,
-      "loss": 0.3088,
-      "step": 15401
-    },
-    {
-      "epoch": 1.0066008757597542,
-      "grad_norm": 0.4392012059688568,
-      "learning_rate": 7.653102908668447e-06,
-      "loss": 0.3307,
-      "step": 15402
-    },
-    {
-      "epoch": 1.0066662309652963,
-      "grad_norm": 0.42826029658317566,
-      "learning_rate": 7.652806925445175e-06,
-      "loss": 0.3196,
-      "step": 15403
-    },
-    {
-      "epoch": 1.0067315861708386,
-      "grad_norm": 0.4623169004917145,
-      "learning_rate": 7.652510929283244e-06,
-      "loss": 0.3779,
-      "step": 15404
-    },
-    {
-      "epoch": 1.0067969413763806,
-      "grad_norm": 0.48810315132141113,
-      "learning_rate": 7.6522149201841e-06,
-      "loss": 0.4204,
-      "step": 15405
-    },
-    {
-      "epoch": 1.0068622965819227,
-      "grad_norm": 0.4800114631652832,
-      "learning_rate": 7.651918898149185e-06,
-      "loss": 0.3685,
-      "step": 15406
-    },
-    {
-      "epoch": 1.0069276517874648,
-      "grad_norm": 0.44584551453590393,
-      "learning_rate": 7.651622863179941e-06,
-      "loss": 0.355,
-      "step": 15407
-    },
-    {
-      "epoch": 1.006993006993007,
-      "grad_norm": 0.5116623640060425,
-      "learning_rate": 7.651326815277816e-06,
-      "loss": 0.3904,
-      "step": 15408
-    },
-    {
-      "epoch": 1.0070583621985492,
-      "grad_norm": 0.4597768485546112,
-      "learning_rate": 7.65103075444425e-06,
-      "loss": 0.3345,
-      "step": 15409
-    },
-    {
-      "epoch": 1.0071237174040912,
-      "grad_norm": 0.4301532506942749,
-      "learning_rate": 7.65073468068069e-06,
-      "loss": 0.3144,
-      "step": 15410
-    },
-    {
-      "epoch": 1.0071890726096333,
-      "grad_norm": 0.42474237084388733,
-      "learning_rate": 7.650438593988577e-06,
-      "loss": 0.3291,
-      "step": 15411
-    },
-    {
-      "epoch": 1.0072544278151754,
-      "grad_norm": 0.4560867249965668,
-      "learning_rate": 7.650142494369356e-06,
-      "loss": 0.372,
-      "step": 15412
-    },
-    {
-      "epoch": 1.0073197830207177,
-      "grad_norm": 0.46581974625587463,
-      "learning_rate": 7.649846381824475e-06,
-      "loss": 0.3413,
-      "step": 15413
-    },
-    {
-      "epoch": 1.0073851382262597,
-      "grad_norm": 0.489432692527771,
-      "learning_rate": 7.649550256355374e-06,
-      "loss": 0.4066,
-      "step": 15414
-    },
-    {
-      "epoch": 1.0074504934318018,
-      "grad_norm": 0.44276174902915955,
-      "learning_rate": 7.649254117963498e-06,
-      "loss": 0.2954,
-      "step": 15415
-    },
-    {
-      "epoch": 1.007515848637344,
-      "grad_norm": 0.4767237901687622,
-      "learning_rate": 7.648957966650293e-06,
-      "loss": 0.3718,
-      "step": 15416
-    },
-    {
-      "epoch": 1.0075812038428862,
-      "grad_norm": 0.4617912173271179,
-      "learning_rate": 7.648661802417201e-06,
-      "loss": 0.3705,
-      "step": 15417
-    },
-    {
-      "epoch": 1.0076465590484283,
-      "grad_norm": 0.46661320328712463,
-      "learning_rate": 7.648365625265668e-06,
-      "loss": 0.3902,
-      "step": 15418
-    },
-    {
-      "epoch": 1.0077119142539703,
-      "grad_norm": 0.4534335136413574,
-      "learning_rate": 7.648069435197138e-06,
-      "loss": 0.3621,
-      "step": 15419
-    },
-    {
-      "epoch": 1.0077772694595124,
-      "grad_norm": 0.42317673563957214,
-      "learning_rate": 7.647773232213057e-06,
-      "loss": 0.2892,
-      "step": 15420
-    },
-    {
-      "epoch": 1.0078426246650545,
-      "grad_norm": 0.46837151050567627,
-      "learning_rate": 7.647477016314869e-06,
-      "loss": 0.3836,
-      "step": 15421
-    },
-    {
-      "epoch": 1.0079079798705968,
-      "grad_norm": 0.43539056181907654,
-      "learning_rate": 7.647180787504016e-06,
-      "loss": 0.3194,
-      "step": 15422
-    },
-    {
-      "epoch": 1.0079733350761388,
-      "grad_norm": 0.47065645456314087,
-      "learning_rate": 7.646884545781947e-06,
-      "loss": 0.3834,
-      "step": 15423
-    },
-    {
-      "epoch": 1.008038690281681,
-      "grad_norm": 0.4062655568122864,
-      "learning_rate": 7.646588291150106e-06,
-      "loss": 0.2855,
-      "step": 15424
-    },
-    {
-      "epoch": 1.008104045487223,
-      "grad_norm": 0.4419330954551697,
-      "learning_rate": 7.646292023609935e-06,
-      "loss": 0.3253,
-      "step": 15425
-    },
-    {
-      "epoch": 1.0081694006927653,
-      "grad_norm": 0.4646981358528137,
-      "learning_rate": 7.645995743162882e-06,
-      "loss": 0.3656,
-      "step": 15426
-    },
-    {
-      "epoch": 1.0082347558983074,
-      "grad_norm": 0.45052090287208557,
-      "learning_rate": 7.645699449810391e-06,
-      "loss": 0.3549,
-      "step": 15427
-    },
-    {
-      "epoch": 1.0083001111038494,
-      "grad_norm": 0.43203607201576233,
-      "learning_rate": 7.645403143553908e-06,
-      "loss": 0.3263,
-      "step": 15428
-    },
-    {
-      "epoch": 1.0083654663093915,
-      "grad_norm": 0.4358994662761688,
-      "learning_rate": 7.645106824394878e-06,
-      "loss": 0.3478,
-      "step": 15429
-    },
-    {
-      "epoch": 1.0084308215149336,
-      "grad_norm": 0.4181872010231018,
-      "learning_rate": 7.644810492334744e-06,
-      "loss": 0.327,
-      "step": 15430
-    },
-    {
-      "epoch": 1.0084961767204759,
-      "grad_norm": 0.43759000301361084,
-      "learning_rate": 7.644514147374954e-06,
-      "loss": 0.3424,
-      "step": 15431
-    },
-    {
-      "epoch": 1.008561531926018,
-      "grad_norm": 0.4420819878578186,
-      "learning_rate": 7.64421778951695e-06,
-      "loss": 0.3644,
-      "step": 15432
-    },
-    {
-      "epoch": 1.00862688713156,
-      "grad_norm": 0.4365168809890747,
-      "learning_rate": 7.643921418762183e-06,
-      "loss": 0.3126,
-      "step": 15433
-    },
-    {
-      "epoch": 1.008692242337102,
-      "grad_norm": 0.45583680272102356,
-      "learning_rate": 7.643625035112095e-06,
-      "loss": 0.3818,
-      "step": 15434
-    },
-    {
-      "epoch": 1.0087575975426444,
-      "grad_norm": 0.44000598788261414,
-      "learning_rate": 7.643328638568129e-06,
-      "loss": 0.3492,
-      "step": 15435
-    },
-    {
-      "epoch": 1.0088229527481865,
-      "grad_norm": 0.44959017634391785,
-      "learning_rate": 7.643032229131737e-06,
-      "loss": 0.3672,
-      "step": 15436
-    },
-    {
-      "epoch": 1.0088883079537285,
-      "grad_norm": 0.4667029082775116,
-      "learning_rate": 7.642735806804357e-06,
-      "loss": 0.3256,
-      "step": 15437
-    },
-    {
-      "epoch": 1.0089536631592706,
-      "grad_norm": 0.4460352957248688,
-      "learning_rate": 7.642439371587442e-06,
-      "loss": 0.3574,
-      "step": 15438
-    },
-    {
-      "epoch": 1.0090190183648127,
-      "grad_norm": 0.4368128776550293,
-      "learning_rate": 7.642142923482434e-06,
-      "loss": 0.3556,
-      "step": 15439
-    },
-    {
-      "epoch": 1.009084373570355,
-      "grad_norm": 0.4322856664657593,
-      "learning_rate": 7.641846462490782e-06,
-      "loss": 0.3077,
-      "step": 15440
-    },
-    {
-      "epoch": 1.009149728775897,
-      "grad_norm": 0.4305310845375061,
-      "learning_rate": 7.641549988613925e-06,
-      "loss": 0.319,
-      "step": 15441
-    },
-    {
-      "epoch": 1.009215083981439,
-      "grad_norm": 0.4275602102279663,
-      "learning_rate": 7.641253501853317e-06,
-      "loss": 0.3593,
-      "step": 15442
-    },
-    {
-      "epoch": 1.0092804391869812,
-      "grad_norm": 0.4231933653354645,
-      "learning_rate": 7.6409570022104e-06,
-      "loss": 0.3258,
-      "step": 15443
-    },
-    {
-      "epoch": 1.0093457943925233,
-      "grad_norm": 0.4606062173843384,
-      "learning_rate": 7.64066048968662e-06,
-      "loss": 0.3681,
-      "step": 15444
-    },
-    {
-      "epoch": 1.0094111495980655,
-      "grad_norm": 0.4889924228191376,
-      "learning_rate": 7.640363964283423e-06,
-      "loss": 0.377,
-      "step": 15445
-    },
-    {
-      "epoch": 1.0094765048036076,
-      "grad_norm": 0.4719698131084442,
-      "learning_rate": 7.640067426002257e-06,
-      "loss": 0.3691,
-      "step": 15446
-    },
-    {
-      "epoch": 1.0095418600091497,
-      "grad_norm": 0.44663551449775696,
-      "learning_rate": 7.63977087484457e-06,
-      "loss": 0.3553,
-      "step": 15447
-    },
-    {
-      "epoch": 1.0096072152146918,
-      "grad_norm": 0.5834388732910156,
-      "learning_rate": 7.639474310811804e-06,
-      "loss": 0.3731,
-      "step": 15448
-    },
-    {
-      "epoch": 1.009672570420234,
-      "grad_norm": 0.4788368046283722,
-      "learning_rate": 7.639177733905407e-06,
-      "loss": 0.3708,
-      "step": 15449
-    },
-    {
-      "epoch": 1.0097379256257761,
-      "grad_norm": 0.4483938217163086,
-      "learning_rate": 7.638881144126825e-06,
-      "loss": 0.3378,
-      "step": 15450
-    },
-    {
-      "epoch": 1.0098032808313182,
-      "grad_norm": 0.5010645389556885,
-      "learning_rate": 7.63858454147751e-06,
-      "loss": 0.3979,
-      "step": 15451
-    },
-    {
-      "epoch": 1.0098686360368603,
-      "grad_norm": 0.4544999301433563,
-      "learning_rate": 7.6382879259589e-06,
-      "loss": 0.3321,
-      "step": 15452
-    },
-    {
-      "epoch": 1.0099339912424024,
-      "grad_norm": 0.4219933748245239,
-      "learning_rate": 7.637991297572447e-06,
-      "loss": 0.322,
-      "step": 15453
-    },
-    {
-      "epoch": 1.0099993464479446,
-      "grad_norm": 0.4401116967201233,
-      "learning_rate": 7.637694656319596e-06,
-      "loss": 0.3596,
-      "step": 15454
-    },
-    {
-      "epoch": 1.0100647016534867,
-      "grad_norm": 0.45100435614585876,
-      "learning_rate": 7.637398002201795e-06,
-      "loss": 0.379,
-      "step": 15455
-    },
-    {
-      "epoch": 1.0101300568590288,
-      "grad_norm": 0.4402828514575958,
-      "learning_rate": 7.63710133522049e-06,
-      "loss": 0.326,
-      "step": 15456
-    },
-    {
-      "epoch": 1.0101954120645709,
-      "grad_norm": 0.5119789242744446,
-      "learning_rate": 7.636804655377129e-06,
-      "loss": 0.2976,
-      "step": 15457
-    },
-    {
-      "epoch": 1.0102607672701132,
-      "grad_norm": 0.4435166120529175,
-      "learning_rate": 7.636507962673156e-06,
-      "loss": 0.3631,
-      "step": 15458
-    },
-    {
-      "epoch": 1.0103261224756552,
-      "grad_norm": 0.44343629479408264,
-      "learning_rate": 7.63621125711002e-06,
-      "loss": 0.3082,
-      "step": 15459
-    },
-    {
-      "epoch": 1.0103914776811973,
-      "grad_norm": 0.46621182560920715,
-      "learning_rate": 7.63591453868917e-06,
-      "loss": 0.3447,
-      "step": 15460
-    },
-    {
-      "epoch": 1.0104568328867394,
-      "grad_norm": 0.433179646730423,
-      "learning_rate": 7.635617807412053e-06,
-      "loss": 0.3114,
-      "step": 15461
-    },
-    {
-      "epoch": 1.0105221880922814,
-      "grad_norm": 0.502048671245575,
-      "learning_rate": 7.635321063280114e-06,
-      "loss": 0.385,
-      "step": 15462
-    },
-    {
-      "epoch": 1.0105875432978237,
-      "grad_norm": 0.4555855691432953,
-      "learning_rate": 7.635024306294801e-06,
-      "loss": 0.3222,
-      "step": 15463
-    },
-    {
-      "epoch": 1.0106528985033658,
-      "grad_norm": 0.49316129088401794,
-      "learning_rate": 7.63472753645756e-06,
-      "loss": 0.3508,
-      "step": 15464
-    },
-    {
-      "epoch": 1.0107182537089079,
-      "grad_norm": 0.4633665382862091,
-      "learning_rate": 7.634430753769842e-06,
-      "loss": 0.3432,
-      "step": 15465
-    },
-    {
-      "epoch": 1.01078360891445,
-      "grad_norm": 0.4293590188026428,
-      "learning_rate": 7.634133958233092e-06,
-      "loss": 0.307,
-      "step": 15466
-    },
-    {
-      "epoch": 1.0108489641199923,
-      "grad_norm": 0.4356231987476349,
-      "learning_rate": 7.633837149848759e-06,
-      "loss": 0.337,
-      "step": 15467
-    },
-    {
-      "epoch": 1.0109143193255343,
-      "grad_norm": 0.4453873634338379,
-      "learning_rate": 7.633540328618289e-06,
-      "loss": 0.3537,
-      "step": 15468
-    },
-    {
-      "epoch": 1.0109796745310764,
-      "grad_norm": 0.44065070152282715,
-      "learning_rate": 7.63324349454313e-06,
-      "loss": 0.346,
-      "step": 15469
-    },
-    {
-      "epoch": 1.0110450297366185,
-      "grad_norm": 0.45693904161453247,
-      "learning_rate": 7.63294664762473e-06,
-      "loss": 0.3457,
-      "step": 15470
-    },
-    {
-      "epoch": 1.0111103849421605,
-      "grad_norm": 0.47622743248939514,
-      "learning_rate": 7.63264978786454e-06,
-      "loss": 0.3711,
-      "step": 15471
-    },
-    {
-      "epoch": 1.0111757401477028,
-      "grad_norm": 0.4707341194152832,
-      "learning_rate": 7.632352915264001e-06,
-      "loss": 0.3864,
-      "step": 15472
-    },
-    {
-      "epoch": 1.011241095353245,
-      "grad_norm": 0.4664469361305237,
-      "learning_rate": 7.632056029824566e-06,
-      "loss": 0.357,
-      "step": 15473
-    },
-    {
-      "epoch": 1.011306450558787,
-      "grad_norm": 0.4514675438404083,
-      "learning_rate": 7.631759131547684e-06,
-      "loss": 0.364,
-      "step": 15474
-    },
-    {
-      "epoch": 1.011371805764329,
-      "grad_norm": 0.46300041675567627,
-      "learning_rate": 7.631462220434798e-06,
-      "loss": 0.3668,
-      "step": 15475
-    },
-    {
-      "epoch": 1.0114371609698714,
-      "grad_norm": 0.43007588386535645,
-      "learning_rate": 7.631165296487361e-06,
-      "loss": 0.3153,
-      "step": 15476
-    },
-    {
-      "epoch": 1.0115025161754134,
-      "grad_norm": 0.4446285367012024,
-      "learning_rate": 7.630868359706818e-06,
-      "loss": 0.3435,
-      "step": 15477
-    },
-    {
-      "epoch": 1.0115678713809555,
-      "grad_norm": 0.4426470398902893,
-      "learning_rate": 7.63057141009462e-06,
-      "loss": 0.3393,
-      "step": 15478
-    },
-    {
-      "epoch": 1.0116332265864976,
-      "grad_norm": 0.45051711797714233,
-      "learning_rate": 7.630274447652214e-06,
-      "loss": 0.3619,
-      "step": 15479
-    },
-    {
-      "epoch": 1.0116985817920396,
-      "grad_norm": 0.4386603832244873,
-      "learning_rate": 7.629977472381047e-06,
-      "loss": 0.3058,
-      "step": 15480
-    },
-    {
-      "epoch": 1.011763936997582,
-      "grad_norm": 0.45015019178390503,
-      "learning_rate": 7.62968048428257e-06,
-      "loss": 0.3635,
-      "step": 15481
-    },
-    {
-      "epoch": 1.011829292203124,
-      "grad_norm": 0.4465944170951843,
-      "learning_rate": 7.629383483358229e-06,
-      "loss": 0.3267,
-      "step": 15482
-    },
-    {
-      "epoch": 1.011894647408666,
-      "grad_norm": 0.45590606331825256,
-      "learning_rate": 7.629086469609474e-06,
-      "loss": 0.3687,
-      "step": 15483
-    },
-    {
-      "epoch": 1.0119600026142082,
-      "grad_norm": 0.4663674235343933,
-      "learning_rate": 7.6287894430377544e-06,
-      "loss": 0.3875,
-      "step": 15484
-    },
-    {
-      "epoch": 1.0120253578197504,
-      "grad_norm": 0.403889924287796,
-      "learning_rate": 7.628492403644519e-06,
-      "loss": 0.2805,
-      "step": 15485
-    },
-    {
-      "epoch": 1.0120907130252925,
-      "grad_norm": 0.4771665036678314,
-      "learning_rate": 7.6281953514312125e-06,
-      "loss": 0.3715,
-      "step": 15486
-    },
-    {
-      "epoch": 1.0121560682308346,
-      "grad_norm": 0.44601204991340637,
-      "learning_rate": 7.627898286399289e-06,
-      "loss": 0.3322,
-      "step": 15487
-    },
-    {
-      "epoch": 1.0122214234363767,
-      "grad_norm": 0.4793049097061157,
-      "learning_rate": 7.627601208550195e-06,
-      "loss": 0.3865,
-      "step": 15488
-    },
-    {
-      "epoch": 1.0122867786419187,
-      "grad_norm": 0.45158910751342773,
-      "learning_rate": 7.627304117885379e-06,
-      "loss": 0.3605,
-      "step": 15489
-    },
-    {
-      "epoch": 1.012352133847461,
-      "grad_norm": 0.6445653438568115,
-      "learning_rate": 7.6270070144062914e-06,
-      "loss": 0.3375,
-      "step": 15490
-    },
-    {
-      "epoch": 1.012417489053003,
-      "grad_norm": 0.43588805198669434,
-      "learning_rate": 7.626709898114379e-06,
-      "loss": 0.3352,
-      "step": 15491
-    },
-    {
-      "epoch": 1.0124828442585452,
-      "grad_norm": 0.4423539638519287,
-      "learning_rate": 7.626412769011095e-06,
-      "loss": 0.3427,
-      "step": 15492
-    },
-    {
-      "epoch": 1.0125481994640873,
-      "grad_norm": 0.45310357213020325,
-      "learning_rate": 7.626115627097883e-06,
-      "loss": 0.3426,
-      "step": 15493
-    },
-    {
-      "epoch": 1.0126135546696295,
-      "grad_norm": 0.5024445056915283,
-      "learning_rate": 7.625818472376198e-06,
-      "loss": 0.3781,
-      "step": 15494
-    },
-    {
-      "epoch": 1.0126789098751716,
-      "grad_norm": 0.46750733256340027,
-      "learning_rate": 7.625521304847484e-06,
-      "loss": 0.3281,
-      "step": 15495
-    },
-    {
-      "epoch": 1.0127442650807137,
-      "grad_norm": 0.4888159930706024,
-      "learning_rate": 7.625224124513195e-06,
-      "loss": 0.3946,
-      "step": 15496
-    },
-    {
-      "epoch": 1.0128096202862558,
-      "grad_norm": 0.4495035707950592,
-      "learning_rate": 7.624926931374777e-06,
-      "loss": 0.356,
-      "step": 15497
-    },
-    {
-      "epoch": 1.0128749754917978,
-      "grad_norm": 0.44272077083587646,
-      "learning_rate": 7.624629725433683e-06,
-      "loss": 0.3578,
-      "step": 15498
-    },
-    {
-      "epoch": 1.0129403306973401,
-      "grad_norm": 0.44645029306411743,
-      "learning_rate": 7.62433250669136e-06,
-      "loss": 0.3423,
-      "step": 15499
-    },
-    {
-      "epoch": 1.0130056859028822,
-      "grad_norm": 0.45229387283325195,
-      "learning_rate": 7.624035275149258e-06,
-      "loss": 0.3625,
-      "step": 15500
-    },
-    {
-      "epoch": 1.0130710411084243,
-      "grad_norm": 0.5196219682693481,
-      "learning_rate": 7.623738030808826e-06,
-      "loss": 0.4103,
-      "step": 15501
-    },
-    {
-      "epoch": 1.0131363963139663,
-      "grad_norm": 0.442560613155365,
-      "learning_rate": 7.623440773671514e-06,
-      "loss": 0.3557,
-      "step": 15502
-    },
-    {
-      "epoch": 1.0132017515195084,
-      "grad_norm": 0.4808710515499115,
-      "learning_rate": 7.623143503738774e-06,
-      "loss": 0.3612,
-      "step": 15503
-    },
-    {
-      "epoch": 1.0132671067250507,
-      "grad_norm": 0.45476454496383667,
-      "learning_rate": 7.622846221012055e-06,
-      "loss": 0.3635,
-      "step": 15504
-    },
-    {
-      "epoch": 1.0133324619305928,
-      "grad_norm": 0.49299511313438416,
-      "learning_rate": 7.622548925492805e-06,
-      "loss": 0.3837,
-      "step": 15505
-    },
-    {
-      "epoch": 1.0133978171361349,
-      "grad_norm": 0.4372859299182892,
-      "learning_rate": 7.622251617182477e-06,
-      "loss": 0.3665,
-      "step": 15506
-    },
-    {
-      "epoch": 1.013463172341677,
-      "grad_norm": 0.4558941423892975,
-      "learning_rate": 7.6219542960825165e-06,
-      "loss": 0.3644,
-      "step": 15507
-    },
-    {
-      "epoch": 1.0135285275472192,
-      "grad_norm": 0.4683453440666199,
-      "learning_rate": 7.621656962194379e-06,
-      "loss": 0.379,
-      "step": 15508
-    },
-    {
-      "epoch": 1.0135938827527613,
-      "grad_norm": 0.46672216057777405,
-      "learning_rate": 7.621359615519512e-06,
-      "loss": 0.3532,
-      "step": 15509
-    },
-    {
-      "epoch": 1.0136592379583034,
-      "grad_norm": 0.45947110652923584,
-      "learning_rate": 7.621062256059366e-06,
-      "loss": 0.3599,
-      "step": 15510
-    },
-    {
-      "epoch": 1.0137245931638454,
-      "grad_norm": 0.4729503095149994,
-      "learning_rate": 7.62076488381539e-06,
-      "loss": 0.3569,
-      "step": 15511
-    },
-    {
-      "epoch": 1.0137899483693875,
-      "grad_norm": 0.46331289410591125,
-      "learning_rate": 7.620467498789038e-06,
-      "loss": 0.3518,
-      "step": 15512
-    },
-    {
-      "epoch": 1.0138553035749298,
-      "grad_norm": 0.4517684578895569,
-      "learning_rate": 7.620170100981756e-06,
-      "loss": 0.3554,
-      "step": 15513
-    },
-    {
-      "epoch": 1.0139206587804719,
-      "grad_norm": 0.43650633096694946,
-      "learning_rate": 7.619872690394998e-06,
-      "loss": 0.3244,
-      "step": 15514
-    },
-    {
-      "epoch": 1.013986013986014,
-      "grad_norm": 0.4438323378562927,
-      "learning_rate": 7.619575267030213e-06,
-      "loss": 0.3476,
-      "step": 15515
-    },
-    {
-      "epoch": 1.014051369191556,
-      "grad_norm": 0.44066789746284485,
-      "learning_rate": 7.619277830888851e-06,
-      "loss": 0.3298,
-      "step": 15516
-    },
-    {
-      "epoch": 1.0141167243970983,
-      "grad_norm": 0.4373476803302765,
-      "learning_rate": 7.618980381972365e-06,
-      "loss": 0.342,
-      "step": 15517
-    },
-    {
-      "epoch": 1.0141820796026404,
-      "grad_norm": 0.4374128580093384,
-      "learning_rate": 7.618682920282203e-06,
-      "loss": 0.3365,
-      "step": 15518
-    },
-    {
-      "epoch": 1.0142474348081825,
-      "grad_norm": 0.4472385048866272,
-      "learning_rate": 7.618385445819819e-06,
-      "loss": 0.3203,
-      "step": 15519
-    },
-    {
-      "epoch": 1.0143127900137245,
-      "grad_norm": 0.4670279324054718,
-      "learning_rate": 7.618087958586661e-06,
-      "loss": 0.3559,
-      "step": 15520
-    },
-    {
-      "epoch": 1.0143781452192666,
-      "grad_norm": 0.43676018714904785,
-      "learning_rate": 7.617790458584181e-06,
-      "loss": 0.3416,
-      "step": 15521
-    },
-    {
-      "epoch": 1.014443500424809,
-      "grad_norm": 0.42490100860595703,
-      "learning_rate": 7.617492945813829e-06,
-      "loss": 0.3344,
-      "step": 15522
-    },
-    {
-      "epoch": 1.014508855630351,
-      "grad_norm": 0.43293091654777527,
-      "learning_rate": 7.617195420277059e-06,
-      "loss": 0.3396,
-      "step": 15523
-    },
-    {
-      "epoch": 1.014574210835893,
-      "grad_norm": 0.4176598787307739,
-      "learning_rate": 7.6168978819753204e-06,
-      "loss": 0.31,
-      "step": 15524
-    },
-    {
-      "epoch": 1.0146395660414351,
-      "grad_norm": 0.43753165006637573,
-      "learning_rate": 7.6166003309100626e-06,
-      "loss": 0.3,
-      "step": 15525
-    },
-    {
-      "epoch": 1.0147049212469774,
-      "grad_norm": 0.4823523163795471,
-      "learning_rate": 7.61630276708274e-06,
-      "loss": 0.354,
-      "step": 15526
-    },
-    {
-      "epoch": 1.0147702764525195,
-      "grad_norm": 0.4753963053226471,
-      "learning_rate": 7.616005190494802e-06,
-      "loss": 0.3897,
-      "step": 15527
-    },
-    {
-      "epoch": 1.0148356316580616,
-      "grad_norm": 0.4880515933036804,
-      "learning_rate": 7.615707601147699e-06,
-      "loss": 0.4284,
-      "step": 15528
-    },
-    {
-      "epoch": 1.0149009868636036,
-      "grad_norm": 0.4545324146747589,
-      "learning_rate": 7.615409999042886e-06,
-      "loss": 0.3765,
-      "step": 15529
-    },
-    {
-      "epoch": 1.0149663420691457,
-      "grad_norm": 0.45359712839126587,
-      "learning_rate": 7.615112384181811e-06,
-      "loss": 0.3638,
-      "step": 15530
-    },
-    {
-      "epoch": 1.015031697274688,
-      "grad_norm": 0.40306368470191956,
-      "learning_rate": 7.614814756565927e-06,
-      "loss": 0.2961,
-      "step": 15531
-    },
-    {
-      "epoch": 1.01509705248023,
-      "grad_norm": 0.47215649485588074,
-      "learning_rate": 7.614517116196686e-06,
-      "loss": 0.3792,
-      "step": 15532
-    },
-    {
-      "epoch": 1.0151624076857722,
-      "grad_norm": 0.5271449089050293,
-      "learning_rate": 7.614219463075539e-06,
-      "loss": 0.4131,
-      "step": 15533
-    },
-    {
-      "epoch": 1.0152277628913142,
-      "grad_norm": 0.46284154057502747,
-      "learning_rate": 7.6139217972039395e-06,
-      "loss": 0.3572,
-      "step": 15534
-    },
-    {
-      "epoch": 1.0152931180968565,
-      "grad_norm": 0.435781866312027,
-      "learning_rate": 7.613624118583336e-06,
-      "loss": 0.3287,
-      "step": 15535
-    },
-    {
-      "epoch": 1.0153584733023986,
-      "grad_norm": 0.40714776515960693,
-      "learning_rate": 7.613326427215182e-06,
-      "loss": 0.2723,
-      "step": 15536
-    },
-    {
-      "epoch": 1.0154238285079407,
-      "grad_norm": 0.4516788423061371,
-      "learning_rate": 7.6130287231009315e-06,
-      "loss": 0.2979,
-      "step": 15537
-    },
-    {
-      "epoch": 1.0154891837134827,
-      "grad_norm": 0.4320693016052246,
-      "learning_rate": 7.612731006242034e-06,
-      "loss": 0.3485,
-      "step": 15538
-    },
-    {
-      "epoch": 1.0155545389190248,
-      "grad_norm": 0.4644910395145416,
-      "learning_rate": 7.612433276639941e-06,
-      "loss": 0.3736,
-      "step": 15539
-    },
-    {
-      "epoch": 1.015619894124567,
-      "grad_norm": 0.53074711561203,
-      "learning_rate": 7.6121355342961065e-06,
-      "loss": 0.3927,
-      "step": 15540
-    },
-    {
-      "epoch": 1.0156852493301092,
-      "grad_norm": 0.4280640184879303,
-      "learning_rate": 7.611837779211982e-06,
-      "loss": 0.3167,
-      "step": 15541
-    },
-    {
-      "epoch": 1.0157506045356512,
-      "grad_norm": 0.44264957308769226,
-      "learning_rate": 7.6115400113890205e-06,
-      "loss": 0.338,
-      "step": 15542
-    },
-    {
-      "epoch": 1.0158159597411933,
-      "grad_norm": 0.46898049116134644,
-      "learning_rate": 7.611242230828673e-06,
-      "loss": 0.3833,
-      "step": 15543
-    },
-    {
-      "epoch": 1.0158813149467356,
-      "grad_norm": 0.46865010261535645,
-      "learning_rate": 7.610944437532393e-06,
-      "loss": 0.3781,
-      "step": 15544
-    },
-    {
-      "epoch": 1.0159466701522777,
-      "grad_norm": 0.46959206461906433,
-      "learning_rate": 7.610646631501632e-06,
-      "loss": 0.3463,
-      "step": 15545
-    },
-    {
-      "epoch": 1.0160120253578198,
-      "grad_norm": 0.41981449723243713,
-      "learning_rate": 7.610348812737845e-06,
-      "loss": 0.3069,
-      "step": 15546
-    },
-    {
-      "epoch": 1.0160773805633618,
-      "grad_norm": 0.47139954566955566,
-      "learning_rate": 7.610050981242479e-06,
-      "loss": 0.3744,
-      "step": 15547
-    },
-    {
-      "epoch": 1.016142735768904,
-      "grad_norm": 0.4501635432243347,
-      "learning_rate": 7.609753137016993e-06,
-      "loss": 0.3554,
-      "step": 15548
-    },
-    {
-      "epoch": 1.0162080909744462,
-      "grad_norm": 0.44787856936454773,
-      "learning_rate": 7.609455280062834e-06,
-      "loss": 0.3349,
-      "step": 15549
-    },
-    {
-      "epoch": 1.0162734461799883,
-      "grad_norm": 0.43762820959091187,
-      "learning_rate": 7.609157410381461e-06,
-      "loss": 0.3561,
-      "step": 15550
-    },
-    {
-      "epoch": 1.0163388013855303,
-      "grad_norm": 0.4427582621574402,
-      "learning_rate": 7.60885952797432e-06,
-      "loss": 0.3086,
-      "step": 15551
-    },
-    {
-      "epoch": 1.0164041565910724,
-      "grad_norm": 0.4594787061214447,
-      "learning_rate": 7.60856163284287e-06,
-      "loss": 0.3906,
-      "step": 15552
-    },
-    {
-      "epoch": 1.0164695117966147,
-      "grad_norm": 0.44954901933670044,
-      "learning_rate": 7.60826372498856e-06,
-      "loss": 0.322,
-      "step": 15553
-    },
-    {
-      "epoch": 1.0165348670021568,
-      "grad_norm": 0.4757879674434662,
-      "learning_rate": 7.607965804412845e-06,
-      "loss": 0.4043,
-      "step": 15554
-    },
-    {
-      "epoch": 1.0166002222076989,
-      "grad_norm": 0.4479594826698303,
-      "learning_rate": 7.607667871117177e-06,
-      "loss": 0.358,
-      "step": 15555
-    },
-    {
-      "epoch": 1.016665577413241,
-      "grad_norm": 0.42708635330200195,
-      "learning_rate": 7.60736992510301e-06,
-      "loss": 0.3367,
-      "step": 15556
-    },
-    {
-      "epoch": 1.016730932618783,
-      "grad_norm": 0.44823840260505676,
-      "learning_rate": 7.6070719663717955e-06,
-      "loss": 0.3242,
-      "step": 15557
-    },
-    {
-      "epoch": 1.0167962878243253,
-      "grad_norm": 0.4422805905342102,
-      "learning_rate": 7.606773994924987e-06,
-      "loss": 0.3226,
-      "step": 15558
-    },
-    {
-      "epoch": 1.0168616430298674,
-      "grad_norm": 0.4667592942714691,
-      "learning_rate": 7.60647601076404e-06,
-      "loss": 0.3558,
-      "step": 15559
-    },
-    {
-      "epoch": 1.0169269982354094,
-      "grad_norm": 0.4609525203704834,
-      "learning_rate": 7.606178013890406e-06,
-      "loss": 0.3569,
-      "step": 15560
-    },
-    {
-      "epoch": 1.0169923534409515,
-      "grad_norm": 0.42644402384757996,
-      "learning_rate": 7.605880004305539e-06,
-      "loss": 0.3427,
-      "step": 15561
-    },
-    {
-      "epoch": 1.0170577086464936,
-      "grad_norm": 0.4584433138370514,
-      "learning_rate": 7.6055819820108926e-06,
-      "loss": 0.3281,
-      "step": 15562
-    },
-    {
-      "epoch": 1.0171230638520359,
-      "grad_norm": 0.4662013649940491,
-      "learning_rate": 7.605283947007921e-06,
-      "loss": 0.3769,
-      "step": 15563
-    },
-    {
-      "epoch": 1.017188419057578,
-      "grad_norm": 0.5001772046089172,
-      "learning_rate": 7.604985899298077e-06,
-      "loss": 0.3784,
-      "step": 15564
-    },
-    {
-      "epoch": 1.01725377426312,
-      "grad_norm": 0.4466564655303955,
-      "learning_rate": 7.604687838882813e-06,
-      "loss": 0.3573,
-      "step": 15565
-    },
-    {
-      "epoch": 1.017319129468662,
-      "grad_norm": 0.45362588763237,
-      "learning_rate": 7.6043897657635846e-06,
-      "loss": 0.3724,
-      "step": 15566
-    },
-    {
-      "epoch": 1.0173844846742044,
-      "grad_norm": 0.44119858741760254,
-      "learning_rate": 7.604091679941845e-06,
-      "loss": 0.3187,
-      "step": 15567
-    },
-    {
-      "epoch": 1.0174498398797465,
-      "grad_norm": 0.47164979577064514,
-      "learning_rate": 7.603793581419048e-06,
-      "loss": 0.3772,
-      "step": 15568
-    },
-    {
-      "epoch": 1.0175151950852885,
-      "grad_norm": 0.4713986814022064,
-      "learning_rate": 7.603495470196648e-06,
-      "loss": 0.3581,
-      "step": 15569
-    },
-    {
-      "epoch": 1.0175805502908306,
-      "grad_norm": 0.4711362421512604,
-      "learning_rate": 7.603197346276099e-06,
-      "loss": 0.3672,
-      "step": 15570
-    },
-    {
-      "epoch": 1.0176459054963727,
-      "grad_norm": 0.43596675992012024,
-      "learning_rate": 7.602899209658854e-06,
-      "loss": 0.3263,
-      "step": 15571
-    },
-    {
-      "epoch": 1.017711260701915,
-      "grad_norm": 0.470787912607193,
-      "learning_rate": 7.602601060346369e-06,
-      "loss": 0.313,
-      "step": 15572
-    },
-    {
-      "epoch": 1.017776615907457,
-      "grad_norm": 0.4555983245372772,
-      "learning_rate": 7.602302898340096e-06,
-      "loss": 0.3734,
-      "step": 15573
-    },
-    {
-      "epoch": 1.0178419711129991,
-      "grad_norm": 0.4501747488975525,
-      "learning_rate": 7.60200472364149e-06,
-      "loss": 0.3384,
-      "step": 15574
-    },
-    {
-      "epoch": 1.0179073263185412,
-      "grad_norm": 0.4738941490650177,
-      "learning_rate": 7.601706536252007e-06,
-      "loss": 0.3629,
-      "step": 15575
-    },
-    {
-      "epoch": 1.0179726815240835,
-      "grad_norm": 0.44299745559692383,
-      "learning_rate": 7.601408336173098e-06,
-      "loss": 0.3284,
-      "step": 15576
-    },
-    {
-      "epoch": 1.0180380367296256,
-      "grad_norm": 0.4432782530784607,
-      "learning_rate": 7.601110123406221e-06,
-      "loss": 0.3249,
-      "step": 15577
-    },
-    {
-      "epoch": 1.0181033919351676,
-      "grad_norm": 0.44111546874046326,
-      "learning_rate": 7.600811897952827e-06,
-      "loss": 0.3534,
-      "step": 15578
-    },
-    {
-      "epoch": 1.0181687471407097,
-      "grad_norm": 0.46954017877578735,
-      "learning_rate": 7.600513659814375e-06,
-      "loss": 0.3831,
-      "step": 15579
-    },
-    {
-      "epoch": 1.0182341023462518,
-      "grad_norm": 0.4363420903682709,
-      "learning_rate": 7.600215408992314e-06,
-      "loss": 0.3443,
-      "step": 15580
-    },
-    {
-      "epoch": 1.018299457551794,
-      "grad_norm": 0.4859357178211212,
-      "learning_rate": 7.5999171454881024e-06,
-      "loss": 0.4015,
-      "step": 15581
-    },
-    {
-      "epoch": 1.0183648127573361,
-      "grad_norm": 0.43814557790756226,
-      "learning_rate": 7.599618869303196e-06,
-      "loss": 0.3096,
-      "step": 15582
-    },
-    {
-      "epoch": 1.0184301679628782,
-      "grad_norm": 0.44182729721069336,
-      "learning_rate": 7.599320580439047e-06,
-      "loss": 0.331,
-      "step": 15583
-    },
-    {
-      "epoch": 1.0184955231684203,
-      "grad_norm": 0.45808476209640503,
-      "learning_rate": 7.5990222788971105e-06,
-      "loss": 0.3625,
-      "step": 15584
-    },
-    {
-      "epoch": 1.0185608783739626,
-      "grad_norm": 0.45872604846954346,
-      "learning_rate": 7.59872396467884e-06,
-      "loss": 0.3723,
-      "step": 15585
-    },
-    {
-      "epoch": 1.0186262335795047,
-      "grad_norm": 0.44794178009033203,
-      "learning_rate": 7.598425637785696e-06,
-      "loss": 0.3518,
-      "step": 15586
-    },
-    {
-      "epoch": 1.0186915887850467,
-      "grad_norm": 0.4414597153663635,
-      "learning_rate": 7.598127298219128e-06,
-      "loss": 0.349,
-      "step": 15587
-    },
-    {
-      "epoch": 1.0187569439905888,
-      "grad_norm": 0.42647597193717957,
-      "learning_rate": 7.5978289459805946e-06,
-      "loss": 0.3371,
-      "step": 15588
-    },
-    {
-      "epoch": 1.0188222991961309,
-      "grad_norm": 0.47627365589141846,
-      "learning_rate": 7.5975305810715476e-06,
-      "loss": 0.4082,
-      "step": 15589
-    },
-    {
-      "epoch": 1.0188876544016732,
-      "grad_norm": 0.4444546103477478,
-      "learning_rate": 7.597232203493446e-06,
-      "loss": 0.3418,
-      "step": 15590
-    },
-    {
-      "epoch": 1.0189530096072152,
-      "grad_norm": 0.43940412998199463,
-      "learning_rate": 7.596933813247741e-06,
-      "loss": 0.3608,
-      "step": 15591
-    },
-    {
-      "epoch": 1.0190183648127573,
-      "grad_norm": 0.4503110349178314,
-      "learning_rate": 7.596635410335891e-06,
-      "loss": 0.35,
-      "step": 15592
-    },
-    {
-      "epoch": 1.0190837200182994,
-      "grad_norm": 0.4423142373561859,
-      "learning_rate": 7.596336994759351e-06,
-      "loss": 0.3254,
-      "step": 15593
-    },
-    {
-      "epoch": 1.0191490752238417,
-      "grad_norm": 0.43201929330825806,
-      "learning_rate": 7.596038566519574e-06,
-      "loss": 0.3319,
-      "step": 15594
-    },
-    {
-      "epoch": 1.0192144304293838,
-      "grad_norm": 0.471842885017395,
-      "learning_rate": 7.59574012561802e-06,
-      "loss": 0.3944,
-      "step": 15595
-    },
-    {
-      "epoch": 1.0192797856349258,
-      "grad_norm": 0.4996616840362549,
-      "learning_rate": 7.595441672056143e-06,
-      "loss": 0.2886,
-      "step": 15596
-    },
-    {
-      "epoch": 1.019345140840468,
-      "grad_norm": 0.4460655748844147,
-      "learning_rate": 7.595143205835396e-06,
-      "loss": 0.3637,
-      "step": 15597
-    },
-    {
-      "epoch": 1.01941049604601,
-      "grad_norm": 0.4370627999305725,
-      "learning_rate": 7.5948447269572365e-06,
-      "loss": 0.3552,
-      "step": 15598
-    },
-    {
-      "epoch": 1.0194758512515523,
-      "grad_norm": 0.43534135818481445,
-      "learning_rate": 7.594546235423122e-06,
-      "loss": 0.3304,
-      "step": 15599
-    },
-    {
-      "epoch": 1.0195412064570943,
-      "grad_norm": 0.44917362928390503,
-      "learning_rate": 7.594247731234504e-06,
-      "loss": 0.3691,
-      "step": 15600
-    },
-    {
-      "epoch": 1.0196065616626364,
-      "grad_norm": 0.4325159192085266,
-      "learning_rate": 7.593949214392843e-06,
-      "loss": 0.3009,
-      "step": 15601
-    },
-    {
-      "epoch": 1.0196719168681785,
-      "grad_norm": 0.4442124664783478,
-      "learning_rate": 7.593650684899593e-06,
-      "loss": 0.3392,
-      "step": 15602
-    },
-    {
-      "epoch": 1.0197372720737208,
-      "grad_norm": 0.4569205939769745,
-      "learning_rate": 7.593352142756209e-06,
-      "loss": 0.3585,
-      "step": 15603
-    },
-    {
-      "epoch": 1.0198026272792629,
-      "grad_norm": 0.42676761746406555,
-      "learning_rate": 7.59305358796415e-06,
-      "loss": 0.2972,
-      "step": 15604
-    },
-    {
-      "epoch": 1.019867982484805,
-      "grad_norm": 0.4391583204269409,
-      "learning_rate": 7.592755020524867e-06,
-      "loss": 0.3521,
-      "step": 15605
-    },
-    {
-      "epoch": 1.019933337690347,
-      "grad_norm": 0.45182445645332336,
-      "learning_rate": 7.592456440439823e-06,
-      "loss": 0.322,
-      "step": 15606
-    },
-    {
-      "epoch": 1.019998692895889,
-      "grad_norm": 0.45521080493927,
-      "learning_rate": 7.592157847710468e-06,
-      "loss": 0.3542,
-      "step": 15607
-    },
-    {
-      "epoch": 1.0200640481014314,
-      "grad_norm": 0.4373794198036194,
-      "learning_rate": 7.591859242338265e-06,
-      "loss": 0.3233,
-      "step": 15608
-    },
-    {
-      "epoch": 1.0201294033069734,
-      "grad_norm": 0.5093589425086975,
-      "learning_rate": 7.591560624324662e-06,
-      "loss": 0.3061,
-      "step": 15609
-    },
-    {
-      "epoch": 1.0201947585125155,
-      "grad_norm": 0.4563637673854828,
-      "learning_rate": 7.591261993671122e-06,
-      "loss": 0.3382,
-      "step": 15610
-    },
-    {
-      "epoch": 1.0202601137180576,
-      "grad_norm": 0.4290340542793274,
-      "learning_rate": 7.5909633503791015e-06,
-      "loss": 0.3262,
-      "step": 15611
-    },
-    {
-      "epoch": 1.0203254689235997,
-      "grad_norm": 0.43287280201911926,
-      "learning_rate": 7.5906646944500515e-06,
-      "loss": 0.3256,
-      "step": 15612
-    },
-    {
-      "epoch": 1.020390824129142,
-      "grad_norm": 0.46789512038230896,
-      "learning_rate": 7.590366025885435e-06,
-      "loss": 0.3841,
-      "step": 15613
-    },
-    {
-      "epoch": 1.020456179334684,
-      "grad_norm": 0.4521138668060303,
-      "learning_rate": 7.5900673446867045e-06,
-      "loss": 0.325,
-      "step": 15614
-    },
-    {
-      "epoch": 1.020521534540226,
-      "grad_norm": 0.44588905572891235,
-      "learning_rate": 7.5897686508553205e-06,
-      "loss": 0.3001,
-      "step": 15615
-    },
-    {
-      "epoch": 1.0205868897457682,
-      "grad_norm": 0.4593271017074585,
-      "learning_rate": 7.5894699443927355e-06,
-      "loss": 0.3669,
-      "step": 15616
-    },
-    {
-      "epoch": 1.0206522449513105,
-      "grad_norm": 0.427633672952652,
-      "learning_rate": 7.589171225300409e-06,
-      "loss": 0.3014,
-      "step": 15617
-    },
-    {
-      "epoch": 1.0207176001568525,
-      "grad_norm": 0.4613807797431946,
-      "learning_rate": 7.588872493579798e-06,
-      "loss": 0.337,
-      "step": 15618
-    },
-    {
-      "epoch": 1.0207829553623946,
-      "grad_norm": 0.458882600069046,
-      "learning_rate": 7.588573749232359e-06,
-      "loss": 0.3357,
-      "step": 15619
-    },
-    {
-      "epoch": 1.0208483105679367,
-      "grad_norm": 0.473646342754364,
-      "learning_rate": 7.588274992259548e-06,
-      "loss": 0.3369,
-      "step": 15620
-    },
-    {
-      "epoch": 1.0209136657734788,
-      "grad_norm": 0.451913058757782,
-      "learning_rate": 7.587976222662824e-06,
-      "loss": 0.3612,
-      "step": 15621
-    },
-    {
-      "epoch": 1.020979020979021,
-      "grad_norm": 0.4600163400173187,
-      "learning_rate": 7.587677440443643e-06,
-      "loss": 0.3463,
-      "step": 15622
-    },
-    {
-      "epoch": 1.0210443761845631,
-      "grad_norm": 0.43435609340667725,
-      "learning_rate": 7.587378645603463e-06,
-      "loss": 0.3294,
-      "step": 15623
-    },
-    {
-      "epoch": 1.0211097313901052,
-      "grad_norm": 0.43905115127563477,
-      "learning_rate": 7.58707983814374e-06,
-      "loss": 0.3392,
-      "step": 15624
-    },
-    {
-      "epoch": 1.0211750865956473,
-      "grad_norm": 0.4720829725265503,
-      "learning_rate": 7.586781018065934e-06,
-      "loss": 0.349,
-      "step": 15625
-    },
-    {
-      "epoch": 1.0212404418011896,
-      "grad_norm": 0.46056434512138367,
-      "learning_rate": 7.5864821853714995e-06,
-      "loss": 0.362,
-      "step": 15626
-    },
-    {
-      "epoch": 1.0213057970067316,
-      "grad_norm": 0.44007593393325806,
-      "learning_rate": 7.5861833400618965e-06,
-      "loss": 0.3156,
-      "step": 15627
-    },
-    {
-      "epoch": 1.0213711522122737,
-      "grad_norm": 0.4564206898212433,
-      "learning_rate": 7.585884482138581e-06,
-      "loss": 0.3777,
-      "step": 15628
-    },
-    {
-      "epoch": 1.0214365074178158,
-      "grad_norm": 0.45567262172698975,
-      "learning_rate": 7.58558561160301e-06,
-      "loss": 0.3566,
-      "step": 15629
-    },
-    {
-      "epoch": 1.0215018626233578,
-      "grad_norm": 0.4550042748451233,
-      "learning_rate": 7.585286728456643e-06,
-      "loss": 0.3754,
-      "step": 15630
-    },
-    {
-      "epoch": 1.0215672178289001,
-      "grad_norm": 0.4357951879501343,
-      "learning_rate": 7.5849878327009385e-06,
-      "loss": 0.333,
-      "step": 15631
-    },
-    {
-      "epoch": 1.0216325730344422,
-      "grad_norm": 0.47056224942207336,
-      "learning_rate": 7.58468892433735e-06,
-      "loss": 0.3795,
-      "step": 15632
-    },
-    {
-      "epoch": 1.0216979282399843,
-      "grad_norm": 0.469344824552536,
-      "learning_rate": 7.58439000336734e-06,
-      "loss": 0.3319,
-      "step": 15633
-    },
-    {
-      "epoch": 1.0217632834455264,
-      "grad_norm": 0.42325741052627563,
-      "learning_rate": 7.584091069792363e-06,
-      "loss": 0.3211,
-      "step": 15634
-    },
-    {
-      "epoch": 1.0218286386510687,
-      "grad_norm": 0.47051575779914856,
-      "learning_rate": 7.583792123613881e-06,
-      "loss": 0.3919,
-      "step": 15635
-    },
-    {
-      "epoch": 1.0218939938566107,
-      "grad_norm": 0.4938119053840637,
-      "learning_rate": 7.583493164833349e-06,
-      "loss": 0.3704,
-      "step": 15636
-    },
-    {
-      "epoch": 1.0219593490621528,
-      "grad_norm": 0.48237931728363037,
-      "learning_rate": 7.583194193452224e-06,
-      "loss": 0.3793,
-      "step": 15637
-    },
-    {
-      "epoch": 1.0220247042676949,
-      "grad_norm": 0.4226386845111847,
-      "learning_rate": 7.582895209471969e-06,
-      "loss": 0.3399,
-      "step": 15638
-    },
-    {
-      "epoch": 1.022090059473237,
-      "grad_norm": 0.4548985958099365,
-      "learning_rate": 7.582596212894038e-06,
-      "loss": 0.3411,
-      "step": 15639
-    },
-    {
-      "epoch": 1.0221554146787792,
-      "grad_norm": 0.47761985659599304,
-      "learning_rate": 7.58229720371989e-06,
-      "loss": 0.3662,
-      "step": 15640
-    },
-    {
-      "epoch": 1.0222207698843213,
-      "grad_norm": 0.4808357059955597,
-      "learning_rate": 7.581998181950985e-06,
-      "loss": 0.3728,
-      "step": 15641
-    },
-    {
-      "epoch": 1.0222861250898634,
-      "grad_norm": 0.4505424201488495,
-      "learning_rate": 7.5816991475887795e-06,
-      "loss": 0.3426,
-      "step": 15642
-    },
-    {
-      "epoch": 1.0223514802954055,
-      "grad_norm": 0.4237038493156433,
-      "learning_rate": 7.581400100634732e-06,
-      "loss": 0.2975,
-      "step": 15643
-    },
-    {
-      "epoch": 1.0224168355009478,
-      "grad_norm": 0.5180762410163879,
-      "learning_rate": 7.581101041090303e-06,
-      "loss": 0.4273,
-      "step": 15644
-    },
-    {
-      "epoch": 1.0224821907064898,
-      "grad_norm": 0.45106804370880127,
-      "learning_rate": 7.58080196895695e-06,
-      "loss": 0.3688,
-      "step": 15645
-    },
-    {
-      "epoch": 1.022547545912032,
-      "grad_norm": 0.44238874316215515,
-      "learning_rate": 7.580502884236132e-06,
-      "loss": 0.3661,
-      "step": 15646
-    },
-    {
-      "epoch": 1.022612901117574,
-      "grad_norm": 0.4164060652256012,
-      "learning_rate": 7.580203786929305e-06,
-      "loss": 0.3183,
-      "step": 15647
-    },
-    {
-      "epoch": 1.022678256323116,
-      "grad_norm": 0.4736102223396301,
-      "learning_rate": 7.5799046770379335e-06,
-      "loss": 0.3742,
-      "step": 15648
-    },
-    {
-      "epoch": 1.0227436115286583,
-      "grad_norm": 0.45015087723731995,
-      "learning_rate": 7.5796055545634716e-06,
-      "loss": 0.3235,
-      "step": 15649
-    },
-    {
-      "epoch": 1.0228089667342004,
-      "grad_norm": 0.4537907540798187,
-      "learning_rate": 7.5793064195073805e-06,
-      "loss": 0.3773,
-      "step": 15650
-    },
-    {
-      "epoch": 1.0228743219397425,
-      "grad_norm": 0.4417157769203186,
-      "learning_rate": 7.579007271871118e-06,
-      "loss": 0.3503,
-      "step": 15651
-    },
-    {
-      "epoch": 1.0229396771452846,
-      "grad_norm": 0.47209954261779785,
-      "learning_rate": 7.578708111656143e-06,
-      "loss": 0.3604,
-      "step": 15652
-    },
-    {
-      "epoch": 1.0230050323508268,
-      "grad_norm": 0.4348152279853821,
-      "learning_rate": 7.578408938863917e-06,
-      "loss": 0.2944,
-      "step": 15653
-    },
-    {
-      "epoch": 1.023070387556369,
-      "grad_norm": 0.4781006872653961,
-      "learning_rate": 7.578109753495895e-06,
-      "loss": 0.3449,
-      "step": 15654
-    },
-    {
-      "epoch": 1.023135742761911,
-      "grad_norm": 0.45873141288757324,
-      "learning_rate": 7.577810555553539e-06,
-      "loss": 0.3276,
-      "step": 15655
-    },
-    {
-      "epoch": 1.023201097967453,
-      "grad_norm": 0.454061895608902,
-      "learning_rate": 7.577511345038311e-06,
-      "loss": 0.3653,
-      "step": 15656
-    },
-    {
-      "epoch": 1.0232664531729951,
-      "grad_norm": 0.4541439116001129,
-      "learning_rate": 7.577212121951664e-06,
-      "loss": 0.3379,
-      "step": 15657
-    },
-    {
-      "epoch": 1.0233318083785374,
-      "grad_norm": 0.43676990270614624,
-      "learning_rate": 7.576912886295063e-06,
-      "loss": 0.3348,
-      "step": 15658
-    },
-    {
-      "epoch": 1.0233971635840795,
-      "grad_norm": 0.43730729818344116,
-      "learning_rate": 7.5766136380699625e-06,
-      "loss": 0.3143,
-      "step": 15659
-    },
-    {
-      "epoch": 1.0234625187896216,
-      "grad_norm": 0.459166944026947,
-      "learning_rate": 7.5763143772778265e-06,
-      "loss": 0.3468,
-      "step": 15660
-    },
-    {
-      "epoch": 1.0235278739951637,
-      "grad_norm": 0.4604553282260895,
-      "learning_rate": 7.576015103920111e-06,
-      "loss": 0.37,
-      "step": 15661
-    },
-    {
-      "epoch": 1.023593229200706,
-      "grad_norm": 0.42918458580970764,
-      "learning_rate": 7.575715817998279e-06,
-      "loss": 0.3299,
-      "step": 15662
-    },
-    {
-      "epoch": 1.023658584406248,
-      "grad_norm": 0.45924463868141174,
-      "learning_rate": 7.575416519513787e-06,
-      "loss": 0.3286,
-      "step": 15663
-    },
-    {
-      "epoch": 1.02372393961179,
-      "grad_norm": 0.4635273814201355,
-      "learning_rate": 7.575117208468099e-06,
-      "loss": 0.3547,
-      "step": 15664
-    },
-    {
-      "epoch": 1.0237892948173322,
-      "grad_norm": 0.5107904672622681,
-      "learning_rate": 7.574817884862671e-06,
-      "loss": 0.4042,
-      "step": 15665
-    },
-    {
-      "epoch": 1.0238546500228742,
-      "grad_norm": 0.4265628457069397,
-      "learning_rate": 7.574518548698964e-06,
-      "loss": 0.3104,
-      "step": 15666
-    },
-    {
-      "epoch": 1.0239200052284165,
-      "grad_norm": 0.44664040207862854,
-      "learning_rate": 7.574219199978438e-06,
-      "loss": 0.3492,
-      "step": 15667
-    },
-    {
-      "epoch": 1.0239853604339586,
-      "grad_norm": 0.4648476541042328,
-      "learning_rate": 7.573919838702553e-06,
-      "loss": 0.3439,
-      "step": 15668
-    },
-    {
-      "epoch": 1.0240507156395007,
-      "grad_norm": 0.46731606125831604,
-      "learning_rate": 7.57362046487277e-06,
-      "loss": 0.3724,
-      "step": 15669
-    },
-    {
-      "epoch": 1.0241160708450427,
-      "grad_norm": 0.39791247248649597,
-      "learning_rate": 7.573321078490548e-06,
-      "loss": 0.2588,
-      "step": 15670
-    },
-    {
-      "epoch": 1.0241814260505848,
-      "grad_norm": 0.4004862308502197,
-      "learning_rate": 7.573021679557347e-06,
-      "loss": 0.3032,
-      "step": 15671
-    },
-    {
-      "epoch": 1.0242467812561271,
-      "grad_norm": 0.47330042719841003,
-      "learning_rate": 7.572722268074628e-06,
-      "loss": 0.3475,
-      "step": 15672
-    },
-    {
-      "epoch": 1.0243121364616692,
-      "grad_norm": 0.45570123195648193,
-      "learning_rate": 7.572422844043852e-06,
-      "loss": 0.3452,
-      "step": 15673
-    },
-    {
-      "epoch": 1.0243774916672113,
-      "grad_norm": 0.4535650908946991,
-      "learning_rate": 7.572123407466479e-06,
-      "loss": 0.3785,
-      "step": 15674
-    },
-    {
-      "epoch": 1.0244428468727533,
-      "grad_norm": 0.4466755986213684,
-      "learning_rate": 7.571823958343968e-06,
-      "loss": 0.35,
-      "step": 15675
-    },
-    {
-      "epoch": 1.0245082020782956,
-      "grad_norm": 0.4653805196285248,
-      "learning_rate": 7.5715244966777816e-06,
-      "loss": 0.3861,
-      "step": 15676
-    },
-    {
-      "epoch": 1.0245735572838377,
-      "grad_norm": 0.4834197163581848,
-      "learning_rate": 7.571225022469377e-06,
-      "loss": 0.3795,
-      "step": 15677
-    },
-    {
-      "epoch": 1.0246389124893798,
-      "grad_norm": 0.46364825963974,
-      "learning_rate": 7.570925535720221e-06,
-      "loss": 0.3252,
-      "step": 15678
-    },
-    {
-      "epoch": 1.0247042676949218,
-      "grad_norm": 0.4515097737312317,
-      "learning_rate": 7.570626036431767e-06,
-      "loss": 0.3314,
-      "step": 15679
-    },
-    {
-      "epoch": 1.024769622900464,
-      "grad_norm": 0.4300325810909271,
-      "learning_rate": 7.570326524605482e-06,
-      "loss": 0.3063,
-      "step": 15680
-    },
-    {
-      "epoch": 1.0248349781060062,
-      "grad_norm": 0.4211832582950592,
-      "learning_rate": 7.570027000242823e-06,
-      "loss": 0.2805,
-      "step": 15681
-    },
-    {
-      "epoch": 1.0249003333115483,
-      "grad_norm": 0.4727987051010132,
-      "learning_rate": 7.569727463345251e-06,
-      "loss": 0.3884,
-      "step": 15682
-    },
-    {
-      "epoch": 1.0249656885170904,
-      "grad_norm": 0.4563107490539551,
-      "learning_rate": 7.569427913914229e-06,
-      "loss": 0.3434,
-      "step": 15683
-    },
-    {
-      "epoch": 1.0250310437226324,
-      "grad_norm": 0.5002644062042236,
-      "learning_rate": 7.569128351951217e-06,
-      "loss": 0.3987,
-      "step": 15684
-    },
-    {
-      "epoch": 1.0250963989281747,
-      "grad_norm": 0.4213293790817261,
-      "learning_rate": 7.5688287774576756e-06,
-      "loss": 0.2893,
-      "step": 15685
-    },
-    {
-      "epoch": 1.0251617541337168,
-      "grad_norm": 0.45721110701560974,
-      "learning_rate": 7.568529190435066e-06,
-      "loss": 0.3567,
-      "step": 15686
-    },
-    {
-      "epoch": 1.0252271093392589,
-      "grad_norm": 0.43511372804641724,
-      "learning_rate": 7.56822959088485e-06,
-      "loss": 0.3365,
-      "step": 15687
-    },
-    {
-      "epoch": 1.025292464544801,
-      "grad_norm": 0.4605725109577179,
-      "learning_rate": 7.567929978808488e-06,
-      "loss": 0.377,
-      "step": 15688
-    },
-    {
-      "epoch": 1.025357819750343,
-      "grad_norm": 0.4372442662715912,
-      "learning_rate": 7.567630354207443e-06,
-      "loss": 0.3495,
-      "step": 15689
-    },
-    {
-      "epoch": 1.0254231749558853,
-      "grad_norm": 0.4581087827682495,
-      "learning_rate": 7.567330717083174e-06,
-      "loss": 0.34,
-      "step": 15690
-    },
-    {
-      "epoch": 1.0254885301614274,
-      "grad_norm": 0.44667884707450867,
-      "learning_rate": 7.567031067437146e-06,
-      "loss": 0.3298,
-      "step": 15691
-    },
-    {
-      "epoch": 1.0255538853669695,
-      "grad_norm": 0.4301671087741852,
-      "learning_rate": 7.566731405270815e-06,
-      "loss": 0.3281,
-      "step": 15692
-    },
-    {
-      "epoch": 1.0256192405725115,
-      "grad_norm": 0.44388964772224426,
-      "learning_rate": 7.5664317305856485e-06,
-      "loss": 0.3654,
-      "step": 15693
-    },
-    {
-      "epoch": 1.0256845957780538,
-      "grad_norm": 0.4563218355178833,
-      "learning_rate": 7.566132043383105e-06,
-      "loss": 0.3612,
-      "step": 15694
-    },
-    {
-      "epoch": 1.025749950983596,
-      "grad_norm": 0.4254835546016693,
-      "learning_rate": 7.565832343664645e-06,
-      "loss": 0.322,
-      "step": 15695
-    },
-    {
-      "epoch": 1.025815306189138,
-      "grad_norm": 0.4602677822113037,
-      "learning_rate": 7.565532631431734e-06,
-      "loss": 0.3146,
-      "step": 15696
-    },
-    {
-      "epoch": 1.02588066139468,
-      "grad_norm": 0.4197689890861511,
-      "learning_rate": 7.565232906685829e-06,
-      "loss": 0.3123,
-      "step": 15697
-    },
-    {
-      "epoch": 1.025946016600222,
-      "grad_norm": 0.4866635501384735,
-      "learning_rate": 7.564933169428396e-06,
-      "loss": 0.3787,
-      "step": 15698
-    },
-    {
-      "epoch": 1.0260113718057644,
-      "grad_norm": 0.4042617976665497,
-      "learning_rate": 7.564633419660894e-06,
-      "loss": 0.2913,
-      "step": 15699
-    },
-    {
-      "epoch": 1.0260767270113065,
-      "grad_norm": 0.45293474197387695,
-      "learning_rate": 7.564333657384788e-06,
-      "loss": 0.3151,
-      "step": 15700
-    },
-    {
-      "epoch": 1.0261420822168486,
-      "grad_norm": 0.5300982594490051,
-      "learning_rate": 7.564033882601538e-06,
-      "loss": 0.4056,
-      "step": 15701
-    },
-    {
-      "epoch": 1.0262074374223906,
-      "grad_norm": 0.4827274680137634,
-      "learning_rate": 7.563734095312606e-06,
-      "loss": 0.3787,
-      "step": 15702
-    },
-    {
-      "epoch": 1.026272792627933,
-      "grad_norm": 0.44669026136398315,
-      "learning_rate": 7.563434295519454e-06,
-      "loss": 0.3386,
-      "step": 15703
-    },
-    {
-      "epoch": 1.026338147833475,
-      "grad_norm": 0.4354687035083771,
-      "learning_rate": 7.563134483223547e-06,
-      "loss": 0.3434,
-      "step": 15704
-    },
-    {
-      "epoch": 1.026403503039017,
-      "grad_norm": 0.4427279829978943,
-      "learning_rate": 7.562834658426342e-06,
-      "loss": 0.3398,
-      "step": 15705
-    },
-    {
-      "epoch": 1.0264688582445591,
-      "grad_norm": 0.43261969089508057,
-      "learning_rate": 7.562534821129307e-06,
-      "loss": 0.3103,
-      "step": 15706
-    },
-    {
-      "epoch": 1.0265342134501012,
-      "grad_norm": 0.4481985867023468,
-      "learning_rate": 7.562234971333901e-06,
-      "loss": 0.3434,
-      "step": 15707
-    },
-    {
-      "epoch": 1.0265995686556435,
-      "grad_norm": 0.48565202951431274,
-      "learning_rate": 7.561935109041588e-06,
-      "loss": 0.3725,
-      "step": 15708
-    },
-    {
-      "epoch": 1.0266649238611856,
-      "grad_norm": 0.46509653329849243,
-      "learning_rate": 7.56163523425383e-06,
-      "loss": 0.3693,
-      "step": 15709
-    },
-    {
-      "epoch": 1.0267302790667276,
-      "grad_norm": 0.41026821732521057,
-      "learning_rate": 7.561335346972088e-06,
-      "loss": 0.305,
-      "step": 15710
-    },
-    {
-      "epoch": 1.0267956342722697,
-      "grad_norm": 0.411944180727005,
-      "learning_rate": 7.561035447197828e-06,
-      "loss": 0.2877,
-      "step": 15711
-    },
-    {
-      "epoch": 1.026860989477812,
-      "grad_norm": 0.43637987971305847,
-      "learning_rate": 7.56073553493251e-06,
-      "loss": 0.3183,
-      "step": 15712
-    },
-    {
-      "epoch": 1.026926344683354,
-      "grad_norm": 0.4577435255050659,
-      "learning_rate": 7.560435610177599e-06,
-      "loss": 0.3603,
-      "step": 15713
-    },
-    {
-      "epoch": 1.0269916998888962,
-      "grad_norm": 0.4846012592315674,
-      "learning_rate": 7.560135672934554e-06,
-      "loss": 0.3273,
-      "step": 15714
-    },
-    {
-      "epoch": 1.0270570550944382,
-      "grad_norm": 0.4223800003528595,
-      "learning_rate": 7.559835723204842e-06,
-      "loss": 0.3203,
-      "step": 15715
-    },
-    {
-      "epoch": 1.0271224102999803,
-      "grad_norm": 0.44712093472480774,
-      "learning_rate": 7.559535760989924e-06,
-      "loss": 0.3608,
-      "step": 15716
-    },
-    {
-      "epoch": 1.0271877655055226,
-      "grad_norm": 0.4641527235507965,
-      "learning_rate": 7.559235786291264e-06,
-      "loss": 0.3396,
-      "step": 15717
-    },
-    {
-      "epoch": 1.0272531207110647,
-      "grad_norm": 0.4962295591831207,
-      "learning_rate": 7.558935799110324e-06,
-      "loss": 0.3735,
-      "step": 15718
-    },
-    {
-      "epoch": 1.0273184759166067,
-      "grad_norm": 0.4628616273403168,
-      "learning_rate": 7.558635799448567e-06,
-      "loss": 0.3774,
-      "step": 15719
-    },
-    {
-      "epoch": 1.0273838311221488,
-      "grad_norm": 0.4333202838897705,
-      "learning_rate": 7.558335787307458e-06,
-      "loss": 0.3188,
-      "step": 15720
-    },
-    {
-      "epoch": 1.0274491863276909,
-      "grad_norm": 0.4812498688697815,
-      "learning_rate": 7.5580357626884584e-06,
-      "loss": 0.36,
-      "step": 15721
-    },
-    {
-      "epoch": 1.0275145415332332,
-      "grad_norm": 0.45785200595855713,
-      "learning_rate": 7.5577357255930336e-06,
-      "loss": 0.3638,
-      "step": 15722
-    },
-    {
-      "epoch": 1.0275798967387753,
-      "grad_norm": 0.4334566593170166,
-      "learning_rate": 7.557435676022643e-06,
-      "loss": 0.3177,
-      "step": 15723
-    },
-    {
-      "epoch": 1.0276452519443173,
-      "grad_norm": 0.4395442306995392,
-      "learning_rate": 7.5571356139787546e-06,
-      "loss": 0.3307,
-      "step": 15724
-    },
-    {
-      "epoch": 1.0277106071498594,
-      "grad_norm": 0.44063910841941833,
-      "learning_rate": 7.55683553946283e-06,
-      "loss": 0.3013,
-      "step": 15725
-    },
-    {
-      "epoch": 1.0277759623554017,
-      "grad_norm": 0.3947782814502716,
-      "learning_rate": 7.5565354524763305e-06,
-      "loss": 0.251,
-      "step": 15726
-    },
-    {
-      "epoch": 1.0278413175609438,
-      "grad_norm": 0.45565927028656006,
-      "learning_rate": 7.556235353020725e-06,
-      "loss": 0.3636,
-      "step": 15727
-    },
-    {
-      "epoch": 1.0279066727664858,
-      "grad_norm": 0.437971830368042,
-      "learning_rate": 7.555935241097472e-06,
-      "loss": 0.328,
-      "step": 15728
-    },
-    {
-      "epoch": 1.027972027972028,
-      "grad_norm": 0.4572085440158844,
-      "learning_rate": 7.555635116708037e-06,
-      "loss": 0.3514,
-      "step": 15729
-    },
-    {
-      "epoch": 1.02803738317757,
-      "grad_norm": 0.4883074164390564,
-      "learning_rate": 7.555334979853886e-06,
-      "loss": 0.3838,
-      "step": 15730
-    },
-    {
-      "epoch": 1.0281027383831123,
-      "grad_norm": 0.45139598846435547,
-      "learning_rate": 7.555034830536479e-06,
-      "loss": 0.3583,
-      "step": 15731
-    },
-    {
-      "epoch": 1.0281680935886544,
-      "grad_norm": 0.461537629365921,
-      "learning_rate": 7.554734668757282e-06,
-      "loss": 0.3387,
-      "step": 15732
-    },
-    {
-      "epoch": 1.0282334487941964,
-      "grad_norm": 0.422015905380249,
-      "learning_rate": 7.554434494517762e-06,
-      "loss": 0.3131,
-      "step": 15733
-    },
-    {
-      "epoch": 1.0282988039997385,
-      "grad_norm": 0.4193853437900543,
-      "learning_rate": 7.5541343078193784e-06,
-      "loss": 0.3254,
-      "step": 15734
-    },
-    {
-      "epoch": 1.0283641592052808,
-      "grad_norm": 0.48123207688331604,
-      "learning_rate": 7.553834108663596e-06,
-      "loss": 0.3688,
-      "step": 15735
-    },
-    {
-      "epoch": 1.0284295144108229,
-      "grad_norm": 0.4750046730041504,
-      "learning_rate": 7.55353389705188e-06,
-      "loss": 0.3578,
-      "step": 15736
-    },
-    {
-      "epoch": 1.028494869616365,
-      "grad_norm": 0.44931864738464355,
-      "learning_rate": 7.553233672985695e-06,
-      "loss": 0.3192,
-      "step": 15737
-    },
-    {
-      "epoch": 1.028560224821907,
-      "grad_norm": 0.4558142125606537,
-      "learning_rate": 7.552933436466505e-06,
-      "loss": 0.3411,
-      "step": 15738
-    },
-    {
-      "epoch": 1.028625580027449,
-      "grad_norm": 0.4375033378601074,
-      "learning_rate": 7.552633187495774e-06,
-      "loss": 0.2989,
-      "step": 15739
-    },
-    {
-      "epoch": 1.0286909352329914,
-      "grad_norm": 0.45783933997154236,
-      "learning_rate": 7.5523329260749665e-06,
-      "loss": 0.3227,
-      "step": 15740
-    },
-    {
-      "epoch": 1.0287562904385334,
-      "grad_norm": 0.41327139735221863,
-      "learning_rate": 7.552032652205548e-06,
-      "loss": 0.2733,
-      "step": 15741
-    },
-    {
-      "epoch": 1.0288216456440755,
-      "grad_norm": 0.47870659828186035,
-      "learning_rate": 7.551732365888982e-06,
-      "loss": 0.3487,
-      "step": 15742
-    },
-    {
-      "epoch": 1.0288870008496176,
-      "grad_norm": 0.4363199770450592,
-      "learning_rate": 7.551432067126732e-06,
-      "loss": 0.3411,
-      "step": 15743
-    },
-    {
-      "epoch": 1.02895235605516,
-      "grad_norm": 0.4649825096130371,
-      "learning_rate": 7.5511317559202646e-06,
-      "loss": 0.3686,
-      "step": 15744
-    },
-    {
-      "epoch": 1.029017711260702,
-      "grad_norm": 0.44911760091781616,
-      "learning_rate": 7.550831432271045e-06,
-      "loss": 0.3485,
-      "step": 15745
-    },
-    {
-      "epoch": 1.029083066466244,
-      "grad_norm": 0.4803098440170288,
-      "learning_rate": 7.550531096180536e-06,
-      "loss": 0.4048,
-      "step": 15746
-    },
-    {
-      "epoch": 1.029148421671786,
-      "grad_norm": 0.4453189969062805,
-      "learning_rate": 7.550230747650202e-06,
-      "loss": 0.3577,
-      "step": 15747
-    },
-    {
-      "epoch": 1.0292137768773282,
-      "grad_norm": 0.4523233473300934,
-      "learning_rate": 7.54993038668151e-06,
-      "loss": 0.3569,
-      "step": 15748
-    },
-    {
-      "epoch": 1.0292791320828705,
-      "grad_norm": 0.4204012453556061,
-      "learning_rate": 7.549630013275924e-06,
-      "loss": 0.3072,
-      "step": 15749
-    },
-    {
-      "epoch": 1.0293444872884125,
-      "grad_norm": 0.44491085410118103,
-      "learning_rate": 7.549329627434909e-06,
-      "loss": 0.3342,
-      "step": 15750
-    },
-    {
-      "epoch": 1.0294098424939546,
-      "grad_norm": 0.4380131959915161,
-      "learning_rate": 7.549029229159932e-06,
-      "loss": 0.3425,
-      "step": 15751
-    },
-    {
-      "epoch": 1.0294751976994967,
-      "grad_norm": 0.42915481328964233,
-      "learning_rate": 7.548728818452456e-06,
-      "loss": 0.3047,
-      "step": 15752
-    },
-    {
-      "epoch": 1.029540552905039,
-      "grad_norm": 0.4336947500705719,
-      "learning_rate": 7.5484283953139445e-06,
-      "loss": 0.3354,
-      "step": 15753
-    },
-    {
-      "epoch": 1.029605908110581,
-      "grad_norm": 0.4183482527732849,
-      "learning_rate": 7.548127959745866e-06,
-      "loss": 0.2959,
-      "step": 15754
-    },
-    {
-      "epoch": 1.0296712633161231,
-      "grad_norm": 0.49265751242637634,
-      "learning_rate": 7.547827511749684e-06,
-      "loss": 0.4263,
-      "step": 15755
-    },
-    {
-      "epoch": 1.0297366185216652,
-      "grad_norm": 0.4421324133872986,
-      "learning_rate": 7.547527051326866e-06,
-      "loss": 0.308,
-      "step": 15756
-    },
-    {
-      "epoch": 1.0298019737272073,
-      "grad_norm": 0.442550927400589,
-      "learning_rate": 7.547226578478874e-06,
-      "loss": 0.3369,
-      "step": 15757
-    },
-    {
-      "epoch": 1.0298673289327496,
-      "grad_norm": 0.4535270929336548,
-      "learning_rate": 7.546926093207178e-06,
-      "loss": 0.3429,
-      "step": 15758
-    },
-    {
-      "epoch": 1.0299326841382916,
-      "grad_norm": 0.46581920981407166,
-      "learning_rate": 7.54662559551324e-06,
-      "loss": 0.3119,
-      "step": 15759
-    },
-    {
-      "epoch": 1.0299980393438337,
-      "grad_norm": 0.43248432874679565,
-      "learning_rate": 7.546325085398526e-06,
-      "loss": 0.3178,
-      "step": 15760
-    },
-    {
-      "epoch": 1.0300633945493758,
-      "grad_norm": 0.46498411893844604,
-      "learning_rate": 7.546024562864503e-06,
-      "loss": 0.383,
-      "step": 15761
-    },
-    {
-      "epoch": 1.030128749754918,
-      "grad_norm": 0.45315492153167725,
-      "learning_rate": 7.545724027912635e-06,
-      "loss": 0.3368,
-      "step": 15762
-    },
-    {
-      "epoch": 1.0301941049604602,
-      "grad_norm": 0.47556623816490173,
-      "learning_rate": 7.545423480544392e-06,
-      "loss": 0.3694,
-      "step": 15763
-    },
-    {
-      "epoch": 1.0302594601660022,
-      "grad_norm": 0.4934963583946228,
-      "learning_rate": 7.545122920761235e-06,
-      "loss": 0.3796,
-      "step": 15764
-    },
-    {
-      "epoch": 1.0303248153715443,
-      "grad_norm": 0.45008614659309387,
-      "learning_rate": 7.544822348564633e-06,
-      "loss": 0.3567,
-      "step": 15765
-    },
-    {
-      "epoch": 1.0303901705770864,
-      "grad_norm": 0.48054632544517517,
-      "learning_rate": 7.544521763956048e-06,
-      "loss": 0.3848,
-      "step": 15766
-    },
-    {
-      "epoch": 1.0304555257826287,
-      "grad_norm": 0.4531192183494568,
-      "learning_rate": 7.544221166936951e-06,
-      "loss": 0.38,
-      "step": 15767
-    },
-    {
-      "epoch": 1.0305208809881707,
-      "grad_norm": 0.47871488332748413,
-      "learning_rate": 7.543920557508806e-06,
-      "loss": 0.2815,
-      "step": 15768
-    },
-    {
-      "epoch": 1.0305862361937128,
-      "grad_norm": 0.4792271554470062,
-      "learning_rate": 7.543619935673079e-06,
-      "loss": 0.3528,
-      "step": 15769
-    },
-    {
-      "epoch": 1.0306515913992549,
-      "grad_norm": 0.4431339204311371,
-      "learning_rate": 7.543319301431235e-06,
-      "loss": 0.3456,
-      "step": 15770
-    },
-    {
-      "epoch": 1.0307169466047972,
-      "grad_norm": 0.45511385798454285,
-      "learning_rate": 7.543018654784743e-06,
-      "loss": 0.3498,
-      "step": 15771
-    },
-    {
-      "epoch": 1.0307823018103393,
-      "grad_norm": 0.44410762190818787,
-      "learning_rate": 7.542717995735068e-06,
-      "loss": 0.3237,
-      "step": 15772
-    },
-    {
-      "epoch": 1.0308476570158813,
-      "grad_norm": 0.43734708428382874,
-      "learning_rate": 7.542417324283675e-06,
-      "loss": 0.3369,
-      "step": 15773
-    },
-    {
-      "epoch": 1.0309130122214234,
-      "grad_norm": 0.42639923095703125,
-      "learning_rate": 7.542116640432035e-06,
-      "loss": 0.3141,
-      "step": 15774
-    },
-    {
-      "epoch": 1.0309783674269655,
-      "grad_norm": 0.4694978594779968,
-      "learning_rate": 7.541815944181609e-06,
-      "loss": 0.3422,
-      "step": 15775
-    },
-    {
-      "epoch": 1.0310437226325078,
-      "grad_norm": 0.45306161046028137,
-      "learning_rate": 7.541515235533866e-06,
-      "loss": 0.3126,
-      "step": 15776
-    },
-    {
-      "epoch": 1.0311090778380498,
-      "grad_norm": 0.4308871626853943,
-      "learning_rate": 7.5412145144902735e-06,
-      "loss": 0.3519,
-      "step": 15777
-    },
-    {
-      "epoch": 1.031174433043592,
-      "grad_norm": 0.4713728427886963,
-      "learning_rate": 7.540913781052297e-06,
-      "loss": 0.339,
-      "step": 15778
-    },
-    {
-      "epoch": 1.031239788249134,
-      "grad_norm": 0.49831151962280273,
-      "learning_rate": 7.5406130352214045e-06,
-      "loss": 0.3671,
-      "step": 15779
-    },
-    {
-      "epoch": 1.0313051434546763,
-      "grad_norm": 0.43398699164390564,
-      "learning_rate": 7.540312276999062e-06,
-      "loss": 0.3094,
-      "step": 15780
-    },
-    {
-      "epoch": 1.0313704986602183,
-      "grad_norm": 0.42147430777549744,
-      "learning_rate": 7.540011506386736e-06,
-      "loss": 0.3189,
-      "step": 15781
-    },
-    {
-      "epoch": 1.0314358538657604,
-      "grad_norm": 0.4849579632282257,
-      "learning_rate": 7.539710723385894e-06,
-      "loss": 0.3753,
-      "step": 15782
-    },
-    {
-      "epoch": 1.0315012090713025,
-      "grad_norm": 0.4356859028339386,
-      "learning_rate": 7.5394099279980025e-06,
-      "loss": 0.3375,
-      "step": 15783
-    },
-    {
-      "epoch": 1.0315665642768446,
-      "grad_norm": 0.43548744916915894,
-      "learning_rate": 7.539109120224529e-06,
-      "loss": 0.3086,
-      "step": 15784
-    },
-    {
-      "epoch": 1.0316319194823869,
-      "grad_norm": 0.40212714672088623,
-      "learning_rate": 7.538808300066943e-06,
-      "loss": 0.2989,
-      "step": 15785
-    },
-    {
-      "epoch": 1.031697274687929,
-      "grad_norm": 0.42388996481895447,
-      "learning_rate": 7.538507467526708e-06,
-      "loss": 0.3054,
-      "step": 15786
-    },
-    {
-      "epoch": 1.031762629893471,
-      "grad_norm": 0.4657570421695709,
-      "learning_rate": 7.538206622605292e-06,
-      "loss": 0.3469,
-      "step": 15787
-    },
-    {
-      "epoch": 1.031827985099013,
-      "grad_norm": 0.4387129545211792,
-      "learning_rate": 7.5379057653041635e-06,
-      "loss": 0.35,
-      "step": 15788
-    },
-    {
-      "epoch": 1.0318933403045552,
-      "grad_norm": 0.41033414006233215,
-      "learning_rate": 7.5376048956247886e-06,
-      "loss": 0.2854,
-      "step": 15789
-    },
-    {
-      "epoch": 1.0319586955100974,
-      "grad_norm": 0.48620110750198364,
-      "learning_rate": 7.537304013568636e-06,
-      "loss": 0.378,
-      "step": 15790
-    },
-    {
-      "epoch": 1.0320240507156395,
-      "grad_norm": 0.43816041946411133,
-      "learning_rate": 7.537003119137173e-06,
-      "loss": 0.3215,
-      "step": 15791
-    },
-    {
-      "epoch": 1.0320894059211816,
-      "grad_norm": 0.4366806745529175,
-      "learning_rate": 7.536702212331867e-06,
-      "loss": 0.3555,
-      "step": 15792
-    },
-    {
-      "epoch": 1.0321547611267237,
-      "grad_norm": 0.433204710483551,
-      "learning_rate": 7.5364012931541844e-06,
-      "loss": 0.3106,
-      "step": 15793
-    },
-    {
-      "epoch": 1.032220116332266,
-      "grad_norm": 0.434613436460495,
-      "learning_rate": 7.536100361605595e-06,
-      "loss": 0.325,
-      "step": 15794
-    },
-    {
-      "epoch": 1.032285471537808,
-      "grad_norm": 0.4877561926841736,
-      "learning_rate": 7.535799417687565e-06,
-      "loss": 0.3829,
-      "step": 15795
-    },
-    {
-      "epoch": 1.03235082674335,
-      "grad_norm": 0.468234658241272,
-      "learning_rate": 7.535498461401563e-06,
-      "loss": 0.3629,
-      "step": 15796
-    },
-    {
-      "epoch": 1.0324161819488922,
-      "grad_norm": 0.43487778306007385,
-      "learning_rate": 7.535197492749057e-06,
-      "loss": 0.3268,
-      "step": 15797
-    },
-    {
-      "epoch": 1.0324815371544342,
-      "grad_norm": 0.45069482922554016,
-      "learning_rate": 7.534896511731514e-06,
-      "loss": 0.3265,
-      "step": 15798
-    },
-    {
-      "epoch": 1.0325468923599765,
-      "grad_norm": 0.4581204056739807,
-      "learning_rate": 7.534595518350403e-06,
-      "loss": 0.3389,
-      "step": 15799
-    },
-    {
-      "epoch": 1.0326122475655186,
-      "grad_norm": 0.43256765604019165,
-      "learning_rate": 7.534294512607191e-06,
-      "loss": 0.3505,
-      "step": 15800
-    },
-    {
-      "epoch": 1.0326776027710607,
-      "grad_norm": 0.5151135921478271,
-      "learning_rate": 7.533993494503347e-06,
-      "loss": 0.41,
-      "step": 15801
-    },
-    {
-      "epoch": 1.0327429579766028,
-      "grad_norm": 0.4975196421146393,
-      "learning_rate": 7.533692464040338e-06,
-      "loss": 0.369,
-      "step": 15802
-    },
-    {
-      "epoch": 1.032808313182145,
-      "grad_norm": 0.4950621724128723,
-      "learning_rate": 7.5333914212196355e-06,
-      "loss": 0.359,
-      "step": 15803
-    },
-    {
-      "epoch": 1.0328736683876871,
-      "grad_norm": 0.46530085802078247,
-      "learning_rate": 7.533090366042703e-06,
-      "loss": 0.3691,
-      "step": 15804
-    },
-    {
-      "epoch": 1.0329390235932292,
-      "grad_norm": 0.4382416307926178,
-      "learning_rate": 7.532789298511013e-06,
-      "loss": 0.3219,
-      "step": 15805
-    },
-    {
-      "epoch": 1.0330043787987713,
-      "grad_norm": 0.4749968647956848,
-      "learning_rate": 7.532488218626032e-06,
-      "loss": 0.3661,
-      "step": 15806
-    },
-    {
-      "epoch": 1.0330697340043133,
-      "grad_norm": 0.4506654739379883,
-      "learning_rate": 7.5321871263892275e-06,
-      "loss": 0.3336,
-      "step": 15807
-    },
-    {
-      "epoch": 1.0331350892098556,
-      "grad_norm": 0.43492308259010315,
-      "learning_rate": 7.53188602180207e-06,
-      "loss": 0.3099,
-      "step": 15808
-    },
-    {
-      "epoch": 1.0332004444153977,
-      "grad_norm": 0.4308479130268097,
-      "learning_rate": 7.531584904866027e-06,
-      "loss": 0.3346,
-      "step": 15809
-    },
-    {
-      "epoch": 1.0332657996209398,
-      "grad_norm": 0.4674423336982727,
-      "learning_rate": 7.5312837755825676e-06,
-      "loss": 0.3869,
-      "step": 15810
-    },
-    {
-      "epoch": 1.0333311548264819,
-      "grad_norm": 0.45534729957580566,
-      "learning_rate": 7.530982633953159e-06,
-      "loss": 0.3499,
-      "step": 15811
-    },
-    {
-      "epoch": 1.0333965100320242,
-      "grad_norm": 0.49120113253593445,
-      "learning_rate": 7.530681479979273e-06,
-      "loss": 0.3672,
-      "step": 15812
-    },
-    {
-      "epoch": 1.0334618652375662,
-      "grad_norm": 0.44581127166748047,
-      "learning_rate": 7.5303803136623755e-06,
-      "loss": 0.332,
-      "step": 15813
-    },
-    {
-      "epoch": 1.0335272204431083,
-      "grad_norm": 0.42955002188682556,
-      "learning_rate": 7.530079135003937e-06,
-      "loss": 0.3419,
-      "step": 15814
-    },
-    {
-      "epoch": 1.0335925756486504,
-      "grad_norm": 0.45427507162094116,
-      "learning_rate": 7.529777944005425e-06,
-      "loss": 0.3561,
-      "step": 15815
-    },
-    {
-      "epoch": 1.0336579308541924,
-      "grad_norm": 0.45866936445236206,
-      "learning_rate": 7.529476740668311e-06,
-      "loss": 0.3328,
-      "step": 15816
-    },
-    {
-      "epoch": 1.0337232860597347,
-      "grad_norm": 0.42922618985176086,
-      "learning_rate": 7.529175524994063e-06,
-      "loss": 0.3147,
-      "step": 15817
-    },
-    {
-      "epoch": 1.0337886412652768,
-      "grad_norm": 0.46843546628952026,
-      "learning_rate": 7.528874296984149e-06,
-      "loss": 0.3595,
-      "step": 15818
-    },
-    {
-      "epoch": 1.0338539964708189,
-      "grad_norm": 0.4613758325576782,
-      "learning_rate": 7.528573056640039e-06,
-      "loss": 0.3531,
-      "step": 15819
-    },
-    {
-      "epoch": 1.033919351676361,
-      "grad_norm": 0.4563542902469635,
-      "learning_rate": 7.528271803963202e-06,
-      "loss": 0.3467,
-      "step": 15820
-    },
-    {
-      "epoch": 1.0339847068819032,
-      "grad_norm": 0.41521739959716797,
-      "learning_rate": 7.527970538955109e-06,
-      "loss": 0.3076,
-      "step": 15821
-    },
-    {
-      "epoch": 1.0340500620874453,
-      "grad_norm": 0.4420032203197479,
-      "learning_rate": 7.5276692616172254e-06,
-      "loss": 0.3262,
-      "step": 15822
-    },
-    {
-      "epoch": 1.0341154172929874,
-      "grad_norm": 0.48577195405960083,
-      "learning_rate": 7.527367971951025e-06,
-      "loss": 0.3784,
-      "step": 15823
-    },
-    {
-      "epoch": 1.0341807724985295,
-      "grad_norm": 0.47907888889312744,
-      "learning_rate": 7.527066669957974e-06,
-      "loss": 0.3493,
-      "step": 15824
-    },
-    {
-      "epoch": 1.0342461277040715,
-      "grad_norm": 0.48235344886779785,
-      "learning_rate": 7.526765355639545e-06,
-      "loss": 0.3887,
-      "step": 15825
-    },
-    {
-      "epoch": 1.0343114829096138,
-      "grad_norm": 0.4414643347263336,
-      "learning_rate": 7.5264640289972045e-06,
-      "loss": 0.3285,
-      "step": 15826
-    },
-    {
-      "epoch": 1.034376838115156,
-      "grad_norm": 0.4412213861942291,
-      "learning_rate": 7.526162690032426e-06,
-      "loss": 0.346,
-      "step": 15827
-    },
-    {
-      "epoch": 1.034442193320698,
-      "grad_norm": 0.44358623027801514,
-      "learning_rate": 7.525861338746676e-06,
-      "loss": 0.3407,
-      "step": 15828
-    },
-    {
-      "epoch": 1.03450754852624,
-      "grad_norm": 0.45164433121681213,
-      "learning_rate": 7.525559975141423e-06,
-      "loss": 0.3273,
-      "step": 15829
-    },
-    {
-      "epoch": 1.0345729037317821,
-      "grad_norm": 0.4686276614665985,
-      "learning_rate": 7.5252585992181415e-06,
-      "loss": 0.363,
-      "step": 15830
-    },
-    {
-      "epoch": 1.0346382589373244,
-      "grad_norm": 0.5003420114517212,
-      "learning_rate": 7.524957210978297e-06,
-      "loss": 0.3692,
-      "step": 15831
-    },
-    {
-      "epoch": 1.0347036141428665,
-      "grad_norm": 0.4188174307346344,
-      "learning_rate": 7.524655810423364e-06,
-      "loss": 0.327,
-      "step": 15832
-    },
-    {
-      "epoch": 1.0347689693484086,
-      "grad_norm": 0.4743111729621887,
-      "learning_rate": 7.524354397554807e-06,
-      "loss": 0.3685,
-      "step": 15833
-    },
-    {
-      "epoch": 1.0348343245539506,
-      "grad_norm": 0.4771732687950134,
-      "learning_rate": 7.524052972374102e-06,
-      "loss": 0.3671,
-      "step": 15834
-    },
-    {
-      "epoch": 1.034899679759493,
-      "grad_norm": 0.4349338412284851,
-      "learning_rate": 7.523751534882714e-06,
-      "loss": 0.3602,
-      "step": 15835
-    },
-    {
-      "epoch": 1.034965034965035,
-      "grad_norm": 0.5162888169288635,
-      "learning_rate": 7.523450085082117e-06,
-      "loss": 0.3988,
-      "step": 15836
-    },
-    {
-      "epoch": 1.035030390170577,
-      "grad_norm": 0.44995149970054626,
-      "learning_rate": 7.523148622973779e-06,
-      "loss": 0.3276,
-      "step": 15837
-    },
-    {
-      "epoch": 1.0350957453761191,
-      "grad_norm": 0.47283655405044556,
-      "learning_rate": 7.522847148559171e-06,
-      "loss": 0.3675,
-      "step": 15838
-    },
-    {
-      "epoch": 1.0351611005816612,
-      "grad_norm": 0.4758037030696869,
-      "learning_rate": 7.5225456618397645e-06,
-      "loss": 0.3598,
-      "step": 15839
-    },
-    {
-      "epoch": 1.0352264557872035,
-      "grad_norm": 0.42274487018585205,
-      "learning_rate": 7.522244162817027e-06,
-      "loss": 0.2811,
-      "step": 15840
-    },
-    {
-      "epoch": 1.0352918109927456,
-      "grad_norm": 0.43868711590766907,
-      "learning_rate": 7.521942651492432e-06,
-      "loss": 0.355,
-      "step": 15841
-    },
-    {
-      "epoch": 1.0353571661982877,
-      "grad_norm": 0.46339425444602966,
-      "learning_rate": 7.521641127867448e-06,
-      "loss": 0.3626,
-      "step": 15842
-    },
-    {
-      "epoch": 1.0354225214038297,
-      "grad_norm": 0.4463191330432892,
-      "learning_rate": 7.5213395919435486e-06,
-      "loss": 0.3418,
-      "step": 15843
-    },
-    {
-      "epoch": 1.035487876609372,
-      "grad_norm": 0.45070981979370117,
-      "learning_rate": 7.521038043722202e-06,
-      "loss": 0.3415,
-      "step": 15844
-    },
-    {
-      "epoch": 1.035553231814914,
-      "grad_norm": 0.44040626287460327,
-      "learning_rate": 7.520736483204878e-06,
-      "loss": 0.3169,
-      "step": 15845
-    },
-    {
-      "epoch": 1.0356185870204562,
-      "grad_norm": 0.48595234751701355,
-      "learning_rate": 7.520434910393051e-06,
-      "loss": 0.3627,
-      "step": 15846
-    },
-    {
-      "epoch": 1.0356839422259982,
-      "grad_norm": 0.424405038356781,
-      "learning_rate": 7.5201333252881884e-06,
-      "loss": 0.2826,
-      "step": 15847
-    },
-    {
-      "epoch": 1.0357492974315403,
-      "grad_norm": 0.47548767924308777,
-      "learning_rate": 7.519831727891763e-06,
-      "loss": 0.3796,
-      "step": 15848
-    },
-    {
-      "epoch": 1.0358146526370826,
-      "grad_norm": 0.4742758572101593,
-      "learning_rate": 7.5195301182052445e-06,
-      "loss": 0.3719,
-      "step": 15849
-    },
-    {
-      "epoch": 1.0358800078426247,
-      "grad_norm": 0.463164359331131,
-      "learning_rate": 7.519228496230107e-06,
-      "loss": 0.3471,
-      "step": 15850
-    },
-    {
-      "epoch": 1.0359453630481668,
-      "grad_norm": 0.43213051557540894,
-      "learning_rate": 7.5189268619678165e-06,
-      "loss": 0.328,
-      "step": 15851
-    },
-    {
-      "epoch": 1.0360107182537088,
-      "grad_norm": 0.4856230914592743,
-      "learning_rate": 7.518625215419848e-06,
-      "loss": 0.3789,
-      "step": 15852
-    },
-    {
-      "epoch": 1.0360760734592511,
-      "grad_norm": 0.47976210713386536,
-      "learning_rate": 7.518323556587672e-06,
-      "loss": 0.3589,
-      "step": 15853
-    },
-    {
-      "epoch": 1.0361414286647932,
-      "grad_norm": 0.44848567247390747,
-      "learning_rate": 7.518021885472759e-06,
-      "loss": 0.3257,
-      "step": 15854
-    },
-    {
-      "epoch": 1.0362067838703353,
-      "grad_norm": 0.4461570680141449,
-      "learning_rate": 7.517720202076583e-06,
-      "loss": 0.3669,
-      "step": 15855
-    },
-    {
-      "epoch": 1.0362721390758773,
-      "grad_norm": 0.4441314935684204,
-      "learning_rate": 7.517418506400611e-06,
-      "loss": 0.3579,
-      "step": 15856
-    },
-    {
-      "epoch": 1.0363374942814194,
-      "grad_norm": 0.3970010280609131,
-      "learning_rate": 7.5171167984463175e-06,
-      "loss": 0.3013,
-      "step": 15857
-    },
-    {
-      "epoch": 1.0364028494869617,
-      "grad_norm": 0.4426477551460266,
-      "learning_rate": 7.516815078215174e-06,
-      "loss": 0.3418,
-      "step": 15858
-    },
-    {
-      "epoch": 1.0364682046925038,
-      "grad_norm": 0.4288714826107025,
-      "learning_rate": 7.516513345708651e-06,
-      "loss": 0.324,
-      "step": 15859
-    },
-    {
-      "epoch": 1.0365335598980459,
-      "grad_norm": 0.45848506689071655,
-      "learning_rate": 7.51621160092822e-06,
-      "loss": 0.3737,
-      "step": 15860
-    },
-    {
-      "epoch": 1.036598915103588,
-      "grad_norm": 0.4246535301208496,
-      "learning_rate": 7.515909843875355e-06,
-      "loss": 0.35,
-      "step": 15861
-    },
-    {
-      "epoch": 1.0366642703091302,
-      "grad_norm": 0.4486737549304962,
-      "learning_rate": 7.515608074551525e-06,
-      "loss": 0.3268,
-      "step": 15862
-    },
-    {
-      "epoch": 1.0367296255146723,
-      "grad_norm": 0.43175071477890015,
-      "learning_rate": 7.515306292958203e-06,
-      "loss": 0.3326,
-      "step": 15863
-    },
-    {
-      "epoch": 1.0367949807202144,
-      "grad_norm": 0.4423149824142456,
-      "learning_rate": 7.51500449909686e-06,
-      "loss": 0.3277,
-      "step": 15864
-    },
-    {
-      "epoch": 1.0368603359257564,
-      "grad_norm": 0.4635217487812042,
-      "learning_rate": 7.51470269296897e-06,
-      "loss": 0.376,
-      "step": 15865
-    },
-    {
-      "epoch": 1.0369256911312985,
-      "grad_norm": 0.4465751349925995,
-      "learning_rate": 7.514400874576004e-06,
-      "loss": 0.3325,
-      "step": 15866
-    },
-    {
-      "epoch": 1.0369910463368408,
-      "grad_norm": 0.4463614225387573,
-      "learning_rate": 7.514099043919433e-06,
-      "loss": 0.3564,
-      "step": 15867
-    },
-    {
-      "epoch": 1.0370564015423829,
-      "grad_norm": 0.4268401265144348,
-      "learning_rate": 7.513797201000731e-06,
-      "loss": 0.3269,
-      "step": 15868
-    },
-    {
-      "epoch": 1.037121756747925,
-      "grad_norm": 0.46786606311798096,
-      "learning_rate": 7.513495345821369e-06,
-      "loss": 0.3661,
-      "step": 15869
-    },
-    {
-      "epoch": 1.037187111953467,
-      "grad_norm": 0.4640296995639801,
-      "learning_rate": 7.513193478382819e-06,
-      "loss": 0.3499,
-      "step": 15870
-    },
-    {
-      "epoch": 1.0372524671590093,
-      "grad_norm": 0.4440566897392273,
-      "learning_rate": 7.512891598686554e-06,
-      "loss": 0.3424,
-      "step": 15871
-    },
-    {
-      "epoch": 1.0373178223645514,
-      "grad_norm": 0.48283493518829346,
-      "learning_rate": 7.512589706734046e-06,
-      "loss": 0.3586,
-      "step": 15872
-    },
-    {
-      "epoch": 1.0373831775700935,
-      "grad_norm": 0.43852174282073975,
-      "learning_rate": 7.512287802526765e-06,
-      "loss": 0.3199,
-      "step": 15873
-    },
-    {
-      "epoch": 1.0374485327756355,
-      "grad_norm": 0.45028823614120483,
-      "learning_rate": 7.51198588606619e-06,
-      "loss": 0.3405,
-      "step": 15874
-    },
-    {
-      "epoch": 1.0375138879811776,
-      "grad_norm": 0.43766912817955017,
-      "learning_rate": 7.5116839573537885e-06,
-      "loss": 0.3134,
-      "step": 15875
-    },
-    {
-      "epoch": 1.03757924318672,
-      "grad_norm": 0.4694659411907196,
-      "learning_rate": 7.511382016391033e-06,
-      "loss": 0.3764,
-      "step": 15876
-    },
-    {
-      "epoch": 1.037644598392262,
-      "grad_norm": 0.4550599455833435,
-      "learning_rate": 7.511080063179399e-06,
-      "loss": 0.3861,
-      "step": 15877
-    },
-    {
-      "epoch": 1.037709953597804,
-      "grad_norm": 0.43722736835479736,
-      "learning_rate": 7.510778097720355e-06,
-      "loss": 0.3501,
-      "step": 15878
-    },
-    {
-      "epoch": 1.0377753088033461,
-      "grad_norm": 0.45036008954048157,
-      "learning_rate": 7.51047612001538e-06,
-      "loss": 0.3378,
-      "step": 15879
-    },
-    {
-      "epoch": 1.0378406640088884,
-      "grad_norm": 0.4425663948059082,
-      "learning_rate": 7.5101741300659395e-06,
-      "loss": 0.3491,
-      "step": 15880
-    },
-    {
-      "epoch": 1.0379060192144305,
-      "grad_norm": 0.43346884846687317,
-      "learning_rate": 7.509872127873512e-06,
-      "loss": 0.3115,
-      "step": 15881
-    },
-    {
-      "epoch": 1.0379713744199726,
-      "grad_norm": 0.4575836658477783,
-      "learning_rate": 7.509570113439571e-06,
-      "loss": 0.3651,
-      "step": 15882
-    },
-    {
-      "epoch": 1.0380367296255146,
-      "grad_norm": 0.4507502019405365,
-      "learning_rate": 7.509268086765583e-06,
-      "loss": 0.3471,
-      "step": 15883
-    },
-    {
-      "epoch": 1.0381020848310567,
-      "grad_norm": 0.4554588496685028,
-      "learning_rate": 7.508966047853028e-06,
-      "loss": 0.3509,
-      "step": 15884
-    },
-    {
-      "epoch": 1.038167440036599,
-      "grad_norm": 0.46485376358032227,
-      "learning_rate": 7.5086639967033745e-06,
-      "loss": 0.3446,
-      "step": 15885
-    },
-    {
-      "epoch": 1.038232795242141,
-      "grad_norm": 0.46044859290122986,
-      "learning_rate": 7.508361933318099e-06,
-      "loss": 0.3594,
-      "step": 15886
-    },
-    {
-      "epoch": 1.0382981504476831,
-      "grad_norm": 0.4197836220264435,
-      "learning_rate": 7.508059857698673e-06,
-      "loss": 0.3052,
-      "step": 15887
-    },
-    {
-      "epoch": 1.0383635056532252,
-      "grad_norm": 0.44064804911613464,
-      "learning_rate": 7.5077577698465696e-06,
-      "loss": 0.3232,
-      "step": 15888
-    },
-    {
-      "epoch": 1.0384288608587675,
-      "grad_norm": 0.4554585814476013,
-      "learning_rate": 7.507455669763263e-06,
-      "loss": 0.3554,
-      "step": 15889
-    },
-    {
-      "epoch": 1.0384942160643096,
-      "grad_norm": 0.4691154360771179,
-      "learning_rate": 7.507153557450228e-06,
-      "loss": 0.3761,
-      "step": 15890
-    },
-    {
-      "epoch": 1.0385595712698517,
-      "grad_norm": 0.44869810342788696,
-      "learning_rate": 7.506851432908935e-06,
-      "loss": 0.3441,
-      "step": 15891
-    },
-    {
-      "epoch": 1.0386249264753937,
-      "grad_norm": 0.534701406955719,
-      "learning_rate": 7.506549296140859e-06,
-      "loss": 0.4173,
-      "step": 15892
-    },
-    {
-      "epoch": 1.0386902816809358,
-      "grad_norm": 0.46809908747673035,
-      "learning_rate": 7.506247147147474e-06,
-      "loss": 0.3832,
-      "step": 15893
-    },
-    {
-      "epoch": 1.038755636886478,
-      "grad_norm": 0.4488159418106079,
-      "learning_rate": 7.5059449859302535e-06,
-      "loss": 0.3458,
-      "step": 15894
-    },
-    {
-      "epoch": 1.0388209920920202,
-      "grad_norm": 0.491122841835022,
-      "learning_rate": 7.505642812490672e-06,
-      "loss": 0.3664,
-      "step": 15895
-    },
-    {
-      "epoch": 1.0388863472975622,
-      "grad_norm": 0.3987635374069214,
-      "learning_rate": 7.505340626830202e-06,
-      "loss": 0.2887,
-      "step": 15896
-    },
-    {
-      "epoch": 1.0389517025031043,
-      "grad_norm": 0.4556143581867218,
-      "learning_rate": 7.505038428950318e-06,
-      "loss": 0.3274,
-      "step": 15897
-    },
-    {
-      "epoch": 1.0390170577086466,
-      "grad_norm": 0.4318532347679138,
-      "learning_rate": 7.504736218852492e-06,
-      "loss": 0.3048,
-      "step": 15898
-    },
-    {
-      "epoch": 1.0390824129141887,
-      "grad_norm": 0.41274791955947876,
-      "learning_rate": 7.504433996538201e-06,
-      "loss": 0.3012,
-      "step": 15899
-    },
-    {
-      "epoch": 1.0391477681197308,
-      "grad_norm": 0.42034855484962463,
-      "learning_rate": 7.504131762008917e-06,
-      "loss": 0.3239,
-      "step": 15900
-    },
-    {
-      "epoch": 1.0392131233252728,
-      "grad_norm": 0.44173717498779297,
-      "learning_rate": 7.503829515266117e-06,
-      "loss": 0.3506,
-      "step": 15901
-    },
-    {
-      "epoch": 1.039278478530815,
-      "grad_norm": 0.45919856429100037,
-      "learning_rate": 7.503527256311272e-06,
-      "loss": 0.3474,
-      "step": 15902
-    },
-    {
-      "epoch": 1.0393438337363572,
-      "grad_norm": 0.40479904413223267,
-      "learning_rate": 7.5032249851458564e-06,
-      "loss": 0.3053,
-      "step": 15903
-    },
-    {
-      "epoch": 1.0394091889418993,
-      "grad_norm": 0.48598939180374146,
-      "learning_rate": 7.502922701771345e-06,
-      "loss": 0.4101,
-      "step": 15904
-    },
-    {
-      "epoch": 1.0394745441474413,
-      "grad_norm": 0.47933247685432434,
-      "learning_rate": 7.502620406189214e-06,
-      "loss": 0.383,
-      "step": 15905
-    },
-    {
-      "epoch": 1.0395398993529834,
-      "grad_norm": 0.4721432626247406,
-      "learning_rate": 7.502318098400936e-06,
-      "loss": 0.3682,
-      "step": 15906
-    },
-    {
-      "epoch": 1.0396052545585255,
-      "grad_norm": 0.4942317008972168,
-      "learning_rate": 7.502015778407985e-06,
-      "loss": 0.3913,
-      "step": 15907
-    },
-    {
-      "epoch": 1.0396706097640678,
-      "grad_norm": 0.4402003288269043,
-      "learning_rate": 7.5017134462118366e-06,
-      "loss": 0.3078,
-      "step": 15908
-    },
-    {
-      "epoch": 1.0397359649696098,
-      "grad_norm": 0.4501919448375702,
-      "learning_rate": 7.501411101813966e-06,
-      "loss": 0.3358,
-      "step": 15909
-    },
-    {
-      "epoch": 1.039801320175152,
-      "grad_norm": 0.4389926493167877,
-      "learning_rate": 7.501108745215845e-06,
-      "loss": 0.3166,
-      "step": 15910
-    },
-    {
-      "epoch": 1.039866675380694,
-      "grad_norm": 0.43391063809394836,
-      "learning_rate": 7.500806376418952e-06,
-      "loss": 0.3236,
-      "step": 15911
-    },
-    {
-      "epoch": 1.0399320305862363,
-      "grad_norm": 0.4366278350353241,
-      "learning_rate": 7.50050399542476e-06,
-      "loss": 0.3254,
-      "step": 15912
-    },
-    {
-      "epoch": 1.0399973857917784,
-      "grad_norm": 0.4802301228046417,
-      "learning_rate": 7.500201602234743e-06,
-      "loss": 0.377,
-      "step": 15913
-    },
-    {
-      "epoch": 1.0400627409973204,
-      "grad_norm": 0.47257718443870544,
-      "learning_rate": 7.499899196850375e-06,
-      "loss": 0.3135,
-      "step": 15914
-    },
-    {
-      "epoch": 1.0401280962028625,
-      "grad_norm": 0.48913314938545227,
-      "learning_rate": 7.499596779273135e-06,
-      "loss": 0.368,
-      "step": 15915
-    },
-    {
-      "epoch": 1.0401934514084046,
-      "grad_norm": 0.43871310353279114,
-      "learning_rate": 7.499294349504494e-06,
-      "loss": 0.3225,
-      "step": 15916
-    },
-    {
-      "epoch": 1.0402588066139469,
-      "grad_norm": 0.46505528688430786,
-      "learning_rate": 7.49899190754593e-06,
-      "loss": 0.343,
-      "step": 15917
-    },
-    {
-      "epoch": 1.040324161819489,
-      "grad_norm": 0.4820912480354309,
-      "learning_rate": 7.498689453398915e-06,
-      "loss": 0.3418,
-      "step": 15918
-    },
-    {
-      "epoch": 1.040389517025031,
-      "grad_norm": 0.4644339680671692,
-      "learning_rate": 7.498386987064927e-06,
-      "loss": 0.3709,
-      "step": 15919
-    },
-    {
-      "epoch": 1.040454872230573,
-      "grad_norm": 0.41147926449775696,
-      "learning_rate": 7.4980845085454405e-06,
-      "loss": 0.285,
-      "step": 15920
-    },
-    {
-      "epoch": 1.0405202274361154,
-      "grad_norm": 0.45514875650405884,
-      "learning_rate": 7.49778201784193e-06,
-      "loss": 0.3397,
-      "step": 15921
-    },
-    {
-      "epoch": 1.0405855826416575,
-      "grad_norm": 0.429605096578598,
-      "learning_rate": 7.497479514955872e-06,
-      "loss": 0.3036,
-      "step": 15922
-    },
-    {
-      "epoch": 1.0406509378471995,
-      "grad_norm": 0.45561668276786804,
-      "learning_rate": 7.49717699988874e-06,
-      "loss": 0.3491,
-      "step": 15923
-    },
-    {
-      "epoch": 1.0407162930527416,
-      "grad_norm": 0.44797518849372864,
-      "learning_rate": 7.49687447264201e-06,
-      "loss": 0.3723,
-      "step": 15924
-    },
-    {
-      "epoch": 1.0407816482582837,
-      "grad_norm": 0.4650128483772278,
-      "learning_rate": 7.496571933217159e-06,
-      "loss": 0.3712,
-      "step": 15925
-    },
-    {
-      "epoch": 1.040847003463826,
-      "grad_norm": 0.43579956889152527,
-      "learning_rate": 7.496269381615664e-06,
-      "loss": 0.3409,
-      "step": 15926
-    },
-    {
-      "epoch": 1.040912358669368,
-      "grad_norm": 0.4608747661113739,
-      "learning_rate": 7.4959668178389956e-06,
-      "loss": 0.3508,
-      "step": 15927
-    },
-    {
-      "epoch": 1.0409777138749101,
-      "grad_norm": 0.44788262248039246,
-      "learning_rate": 7.4956642418886336e-06,
-      "loss": 0.3299,
-      "step": 15928
-    },
-    {
-      "epoch": 1.0410430690804522,
-      "grad_norm": 0.4167988896369934,
-      "learning_rate": 7.495361653766052e-06,
-      "loss": 0.3054,
-      "step": 15929
-    },
-    {
-      "epoch": 1.0411084242859945,
-      "grad_norm": 0.47324731945991516,
-      "learning_rate": 7.4950590534727274e-06,
-      "loss": 0.3878,
-      "step": 15930
-    },
-    {
-      "epoch": 1.0411737794915366,
-      "grad_norm": 0.47262775897979736,
-      "learning_rate": 7.494756441010136e-06,
-      "loss": 0.3687,
-      "step": 15931
-    },
-    {
-      "epoch": 1.0412391346970786,
-      "grad_norm": 0.44201600551605225,
-      "learning_rate": 7.494453816379752e-06,
-      "loss": 0.3237,
-      "step": 15932
-    },
-    {
-      "epoch": 1.0413044899026207,
-      "grad_norm": 0.45032525062561035,
-      "learning_rate": 7.494151179583054e-06,
-      "loss": 0.3654,
-      "step": 15933
-    },
-    {
-      "epoch": 1.0413698451081628,
-      "grad_norm": 0.47340017557144165,
-      "learning_rate": 7.4938485306215145e-06,
-      "loss": 0.3374,
-      "step": 15934
-    },
-    {
-      "epoch": 1.041435200313705,
-      "grad_norm": 0.49045827984809875,
-      "learning_rate": 7.493545869496614e-06,
-      "loss": 0.4076,
-      "step": 15935
-    },
-    {
-      "epoch": 1.0415005555192471,
-      "grad_norm": 0.42957648634910583,
-      "learning_rate": 7.493243196209825e-06,
-      "loss": 0.3085,
-      "step": 15936
-    },
-    {
-      "epoch": 1.0415659107247892,
-      "grad_norm": 0.4105693995952606,
-      "learning_rate": 7.492940510762627e-06,
-      "loss": 0.2988,
-      "step": 15937
-    },
-    {
-      "epoch": 1.0416312659303313,
-      "grad_norm": 0.4495973587036133,
-      "learning_rate": 7.492637813156492e-06,
-      "loss": 0.351,
-      "step": 15938
-    },
-    {
-      "epoch": 1.0416966211358736,
-      "grad_norm": 0.4780735969543457,
-      "learning_rate": 7.492335103392901e-06,
-      "loss": 0.377,
-      "step": 15939
-    },
-    {
-      "epoch": 1.0417619763414157,
-      "grad_norm": 0.427325040102005,
-      "learning_rate": 7.492032381473326e-06,
-      "loss": 0.3408,
-      "step": 15940
-    },
-    {
-      "epoch": 1.0418273315469577,
-      "grad_norm": 0.46285080909729004,
-      "learning_rate": 7.4917296473992476e-06,
-      "loss": 0.3731,
-      "step": 15941
-    },
-    {
-      "epoch": 1.0418926867524998,
-      "grad_norm": 0.45030486583709717,
-      "learning_rate": 7.49142690117214e-06,
-      "loss": 0.3405,
-      "step": 15942
-    },
-    {
-      "epoch": 1.0419580419580419,
-      "grad_norm": 0.41049444675445557,
-      "learning_rate": 7.49112414279348e-06,
-      "loss": 0.2875,
-      "step": 15943
-    },
-    {
-      "epoch": 1.0420233971635842,
-      "grad_norm": 0.4690943956375122,
-      "learning_rate": 7.490821372264745e-06,
-      "loss": 0.3651,
-      "step": 15944
-    },
-    {
-      "epoch": 1.0420887523691262,
-      "grad_norm": 0.4470690190792084,
-      "learning_rate": 7.49051858958741e-06,
-      "loss": 0.382,
-      "step": 15945
-    },
-    {
-      "epoch": 1.0421541075746683,
-      "grad_norm": 0.4730600416660309,
-      "learning_rate": 7.490215794762955e-06,
-      "loss": 0.3601,
-      "step": 15946
-    },
-    {
-      "epoch": 1.0422194627802104,
-      "grad_norm": 0.41757455468177795,
-      "learning_rate": 7.489912987792853e-06,
-      "loss": 0.3122,
-      "step": 15947
-    },
-    {
-      "epoch": 1.0422848179857525,
-      "grad_norm": 0.40229374170303345,
-      "learning_rate": 7.489610168678585e-06,
-      "loss": 0.3071,
-      "step": 15948
-    },
-    {
-      "epoch": 1.0423501731912947,
-      "grad_norm": 0.4292026162147522,
-      "learning_rate": 7.4893073374216245e-06,
-      "loss": 0.3465,
-      "step": 15949
-    },
-    {
-      "epoch": 1.0424155283968368,
-      "grad_norm": 0.46020931005477905,
-      "learning_rate": 7.489004494023449e-06,
-      "loss": 0.3576,
-      "step": 15950
-    },
-    {
-      "epoch": 1.042480883602379,
-      "grad_norm": 0.4452332854270935,
-      "learning_rate": 7.488701638485537e-06,
-      "loss": 0.3183,
-      "step": 15951
-    },
-    {
-      "epoch": 1.042546238807921,
-      "grad_norm": 0.47011852264404297,
-      "learning_rate": 7.488398770809364e-06,
-      "loss": 0.3887,
-      "step": 15952
-    },
-    {
-      "epoch": 1.0426115940134633,
-      "grad_norm": 0.395666241645813,
-      "learning_rate": 7.488095890996411e-06,
-      "loss": 0.2706,
-      "step": 15953
-    },
-    {
-      "epoch": 1.0426769492190053,
-      "grad_norm": 0.44158506393432617,
-      "learning_rate": 7.487792999048149e-06,
-      "loss": 0.3308,
-      "step": 15954
-    },
-    {
-      "epoch": 1.0427423044245474,
-      "grad_norm": 0.44987404346466064,
-      "learning_rate": 7.487490094966061e-06,
-      "loss": 0.3301,
-      "step": 15955
-    },
-    {
-      "epoch": 1.0428076596300895,
-      "grad_norm": 0.43153107166290283,
-      "learning_rate": 7.4871871787516206e-06,
-      "loss": 0.3159,
-      "step": 15956
-    },
-    {
-      "epoch": 1.0428730148356316,
-      "grad_norm": 0.44239094853401184,
-      "learning_rate": 7.486884250406308e-06,
-      "loss": 0.3351,
-      "step": 15957
-    },
-    {
-      "epoch": 1.0429383700411738,
-      "grad_norm": 0.44070103764533997,
-      "learning_rate": 7.486581309931598e-06,
-      "loss": 0.3244,
-      "step": 15958
-    },
-    {
-      "epoch": 1.043003725246716,
-      "grad_norm": 0.4654008150100708,
-      "learning_rate": 7.486278357328971e-06,
-      "loss": 0.3834,
-      "step": 15959
-    },
-    {
-      "epoch": 1.043069080452258,
-      "grad_norm": 0.42152294516563416,
-      "learning_rate": 7.485975392599902e-06,
-      "loss": 0.2991,
-      "step": 15960
-    },
-    {
-      "epoch": 1.0431344356578,
-      "grad_norm": 0.47432637214660645,
-      "learning_rate": 7.485672415745869e-06,
-      "loss": 0.3729,
-      "step": 15961
-    },
-    {
-      "epoch": 1.0431997908633424,
-      "grad_norm": 0.44445574283599854,
-      "learning_rate": 7.485369426768353e-06,
-      "loss": 0.3265,
-      "step": 15962
-    },
-    {
-      "epoch": 1.0432651460688844,
-      "grad_norm": 0.4614851772785187,
-      "learning_rate": 7.485066425668827e-06,
-      "loss": 0.3501,
-      "step": 15963
-    },
-    {
-      "epoch": 1.0433305012744265,
-      "grad_norm": 0.46275201439857483,
-      "learning_rate": 7.484763412448772e-06,
-      "loss": 0.3407,
-      "step": 15964
-    },
-    {
-      "epoch": 1.0433958564799686,
-      "grad_norm": 0.44269418716430664,
-      "learning_rate": 7.484460387109664e-06,
-      "loss": 0.3365,
-      "step": 15965
-    },
-    {
-      "epoch": 1.0434612116855106,
-      "grad_norm": 0.46993061900138855,
-      "learning_rate": 7.484157349652984e-06,
-      "loss": 0.3623,
-      "step": 15966
-    },
-    {
-      "epoch": 1.043526566891053,
-      "grad_norm": 0.43421483039855957,
-      "learning_rate": 7.4838543000802075e-06,
-      "loss": 0.3135,
-      "step": 15967
-    },
-    {
-      "epoch": 1.043591922096595,
-      "grad_norm": 0.454470694065094,
-      "learning_rate": 7.4835512383928125e-06,
-      "loss": 0.337,
-      "step": 15968
-    },
-    {
-      "epoch": 1.043657277302137,
-      "grad_norm": 0.43242812156677246,
-      "learning_rate": 7.483248164592278e-06,
-      "loss": 0.349,
-      "step": 15969
-    },
-    {
-      "epoch": 1.0437226325076792,
-      "grad_norm": 0.48517265915870667,
-      "learning_rate": 7.482945078680081e-06,
-      "loss": 0.3454,
-      "step": 15970
-    },
-    {
-      "epoch": 1.0437879877132215,
-      "grad_norm": 0.44539085030555725,
-      "learning_rate": 7.482641980657702e-06,
-      "loss": 0.3283,
-      "step": 15971
-    },
-    {
-      "epoch": 1.0438533429187635,
-      "grad_norm": 0.45339563488960266,
-      "learning_rate": 7.482338870526617e-06,
-      "loss": 0.3659,
-      "step": 15972
-    },
-    {
-      "epoch": 1.0439186981243056,
-      "grad_norm": 0.4391777813434601,
-      "learning_rate": 7.482035748288306e-06,
-      "loss": 0.3239,
-      "step": 15973
-    },
-    {
-      "epoch": 1.0439840533298477,
-      "grad_norm": 0.4881090521812439,
-      "learning_rate": 7.481732613944247e-06,
-      "loss": 0.3715,
-      "step": 15974
-    },
-    {
-      "epoch": 1.0440494085353897,
-      "grad_norm": 0.4271552860736847,
-      "learning_rate": 7.481429467495919e-06,
-      "loss": 0.3305,
-      "step": 15975
-    },
-    {
-      "epoch": 1.044114763740932,
-      "grad_norm": 0.45213812589645386,
-      "learning_rate": 7.481126308944797e-06,
-      "loss": 0.3603,
-      "step": 15976
-    },
-    {
-      "epoch": 1.0441801189464741,
-      "grad_norm": 0.46165549755096436,
-      "learning_rate": 7.480823138292365e-06,
-      "loss": 0.3607,
-      "step": 15977
-    },
-    {
-      "epoch": 1.0442454741520162,
-      "grad_norm": 0.44902893900871277,
-      "learning_rate": 7.480519955540099e-06,
-      "loss": 0.315,
-      "step": 15978
-    },
-    {
-      "epoch": 1.0443108293575583,
-      "grad_norm": 0.4467196762561798,
-      "learning_rate": 7.480216760689476e-06,
-      "loss": 0.3216,
-      "step": 15979
-    },
-    {
-      "epoch": 1.0443761845631006,
-      "grad_norm": 0.5242639780044556,
-      "learning_rate": 7.4799135537419785e-06,
-      "loss": 0.3875,
-      "step": 15980
-    },
-    {
-      "epoch": 1.0444415397686426,
-      "grad_norm": 0.4836612939834595,
-      "learning_rate": 7.4796103346990825e-06,
-      "loss": 0.3904,
-      "step": 15981
-    },
-    {
-      "epoch": 1.0445068949741847,
-      "grad_norm": 0.45392730832099915,
-      "learning_rate": 7.479307103562268e-06,
-      "loss": 0.3063,
-      "step": 15982
-    },
-    {
-      "epoch": 1.0445722501797268,
-      "grad_norm": 0.43818655610084534,
-      "learning_rate": 7.479003860333014e-06,
-      "loss": 0.3209,
-      "step": 15983
-    },
-    {
-      "epoch": 1.0446376053852688,
-      "grad_norm": 0.521374523639679,
-      "learning_rate": 7.478700605012798e-06,
-      "loss": 0.4192,
-      "step": 15984
-    },
-    {
-      "epoch": 1.0447029605908111,
-      "grad_norm": 0.4648186266422272,
-      "learning_rate": 7.478397337603103e-06,
-      "loss": 0.3595,
-      "step": 15985
-    },
-    {
-      "epoch": 1.0447683157963532,
-      "grad_norm": 0.4561239778995514,
-      "learning_rate": 7.478094058105404e-06,
-      "loss": 0.3624,
-      "step": 15986
-    },
-    {
-      "epoch": 1.0448336710018953,
-      "grad_norm": 0.4138014018535614,
-      "learning_rate": 7.477790766521182e-06,
-      "loss": 0.3176,
-      "step": 15987
-    },
-    {
-      "epoch": 1.0448990262074374,
-      "grad_norm": 0.5003045797348022,
-      "learning_rate": 7.477487462851916e-06,
-      "loss": 0.418,
-      "step": 15988
-    },
-    {
-      "epoch": 1.0449643814129796,
-      "grad_norm": 0.4711742699146271,
-      "learning_rate": 7.4771841470990854e-06,
-      "loss": 0.3621,
-      "step": 15989
-    },
-    {
-      "epoch": 1.0450297366185217,
-      "grad_norm": 0.4732164740562439,
-      "learning_rate": 7.47688081926417e-06,
-      "loss": 0.3575,
-      "step": 15990
-    },
-    {
-      "epoch": 1.0450950918240638,
-      "grad_norm": 0.4564274847507477,
-      "learning_rate": 7.476577479348649e-06,
-      "loss": 0.352,
-      "step": 15991
-    },
-    {
-      "epoch": 1.0451604470296059,
-      "grad_norm": 0.421882688999176,
-      "learning_rate": 7.476274127353999e-06,
-      "loss": 0.335,
-      "step": 15992
-    },
-    {
-      "epoch": 1.045225802235148,
-      "grad_norm": 0.3941395878791809,
-      "learning_rate": 7.475970763281705e-06,
-      "loss": 0.2798,
-      "step": 15993
-    },
-    {
-      "epoch": 1.0452911574406902,
-      "grad_norm": 0.42184966802597046,
-      "learning_rate": 7.475667387133244e-06,
-      "loss": 0.3187,
-      "step": 15994
-    },
-    {
-      "epoch": 1.0453565126462323,
-      "grad_norm": 0.4963744878768921,
-      "learning_rate": 7.475363998910096e-06,
-      "loss": 0.3281,
-      "step": 15995
-    },
-    {
-      "epoch": 1.0454218678517744,
-      "grad_norm": 0.4685758054256439,
-      "learning_rate": 7.475060598613738e-06,
-      "loss": 0.3775,
-      "step": 15996
-    },
-    {
-      "epoch": 1.0454872230573165,
-      "grad_norm": 0.4629594385623932,
-      "learning_rate": 7.4747571862456534e-06,
-      "loss": 0.3707,
-      "step": 15997
-    },
-    {
-      "epoch": 1.0455525782628587,
-      "grad_norm": 0.4789305329322815,
-      "learning_rate": 7.4744537618073194e-06,
-      "loss": 0.3796,
-      "step": 15998
-    },
-    {
-      "epoch": 1.0456179334684008,
-      "grad_norm": 0.4757988452911377,
-      "learning_rate": 7.474150325300218e-06,
-      "loss": 0.3859,
-      "step": 15999
-    },
-    {
-      "epoch": 1.045683288673943,
-      "grad_norm": 0.4701661169528961,
-      "learning_rate": 7.473846876725829e-06,
-      "loss": 0.3355,
-      "step": 16000
-    },
-    {
-      "epoch": 1.045748643879485,
-      "grad_norm": 0.44279995560646057,
-      "learning_rate": 7.47354341608563e-06,
-      "loss": 0.3167,
-      "step": 16001
-    },
-    {
-      "epoch": 1.045813999085027,
-      "grad_norm": 0.4664006531238556,
-      "learning_rate": 7.473239943381104e-06,
-      "loss": 0.3415,
-      "step": 16002
-    },
-    {
-      "epoch": 1.0458793542905693,
-      "grad_norm": 0.4168444573879242,
-      "learning_rate": 7.4729364586137295e-06,
-      "loss": 0.2826,
-      "step": 16003
-    },
-    {
-      "epoch": 1.0459447094961114,
-      "grad_norm": 0.43262356519699097,
-      "learning_rate": 7.472632961784988e-06,
-      "loss": 0.2785,
-      "step": 16004
-    },
-    {
-      "epoch": 1.0460100647016535,
-      "grad_norm": 0.44316473603248596,
-      "learning_rate": 7.472329452896358e-06,
-      "loss": 0.3253,
-      "step": 16005
-    },
-    {
-      "epoch": 1.0460754199071955,
-      "grad_norm": 0.42974090576171875,
-      "learning_rate": 7.472025931949321e-06,
-      "loss": 0.3071,
-      "step": 16006
-    },
-    {
-      "epoch": 1.0461407751127378,
-      "grad_norm": 0.450324147939682,
-      "learning_rate": 7.471722398945358e-06,
-      "loss": 0.3471,
-      "step": 16007
-    },
-    {
-      "epoch": 1.04620613031828,
-      "grad_norm": 0.4668424725532532,
-      "learning_rate": 7.471418853885948e-06,
-      "loss": 0.3629,
-      "step": 16008
-    },
-    {
-      "epoch": 1.046271485523822,
-      "grad_norm": 0.4546244442462921,
-      "learning_rate": 7.471115296772572e-06,
-      "loss": 0.3608,
-      "step": 16009
-    },
-    {
-      "epoch": 1.046336840729364,
-      "grad_norm": 0.44972726702690125,
-      "learning_rate": 7.470811727606711e-06,
-      "loss": 0.3624,
-      "step": 16010
-    },
-    {
-      "epoch": 1.0464021959349061,
-      "grad_norm": 0.4652496576309204,
-      "learning_rate": 7.470508146389844e-06,
-      "loss": 0.3359,
-      "step": 16011
-    },
-    {
-      "epoch": 1.0464675511404484,
-      "grad_norm": 0.46383875608444214,
-      "learning_rate": 7.470204553123453e-06,
-      "loss": 0.39,
-      "step": 16012
-    },
-    {
-      "epoch": 1.0465329063459905,
-      "grad_norm": 0.4285947382450104,
-      "learning_rate": 7.469900947809021e-06,
-      "loss": 0.3534,
-      "step": 16013
-    },
-    {
-      "epoch": 1.0465982615515326,
-      "grad_norm": 0.4756484031677246,
-      "learning_rate": 7.469597330448025e-06,
-      "loss": 0.3788,
-      "step": 16014
-    },
-    {
-      "epoch": 1.0466636167570746,
-      "grad_norm": 0.4348495304584503,
-      "learning_rate": 7.4692937010419465e-06,
-      "loss": 0.3232,
-      "step": 16015
-    },
-    {
-      "epoch": 1.0467289719626167,
-      "grad_norm": 0.4852817952632904,
-      "learning_rate": 7.46899005959227e-06,
-      "loss": 0.3604,
-      "step": 16016
-    },
-    {
-      "epoch": 1.046794327168159,
-      "grad_norm": 0.44126904010772705,
-      "learning_rate": 7.468686406100469e-06,
-      "loss": 0.2922,
-      "step": 16017
-    },
-    {
-      "epoch": 1.046859682373701,
-      "grad_norm": 0.44158703088760376,
-      "learning_rate": 7.468382740568033e-06,
-      "loss": 0.3494,
-      "step": 16018
-    },
-    {
-      "epoch": 1.0469250375792432,
-      "grad_norm": 0.42098841071128845,
-      "learning_rate": 7.468079062996437e-06,
-      "loss": 0.3307,
-      "step": 16019
-    },
-    {
-      "epoch": 1.0469903927847852,
-      "grad_norm": 0.46187469363212585,
-      "learning_rate": 7.467775373387165e-06,
-      "loss": 0.3707,
-      "step": 16020
-    },
-    {
-      "epoch": 1.0470557479903275,
-      "grad_norm": 0.45330387353897095,
-      "learning_rate": 7.467471671741697e-06,
-      "loss": 0.3111,
-      "step": 16021
-    },
-    {
-      "epoch": 1.0471211031958696,
-      "grad_norm": 0.447543203830719,
-      "learning_rate": 7.467167958061516e-06,
-      "loss": 0.3418,
-      "step": 16022
-    },
-    {
-      "epoch": 1.0471864584014117,
-      "grad_norm": 0.4412584602832794,
-      "learning_rate": 7.466864232348102e-06,
-      "loss": 0.3129,
-      "step": 16023
-    },
-    {
-      "epoch": 1.0472518136069537,
-      "grad_norm": 0.47823694348335266,
-      "learning_rate": 7.4665604946029365e-06,
-      "loss": 0.3881,
-      "step": 16024
-    },
-    {
-      "epoch": 1.0473171688124958,
-      "grad_norm": 0.4403769075870514,
-      "learning_rate": 7.466256744827501e-06,
-      "loss": 0.3291,
-      "step": 16025
-    },
-    {
-      "epoch": 1.047382524018038,
-      "grad_norm": 0.5087055563926697,
-      "learning_rate": 7.465952983023277e-06,
-      "loss": 0.4303,
-      "step": 16026
-    },
-    {
-      "epoch": 1.0474478792235802,
-      "grad_norm": 0.44996869564056396,
-      "learning_rate": 7.465649209191746e-06,
-      "loss": 0.3593,
-      "step": 16027
-    },
-    {
-      "epoch": 1.0475132344291223,
-      "grad_norm": 0.44411277770996094,
-      "learning_rate": 7.465345423334389e-06,
-      "loss": 0.3351,
-      "step": 16028
-    },
-    {
-      "epoch": 1.0475785896346643,
-      "grad_norm": 0.4417063891887665,
-      "learning_rate": 7.465041625452689e-06,
-      "loss": 0.3357,
-      "step": 16029
-    },
-    {
-      "epoch": 1.0476439448402066,
-      "grad_norm": 0.4979275166988373,
-      "learning_rate": 7.464737815548126e-06,
-      "loss": 0.3533,
-      "step": 16030
-    },
-    {
-      "epoch": 1.0477093000457487,
-      "grad_norm": 0.44359245896339417,
-      "learning_rate": 7.464433993622185e-06,
-      "loss": 0.3252,
-      "step": 16031
-    },
-    {
-      "epoch": 1.0477746552512908,
-      "grad_norm": 0.4697877764701843,
-      "learning_rate": 7.464130159676344e-06,
-      "loss": 0.3826,
-      "step": 16032
-    },
-    {
-      "epoch": 1.0478400104568328,
-      "grad_norm": 0.48326537013053894,
-      "learning_rate": 7.463826313712086e-06,
-      "loss": 0.3844,
-      "step": 16033
-    },
-    {
-      "epoch": 1.047905365662375,
-      "grad_norm": 0.45493197441101074,
-      "learning_rate": 7.463522455730894e-06,
-      "loss": 0.336,
-      "step": 16034
-    },
-    {
-      "epoch": 1.0479707208679172,
-      "grad_norm": 0.4514913856983185,
-      "learning_rate": 7.463218585734249e-06,
-      "loss": 0.3621,
-      "step": 16035
-    },
-    {
-      "epoch": 1.0480360760734593,
-      "grad_norm": 0.4070689082145691,
-      "learning_rate": 7.462914703723635e-06,
-      "loss": 0.2869,
-      "step": 16036
-    },
-    {
-      "epoch": 1.0481014312790013,
-      "grad_norm": 0.46747660636901855,
-      "learning_rate": 7.462610809700533e-06,
-      "loss": 0.36,
-      "step": 16037
-    },
-    {
-      "epoch": 1.0481667864845434,
-      "grad_norm": 0.44843965768814087,
-      "learning_rate": 7.462306903666424e-06,
-      "loss": 0.3083,
-      "step": 16038
-    },
-    {
-      "epoch": 1.0482321416900857,
-      "grad_norm": 0.458403080701828,
-      "learning_rate": 7.46200298562279e-06,
-      "loss": 0.3547,
-      "step": 16039
-    },
-    {
-      "epoch": 1.0482974968956278,
-      "grad_norm": 0.4130711555480957,
-      "learning_rate": 7.461699055571117e-06,
-      "loss": 0.3083,
-      "step": 16040
-    },
-    {
-      "epoch": 1.0483628521011699,
-      "grad_norm": 0.46022409200668335,
-      "learning_rate": 7.461395113512883e-06,
-      "loss": 0.3379,
-      "step": 16041
-    },
-    {
-      "epoch": 1.048428207306712,
-      "grad_norm": 0.48253992199897766,
-      "learning_rate": 7.461091159449574e-06,
-      "loss": 0.351,
-      "step": 16042
-    },
-    {
-      "epoch": 1.048493562512254,
-      "grad_norm": 0.4672299325466156,
-      "learning_rate": 7.46078719338267e-06,
-      "loss": 0.345,
-      "step": 16043
-    },
-    {
-      "epoch": 1.0485589177177963,
-      "grad_norm": 0.4777798354625702,
-      "learning_rate": 7.460483215313653e-06,
-      "loss": 0.398,
-      "step": 16044
-    },
-    {
-      "epoch": 1.0486242729233384,
-      "grad_norm": 0.4742797315120697,
-      "learning_rate": 7.460179225244009e-06,
-      "loss": 0.3299,
-      "step": 16045
-    },
-    {
-      "epoch": 1.0486896281288804,
-      "grad_norm": 0.4599107503890991,
-      "learning_rate": 7.459875223175217e-06,
-      "loss": 0.2892,
-      "step": 16046
-    },
-    {
-      "epoch": 1.0487549833344225,
-      "grad_norm": 0.4627104699611664,
-      "learning_rate": 7.459571209108762e-06,
-      "loss": 0.3515,
-      "step": 16047
-    },
-    {
-      "epoch": 1.0488203385399648,
-      "grad_norm": 0.42812010645866394,
-      "learning_rate": 7.459267183046126e-06,
-      "loss": 0.2939,
-      "step": 16048
-    },
-    {
-      "epoch": 1.0488856937455069,
-      "grad_norm": 0.4522375762462616,
-      "learning_rate": 7.4589631449887934e-06,
-      "loss": 0.3608,
-      "step": 16049
-    },
-    {
-      "epoch": 1.048951048951049,
-      "grad_norm": 0.40866246819496155,
-      "learning_rate": 7.4586590949382435e-06,
-      "loss": 0.2941,
-      "step": 16050
-    },
-    {
-      "epoch": 1.049016404156591,
-      "grad_norm": 0.4577885866165161,
-      "learning_rate": 7.4583550328959635e-06,
-      "loss": 0.3588,
-      "step": 16051
-    },
-    {
-      "epoch": 1.049081759362133,
-      "grad_norm": 0.4313019812107086,
-      "learning_rate": 7.458050958863433e-06,
-      "loss": 0.3215,
-      "step": 16052
-    },
-    {
-      "epoch": 1.0491471145676754,
-      "grad_norm": 0.45992523431777954,
-      "learning_rate": 7.457746872842137e-06,
-      "loss": 0.3674,
-      "step": 16053
-    },
-    {
-      "epoch": 1.0492124697732175,
-      "grad_norm": 0.438744455575943,
-      "learning_rate": 7.457442774833558e-06,
-      "loss": 0.3452,
-      "step": 16054
-    },
-    {
-      "epoch": 1.0492778249787595,
-      "grad_norm": 0.4470614194869995,
-      "learning_rate": 7.457138664839178e-06,
-      "loss": 0.3629,
-      "step": 16055
-    },
-    {
-      "epoch": 1.0493431801843016,
-      "grad_norm": 0.4344133734703064,
-      "learning_rate": 7.456834542860483e-06,
-      "loss": 0.3384,
-      "step": 16056
-    },
-    {
-      "epoch": 1.0494085353898437,
-      "grad_norm": 0.4609927237033844,
-      "learning_rate": 7.456530408898954e-06,
-      "loss": 0.3581,
-      "step": 16057
-    },
-    {
-      "epoch": 1.049473890595386,
-      "grad_norm": 0.44912052154541016,
-      "learning_rate": 7.456226262956077e-06,
-      "loss": 0.3278,
-      "step": 16058
-    },
-    {
-      "epoch": 1.049539245800928,
-      "grad_norm": 0.4931497573852539,
-      "learning_rate": 7.455922105033331e-06,
-      "loss": 0.3756,
-      "step": 16059
-    },
-    {
-      "epoch": 1.0496046010064701,
-      "grad_norm": 0.45639166235923767,
-      "learning_rate": 7.455617935132205e-06,
-      "loss": 0.3309,
-      "step": 16060
-    },
-    {
-      "epoch": 1.0496699562120122,
-      "grad_norm": 0.4729221761226654,
-      "learning_rate": 7.455313753254177e-06,
-      "loss": 0.3772,
-      "step": 16061
-    },
-    {
-      "epoch": 1.0497353114175545,
-      "grad_norm": 0.4363824427127838,
-      "learning_rate": 7.455009559400733e-06,
-      "loss": 0.3135,
-      "step": 16062
-    },
-    {
-      "epoch": 1.0498006666230966,
-      "grad_norm": 0.466463565826416,
-      "learning_rate": 7.454705353573359e-06,
-      "loss": 0.366,
-      "step": 16063
-    },
-    {
-      "epoch": 1.0498660218286386,
-      "grad_norm": 0.45507103204727173,
-      "learning_rate": 7.454401135773535e-06,
-      "loss": 0.3462,
-      "step": 16064
-    },
-    {
-      "epoch": 1.0499313770341807,
-      "grad_norm": 0.5240984559059143,
-      "learning_rate": 7.454096906002747e-06,
-      "loss": 0.3706,
-      "step": 16065
-    },
-    {
-      "epoch": 1.0499967322397228,
-      "grad_norm": 0.46580490469932556,
-      "learning_rate": 7.453792664262478e-06,
-      "loss": 0.362,
-      "step": 16066
-    },
-    {
-      "epoch": 1.050062087445265,
-      "grad_norm": 0.46701693534851074,
-      "learning_rate": 7.453488410554213e-06,
-      "loss": 0.384,
-      "step": 16067
-    },
-    {
-      "epoch": 1.0501274426508072,
-      "grad_norm": 0.4378340244293213,
-      "learning_rate": 7.453184144879433e-06,
-      "loss": 0.3146,
-      "step": 16068
-    },
-    {
-      "epoch": 1.0501927978563492,
-      "grad_norm": 0.454223096370697,
-      "learning_rate": 7.452879867239627e-06,
-      "loss": 0.3054,
-      "step": 16069
-    },
-    {
-      "epoch": 1.0502581530618913,
-      "grad_norm": 0.43760383129119873,
-      "learning_rate": 7.452575577636274e-06,
-      "loss": 0.3388,
-      "step": 16070
-    },
-    {
-      "epoch": 1.0503235082674336,
-      "grad_norm": 0.42061540484428406,
-      "learning_rate": 7.45227127607086e-06,
-      "loss": 0.3195,
-      "step": 16071
-    },
-    {
-      "epoch": 1.0503888634729757,
-      "grad_norm": 0.4589472711086273,
-      "learning_rate": 7.45196696254487e-06,
-      "loss": 0.3657,
-      "step": 16072
-    },
-    {
-      "epoch": 1.0504542186785177,
-      "grad_norm": 0.46775344014167786,
-      "learning_rate": 7.451662637059788e-06,
-      "loss": 0.3756,
-      "step": 16073
-    },
-    {
-      "epoch": 1.0505195738840598,
-      "grad_norm": 0.46782544255256653,
-      "learning_rate": 7.451358299617097e-06,
-      "loss": 0.3192,
-      "step": 16074
-    },
-    {
-      "epoch": 1.0505849290896019,
-      "grad_norm": 0.43963491916656494,
-      "learning_rate": 7.4510539502182835e-06,
-      "loss": 0.3565,
-      "step": 16075
-    },
-    {
-      "epoch": 1.0506502842951442,
-      "grad_norm": 0.48291751742362976,
-      "learning_rate": 7.45074958886483e-06,
-      "loss": 0.3714,
-      "step": 16076
-    },
-    {
-      "epoch": 1.0507156395006862,
-      "grad_norm": 0.4731026887893677,
-      "learning_rate": 7.450445215558222e-06,
-      "loss": 0.3431,
-      "step": 16077
-    },
-    {
-      "epoch": 1.0507809947062283,
-      "grad_norm": 0.4549613296985626,
-      "learning_rate": 7.450140830299945e-06,
-      "loss": 0.3671,
-      "step": 16078
-    },
-    {
-      "epoch": 1.0508463499117704,
-      "grad_norm": 0.41363072395324707,
-      "learning_rate": 7.449836433091481e-06,
-      "loss": 0.2931,
-      "step": 16079
-    },
-    {
-      "epoch": 1.0509117051173127,
-      "grad_norm": 0.4579743444919586,
-      "learning_rate": 7.449532023934316e-06,
-      "loss": 0.3524,
-      "step": 16080
-    },
-    {
-      "epoch": 1.0509770603228548,
-      "grad_norm": 0.43860164284706116,
-      "learning_rate": 7.449227602829936e-06,
-      "loss": 0.3211,
-      "step": 16081
-    },
-    {
-      "epoch": 1.0510424155283968,
-      "grad_norm": 0.5000020861625671,
-      "learning_rate": 7.448923169779822e-06,
-      "loss": 0.392,
-      "step": 16082
-    },
-    {
-      "epoch": 1.051107770733939,
-      "grad_norm": 0.46752145886421204,
-      "learning_rate": 7.448618724785464e-06,
-      "loss": 0.3347,
-      "step": 16083
-    },
-    {
-      "epoch": 1.051173125939481,
-      "grad_norm": 0.46789035201072693,
-      "learning_rate": 7.448314267848342e-06,
-      "loss": 0.395,
-      "step": 16084
-    },
-    {
-      "epoch": 1.0512384811450233,
-      "grad_norm": 0.44531866908073425,
-      "learning_rate": 7.448009798969945e-06,
-      "loss": 0.3261,
-      "step": 16085
-    },
-    {
-      "epoch": 1.0513038363505653,
-      "grad_norm": 0.48364758491516113,
-      "learning_rate": 7.447705318151754e-06,
-      "loss": 0.3822,
-      "step": 16086
-    },
-    {
-      "epoch": 1.0513691915561074,
-      "grad_norm": 0.45722150802612305,
-      "learning_rate": 7.447400825395259e-06,
-      "loss": 0.3327,
-      "step": 16087
-    },
-    {
-      "epoch": 1.0514345467616495,
-      "grad_norm": 0.4609781801700592,
-      "learning_rate": 7.44709632070194e-06,
-      "loss": 0.3468,
-      "step": 16088
-    },
-    {
-      "epoch": 1.0514999019671918,
-      "grad_norm": 0.44670552015304565,
-      "learning_rate": 7.446791804073285e-06,
-      "loss": 0.3474,
-      "step": 16089
-    },
-    {
-      "epoch": 1.0515652571727339,
-      "grad_norm": 0.44275641441345215,
-      "learning_rate": 7.44648727551078e-06,
-      "loss": 0.3545,
-      "step": 16090
-    },
-    {
-      "epoch": 1.051630612378276,
-      "grad_norm": 0.497925728559494,
-      "learning_rate": 7.446182735015908e-06,
-      "loss": 0.3845,
-      "step": 16091
-    },
-    {
-      "epoch": 1.051695967583818,
-      "grad_norm": 0.49487441778182983,
-      "learning_rate": 7.445878182590155e-06,
-      "loss": 0.3849,
-      "step": 16092
-    },
-    {
-      "epoch": 1.05176132278936,
-      "grad_norm": 0.48519906401634216,
-      "learning_rate": 7.445573618235007e-06,
-      "loss": 0.3539,
-      "step": 16093
-    },
-    {
-      "epoch": 1.0518266779949024,
-      "grad_norm": 0.4065033495426178,
-      "learning_rate": 7.445269041951949e-06,
-      "loss": 0.2864,
-      "step": 16094
-    },
-    {
-      "epoch": 1.0518920332004444,
-      "grad_norm": 0.4485379755496979,
-      "learning_rate": 7.444964453742467e-06,
-      "loss": 0.3567,
-      "step": 16095
-    },
-    {
-      "epoch": 1.0519573884059865,
-      "grad_norm": 0.42287561297416687,
-      "learning_rate": 7.444659853608047e-06,
-      "loss": 0.3125,
-      "step": 16096
-    },
-    {
-      "epoch": 1.0520227436115286,
-      "grad_norm": 0.41896817088127136,
-      "learning_rate": 7.444355241550174e-06,
-      "loss": 0.2971,
-      "step": 16097
-    },
-    {
-      "epoch": 1.0520880988170709,
-      "grad_norm": 0.45448678731918335,
-      "learning_rate": 7.444050617570332e-06,
-      "loss": 0.3442,
-      "step": 16098
-    },
-    {
-      "epoch": 1.052153454022613,
-      "grad_norm": 0.48453983664512634,
-      "learning_rate": 7.44374598167001e-06,
-      "loss": 0.374,
-      "step": 16099
-    },
-    {
-      "epoch": 1.052218809228155,
-      "grad_norm": 0.44335755705833435,
-      "learning_rate": 7.443441333850693e-06,
-      "loss": 0.3608,
-      "step": 16100
-    },
-    {
-      "epoch": 1.052284164433697,
-      "grad_norm": 0.44495922327041626,
-      "learning_rate": 7.443136674113865e-06,
-      "loss": 0.3715,
-      "step": 16101
-    },
-    {
-      "epoch": 1.0523495196392392,
-      "grad_norm": 0.46403542160987854,
-      "learning_rate": 7.442832002461012e-06,
-      "loss": 0.3703,
-      "step": 16102
-    },
-    {
-      "epoch": 1.0524148748447815,
-      "grad_norm": 0.45666810870170593,
-      "learning_rate": 7.442527318893623e-06,
-      "loss": 0.3269,
-      "step": 16103
-    },
-    {
-      "epoch": 1.0524802300503235,
-      "grad_norm": 0.46022355556488037,
-      "learning_rate": 7.44222262341318e-06,
-      "loss": 0.3435,
-      "step": 16104
-    },
-    {
-      "epoch": 1.0525455852558656,
-      "grad_norm": 0.4367315471172333,
-      "learning_rate": 7.441917916021173e-06,
-      "loss": 0.2846,
-      "step": 16105
-    },
-    {
-      "epoch": 1.0526109404614077,
-      "grad_norm": 0.42847898602485657,
-      "learning_rate": 7.441613196719085e-06,
-      "loss": 0.3082,
-      "step": 16106
-    },
-    {
-      "epoch": 1.05267629566695,
-      "grad_norm": 0.41425612568855286,
-      "learning_rate": 7.441308465508405e-06,
-      "loss": 0.2609,
-      "step": 16107
-    },
-    {
-      "epoch": 1.052741650872492,
-      "grad_norm": 0.4618186950683594,
-      "learning_rate": 7.441003722390617e-06,
-      "loss": 0.3516,
-      "step": 16108
-    },
-    {
-      "epoch": 1.0528070060780341,
-      "grad_norm": 0.49293121695518494,
-      "learning_rate": 7.440698967367208e-06,
-      "loss": 0.3753,
-      "step": 16109
-    },
-    {
-      "epoch": 1.0528723612835762,
-      "grad_norm": 0.4523603320121765,
-      "learning_rate": 7.440394200439665e-06,
-      "loss": 0.3325,
-      "step": 16110
-    },
-    {
-      "epoch": 1.0529377164891183,
-      "grad_norm": 0.43645426630973816,
-      "learning_rate": 7.440089421609475e-06,
-      "loss": 0.329,
-      "step": 16111
-    },
-    {
-      "epoch": 1.0530030716946606,
-      "grad_norm": 0.46498623490333557,
-      "learning_rate": 7.4397846308781214e-06,
-      "loss": 0.3477,
-      "step": 16112
-    },
-    {
-      "epoch": 1.0530684269002026,
-      "grad_norm": 0.43298086524009705,
-      "learning_rate": 7.439479828247094e-06,
-      "loss": 0.327,
-      "step": 16113
-    },
-    {
-      "epoch": 1.0531337821057447,
-      "grad_norm": 0.4323459267616272,
-      "learning_rate": 7.439175013717879e-06,
-      "loss": 0.3293,
-      "step": 16114
-    },
-    {
-      "epoch": 1.0531991373112868,
-      "grad_norm": 0.4473608136177063,
-      "learning_rate": 7.438870187291961e-06,
-      "loss": 0.3665,
-      "step": 16115
-    },
-    {
-      "epoch": 1.053264492516829,
-      "grad_norm": 0.39610910415649414,
-      "learning_rate": 7.43856534897083e-06,
-      "loss": 0.2983,
-      "step": 16116
-    },
-    {
-      "epoch": 1.0533298477223711,
-      "grad_norm": 0.4124915301799774,
-      "learning_rate": 7.43826049875597e-06,
-      "loss": 0.2992,
-      "step": 16117
-    },
-    {
-      "epoch": 1.0533952029279132,
-      "grad_norm": 0.4507128596305847,
-      "learning_rate": 7.437955636648868e-06,
-      "loss": 0.3268,
-      "step": 16118
-    },
-    {
-      "epoch": 1.0534605581334553,
-      "grad_norm": 0.44693493843078613,
-      "learning_rate": 7.437650762651014e-06,
-      "loss": 0.3318,
-      "step": 16119
-    },
-    {
-      "epoch": 1.0535259133389974,
-      "grad_norm": 0.45532897114753723,
-      "learning_rate": 7.4373458767638915e-06,
-      "loss": 0.3558,
-      "step": 16120
-    },
-    {
-      "epoch": 1.0535912685445397,
-      "grad_norm": 0.47934550046920776,
-      "learning_rate": 7.43704097898899e-06,
-      "loss": 0.3719,
-      "step": 16121
-    },
-    {
-      "epoch": 1.0536566237500817,
-      "grad_norm": 0.46038565039634705,
-      "learning_rate": 7.436736069327792e-06,
-      "loss": 0.3473,
-      "step": 16122
-    },
-    {
-      "epoch": 1.0537219789556238,
-      "grad_norm": 0.4664520025253296,
-      "learning_rate": 7.436431147781791e-06,
-      "loss": 0.3668,
-      "step": 16123
-    },
-    {
-      "epoch": 1.0537873341611659,
-      "grad_norm": 0.41246747970581055,
-      "learning_rate": 7.43612621435247e-06,
-      "loss": 0.2976,
-      "step": 16124
-    },
-    {
-      "epoch": 1.053852689366708,
-      "grad_norm": 0.4643338620662689,
-      "learning_rate": 7.435821269041319e-06,
-      "loss": 0.3622,
-      "step": 16125
-    },
-    {
-      "epoch": 1.0539180445722502,
-      "grad_norm": 0.4554542005062103,
-      "learning_rate": 7.435516311849822e-06,
-      "loss": 0.3613,
-      "step": 16126
-    },
-    {
-      "epoch": 1.0539833997777923,
-      "grad_norm": 0.4273832440376282,
-      "learning_rate": 7.43521134277947e-06,
-      "loss": 0.3206,
-      "step": 16127
-    },
-    {
-      "epoch": 1.0540487549833344,
-      "grad_norm": 0.4744766354560852,
-      "learning_rate": 7.434906361831746e-06,
-      "loss": 0.3163,
-      "step": 16128
-    },
-    {
-      "epoch": 1.0541141101888765,
-      "grad_norm": 0.42165735363960266,
-      "learning_rate": 7.434601369008142e-06,
-      "loss": 0.2951,
-      "step": 16129
-    },
-    {
-      "epoch": 1.0541794653944188,
-      "grad_norm": 0.442953884601593,
-      "learning_rate": 7.434296364310144e-06,
-      "loss": 0.3424,
-      "step": 16130
-    },
-    {
-      "epoch": 1.0542448205999608,
-      "grad_norm": 0.4295116364955902,
-      "learning_rate": 7.433991347739238e-06,
-      "loss": 0.3412,
-      "step": 16131
-    },
-    {
-      "epoch": 1.054310175805503,
-      "grad_norm": 0.44163334369659424,
-      "learning_rate": 7.4336863192969135e-06,
-      "loss": 0.348,
-      "step": 16132
-    },
-    {
-      "epoch": 1.054375531011045,
-      "grad_norm": 0.43274348974227905,
-      "learning_rate": 7.433381278984657e-06,
-      "loss": 0.3264,
-      "step": 16133
-    },
-    {
-      "epoch": 1.054440886216587,
-      "grad_norm": 0.4627331793308258,
-      "learning_rate": 7.433076226803959e-06,
-      "loss": 0.3732,
-      "step": 16134
-    },
-    {
-      "epoch": 1.0545062414221293,
-      "grad_norm": 0.4267916679382324,
-      "learning_rate": 7.432771162756305e-06,
-      "loss": 0.3559,
-      "step": 16135
-    },
-    {
-      "epoch": 1.0545715966276714,
-      "grad_norm": 0.5071887373924255,
-      "learning_rate": 7.432466086843182e-06,
-      "loss": 0.3824,
-      "step": 16136
-    },
-    {
-      "epoch": 1.0546369518332135,
-      "grad_norm": 0.4612812101840973,
-      "learning_rate": 7.432160999066079e-06,
-      "loss": 0.3652,
-      "step": 16137
-    },
-    {
-      "epoch": 1.0547023070387556,
-      "grad_norm": 0.45583727955818176,
-      "learning_rate": 7.431855899426485e-06,
-      "loss": 0.3629,
-      "step": 16138
-    },
-    {
-      "epoch": 1.0547676622442979,
-      "grad_norm": 0.4504631459712982,
-      "learning_rate": 7.431550787925887e-06,
-      "loss": 0.3415,
-      "step": 16139
-    },
-    {
-      "epoch": 1.05483301744984,
-      "grad_norm": 0.4481804370880127,
-      "learning_rate": 7.431245664565774e-06,
-      "loss": 0.3423,
-      "step": 16140
-    },
-    {
-      "epoch": 1.054898372655382,
-      "grad_norm": 0.4327700734138489,
-      "learning_rate": 7.4309405293476344e-06,
-      "loss": 0.3375,
-      "step": 16141
-    },
-    {
-      "epoch": 1.054963727860924,
-      "grad_norm": 0.46866220235824585,
-      "learning_rate": 7.430635382272954e-06,
-      "loss": 0.3837,
-      "step": 16142
-    },
-    {
-      "epoch": 1.0550290830664661,
-      "grad_norm": 0.452713280916214,
-      "learning_rate": 7.430330223343223e-06,
-      "loss": 0.3364,
-      "step": 16143
-    },
-    {
-      "epoch": 1.0550944382720084,
-      "grad_norm": 0.450216144323349,
-      "learning_rate": 7.430025052559929e-06,
-      "loss": 0.3799,
-      "step": 16144
-    },
-    {
-      "epoch": 1.0551597934775505,
-      "grad_norm": 0.45943623781204224,
-      "learning_rate": 7.429719869924563e-06,
-      "loss": 0.3693,
-      "step": 16145
-    },
-    {
-      "epoch": 1.0552251486830926,
-      "grad_norm": 0.42963409423828125,
-      "learning_rate": 7.42941467543861e-06,
-      "loss": 0.3046,
-      "step": 16146
-    },
-    {
-      "epoch": 1.0552905038886347,
-      "grad_norm": 0.42389383912086487,
-      "learning_rate": 7.42910946910356e-06,
-      "loss": 0.3167,
-      "step": 16147
-    },
-    {
-      "epoch": 1.055355859094177,
-      "grad_norm": 0.4476156532764435,
-      "learning_rate": 7.4288042509209026e-06,
-      "loss": 0.3475,
-      "step": 16148
-    },
-    {
-      "epoch": 1.055421214299719,
-      "grad_norm": 0.4937572479248047,
-      "learning_rate": 7.428499020892123e-06,
-      "loss": 0.3817,
-      "step": 16149
-    },
-    {
-      "epoch": 1.055486569505261,
-      "grad_norm": 0.4730284512042999,
-      "learning_rate": 7.428193779018715e-06,
-      "loss": 0.3641,
-      "step": 16150
-    },
-    {
-      "epoch": 1.0555519247108032,
-      "grad_norm": 0.43692266941070557,
-      "learning_rate": 7.427888525302164e-06,
-      "loss": 0.3665,
-      "step": 16151
-    },
-    {
-      "epoch": 1.0556172799163452,
-      "grad_norm": 0.4530215263366699,
-      "learning_rate": 7.42758325974396e-06,
-      "loss": 0.349,
-      "step": 16152
-    },
-    {
-      "epoch": 1.0556826351218875,
-      "grad_norm": 0.4344078302383423,
-      "learning_rate": 7.427277982345591e-06,
-      "loss": 0.3323,
-      "step": 16153
-    },
-    {
-      "epoch": 1.0557479903274296,
-      "grad_norm": 0.4584748446941376,
-      "learning_rate": 7.426972693108547e-06,
-      "loss": 0.3416,
-      "step": 16154
-    },
-    {
-      "epoch": 1.0558133455329717,
-      "grad_norm": 0.41894954442977905,
-      "learning_rate": 7.426667392034315e-06,
-      "loss": 0.2976,
-      "step": 16155
-    },
-    {
-      "epoch": 1.0558787007385138,
-      "grad_norm": 0.474901020526886,
-      "learning_rate": 7.426362079124385e-06,
-      "loss": 0.4069,
-      "step": 16156
-    },
-    {
-      "epoch": 1.055944055944056,
-      "grad_norm": 0.4707310199737549,
-      "learning_rate": 7.426056754380249e-06,
-      "loss": 0.3668,
-      "step": 16157
-    },
-    {
-      "epoch": 1.0560094111495981,
-      "grad_norm": 0.42840293049812317,
-      "learning_rate": 7.425751417803392e-06,
-      "loss": 0.332,
-      "step": 16158
-    },
-    {
-      "epoch": 1.0560747663551402,
-      "grad_norm": 0.48791441321372986,
-      "learning_rate": 7.4254460693953054e-06,
-      "loss": 0.3722,
-      "step": 16159
-    },
-    {
-      "epoch": 1.0561401215606823,
-      "grad_norm": 0.4557345509529114,
-      "learning_rate": 7.425140709157477e-06,
-      "loss": 0.3654,
-      "step": 16160
-    },
-    {
-      "epoch": 1.0562054767662243,
-      "grad_norm": 0.4374857246875763,
-      "learning_rate": 7.4248353370913985e-06,
-      "loss": 0.3404,
-      "step": 16161
-    },
-    {
-      "epoch": 1.0562708319717666,
-      "grad_norm": 0.46094179153442383,
-      "learning_rate": 7.4245299531985584e-06,
-      "loss": 0.3567,
-      "step": 16162
-    },
-    {
-      "epoch": 1.0563361871773087,
-      "grad_norm": 0.454353392124176,
-      "learning_rate": 7.4242245574804464e-06,
-      "loss": 0.3497,
-      "step": 16163
-    },
-    {
-      "epoch": 1.0564015423828508,
-      "grad_norm": 0.46465417742729187,
-      "learning_rate": 7.423919149938549e-06,
-      "loss": 0.3566,
-      "step": 16164
-    },
-    {
-      "epoch": 1.0564668975883929,
-      "grad_norm": 0.4740929901599884,
-      "learning_rate": 7.4236137305743595e-06,
-      "loss": 0.3601,
-      "step": 16165
-    },
-    {
-      "epoch": 1.056532252793935,
-      "grad_norm": 0.43093177676200867,
-      "learning_rate": 7.423308299389367e-06,
-      "loss": 0.3255,
-      "step": 16166
-    },
-    {
-      "epoch": 1.0565976079994772,
-      "grad_norm": 0.5268555283546448,
-      "learning_rate": 7.423002856385059e-06,
-      "loss": 0.3888,
-      "step": 16167
-    },
-    {
-      "epoch": 1.0566629632050193,
-      "grad_norm": 0.49433204531669617,
-      "learning_rate": 7.422697401562927e-06,
-      "loss": 0.3602,
-      "step": 16168
-    },
-    {
-      "epoch": 1.0567283184105614,
-      "grad_norm": 0.4234844446182251,
-      "learning_rate": 7.42239193492446e-06,
-      "loss": 0.291,
-      "step": 16169
-    },
-    {
-      "epoch": 1.0567936736161034,
-      "grad_norm": 0.4375132918357849,
-      "learning_rate": 7.422086456471149e-06,
-      "loss": 0.3128,
-      "step": 16170
-    },
-    {
-      "epoch": 1.0568590288216457,
-      "grad_norm": 0.4753700792789459,
-      "learning_rate": 7.421780966204483e-06,
-      "loss": 0.3788,
-      "step": 16171
-    },
-    {
-      "epoch": 1.0569243840271878,
-      "grad_norm": 0.4939681887626648,
-      "learning_rate": 7.421475464125954e-06,
-      "loss": 0.4117,
-      "step": 16172
-    },
-    {
-      "epoch": 1.0569897392327299,
-      "grad_norm": 0.4380776286125183,
-      "learning_rate": 7.421169950237047e-06,
-      "loss": 0.3213,
-      "step": 16173
-    },
-    {
-      "epoch": 1.057055094438272,
-      "grad_norm": 0.41436949372291565,
-      "learning_rate": 7.420864424539258e-06,
-      "loss": 0.2979,
-      "step": 16174
-    },
-    {
-      "epoch": 1.057120449643814,
-      "grad_norm": 0.46490415930747986,
-      "learning_rate": 7.420558887034074e-06,
-      "loss": 0.3494,
-      "step": 16175
-    },
-    {
-      "epoch": 1.0571858048493563,
-      "grad_norm": 0.42988309264183044,
-      "learning_rate": 7.4202533377229845e-06,
-      "loss": 0.3515,
-      "step": 16176
-    },
-    {
-      "epoch": 1.0572511600548984,
-      "grad_norm": 0.43051785230636597,
-      "learning_rate": 7.419947776607482e-06,
-      "loss": 0.3153,
-      "step": 16177
-    },
-    {
-      "epoch": 1.0573165152604405,
-      "grad_norm": 0.4236520230770111,
-      "learning_rate": 7.4196422036890545e-06,
-      "loss": 0.3057,
-      "step": 16178
-    },
-    {
-      "epoch": 1.0573818704659825,
-      "grad_norm": 0.4344591498374939,
-      "learning_rate": 7.419336618969196e-06,
-      "loss": 0.3324,
-      "step": 16179
-    },
-    {
-      "epoch": 1.0574472256715248,
-      "grad_norm": 0.44322726130485535,
-      "learning_rate": 7.419031022449393e-06,
-      "loss": 0.3086,
-      "step": 16180
-    },
-    {
-      "epoch": 1.057512580877067,
-      "grad_norm": 0.4004443287849426,
-      "learning_rate": 7.418725414131138e-06,
-      "loss": 0.2773,
-      "step": 16181
-    },
-    {
-      "epoch": 1.057577936082609,
-      "grad_norm": 0.4375944435596466,
-      "learning_rate": 7.418419794015923e-06,
-      "loss": 0.3639,
-      "step": 16182
-    },
-    {
-      "epoch": 1.057643291288151,
-      "grad_norm": 0.47543880343437195,
-      "learning_rate": 7.418114162105236e-06,
-      "loss": 0.3951,
-      "step": 16183
-    },
-    {
-      "epoch": 1.0577086464936931,
-      "grad_norm": 0.4490235447883606,
-      "learning_rate": 7.417808518400566e-06,
-      "loss": 0.3689,
-      "step": 16184
-    },
-    {
-      "epoch": 1.0577740016992354,
-      "grad_norm": 0.455479234457016,
-      "learning_rate": 7.41750286290341e-06,
-      "loss": 0.3619,
-      "step": 16185
-    },
-    {
-      "epoch": 1.0578393569047775,
-      "grad_norm": 0.41816574335098267,
-      "learning_rate": 7.417197195615253e-06,
-      "loss": 0.2951,
-      "step": 16186
-    },
-    {
-      "epoch": 1.0579047121103196,
-      "grad_norm": 0.45505598187446594,
-      "learning_rate": 7.4168915165375875e-06,
-      "loss": 0.3274,
-      "step": 16187
-    },
-    {
-      "epoch": 1.0579700673158616,
-      "grad_norm": 0.44512656331062317,
-      "learning_rate": 7.4165858256719055e-06,
-      "loss": 0.3405,
-      "step": 16188
-    },
-    {
-      "epoch": 1.058035422521404,
-      "grad_norm": 0.45190003514289856,
-      "learning_rate": 7.4162801230196965e-06,
-      "loss": 0.3491,
-      "step": 16189
-    },
-    {
-      "epoch": 1.058100777726946,
-      "grad_norm": 0.4577966630458832,
-      "learning_rate": 7.415974408582454e-06,
-      "loss": 0.312,
-      "step": 16190
-    },
-    {
-      "epoch": 1.058166132932488,
-      "grad_norm": 0.4371005892753601,
-      "learning_rate": 7.415668682361665e-06,
-      "loss": 0.3099,
-      "step": 16191
-    },
-    {
-      "epoch": 1.0582314881380301,
-      "grad_norm": 0.43823009729385376,
-      "learning_rate": 7.415362944358825e-06,
-      "loss": 0.3149,
-      "step": 16192
-    },
-    {
-      "epoch": 1.0582968433435722,
-      "grad_norm": 0.4281201660633087,
-      "learning_rate": 7.415057194575422e-06,
-      "loss": 0.3473,
-      "step": 16193
-    },
-    {
-      "epoch": 1.0583621985491145,
-      "grad_norm": 0.47924160957336426,
-      "learning_rate": 7.414751433012948e-06,
-      "loss": 0.3802,
-      "step": 16194
-    },
-    {
-      "epoch": 1.0584275537546566,
-      "grad_norm": 0.44475483894348145,
-      "learning_rate": 7.414445659672897e-06,
-      "loss": 0.3564,
-      "step": 16195
-    },
-    {
-      "epoch": 1.0584929089601987,
-      "grad_norm": 0.45348167419433594,
-      "learning_rate": 7.414139874556755e-06,
-      "loss": 0.3677,
-      "step": 16196
-    },
-    {
-      "epoch": 1.0585582641657407,
-      "grad_norm": 0.43412622809410095,
-      "learning_rate": 7.4138340776660185e-06,
-      "loss": 0.3366,
-      "step": 16197
-    },
-    {
-      "epoch": 1.058623619371283,
-      "grad_norm": 0.4788056015968323,
-      "learning_rate": 7.413528269002176e-06,
-      "loss": 0.3894,
-      "step": 16198
-    },
-    {
-      "epoch": 1.058688974576825,
-      "grad_norm": 0.4885151982307434,
-      "learning_rate": 7.4132224485667215e-06,
-      "loss": 0.3112,
-      "step": 16199
-    },
-    {
-      "epoch": 1.0587543297823672,
-      "grad_norm": 0.45134130120277405,
-      "learning_rate": 7.4129166163611434e-06,
-      "loss": 0.3681,
-      "step": 16200
-    },
-    {
-      "epoch": 1.0588196849879092,
-      "grad_norm": 0.4581235647201538,
-      "learning_rate": 7.412610772386935e-06,
-      "loss": 0.3533,
-      "step": 16201
-    },
-    {
-      "epoch": 1.0588850401934513,
-      "grad_norm": 0.4625314474105835,
-      "learning_rate": 7.41230491664559e-06,
-      "loss": 0.3772,
-      "step": 16202
-    },
-    {
-      "epoch": 1.0589503953989936,
-      "grad_norm": 0.47362852096557617,
-      "learning_rate": 7.411999049138596e-06,
-      "loss": 0.3575,
-      "step": 16203
-    },
-    {
-      "epoch": 1.0590157506045357,
-      "grad_norm": 0.4151182174682617,
-      "learning_rate": 7.411693169867449e-06,
-      "loss": 0.2837,
-      "step": 16204
-    },
-    {
-      "epoch": 1.0590811058100777,
-      "grad_norm": 0.4111938178539276,
-      "learning_rate": 7.411387278833637e-06,
-      "loss": 0.2965,
-      "step": 16205
-    },
-    {
-      "epoch": 1.0591464610156198,
-      "grad_norm": 0.48952680826187134,
-      "learning_rate": 7.4110813760386555e-06,
-      "loss": 0.378,
-      "step": 16206
-    },
-    {
-      "epoch": 1.0592118162211621,
-      "grad_norm": 0.4351980984210968,
-      "learning_rate": 7.410775461483995e-06,
-      "loss": 0.301,
-      "step": 16207
-    },
-    {
-      "epoch": 1.0592771714267042,
-      "grad_norm": 0.44695401191711426,
-      "learning_rate": 7.410469535171145e-06,
-      "loss": 0.3177,
-      "step": 16208
-    },
-    {
-      "epoch": 1.0593425266322463,
-      "grad_norm": 0.44649437069892883,
-      "learning_rate": 7.410163597101601e-06,
-      "loss": 0.3352,
-      "step": 16209
-    },
-    {
-      "epoch": 1.0594078818377883,
-      "grad_norm": 0.4653000831604004,
-      "learning_rate": 7.4098576472768555e-06,
-      "loss": 0.3736,
-      "step": 16210
-    },
-    {
-      "epoch": 1.0594732370433304,
-      "grad_norm": 0.5027621388435364,
-      "learning_rate": 7.409551685698398e-06,
-      "loss": 0.4325,
-      "step": 16211
-    },
-    {
-      "epoch": 1.0595385922488727,
-      "grad_norm": 0.4689064025878906,
-      "learning_rate": 7.409245712367724e-06,
-      "loss": 0.3849,
-      "step": 16212
-    },
-    {
-      "epoch": 1.0596039474544148,
-      "grad_norm": 0.43116524815559387,
-      "learning_rate": 7.408939727286323e-06,
-      "loss": 0.3181,
-      "step": 16213
-    },
-    {
-      "epoch": 1.0596693026599568,
-      "grad_norm": 0.4424514174461365,
-      "learning_rate": 7.4086337304556875e-06,
-      "loss": 0.3607,
-      "step": 16214
-    },
-    {
-      "epoch": 1.059734657865499,
-      "grad_norm": 0.4705367088317871,
-      "learning_rate": 7.408327721877312e-06,
-      "loss": 0.391,
-      "step": 16215
-    },
-    {
-      "epoch": 1.0598000130710412,
-      "grad_norm": 0.46798431873321533,
-      "learning_rate": 7.408021701552688e-06,
-      "loss": 0.3364,
-      "step": 16216
-    },
-    {
-      "epoch": 1.0598653682765833,
-      "grad_norm": 0.4381452798843384,
-      "learning_rate": 7.407715669483306e-06,
-      "loss": 0.3528,
-      "step": 16217
-    },
-    {
-      "epoch": 1.0599307234821254,
-      "grad_norm": 0.4279400408267975,
-      "learning_rate": 7.407409625670663e-06,
-      "loss": 0.316,
-      "step": 16218
-    },
-    {
-      "epoch": 1.0599960786876674,
-      "grad_norm": 0.47679150104522705,
-      "learning_rate": 7.4071035701162475e-06,
-      "loss": 0.4004,
-      "step": 16219
-    },
-    {
-      "epoch": 1.0600614338932095,
-      "grad_norm": 0.4515838325023651,
-      "learning_rate": 7.4067975028215555e-06,
-      "loss": 0.3322,
-      "step": 16220
-    },
-    {
-      "epoch": 1.0601267890987518,
-      "grad_norm": 0.44531142711639404,
-      "learning_rate": 7.406491423788077e-06,
-      "loss": 0.3397,
-      "step": 16221
-    },
-    {
-      "epoch": 1.0601921443042939,
-      "grad_norm": 0.44834426045417786,
-      "learning_rate": 7.406185333017307e-06,
-      "loss": 0.3311,
-      "step": 16222
-    },
-    {
-      "epoch": 1.060257499509836,
-      "grad_norm": 0.41237568855285645,
-      "learning_rate": 7.405879230510737e-06,
-      "loss": 0.2965,
-      "step": 16223
-    },
-    {
-      "epoch": 1.060322854715378,
-      "grad_norm": 0.48074665665626526,
-      "learning_rate": 7.405573116269861e-06,
-      "loss": 0.3557,
-      "step": 16224
-    },
-    {
-      "epoch": 1.0603882099209203,
-      "grad_norm": 0.46773916482925415,
-      "learning_rate": 7.405266990296172e-06,
-      "loss": 0.3379,
-      "step": 16225
-    },
-    {
-      "epoch": 1.0604535651264624,
-      "grad_norm": 0.4513261914253235,
-      "learning_rate": 7.404960852591162e-06,
-      "loss": 0.3516,
-      "step": 16226
-    },
-    {
-      "epoch": 1.0605189203320045,
-      "grad_norm": 0.43539005517959595,
-      "learning_rate": 7.404654703156324e-06,
-      "loss": 0.3409,
-      "step": 16227
-    },
-    {
-      "epoch": 1.0605842755375465,
-      "grad_norm": 0.45895013213157654,
-      "learning_rate": 7.404348541993152e-06,
-      "loss": 0.3344,
-      "step": 16228
-    },
-    {
-      "epoch": 1.0606496307430886,
-      "grad_norm": 0.4350387752056122,
-      "learning_rate": 7.40404236910314e-06,
-      "loss": 0.3463,
-      "step": 16229
-    },
-    {
-      "epoch": 1.060714985948631,
-      "grad_norm": 0.4167777895927429,
-      "learning_rate": 7.403736184487781e-06,
-      "loss": 0.289,
-      "step": 16230
-    },
-    {
-      "epoch": 1.060780341154173,
-      "grad_norm": 0.4347829520702362,
-      "learning_rate": 7.403429988148567e-06,
-      "loss": 0.328,
-      "step": 16231
-    },
-    {
-      "epoch": 1.060845696359715,
-      "grad_norm": 0.4480380117893219,
-      "learning_rate": 7.403123780086993e-06,
-      "loss": 0.3337,
-      "step": 16232
-    },
-    {
-      "epoch": 1.0609110515652571,
-      "grad_norm": 0.42906734347343445,
-      "learning_rate": 7.402817560304551e-06,
-      "loss": 0.3286,
-      "step": 16233
-    },
-    {
-      "epoch": 1.0609764067707994,
-      "grad_norm": 0.44069650769233704,
-      "learning_rate": 7.402511328802735e-06,
-      "loss": 0.3252,
-      "step": 16234
-    },
-    {
-      "epoch": 1.0610417619763415,
-      "grad_norm": 0.4493826627731323,
-      "learning_rate": 7.4022050855830405e-06,
-      "loss": 0.3774,
-      "step": 16235
-    },
-    {
-      "epoch": 1.0611071171818836,
-      "grad_norm": 0.42076724767684937,
-      "learning_rate": 7.401898830646958e-06,
-      "loss": 0.318,
-      "step": 16236
-    },
-    {
-      "epoch": 1.0611724723874256,
-      "grad_norm": 0.4470517635345459,
-      "learning_rate": 7.401592563995984e-06,
-      "loss": 0.3305,
-      "step": 16237
-    },
-    {
-      "epoch": 1.0612378275929677,
-      "grad_norm": 0.44872480630874634,
-      "learning_rate": 7.40128628563161e-06,
-      "loss": 0.3292,
-      "step": 16238
-    },
-    {
-      "epoch": 1.06130318279851,
-      "grad_norm": 0.4473561644554138,
-      "learning_rate": 7.400979995555332e-06,
-      "loss": 0.3351,
-      "step": 16239
-    },
-    {
-      "epoch": 1.061368538004052,
-      "grad_norm": 0.45642489194869995,
-      "learning_rate": 7.400673693768643e-06,
-      "loss": 0.3534,
-      "step": 16240
-    },
-    {
-      "epoch": 1.0614338932095941,
-      "grad_norm": 0.48431986570358276,
-      "learning_rate": 7.4003673802730345e-06,
-      "loss": 0.3511,
-      "step": 16241
-    },
-    {
-      "epoch": 1.0614992484151362,
-      "grad_norm": 0.43051964044570923,
-      "learning_rate": 7.400061055070005e-06,
-      "loss": 0.3029,
-      "step": 16242
-    },
-    {
-      "epoch": 1.0615646036206783,
-      "grad_norm": 0.47000324726104736,
-      "learning_rate": 7.399754718161045e-06,
-      "loss": 0.3818,
-      "step": 16243
-    },
-    {
-      "epoch": 1.0616299588262206,
-      "grad_norm": 0.5296894311904907,
-      "learning_rate": 7.39944836954765e-06,
-      "loss": 0.4262,
-      "step": 16244
-    },
-    {
-      "epoch": 1.0616953140317626,
-      "grad_norm": 0.436626136302948,
-      "learning_rate": 7.399142009231315e-06,
-      "loss": 0.3277,
-      "step": 16245
-    },
-    {
-      "epoch": 1.0617606692373047,
-      "grad_norm": 0.4629237949848175,
-      "learning_rate": 7.398835637213534e-06,
-      "loss": 0.3524,
-      "step": 16246
-    },
-    {
-      "epoch": 1.0618260244428468,
-      "grad_norm": 0.48072561621665955,
-      "learning_rate": 7.3985292534957986e-06,
-      "loss": 0.3825,
-      "step": 16247
-    },
-    {
-      "epoch": 1.061891379648389,
-      "grad_norm": 0.4785764217376709,
-      "learning_rate": 7.398222858079607e-06,
-      "loss": 0.3943,
-      "step": 16248
-    },
-    {
-      "epoch": 1.0619567348539312,
-      "grad_norm": 0.4860216975212097,
-      "learning_rate": 7.3979164509664494e-06,
-      "loss": 0.3805,
-      "step": 16249
-    },
-    {
-      "epoch": 1.0620220900594732,
-      "grad_norm": 0.4337969124317169,
-      "learning_rate": 7.3976100321578235e-06,
-      "loss": 0.3047,
-      "step": 16250
-    },
-    {
-      "epoch": 1.0620874452650153,
-      "grad_norm": 0.4114070534706116,
-      "learning_rate": 7.397303601655223e-06,
-      "loss": 0.3047,
-      "step": 16251
-    },
-    {
-      "epoch": 1.0621528004705574,
-      "grad_norm": 0.43916887044906616,
-      "learning_rate": 7.396997159460142e-06,
-      "loss": 0.3256,
-      "step": 16252
-    },
-    {
-      "epoch": 1.0622181556760997,
-      "grad_norm": 0.4180777072906494,
-      "learning_rate": 7.396690705574077e-06,
-      "loss": 0.2866,
-      "step": 16253
-    },
-    {
-      "epoch": 1.0622835108816417,
-      "grad_norm": 0.4620179235935211,
-      "learning_rate": 7.39638423999852e-06,
-      "loss": 0.3344,
-      "step": 16254
-    },
-    {
-      "epoch": 1.0623488660871838,
-      "grad_norm": 0.4804803431034088,
-      "learning_rate": 7.396077762734967e-06,
-      "loss": 0.3902,
-      "step": 16255
-    },
-    {
-      "epoch": 1.062414221292726,
-      "grad_norm": 0.4444611370563507,
-      "learning_rate": 7.3957712737849106e-06,
-      "loss": 0.322,
-      "step": 16256
-    },
-    {
-      "epoch": 1.0624795764982682,
-      "grad_norm": 0.4772826135158539,
-      "learning_rate": 7.395464773149851e-06,
-      "loss": 0.3355,
-      "step": 16257
-    },
-    {
-      "epoch": 1.0625449317038103,
-      "grad_norm": 0.5063560605049133,
-      "learning_rate": 7.395158260831279e-06,
-      "loss": 0.4246,
-      "step": 16258
-    },
-    {
-      "epoch": 1.0626102869093523,
-      "grad_norm": 0.4525230824947357,
-      "learning_rate": 7.394851736830688e-06,
-      "loss": 0.3292,
-      "step": 16259
-    },
-    {
-      "epoch": 1.0626756421148944,
-      "grad_norm": 0.44800588488578796,
-      "learning_rate": 7.3945452011495785e-06,
-      "loss": 0.3447,
-      "step": 16260
-    },
-    {
-      "epoch": 1.0627409973204365,
-      "grad_norm": 0.4656405746936798,
-      "learning_rate": 7.3942386537894404e-06,
-      "loss": 0.352,
-      "step": 16261
-    },
-    {
-      "epoch": 1.0628063525259788,
-      "grad_norm": 0.4371345639228821,
-      "learning_rate": 7.3939320947517725e-06,
-      "loss": 0.3089,
-      "step": 16262
-    },
-    {
-      "epoch": 1.0628717077315208,
-      "grad_norm": 0.4875541925430298,
-      "learning_rate": 7.393625524038067e-06,
-      "loss": 0.387,
-      "step": 16263
-    },
-    {
-      "epoch": 1.062937062937063,
-      "grad_norm": 0.4274907112121582,
-      "learning_rate": 7.393318941649822e-06,
-      "loss": 0.3171,
-      "step": 16264
-    },
-    {
-      "epoch": 1.063002418142605,
-      "grad_norm": 0.4773643910884857,
-      "learning_rate": 7.39301234758853e-06,
-      "loss": 0.412,
-      "step": 16265
-    },
-    {
-      "epoch": 1.0630677733481473,
-      "grad_norm": 0.44621041417121887,
-      "learning_rate": 7.392705741855688e-06,
-      "loss": 0.3252,
-      "step": 16266
-    },
-    {
-      "epoch": 1.0631331285536894,
-      "grad_norm": 0.4797162115573883,
-      "learning_rate": 7.392399124452793e-06,
-      "loss": 0.2904,
-      "step": 16267
-    },
-    {
-      "epoch": 1.0631984837592314,
-      "grad_norm": 0.4833502173423767,
-      "learning_rate": 7.392092495381338e-06,
-      "loss": 0.366,
-      "step": 16268
-    },
-    {
-      "epoch": 1.0632638389647735,
-      "grad_norm": 0.4435287117958069,
-      "learning_rate": 7.391785854642819e-06,
-      "loss": 0.3415,
-      "step": 16269
-    },
-    {
-      "epoch": 1.0633291941703156,
-      "grad_norm": 0.4675077199935913,
-      "learning_rate": 7.3914792022387295e-06,
-      "loss": 0.3398,
-      "step": 16270
-    },
-    {
-      "epoch": 1.0633945493758579,
-      "grad_norm": 0.4949605166912079,
-      "learning_rate": 7.39117253817057e-06,
-      "loss": 0.3716,
-      "step": 16271
-    },
-    {
-      "epoch": 1.0634599045814,
-      "grad_norm": 0.47651195526123047,
-      "learning_rate": 7.390865862439832e-06,
-      "loss": 0.3266,
-      "step": 16272
-    },
-    {
-      "epoch": 1.063525259786942,
-      "grad_norm": 0.4458908140659332,
-      "learning_rate": 7.390559175048015e-06,
-      "loss": 0.3515,
-      "step": 16273
-    },
-    {
-      "epoch": 1.063590614992484,
-      "grad_norm": 0.42917072772979736,
-      "learning_rate": 7.390252475996611e-06,
-      "loss": 0.3246,
-      "step": 16274
-    },
-    {
-      "epoch": 1.0636559701980262,
-      "grad_norm": 0.45001745223999023,
-      "learning_rate": 7.389945765287119e-06,
-      "loss": 0.3107,
-      "step": 16275
-    },
-    {
-      "epoch": 1.0637213254035685,
-      "grad_norm": 0.44215458631515503,
-      "learning_rate": 7.389639042921031e-06,
-      "loss": 0.3005,
-      "step": 16276
-    },
-    {
-      "epoch": 1.0637866806091105,
-      "grad_norm": 0.4572295546531677,
-      "learning_rate": 7.3893323088998484e-06,
-      "loss": 0.3516,
-      "step": 16277
-    },
-    {
-      "epoch": 1.0638520358146526,
-      "grad_norm": 0.4386812150478363,
-      "learning_rate": 7.389025563225063e-06,
-      "loss": 0.3306,
-      "step": 16278
-    },
-    {
-      "epoch": 1.0639173910201947,
-      "grad_norm": 0.5081034302711487,
-      "learning_rate": 7.388718805898172e-06,
-      "loss": 0.3801,
-      "step": 16279
-    },
-    {
-      "epoch": 1.063982746225737,
-      "grad_norm": 0.42380234599113464,
-      "learning_rate": 7.3884120369206735e-06,
-      "loss": 0.3291,
-      "step": 16280
-    },
-    {
-      "epoch": 1.064048101431279,
-      "grad_norm": 0.4563099443912506,
-      "learning_rate": 7.38810525629406e-06,
-      "loss": 0.3471,
-      "step": 16281
-    },
-    {
-      "epoch": 1.064113456636821,
-      "grad_norm": 0.4397892951965332,
-      "learning_rate": 7.387798464019831e-06,
-      "loss": 0.3364,
-      "step": 16282
-    },
-    {
-      "epoch": 1.0641788118423632,
-      "grad_norm": 0.46197909116744995,
-      "learning_rate": 7.3874916600994804e-06,
-      "loss": 0.38,
-      "step": 16283
-    },
-    {
-      "epoch": 1.0642441670479053,
-      "grad_norm": 0.44674259424209595,
-      "learning_rate": 7.387184844534507e-06,
-      "loss": 0.3556,
-      "step": 16284
-    },
-    {
-      "epoch": 1.0643095222534475,
-      "grad_norm": 0.4053109288215637,
-      "learning_rate": 7.386878017326407e-06,
-      "loss": 0.2841,
-      "step": 16285
-    },
-    {
-      "epoch": 1.0643748774589896,
-      "grad_norm": 0.42249050736427307,
-      "learning_rate": 7.3865711784766746e-06,
-      "loss": 0.2992,
-      "step": 16286
-    },
-    {
-      "epoch": 1.0644402326645317,
-      "grad_norm": 0.46023961901664734,
-      "learning_rate": 7.386264327986808e-06,
-      "loss": 0.3845,
-      "step": 16287
-    },
-    {
-      "epoch": 1.0645055878700738,
-      "grad_norm": 0.4664514660835266,
-      "learning_rate": 7.385957465858305e-06,
-      "loss": 0.3466,
-      "step": 16288
-    },
-    {
-      "epoch": 1.064570943075616,
-      "grad_norm": 0.4404003620147705,
-      "learning_rate": 7.38565059209266e-06,
-      "loss": 0.3432,
-      "step": 16289
-    },
-    {
-      "epoch": 1.0646362982811581,
-      "grad_norm": 0.4451841413974762,
-      "learning_rate": 7.3853437066913705e-06,
-      "loss": 0.3616,
-      "step": 16290
-    },
-    {
-      "epoch": 1.0647016534867002,
-      "grad_norm": 0.46495670080184937,
-      "learning_rate": 7.3850368096559335e-06,
-      "loss": 0.3601,
-      "step": 16291
-    },
-    {
-      "epoch": 1.0647670086922423,
-      "grad_norm": 0.43297263979911804,
-      "learning_rate": 7.384729900987844e-06,
-      "loss": 0.3304,
-      "step": 16292
-    },
-    {
-      "epoch": 1.0648323638977844,
-      "grad_norm": 0.44345468282699585,
-      "learning_rate": 7.384422980688602e-06,
-      "loss": 0.3415,
-      "step": 16293
-    },
-    {
-      "epoch": 1.0648977191033266,
-      "grad_norm": 0.43672364950180054,
-      "learning_rate": 7.384116048759703e-06,
-      "loss": 0.3134,
-      "step": 16294
-    },
-    {
-      "epoch": 1.0649630743088687,
-      "grad_norm": 0.4559917449951172,
-      "learning_rate": 7.383809105202645e-06,
-      "loss": 0.3181,
-      "step": 16295
-    },
-    {
-      "epoch": 1.0650284295144108,
-      "grad_norm": 0.44162875413894653,
-      "learning_rate": 7.3835021500189245e-06,
-      "loss": 0.3213,
-      "step": 16296
-    },
-    {
-      "epoch": 1.0650937847199529,
-      "grad_norm": 0.40349656343460083,
-      "learning_rate": 7.383195183210036e-06,
-      "loss": 0.2713,
-      "step": 16297
-    },
-    {
-      "epoch": 1.0651591399254952,
-      "grad_norm": 0.4892532229423523,
-      "learning_rate": 7.38288820477748e-06,
-      "loss": 0.2965,
-      "step": 16298
-    },
-    {
-      "epoch": 1.0652244951310372,
-      "grad_norm": 0.45245566964149475,
-      "learning_rate": 7.382581214722753e-06,
-      "loss": 0.3269,
-      "step": 16299
-    },
-    {
-      "epoch": 1.0652898503365793,
-      "grad_norm": 0.476072758436203,
-      "learning_rate": 7.382274213047352e-06,
-      "loss": 0.3661,
-      "step": 16300
-    },
-    {
-      "epoch": 1.0653552055421214,
-      "grad_norm": 0.46941715478897095,
-      "learning_rate": 7.381967199752773e-06,
-      "loss": 0.3666,
-      "step": 16301
-    },
-    {
-      "epoch": 1.0654205607476634,
-      "grad_norm": 0.4570242464542389,
-      "learning_rate": 7.381660174840517e-06,
-      "loss": 0.3335,
-      "step": 16302
-    },
-    {
-      "epoch": 1.0654859159532057,
-      "grad_norm": 0.5833566188812256,
-      "learning_rate": 7.381353138312078e-06,
-      "loss": 0.3668,
-      "step": 16303
-    },
-    {
-      "epoch": 1.0655512711587478,
-      "grad_norm": 0.4997336268424988,
-      "learning_rate": 7.381046090168955e-06,
-      "loss": 0.3706,
-      "step": 16304
-    },
-    {
-      "epoch": 1.0656166263642899,
-      "grad_norm": 0.48211637139320374,
-      "learning_rate": 7.380739030412645e-06,
-      "loss": 0.3203,
-      "step": 16305
-    },
-    {
-      "epoch": 1.065681981569832,
-      "grad_norm": 0.4539371430873871,
-      "learning_rate": 7.380431959044646e-06,
-      "loss": 0.3193,
-      "step": 16306
-    },
-    {
-      "epoch": 1.0657473367753743,
-      "grad_norm": 0.43627384305000305,
-      "learning_rate": 7.380124876066456e-06,
-      "loss": 0.2922,
-      "step": 16307
-    },
-    {
-      "epoch": 1.0658126919809163,
-      "grad_norm": 0.4480058252811432,
-      "learning_rate": 7.379817781479572e-06,
-      "loss": 0.3639,
-      "step": 16308
-    },
-    {
-      "epoch": 1.0658780471864584,
-      "grad_norm": 0.43765416741371155,
-      "learning_rate": 7.379510675285494e-06,
-      "loss": 0.3269,
-      "step": 16309
-    },
-    {
-      "epoch": 1.0659434023920005,
-      "grad_norm": 0.44008341431617737,
-      "learning_rate": 7.379203557485717e-06,
-      "loss": 0.3131,
-      "step": 16310
-    },
-    {
-      "epoch": 1.0660087575975425,
-      "grad_norm": 0.47002214193344116,
-      "learning_rate": 7.3788964280817395e-06,
-      "loss": 0.3586,
-      "step": 16311
-    },
-    {
-      "epoch": 1.0660741128030848,
-      "grad_norm": 0.47891783714294434,
-      "learning_rate": 7.3785892870750596e-06,
-      "loss": 0.3631,
-      "step": 16312
-    },
-    {
-      "epoch": 1.066139468008627,
-      "grad_norm": 0.4591046869754791,
-      "learning_rate": 7.378282134467176e-06,
-      "loss": 0.366,
-      "step": 16313
-    },
-    {
-      "epoch": 1.066204823214169,
-      "grad_norm": 0.4661593437194824,
-      "learning_rate": 7.377974970259587e-06,
-      "loss": 0.3694,
-      "step": 16314
-    },
-    {
-      "epoch": 1.066270178419711,
-      "grad_norm": 0.44471195340156555,
-      "learning_rate": 7.3776677944537915e-06,
-      "loss": 0.3427,
-      "step": 16315
-    },
-    {
-      "epoch": 1.0663355336252534,
-      "grad_norm": 0.481689453125,
-      "learning_rate": 7.377360607051285e-06,
-      "loss": 0.3825,
-      "step": 16316
-    },
-    {
-      "epoch": 1.0664008888307954,
-      "grad_norm": 0.4517310857772827,
-      "learning_rate": 7.377053408053566e-06,
-      "loss": 0.3297,
-      "step": 16317
-    },
-    {
-      "epoch": 1.0664662440363375,
-      "grad_norm": 0.43472006916999817,
-      "learning_rate": 7.376746197462137e-06,
-      "loss": 0.2861,
-      "step": 16318
-    },
-    {
-      "epoch": 1.0665315992418796,
-      "grad_norm": 0.4931611716747284,
-      "learning_rate": 7.376438975278491e-06,
-      "loss": 0.3898,
-      "step": 16319
-    },
-    {
-      "epoch": 1.0665969544474216,
-      "grad_norm": 0.4449213743209839,
-      "learning_rate": 7.376131741504129e-06,
-      "loss": 0.3398,
-      "step": 16320
-    },
-    {
-      "epoch": 1.066662309652964,
-      "grad_norm": 0.44252240657806396,
-      "learning_rate": 7.37582449614055e-06,
-      "loss": 0.303,
-      "step": 16321
-    },
-    {
-      "epoch": 1.066727664858506,
-      "grad_norm": 0.47769948840141296,
-      "learning_rate": 7.375517239189251e-06,
-      "loss": 0.363,
-      "step": 16322
-    },
-    {
-      "epoch": 1.066793020064048,
-      "grad_norm": 0.45678725838661194,
-      "learning_rate": 7.375209970651733e-06,
-      "loss": 0.3468,
-      "step": 16323
-    },
-    {
-      "epoch": 1.0668583752695902,
-      "grad_norm": 0.43321335315704346,
-      "learning_rate": 7.374902690529493e-06,
-      "loss": 0.3491,
-      "step": 16324
-    },
-    {
-      "epoch": 1.0669237304751324,
-      "grad_norm": 0.45774635672569275,
-      "learning_rate": 7.374595398824029e-06,
-      "loss": 0.3771,
-      "step": 16325
-    },
-    {
-      "epoch": 1.0669890856806745,
-      "grad_norm": 0.45558732748031616,
-      "learning_rate": 7.37428809553684e-06,
-      "loss": 0.3552,
-      "step": 16326
-    },
-    {
-      "epoch": 1.0670544408862166,
-      "grad_norm": 0.43057477474212646,
-      "learning_rate": 7.373980780669427e-06,
-      "loss": 0.3136,
-      "step": 16327
-    },
-    {
-      "epoch": 1.0671197960917587,
-      "grad_norm": 0.4441756308078766,
-      "learning_rate": 7.373673454223285e-06,
-      "loss": 0.3574,
-      "step": 16328
-    },
-    {
-      "epoch": 1.0671851512973007,
-      "grad_norm": 0.47489261627197266,
-      "learning_rate": 7.373366116199918e-06,
-      "loss": 0.376,
-      "step": 16329
-    },
-    {
-      "epoch": 1.067250506502843,
-      "grad_norm": 0.46269145607948303,
-      "learning_rate": 7.373058766600821e-06,
-      "loss": 0.3112,
-      "step": 16330
-    },
-    {
-      "epoch": 1.067315861708385,
-      "grad_norm": 0.43337777256965637,
-      "learning_rate": 7.372751405427495e-06,
-      "loss": 0.3225,
-      "step": 16331
-    },
-    {
-      "epoch": 1.0673812169139272,
-      "grad_norm": 0.42257869243621826,
-      "learning_rate": 7.3724440326814376e-06,
-      "loss": 0.2858,
-      "step": 16332
-    },
-    {
-      "epoch": 1.0674465721194693,
-      "grad_norm": 0.4334564805030823,
-      "learning_rate": 7.37213664836415e-06,
-      "loss": 0.3132,
-      "step": 16333
-    },
-    {
-      "epoch": 1.0675119273250115,
-      "grad_norm": 0.46192216873168945,
-      "learning_rate": 7.371829252477127e-06,
-      "loss": 0.3667,
-      "step": 16334
-    },
-    {
-      "epoch": 1.0675772825305536,
-      "grad_norm": 0.4625583589076996,
-      "learning_rate": 7.371521845021874e-06,
-      "loss": 0.379,
-      "step": 16335
-    },
-    {
-      "epoch": 1.0676426377360957,
-      "grad_norm": 0.4332091510295868,
-      "learning_rate": 7.371214425999888e-06,
-      "loss": 0.3264,
-      "step": 16336
-    },
-    {
-      "epoch": 1.0677079929416378,
-      "grad_norm": 0.44488924741744995,
-      "learning_rate": 7.370906995412665e-06,
-      "loss": 0.323,
-      "step": 16337
-    },
-    {
-      "epoch": 1.0677733481471798,
-      "grad_norm": 0.4677157402038574,
-      "learning_rate": 7.370599553261709e-06,
-      "loss": 0.3615,
-      "step": 16338
-    },
-    {
-      "epoch": 1.0678387033527221,
-      "grad_norm": 0.4546584188938141,
-      "learning_rate": 7.370292099548516e-06,
-      "loss": 0.3497,
-      "step": 16339
-    },
-    {
-      "epoch": 1.0679040585582642,
-      "grad_norm": 0.423282653093338,
-      "learning_rate": 7.369984634274589e-06,
-      "loss": 0.3264,
-      "step": 16340
-    },
-    {
-      "epoch": 1.0679694137638063,
-      "grad_norm": 0.46851375699043274,
-      "learning_rate": 7.369677157441425e-06,
-      "loss": 0.3952,
-      "step": 16341
-    },
-    {
-      "epoch": 1.0680347689693483,
-      "grad_norm": 0.4636993706226349,
-      "learning_rate": 7.369369669050526e-06,
-      "loss": 0.333,
-      "step": 16342
-    },
-    {
-      "epoch": 1.0681001241748906,
-      "grad_norm": 0.49464917182922363,
-      "learning_rate": 7.3690621691033895e-06,
-      "loss": 0.3477,
-      "step": 16343
-    },
-    {
-      "epoch": 1.0681654793804327,
-      "grad_norm": 0.4228525757789612,
-      "learning_rate": 7.368754657601516e-06,
-      "loss": 0.3055,
-      "step": 16344
-    },
-    {
-      "epoch": 1.0682308345859748,
-      "grad_norm": 0.48316940665245056,
-      "learning_rate": 7.3684471345464046e-06,
-      "loss": 0.3625,
-      "step": 16345
-    },
-    {
-      "epoch": 1.0682961897915169,
-      "grad_norm": 0.49780622124671936,
-      "learning_rate": 7.368139599939557e-06,
-      "loss": 0.3822,
-      "step": 16346
-    },
-    {
-      "epoch": 1.068361544997059,
-      "grad_norm": 0.4277776777744293,
-      "learning_rate": 7.367832053782471e-06,
-      "loss": 0.3077,
-      "step": 16347
-    },
-    {
-      "epoch": 1.0684269002026012,
-      "grad_norm": 0.4928586781024933,
-      "learning_rate": 7.367524496076648e-06,
-      "loss": 0.3912,
-      "step": 16348
-    },
-    {
-      "epoch": 1.0684922554081433,
-      "grad_norm": 0.4875708520412445,
-      "learning_rate": 7.36721692682359e-06,
-      "loss": 0.3722,
-      "step": 16349
-    },
-    {
-      "epoch": 1.0685576106136854,
-      "grad_norm": 0.42442789673805237,
-      "learning_rate": 7.366909346024793e-06,
-      "loss": 0.2881,
-      "step": 16350
-    },
-    {
-      "epoch": 1.0686229658192274,
-      "grad_norm": 0.4443688988685608,
-      "learning_rate": 7.3666017536817605e-06,
-      "loss": 0.3451,
-      "step": 16351
-    },
-    {
-      "epoch": 1.0686883210247697,
-      "grad_norm": 0.40238940715789795,
-      "learning_rate": 7.36629414979599e-06,
-      "loss": 0.2745,
-      "step": 16352
-    },
-    {
-      "epoch": 1.0687536762303118,
-      "grad_norm": 0.4478525221347809,
-      "learning_rate": 7.3659865343689844e-06,
-      "loss": 0.3602,
-      "step": 16353
-    },
-    {
-      "epoch": 1.0688190314358539,
-      "grad_norm": 0.4733360707759857,
-      "learning_rate": 7.365678907402242e-06,
-      "loss": 0.366,
-      "step": 16354
-    },
-    {
-      "epoch": 1.068884386641396,
-      "grad_norm": 0.49481990933418274,
-      "learning_rate": 7.365371268897263e-06,
-      "loss": 0.3781,
-      "step": 16355
-    },
-    {
-      "epoch": 1.068949741846938,
-      "grad_norm": 0.4631558060646057,
-      "learning_rate": 7.36506361885555e-06,
-      "loss": 0.3767,
-      "step": 16356
-    },
-    {
-      "epoch": 1.0690150970524803,
-      "grad_norm": 0.42552196979522705,
-      "learning_rate": 7.364755957278602e-06,
-      "loss": 0.3118,
-      "step": 16357
-    },
-    {
-      "epoch": 1.0690804522580224,
-      "grad_norm": 0.4559949040412903,
-      "learning_rate": 7.364448284167921e-06,
-      "loss": 0.3312,
-      "step": 16358
-    },
-    {
-      "epoch": 1.0691458074635645,
-      "grad_norm": 0.41191062331199646,
-      "learning_rate": 7.3641405995250045e-06,
-      "loss": 0.3148,
-      "step": 16359
-    },
-    {
-      "epoch": 1.0692111626691065,
-      "grad_norm": 0.4348178207874298,
-      "learning_rate": 7.363832903351358e-06,
-      "loss": 0.3173,
-      "step": 16360
-    },
-    {
-      "epoch": 1.0692765178746486,
-      "grad_norm": 0.4582132399082184,
-      "learning_rate": 7.363525195648477e-06,
-      "loss": 0.3132,
-      "step": 16361
-    },
-    {
-      "epoch": 1.069341873080191,
-      "grad_norm": 0.44204777479171753,
-      "learning_rate": 7.363217476417868e-06,
-      "loss": 0.3175,
-      "step": 16362
-    },
-    {
-      "epoch": 1.069407228285733,
-      "grad_norm": 0.47324949502944946,
-      "learning_rate": 7.3629097456610266e-06,
-      "loss": 0.4177,
-      "step": 16363
-    },
-    {
-      "epoch": 1.069472583491275,
-      "grad_norm": 0.43854910135269165,
-      "learning_rate": 7.362602003379456e-06,
-      "loss": 0.2915,
-      "step": 16364
-    },
-    {
-      "epoch": 1.0695379386968171,
-      "grad_norm": 0.4717777669429779,
-      "learning_rate": 7.362294249574656e-06,
-      "loss": 0.3806,
-      "step": 16365
-    },
-    {
-      "epoch": 1.0696032939023594,
-      "grad_norm": 0.4293909966945648,
-      "learning_rate": 7.3619864842481295e-06,
-      "loss": 0.3067,
-      "step": 16366
-    },
-    {
-      "epoch": 1.0696686491079015,
-      "grad_norm": 0.45948585867881775,
-      "learning_rate": 7.361678707401376e-06,
-      "loss": 0.3194,
-      "step": 16367
-    },
-    {
-      "epoch": 1.0697340043134436,
-      "grad_norm": 0.4923698902130127,
-      "learning_rate": 7.361370919035898e-06,
-      "loss": 0.3417,
-      "step": 16368
-    },
-    {
-      "epoch": 1.0697993595189856,
-      "grad_norm": 0.46501636505126953,
-      "learning_rate": 7.3610631191531955e-06,
-      "loss": 0.3676,
-      "step": 16369
-    },
-    {
-      "epoch": 1.0698647147245277,
-      "grad_norm": 0.4651261866092682,
-      "learning_rate": 7.360755307754771e-06,
-      "loss": 0.3326,
-      "step": 16370
-    },
-    {
-      "epoch": 1.06993006993007,
-      "grad_norm": 0.4757455289363861,
-      "learning_rate": 7.360447484842123e-06,
-      "loss": 0.3405,
-      "step": 16371
-    },
-    {
-      "epoch": 1.069995425135612,
-      "grad_norm": 0.4725337326526642,
-      "learning_rate": 7.360139650416758e-06,
-      "loss": 0.351,
-      "step": 16372
-    },
-    {
-      "epoch": 1.0700607803411541,
-      "grad_norm": 0.454908162355423,
-      "learning_rate": 7.359831804480173e-06,
-      "loss": 0.3071,
-      "step": 16373
-    },
-    {
-      "epoch": 1.0701261355466962,
-      "grad_norm": 0.46444931626319885,
-      "learning_rate": 7.3595239470338705e-06,
-      "loss": 0.3658,
-      "step": 16374
-    },
-    {
-      "epoch": 1.0701914907522385,
-      "grad_norm": 0.4421345293521881,
-      "learning_rate": 7.359216078079352e-06,
-      "loss": 0.3291,
-      "step": 16375
-    },
-    {
-      "epoch": 1.0702568459577806,
-      "grad_norm": 0.4262407124042511,
-      "learning_rate": 7.35890819761812e-06,
-      "loss": 0.2771,
-      "step": 16376
-    },
-    {
-      "epoch": 1.0703222011633227,
-      "grad_norm": 0.45567482709884644,
-      "learning_rate": 7.358600305651674e-06,
-      "loss": 0.3641,
-      "step": 16377
-    },
-    {
-      "epoch": 1.0703875563688647,
-      "grad_norm": 0.4575360119342804,
-      "learning_rate": 7.35829240218152e-06,
-      "loss": 0.3211,
-      "step": 16378
-    },
-    {
-      "epoch": 1.0704529115744068,
-      "grad_norm": 0.4639178514480591,
-      "learning_rate": 7.3579844872091535e-06,
-      "loss": 0.3944,
-      "step": 16379
-    },
-    {
-      "epoch": 1.070518266779949,
-      "grad_norm": 0.41772863268852234,
-      "learning_rate": 7.357676560736082e-06,
-      "loss": 0.2997,
-      "step": 16380
-    },
-    {
-      "epoch": 1.0705836219854912,
-      "grad_norm": 0.4523961544036865,
-      "learning_rate": 7.357368622763805e-06,
-      "loss": 0.3414,
-      "step": 16381
-    },
-    {
-      "epoch": 1.0706489771910332,
-      "grad_norm": 0.45985814929008484,
-      "learning_rate": 7.357060673293824e-06,
-      "loss": 0.3424,
-      "step": 16382
-    },
-    {
-      "epoch": 1.0707143323965753,
-      "grad_norm": 0.4585316777229309,
-      "learning_rate": 7.3567527123276425e-06,
-      "loss": 0.322,
-      "step": 16383
-    },
-    {
-      "epoch": 1.0707796876021174,
-      "grad_norm": 0.4948931634426117,
-      "learning_rate": 7.3564447398667605e-06,
-      "loss": 0.3905,
-      "step": 16384
-    },
-    {
-      "epoch": 1.0708450428076597,
-      "grad_norm": 0.47296464443206787,
-      "learning_rate": 7.356136755912682e-06,
-      "loss": 0.3613,
-      "step": 16385
-    },
-    {
-      "epoch": 1.0709103980132018,
-      "grad_norm": 0.4087711274623871,
-      "learning_rate": 7.3558287604669075e-06,
-      "loss": 0.2598,
-      "step": 16386
-    },
-    {
-      "epoch": 1.0709757532187438,
-      "grad_norm": 0.46858835220336914,
-      "learning_rate": 7.35552075353094e-06,
-      "loss": 0.3376,
-      "step": 16387
-    },
-    {
-      "epoch": 1.071041108424286,
-      "grad_norm": 0.47748062014579773,
-      "learning_rate": 7.355212735106282e-06,
-      "loss": 0.3537,
-      "step": 16388
-    },
-    {
-      "epoch": 1.0711064636298282,
-      "grad_norm": 0.48925772309303284,
-      "learning_rate": 7.354904705194436e-06,
-      "loss": 0.3954,
-      "step": 16389
-    },
-    {
-      "epoch": 1.0711718188353703,
-      "grad_norm": 0.44123828411102295,
-      "learning_rate": 7.354596663796903e-06,
-      "loss": 0.3451,
-      "step": 16390
-    },
-    {
-      "epoch": 1.0712371740409123,
-      "grad_norm": 0.45615312457084656,
-      "learning_rate": 7.354288610915187e-06,
-      "loss": 0.3485,
-      "step": 16391
-    },
-    {
-      "epoch": 1.0713025292464544,
-      "grad_norm": 0.4208703339099884,
-      "learning_rate": 7.35398054655079e-06,
-      "loss": 0.317,
-      "step": 16392
-    },
-    {
-      "epoch": 1.0713678844519965,
-      "grad_norm": 0.42598918080329895,
-      "learning_rate": 7.353672470705216e-06,
-      "loss": 0.3097,
-      "step": 16393
-    },
-    {
-      "epoch": 1.0714332396575388,
-      "grad_norm": 0.47139468789100647,
-      "learning_rate": 7.3533643833799636e-06,
-      "loss": 0.3549,
-      "step": 16394
-    },
-    {
-      "epoch": 1.0714985948630809,
-      "grad_norm": 0.46468374133110046,
-      "learning_rate": 7.353056284576537e-06,
-      "loss": 0.3782,
-      "step": 16395
-    },
-    {
-      "epoch": 1.071563950068623,
-      "grad_norm": 0.4394523501396179,
-      "learning_rate": 7.3527481742964424e-06,
-      "loss": 0.3223,
-      "step": 16396
-    },
-    {
-      "epoch": 1.071629305274165,
-      "grad_norm": 0.47597095370292664,
-      "learning_rate": 7.352440052541178e-06,
-      "loss": 0.4177,
-      "step": 16397
-    },
-    {
-      "epoch": 1.0716946604797073,
-      "grad_norm": 0.4225848615169525,
-      "learning_rate": 7.352131919312248e-06,
-      "loss": 0.3357,
-      "step": 16398
-    },
-    {
-      "epoch": 1.0717600156852494,
-      "grad_norm": 0.44464272260665894,
-      "learning_rate": 7.351823774611158e-06,
-      "loss": 0.3294,
-      "step": 16399
-    },
-    {
-      "epoch": 1.0718253708907914,
-      "grad_norm": 0.4471379220485687,
-      "learning_rate": 7.351515618439406e-06,
-      "loss": 0.3125,
-      "step": 16400
-    },
-    {
-      "epoch": 1.0718907260963335,
-      "grad_norm": 0.4395327866077423,
-      "learning_rate": 7.351207450798501e-06,
-      "loss": 0.2769,
-      "step": 16401
-    },
-    {
-      "epoch": 1.0719560813018756,
-      "grad_norm": 0.4541919231414795,
-      "learning_rate": 7.3508992716899395e-06,
-      "loss": 0.3563,
-      "step": 16402
-    },
-    {
-      "epoch": 1.0720214365074179,
-      "grad_norm": 0.4204372465610504,
-      "learning_rate": 7.35059108111523e-06,
-      "loss": 0.3301,
-      "step": 16403
-    },
-    {
-      "epoch": 1.07208679171296,
-      "grad_norm": 0.47538837790489197,
-      "learning_rate": 7.350282879075872e-06,
-      "loss": 0.3863,
-      "step": 16404
-    },
-    {
-      "epoch": 1.072152146918502,
-      "grad_norm": 0.4320773184299469,
-      "learning_rate": 7.349974665573372e-06,
-      "loss": 0.3314,
-      "step": 16405
-    },
-    {
-      "epoch": 1.072217502124044,
-      "grad_norm": 0.4484153985977173,
-      "learning_rate": 7.34966644060923e-06,
-      "loss": 0.3533,
-      "step": 16406
-    },
-    {
-      "epoch": 1.0722828573295864,
-      "grad_norm": 0.42012518644332886,
-      "learning_rate": 7.349358204184951e-06,
-      "loss": 0.3221,
-      "step": 16407
-    },
-    {
-      "epoch": 1.0723482125351285,
-      "grad_norm": 0.4573407769203186,
-      "learning_rate": 7.349049956302039e-06,
-      "loss": 0.3671,
-      "step": 16408
-    },
-    {
-      "epoch": 1.0724135677406705,
-      "grad_norm": 0.442060649394989,
-      "learning_rate": 7.348741696961995e-06,
-      "loss": 0.33,
-      "step": 16409
-    },
-    {
-      "epoch": 1.0724789229462126,
-      "grad_norm": 0.44182801246643066,
-      "learning_rate": 7.348433426166326e-06,
-      "loss": 0.3029,
-      "step": 16410
-    },
-    {
-      "epoch": 1.0725442781517547,
-      "grad_norm": 0.4631810486316681,
-      "learning_rate": 7.348125143916531e-06,
-      "loss": 0.3541,
-      "step": 16411
-    },
-    {
-      "epoch": 1.072609633357297,
-      "grad_norm": 0.4371936023235321,
-      "learning_rate": 7.347816850214118e-06,
-      "loss": 0.3704,
-      "step": 16412
-    },
-    {
-      "epoch": 1.072674988562839,
-      "grad_norm": 0.44719499349594116,
-      "learning_rate": 7.347508545060589e-06,
-      "loss": 0.3192,
-      "step": 16413
-    },
-    {
-      "epoch": 1.0727403437683811,
-      "grad_norm": 0.44428518414497375,
-      "learning_rate": 7.347200228457447e-06,
-      "loss": 0.323,
-      "step": 16414
-    },
-    {
-      "epoch": 1.0728056989739232,
-      "grad_norm": 0.4393537640571594,
-      "learning_rate": 7.346891900406197e-06,
-      "loss": 0.3456,
-      "step": 16415
-    },
-    {
-      "epoch": 1.0728710541794655,
-      "grad_norm": 0.4477136731147766,
-      "learning_rate": 7.346583560908343e-06,
-      "loss": 0.3341,
-      "step": 16416
-    },
-    {
-      "epoch": 1.0729364093850076,
-      "grad_norm": 0.4214542806148529,
-      "learning_rate": 7.346275209965386e-06,
-      "loss": 0.3476,
-      "step": 16417
-    },
-    {
-      "epoch": 1.0730017645905496,
-      "grad_norm": 0.46620863676071167,
-      "learning_rate": 7.345966847578831e-06,
-      "loss": 0.3528,
-      "step": 16418
-    },
-    {
-      "epoch": 1.0730671197960917,
-      "grad_norm": 0.47876277565956116,
-      "learning_rate": 7.345658473750186e-06,
-      "loss": 0.3961,
-      "step": 16419
-    },
-    {
-      "epoch": 1.0731324750016338,
-      "grad_norm": 0.47290778160095215,
-      "learning_rate": 7.345350088480951e-06,
-      "loss": 0.3804,
-      "step": 16420
-    },
-    {
-      "epoch": 1.073197830207176,
-      "grad_norm": 0.44703933596611023,
-      "learning_rate": 7.34504169177263e-06,
-      "loss": 0.3388,
-      "step": 16421
-    },
-    {
-      "epoch": 1.0732631854127181,
-      "grad_norm": 0.417653352022171,
-      "learning_rate": 7.344733283626728e-06,
-      "loss": 0.3017,
-      "step": 16422
-    },
-    {
-      "epoch": 1.0733285406182602,
-      "grad_norm": 0.48385077714920044,
-      "learning_rate": 7.344424864044751e-06,
-      "loss": 0.3913,
-      "step": 16423
-    },
-    {
-      "epoch": 1.0733938958238023,
-      "grad_norm": 0.4289487600326538,
-      "learning_rate": 7.3441164330282015e-06,
-      "loss": 0.3273,
-      "step": 16424
-    },
-    {
-      "epoch": 1.0734592510293446,
-      "grad_norm": 0.4383748769760132,
-      "learning_rate": 7.343807990578584e-06,
-      "loss": 0.3162,
-      "step": 16425
-    },
-    {
-      "epoch": 1.0735246062348867,
-      "grad_norm": 0.46777331829071045,
-      "learning_rate": 7.343499536697403e-06,
-      "loss": 0.4019,
-      "step": 16426
-    },
-    {
-      "epoch": 1.0735899614404287,
-      "grad_norm": 0.470061331987381,
-      "learning_rate": 7.343191071386162e-06,
-      "loss": 0.3817,
-      "step": 16427
-    },
-    {
-      "epoch": 1.0736553166459708,
-      "grad_norm": 0.4281580150127411,
-      "learning_rate": 7.342882594646368e-06,
-      "loss": 0.3224,
-      "step": 16428
-    },
-    {
-      "epoch": 1.0737206718515129,
-      "grad_norm": 0.5027729868888855,
-      "learning_rate": 7.342574106479522e-06,
-      "loss": 0.4134,
-      "step": 16429
-    },
-    {
-      "epoch": 1.0737860270570552,
-      "grad_norm": 0.4550667703151703,
-      "learning_rate": 7.342265606887132e-06,
-      "loss": 0.388,
-      "step": 16430
-    },
-    {
-      "epoch": 1.0738513822625972,
-      "grad_norm": 0.47943225502967834,
-      "learning_rate": 7.3419570958707e-06,
-      "loss": 0.3665,
-      "step": 16431
-    },
-    {
-      "epoch": 1.0739167374681393,
-      "grad_norm": 0.4488272964954376,
-      "learning_rate": 7.341648573431734e-06,
-      "loss": 0.312,
-      "step": 16432
-    },
-    {
-      "epoch": 1.0739820926736814,
-      "grad_norm": 0.4339866638183594,
-      "learning_rate": 7.341340039571734e-06,
-      "loss": 0.2891,
-      "step": 16433
-    },
-    {
-      "epoch": 1.0740474478792237,
-      "grad_norm": 0.4570108652114868,
-      "learning_rate": 7.341031494292209e-06,
-      "loss": 0.3712,
-      "step": 16434
-    },
-    {
-      "epoch": 1.0741128030847658,
-      "grad_norm": 0.4831475019454956,
-      "learning_rate": 7.340722937594663e-06,
-      "loss": 0.3824,
-      "step": 16435
-    },
-    {
-      "epoch": 1.0741781582903078,
-      "grad_norm": 0.438126802444458,
-      "learning_rate": 7.3404143694806e-06,
-      "loss": 0.3262,
-      "step": 16436
-    },
-    {
-      "epoch": 1.07424351349585,
-      "grad_norm": 0.44917792081832886,
-      "learning_rate": 7.340105789951524e-06,
-      "loss": 0.3042,
-      "step": 16437
-    },
-    {
-      "epoch": 1.074308868701392,
-      "grad_norm": 0.43990835547447205,
-      "learning_rate": 7.339797199008942e-06,
-      "loss": 0.3545,
-      "step": 16438
-    },
-    {
-      "epoch": 1.0743742239069343,
-      "grad_norm": 0.43734100461006165,
-      "learning_rate": 7.33948859665436e-06,
-      "loss": 0.3273,
-      "step": 16439
-    },
-    {
-      "epoch": 1.0744395791124763,
-      "grad_norm": 0.4544733464717865,
-      "learning_rate": 7.33917998288928e-06,
-      "loss": 0.3407,
-      "step": 16440
-    },
-    {
-      "epoch": 1.0745049343180184,
-      "grad_norm": 0.45032691955566406,
-      "learning_rate": 7.3388713577152095e-06,
-      "loss": 0.3571,
-      "step": 16441
-    },
-    {
-      "epoch": 1.0745702895235605,
-      "grad_norm": 0.439270555973053,
-      "learning_rate": 7.338562721133652e-06,
-      "loss": 0.3322,
-      "step": 16442
-    },
-    {
-      "epoch": 1.0746356447291028,
-      "grad_norm": 0.44508621096611023,
-      "learning_rate": 7.338254073146115e-06,
-      "loss": 0.3273,
-      "step": 16443
-    },
-    {
-      "epoch": 1.0747009999346449,
-      "grad_norm": 0.4212568700313568,
-      "learning_rate": 7.3379454137541015e-06,
-      "loss": 0.3178,
-      "step": 16444
-    },
-    {
-      "epoch": 1.074766355140187,
-      "grad_norm": 0.46879836916923523,
-      "learning_rate": 7.33763674295912e-06,
-      "loss": 0.3564,
-      "step": 16445
-    },
-    {
-      "epoch": 1.074831710345729,
-      "grad_norm": 0.4160691201686859,
-      "learning_rate": 7.337328060762674e-06,
-      "loss": 0.3006,
-      "step": 16446
-    },
-    {
-      "epoch": 1.074897065551271,
-      "grad_norm": 0.46859854459762573,
-      "learning_rate": 7.337019367166269e-06,
-      "loss": 0.3387,
-      "step": 16447
-    },
-    {
-      "epoch": 1.0749624207568134,
-      "grad_norm": 0.45349177718162537,
-      "learning_rate": 7.336710662171411e-06,
-      "loss": 0.3477,
-      "step": 16448
-    },
-    {
-      "epoch": 1.0750277759623554,
-      "grad_norm": 0.4608268737792969,
-      "learning_rate": 7.336401945779605e-06,
-      "loss": 0.3461,
-      "step": 16449
-    },
-    {
-      "epoch": 1.0750931311678975,
-      "grad_norm": 0.4648530185222626,
-      "learning_rate": 7.336093217992359e-06,
-      "loss": 0.3703,
-      "step": 16450
-    },
-    {
-      "epoch": 1.0751584863734396,
-      "grad_norm": 0.4413801431655884,
-      "learning_rate": 7.335784478811175e-06,
-      "loss": 0.3433,
-      "step": 16451
-    },
-    {
-      "epoch": 1.0752238415789819,
-      "grad_norm": 0.44498562812805176,
-      "learning_rate": 7.335475728237562e-06,
-      "loss": 0.328,
-      "step": 16452
-    },
-    {
-      "epoch": 1.075289196784524,
-      "grad_norm": 0.4815255403518677,
-      "learning_rate": 7.3351669662730245e-06,
-      "loss": 0.3981,
-      "step": 16453
-    },
-    {
-      "epoch": 1.075354551990066,
-      "grad_norm": 0.44285887479782104,
-      "learning_rate": 7.334858192919069e-06,
-      "loss": 0.3424,
-      "step": 16454
-    },
-    {
-      "epoch": 1.075419907195608,
-      "grad_norm": 0.4815467596054077,
-      "learning_rate": 7.3345494081772005e-06,
-      "loss": 0.3668,
-      "step": 16455
-    },
-    {
-      "epoch": 1.0754852624011502,
-      "grad_norm": 0.429591566324234,
-      "learning_rate": 7.3342406120489275e-06,
-      "loss": 0.3143,
-      "step": 16456
-    },
-    {
-      "epoch": 1.0755506176066925,
-      "grad_norm": 0.488942414522171,
-      "learning_rate": 7.333931804535753e-06,
-      "loss": 0.3861,
-      "step": 16457
-    },
-    {
-      "epoch": 1.0756159728122345,
-      "grad_norm": 0.5187869071960449,
-      "learning_rate": 7.333622985639184e-06,
-      "loss": 0.3928,
-      "step": 16458
-    },
-    {
-      "epoch": 1.0756813280177766,
-      "grad_norm": 0.46916940808296204,
-      "learning_rate": 7.333314155360729e-06,
-      "loss": 0.2882,
-      "step": 16459
-    },
-    {
-      "epoch": 1.0757466832233187,
-      "grad_norm": 0.46073561906814575,
-      "learning_rate": 7.3330053137018905e-06,
-      "loss": 0.3545,
-      "step": 16460
-    },
-    {
-      "epoch": 1.075812038428861,
-      "grad_norm": 0.40526437759399414,
-      "learning_rate": 7.332696460664178e-06,
-      "loss": 0.2692,
-      "step": 16461
-    },
-    {
-      "epoch": 1.075877393634403,
-      "grad_norm": 0.4188133776187897,
-      "learning_rate": 7.332387596249096e-06,
-      "loss": 0.305,
-      "step": 16462
-    },
-    {
-      "epoch": 1.0759427488399451,
-      "grad_norm": 0.4839152991771698,
-      "learning_rate": 7.332078720458154e-06,
-      "loss": 0.3972,
-      "step": 16463
-    },
-    {
-      "epoch": 1.0760081040454872,
-      "grad_norm": 0.4279783368110657,
-      "learning_rate": 7.331769833292853e-06,
-      "loss": 0.3122,
-      "step": 16464
-    },
-    {
-      "epoch": 1.0760734592510293,
-      "grad_norm": 0.41427361965179443,
-      "learning_rate": 7.331460934754704e-06,
-      "loss": 0.3015,
-      "step": 16465
-    },
-    {
-      "epoch": 1.0761388144565716,
-      "grad_norm": 0.4339550733566284,
-      "learning_rate": 7.331152024845214e-06,
-      "loss": 0.3076,
-      "step": 16466
-    },
-    {
-      "epoch": 1.0762041696621136,
-      "grad_norm": 0.41602861881256104,
-      "learning_rate": 7.330843103565885e-06,
-      "loss": 0.2882,
-      "step": 16467
-    },
-    {
-      "epoch": 1.0762695248676557,
-      "grad_norm": 0.43938368558883667,
-      "learning_rate": 7.330534170918229e-06,
-      "loss": 0.3717,
-      "step": 16468
-    },
-    {
-      "epoch": 1.0763348800731978,
-      "grad_norm": 0.46085667610168457,
-      "learning_rate": 7.330225226903749e-06,
-      "loss": 0.3675,
-      "step": 16469
-    },
-    {
-      "epoch": 1.07640023527874,
-      "grad_norm": 0.46619588136672974,
-      "learning_rate": 7.3299162715239536e-06,
-      "loss": 0.3642,
-      "step": 16470
-    },
-    {
-      "epoch": 1.0764655904842821,
-      "grad_norm": 0.4606631398200989,
-      "learning_rate": 7.32960730478035e-06,
-      "loss": 0.3556,
-      "step": 16471
-    },
-    {
-      "epoch": 1.0765309456898242,
-      "grad_norm": 0.4193075895309448,
-      "learning_rate": 7.329298326674444e-06,
-      "loss": 0.3133,
-      "step": 16472
-    },
-    {
-      "epoch": 1.0765963008953663,
-      "grad_norm": 0.4827231168746948,
-      "learning_rate": 7.3289893372077435e-06,
-      "loss": 0.3762,
-      "step": 16473
-    },
-    {
-      "epoch": 1.0766616561009084,
-      "grad_norm": 0.4541878402233124,
-      "learning_rate": 7.328680336381755e-06,
-      "loss": 0.3667,
-      "step": 16474
-    },
-    {
-      "epoch": 1.0767270113064507,
-      "grad_norm": 0.45358648896217346,
-      "learning_rate": 7.328371324197986e-06,
-      "loss": 0.3467,
-      "step": 16475
-    },
-    {
-      "epoch": 1.0767923665119927,
-      "grad_norm": 0.46547931432724,
-      "learning_rate": 7.328062300657943e-06,
-      "loss": 0.3499,
-      "step": 16476
-    },
-    {
-      "epoch": 1.0768577217175348,
-      "grad_norm": 0.45730358362197876,
-      "learning_rate": 7.327753265763136e-06,
-      "loss": 0.3589,
-      "step": 16477
-    },
-    {
-      "epoch": 1.0769230769230769,
-      "grad_norm": 0.47796475887298584,
-      "learning_rate": 7.327444219515066e-06,
-      "loss": 0.3846,
-      "step": 16478
-    },
-    {
-      "epoch": 1.076988432128619,
-      "grad_norm": 0.4266526699066162,
-      "learning_rate": 7.327135161915247e-06,
-      "loss": 0.3193,
-      "step": 16479
-    },
-    {
-      "epoch": 1.0770537873341612,
-      "grad_norm": 0.4600265920162201,
-      "learning_rate": 7.326826092965182e-06,
-      "loss": 0.3449,
-      "step": 16480
-    },
-    {
-      "epoch": 1.0771191425397033,
-      "grad_norm": 0.4514351487159729,
-      "learning_rate": 7.326517012666381e-06,
-      "loss": 0.3345,
-      "step": 16481
-    },
-    {
-      "epoch": 1.0771844977452454,
-      "grad_norm": 0.42739933729171753,
-      "learning_rate": 7.32620792102035e-06,
-      "loss": 0.3231,
-      "step": 16482
-    },
-    {
-      "epoch": 1.0772498529507875,
-      "grad_norm": 0.4579569697380066,
-      "learning_rate": 7.325898818028597e-06,
-      "loss": 0.3518,
-      "step": 16483
-    },
-    {
-      "epoch": 1.0773152081563298,
-      "grad_norm": 0.4700823128223419,
-      "learning_rate": 7.32558970369263e-06,
-      "loss": 0.3788,
-      "step": 16484
-    },
-    {
-      "epoch": 1.0773805633618718,
-      "grad_norm": 0.48954805731773376,
-      "learning_rate": 7.325280578013955e-06,
-      "loss": 0.4094,
-      "step": 16485
-    },
-    {
-      "epoch": 1.077445918567414,
-      "grad_norm": 0.4473172128200531,
-      "learning_rate": 7.3249714409940846e-06,
-      "loss": 0.3053,
-      "step": 16486
-    },
-    {
-      "epoch": 1.077511273772956,
-      "grad_norm": 0.4740552008152008,
-      "learning_rate": 7.32466229263452e-06,
-      "loss": 0.3613,
-      "step": 16487
-    },
-    {
-      "epoch": 1.077576628978498,
-      "grad_norm": 0.466447114944458,
-      "learning_rate": 7.324353132936773e-06,
-      "loss": 0.3444,
-      "step": 16488
-    },
-    {
-      "epoch": 1.0776419841840403,
-      "grad_norm": 0.4432372450828552,
-      "learning_rate": 7.3240439619023495e-06,
-      "loss": 0.3414,
-      "step": 16489
-    },
-    {
-      "epoch": 1.0777073393895824,
-      "grad_norm": 0.4533449113368988,
-      "learning_rate": 7.323734779532761e-06,
-      "loss": 0.3494,
-      "step": 16490
-    },
-    {
-      "epoch": 1.0777726945951245,
-      "grad_norm": 0.4982556998729706,
-      "learning_rate": 7.32342558582951e-06,
-      "loss": 0.3855,
-      "step": 16491
-    },
-    {
-      "epoch": 1.0778380498006666,
-      "grad_norm": 0.47109490633010864,
-      "learning_rate": 7.3231163807941094e-06,
-      "loss": 0.3666,
-      "step": 16492
-    },
-    {
-      "epoch": 1.0779034050062088,
-      "grad_norm": 0.4856020510196686,
-      "learning_rate": 7.322807164428065e-06,
-      "loss": 0.4174,
-      "step": 16493
-    },
-    {
-      "epoch": 1.077968760211751,
-      "grad_norm": 0.4741486608982086,
-      "learning_rate": 7.322497936732885e-06,
-      "loss": 0.3674,
-      "step": 16494
-    },
-    {
-      "epoch": 1.078034115417293,
-      "grad_norm": 0.5163902044296265,
-      "learning_rate": 7.322188697710079e-06,
-      "loss": 0.4159,
-      "step": 16495
-    },
-    {
-      "epoch": 1.078099470622835,
-      "grad_norm": 0.4352622628211975,
-      "learning_rate": 7.3218794473611535e-06,
-      "loss": 0.3349,
-      "step": 16496
-    },
-    {
-      "epoch": 1.0781648258283771,
-      "grad_norm": 0.470027893781662,
-      "learning_rate": 7.321570185687616e-06,
-      "loss": 0.3793,
-      "step": 16497
-    },
-    {
-      "epoch": 1.0782301810339194,
-      "grad_norm": 0.4309051036834717,
-      "learning_rate": 7.321260912690979e-06,
-      "loss": 0.2972,
-      "step": 16498
-    },
-    {
-      "epoch": 1.0782955362394615,
-      "grad_norm": 0.42376551032066345,
-      "learning_rate": 7.320951628372748e-06,
-      "loss": 0.3162,
-      "step": 16499
-    },
-    {
-      "epoch": 1.0783608914450036,
-      "grad_norm": 0.4436875283718109,
-      "learning_rate": 7.3206423327344314e-06,
-      "loss": 0.3284,
-      "step": 16500
-    },
-    {
-      "epoch": 1.0784262466505457,
-      "grad_norm": 0.44389382004737854,
-      "learning_rate": 7.320333025777537e-06,
-      "loss": 0.3438,
-      "step": 16501
-    },
-    {
-      "epoch": 1.0784916018560877,
-      "grad_norm": 0.42851725220680237,
-      "learning_rate": 7.320023707503576e-06,
-      "loss": 0.3202,
-      "step": 16502
-    },
-    {
-      "epoch": 1.07855695706163,
-      "grad_norm": 0.4838922619819641,
-      "learning_rate": 7.319714377914056e-06,
-      "loss": 0.3562,
-      "step": 16503
-    },
-    {
-      "epoch": 1.078622312267172,
-      "grad_norm": 0.47055715322494507,
-      "learning_rate": 7.319405037010484e-06,
-      "loss": 0.3466,
-      "step": 16504
-    },
-    {
-      "epoch": 1.0786876674727142,
-      "grad_norm": 0.4585447311401367,
-      "learning_rate": 7.31909568479437e-06,
-      "loss": 0.3712,
-      "step": 16505
-    },
-    {
-      "epoch": 1.0787530226782562,
-      "grad_norm": 0.4848266839981079,
-      "learning_rate": 7.318786321267224e-06,
-      "loss": 0.3992,
-      "step": 16506
-    },
-    {
-      "epoch": 1.0788183778837985,
-      "grad_norm": 0.5263270139694214,
-      "learning_rate": 7.318476946430553e-06,
-      "loss": 0.4139,
-      "step": 16507
-    },
-    {
-      "epoch": 1.0788837330893406,
-      "grad_norm": 0.42925411462783813,
-      "learning_rate": 7.318167560285868e-06,
-      "loss": 0.294,
-      "step": 16508
-    },
-    {
-      "epoch": 1.0789490882948827,
-      "grad_norm": 0.469771146774292,
-      "learning_rate": 7.317858162834674e-06,
-      "loss": 0.3535,
-      "step": 16509
-    },
-    {
-      "epoch": 1.0790144435004247,
-      "grad_norm": 0.4327908456325531,
-      "learning_rate": 7.317548754078486e-06,
-      "loss": 0.3095,
-      "step": 16510
-    },
-    {
-      "epoch": 1.0790797987059668,
-      "grad_norm": 0.4393147826194763,
-      "learning_rate": 7.317239334018809e-06,
-      "loss": 0.3269,
-      "step": 16511
-    },
-    {
-      "epoch": 1.0791451539115091,
-      "grad_norm": 0.41900089383125305,
-      "learning_rate": 7.316929902657152e-06,
-      "loss": 0.2763,
-      "step": 16512
-    },
-    {
-      "epoch": 1.0792105091170512,
-      "grad_norm": 0.412723183631897,
-      "learning_rate": 7.316620459995027e-06,
-      "loss": 0.2744,
-      "step": 16513
-    },
-    {
-      "epoch": 1.0792758643225933,
-      "grad_norm": 0.4568878412246704,
-      "learning_rate": 7.316311006033938e-06,
-      "loss": 0.3509,
-      "step": 16514
-    },
-    {
-      "epoch": 1.0793412195281353,
-      "grad_norm": 0.4407266080379486,
-      "learning_rate": 7.316001540775401e-06,
-      "loss": 0.3305,
-      "step": 16515
-    },
-    {
-      "epoch": 1.0794065747336776,
-      "grad_norm": 0.4252583086490631,
-      "learning_rate": 7.315692064220921e-06,
-      "loss": 0.3049,
-      "step": 16516
-    },
-    {
-      "epoch": 1.0794719299392197,
-      "grad_norm": 0.5093232989311218,
-      "learning_rate": 7.315382576372009e-06,
-      "loss": 0.3745,
-      "step": 16517
-    },
-    {
-      "epoch": 1.0795372851447618,
-      "grad_norm": 0.43224215507507324,
-      "learning_rate": 7.315073077230173e-06,
-      "loss": 0.3075,
-      "step": 16518
-    },
-    {
-      "epoch": 1.0796026403503038,
-      "grad_norm": 0.4391690194606781,
-      "learning_rate": 7.314763566796925e-06,
-      "loss": 0.3346,
-      "step": 16519
-    },
-    {
-      "epoch": 1.079667995555846,
-      "grad_norm": 0.46438995003700256,
-      "learning_rate": 7.314454045073772e-06,
-      "loss": 0.3624,
-      "step": 16520
-    },
-    {
-      "epoch": 1.0797333507613882,
-      "grad_norm": 0.4262702763080597,
-      "learning_rate": 7.314144512062225e-06,
-      "loss": 0.3174,
-      "step": 16521
-    },
-    {
-      "epoch": 1.0797987059669303,
-      "grad_norm": 0.4562312066555023,
-      "learning_rate": 7.313834967763795e-06,
-      "loss": 0.3395,
-      "step": 16522
-    },
-    {
-      "epoch": 1.0798640611724724,
-      "grad_norm": 0.4660331606864929,
-      "learning_rate": 7.313525412179989e-06,
-      "loss": 0.3577,
-      "step": 16523
-    },
-    {
-      "epoch": 1.0799294163780144,
-      "grad_norm": 0.4537416100502014,
-      "learning_rate": 7.313215845312318e-06,
-      "loss": 0.3696,
-      "step": 16524
-    },
-    {
-      "epoch": 1.0799947715835567,
-      "grad_norm": 0.4765056371688843,
-      "learning_rate": 7.312906267162292e-06,
-      "loss": 0.358,
-      "step": 16525
-    },
-    {
-      "epoch": 1.0800601267890988,
-      "grad_norm": 0.403167724609375,
-      "learning_rate": 7.31259667773142e-06,
-      "loss": 0.256,
-      "step": 16526
-    },
-    {
-      "epoch": 1.0801254819946409,
-      "grad_norm": 0.4833615720272064,
-      "learning_rate": 7.312287077021214e-06,
-      "loss": 0.3468,
-      "step": 16527
-    },
-    {
-      "epoch": 1.080190837200183,
-      "grad_norm": 0.4581005871295929,
-      "learning_rate": 7.311977465033183e-06,
-      "loss": 0.3421,
-      "step": 16528
-    },
-    {
-      "epoch": 1.080256192405725,
-      "grad_norm": 0.4372236430644989,
-      "learning_rate": 7.311667841768837e-06,
-      "loss": 0.3001,
-      "step": 16529
-    },
-    {
-      "epoch": 1.0803215476112673,
-      "grad_norm": 0.4552151560783386,
-      "learning_rate": 7.311358207229685e-06,
-      "loss": 0.3439,
-      "step": 16530
-    },
-    {
-      "epoch": 1.0803869028168094,
-      "grad_norm": 0.40541872382164,
-      "learning_rate": 7.31104856141724e-06,
-      "loss": 0.295,
-      "step": 16531
-    },
-    {
-      "epoch": 1.0804522580223515,
-      "grad_norm": 0.4337131381034851,
-      "learning_rate": 7.310738904333009e-06,
-      "loss": 0.3135,
-      "step": 16532
-    },
-    {
-      "epoch": 1.0805176132278935,
-      "grad_norm": 0.42380663752555847,
-      "learning_rate": 7.310429235978504e-06,
-      "loss": 0.306,
-      "step": 16533
-    },
-    {
-      "epoch": 1.0805829684334358,
-      "grad_norm": 0.4487023651599884,
-      "learning_rate": 7.310119556355235e-06,
-      "loss": 0.342,
-      "step": 16534
-    },
-    {
-      "epoch": 1.080648323638978,
-      "grad_norm": 0.4362575113773346,
-      "learning_rate": 7.309809865464714e-06,
-      "loss": 0.3431,
-      "step": 16535
-    },
-    {
-      "epoch": 1.08071367884452,
-      "grad_norm": 0.4611964523792267,
-      "learning_rate": 7.3095001633084495e-06,
-      "loss": 0.3738,
-      "step": 16536
-    },
-    {
-      "epoch": 1.080779034050062,
-      "grad_norm": 0.43904024362564087,
-      "learning_rate": 7.309190449887953e-06,
-      "loss": 0.3321,
-      "step": 16537
-    },
-    {
-      "epoch": 1.080844389255604,
-      "grad_norm": 0.4666799306869507,
-      "learning_rate": 7.3088807252047345e-06,
-      "loss": 0.3767,
-      "step": 16538
-    },
-    {
-      "epoch": 1.0809097444611464,
-      "grad_norm": 0.43932658433914185,
-      "learning_rate": 7.308570989260304e-06,
-      "loss": 0.3191,
-      "step": 16539
-    },
-    {
-      "epoch": 1.0809750996666885,
-      "grad_norm": 0.469312846660614,
-      "learning_rate": 7.308261242056174e-06,
-      "loss": 0.3545,
-      "step": 16540
-    },
-    {
-      "epoch": 1.0810404548722305,
-      "grad_norm": 0.4795440137386322,
-      "learning_rate": 7.3079514835938536e-06,
-      "loss": 0.3001,
-      "step": 16541
-    },
-    {
-      "epoch": 1.0811058100777726,
-      "grad_norm": 0.41839662194252014,
-      "learning_rate": 7.307641713874855e-06,
-      "loss": 0.3012,
-      "step": 16542
-    },
-    {
-      "epoch": 1.081171165283315,
-      "grad_norm": 0.4089526832103729,
-      "learning_rate": 7.307331932900688e-06,
-      "loss": 0.3355,
-      "step": 16543
-    },
-    {
-      "epoch": 1.081236520488857,
-      "grad_norm": 0.4382639229297638,
-      "learning_rate": 7.307022140672863e-06,
-      "loss": 0.3158,
-      "step": 16544
-    },
-    {
-      "epoch": 1.081301875694399,
-      "grad_norm": 0.49214106798171997,
-      "learning_rate": 7.3067123371928915e-06,
-      "loss": 0.3886,
-      "step": 16545
-    },
-    {
-      "epoch": 1.0813672308999411,
-      "grad_norm": 0.47231751680374146,
-      "learning_rate": 7.306402522462286e-06,
-      "loss": 0.3779,
-      "step": 16546
-    },
-    {
-      "epoch": 1.0814325861054832,
-      "grad_norm": 0.4860769510269165,
-      "learning_rate": 7.306092696482556e-06,
-      "loss": 0.3672,
-      "step": 16547
-    },
-    {
-      "epoch": 1.0814979413110255,
-      "grad_norm": 0.49187856912612915,
-      "learning_rate": 7.305782859255213e-06,
-      "loss": 0.3827,
-      "step": 16548
-    },
-    {
-      "epoch": 1.0815632965165676,
-      "grad_norm": 0.43406373262405396,
-      "learning_rate": 7.305473010781767e-06,
-      "loss": 0.321,
-      "step": 16549
-    },
-    {
-      "epoch": 1.0816286517221096,
-      "grad_norm": 0.4567713439464569,
-      "learning_rate": 7.305163151063733e-06,
-      "loss": 0.3867,
-      "step": 16550
-    },
-    {
-      "epoch": 1.0816940069276517,
-      "grad_norm": 0.43306538462638855,
-      "learning_rate": 7.304853280102618e-06,
-      "loss": 0.3175,
-      "step": 16551
-    },
-    {
-      "epoch": 1.081759362133194,
-      "grad_norm": 0.4578908085823059,
-      "learning_rate": 7.304543397899936e-06,
-      "loss": 0.3499,
-      "step": 16552
-    },
-    {
-      "epoch": 1.081824717338736,
-      "grad_norm": 0.4774589538574219,
-      "learning_rate": 7.304233504457197e-06,
-      "loss": 0.3686,
-      "step": 16553
-    },
-    {
-      "epoch": 1.0818900725442782,
-      "grad_norm": 0.46102818846702576,
-      "learning_rate": 7.303923599775913e-06,
-      "loss": 0.32,
-      "step": 16554
-    },
-    {
-      "epoch": 1.0819554277498202,
-      "grad_norm": 0.4556474983692169,
-      "learning_rate": 7.303613683857596e-06,
-      "loss": 0.3604,
-      "step": 16555
-    },
-    {
-      "epoch": 1.0820207829553623,
-      "grad_norm": 0.4312027394771576,
-      "learning_rate": 7.3033037567037545e-06,
-      "loss": 0.3155,
-      "step": 16556
-    },
-    {
-      "epoch": 1.0820861381609046,
-      "grad_norm": 0.4546574354171753,
-      "learning_rate": 7.302993818315906e-06,
-      "loss": 0.3319,
-      "step": 16557
-    },
-    {
-      "epoch": 1.0821514933664467,
-      "grad_norm": 0.43697255849838257,
-      "learning_rate": 7.302683868695557e-06,
-      "loss": 0.3518,
-      "step": 16558
-    },
-    {
-      "epoch": 1.0822168485719887,
-      "grad_norm": 0.4205278754234314,
-      "learning_rate": 7.30237390784422e-06,
-      "loss": 0.2999,
-      "step": 16559
-    },
-    {
-      "epoch": 1.0822822037775308,
-      "grad_norm": 0.4171440303325653,
-      "learning_rate": 7.3020639357634105e-06,
-      "loss": 0.3081,
-      "step": 16560
-    },
-    {
-      "epoch": 1.082347558983073,
-      "grad_norm": 0.47382208704948425,
-      "learning_rate": 7.301753952454636e-06,
-      "loss": 0.3629,
-      "step": 16561
-    },
-    {
-      "epoch": 1.0824129141886152,
-      "grad_norm": 0.46627259254455566,
-      "learning_rate": 7.301443957919412e-06,
-      "loss": 0.3904,
-      "step": 16562
-    },
-    {
-      "epoch": 1.0824782693941573,
-      "grad_norm": 0.45141178369522095,
-      "learning_rate": 7.301133952159246e-06,
-      "loss": 0.3585,
-      "step": 16563
-    },
-    {
-      "epoch": 1.0825436245996993,
-      "grad_norm": 0.4470028877258301,
-      "learning_rate": 7.300823935175654e-06,
-      "loss": 0.329,
-      "step": 16564
-    },
-    {
-      "epoch": 1.0826089798052414,
-      "grad_norm": 0.4529719352722168,
-      "learning_rate": 7.300513906970146e-06,
-      "loss": 0.3575,
-      "step": 16565
-    },
-    {
-      "epoch": 1.0826743350107837,
-      "grad_norm": 0.46421998739242554,
-      "learning_rate": 7.3002038675442345e-06,
-      "loss": 0.3287,
-      "step": 16566
-    },
-    {
-      "epoch": 1.0827396902163258,
-      "grad_norm": 0.4930346608161926,
-      "learning_rate": 7.299893816899432e-06,
-      "loss": 0.3565,
-      "step": 16567
-    },
-    {
-      "epoch": 1.0828050454218678,
-      "grad_norm": 0.4851548969745636,
-      "learning_rate": 7.299583755037251e-06,
-      "loss": 0.3723,
-      "step": 16568
-    },
-    {
-      "epoch": 1.08287040062741,
-      "grad_norm": 0.44411322474479675,
-      "learning_rate": 7.299273681959202e-06,
-      "loss": 0.3256,
-      "step": 16569
-    },
-    {
-      "epoch": 1.0829357558329522,
-      "grad_norm": 0.4573776125907898,
-      "learning_rate": 7.2989635976668e-06,
-      "loss": 0.3408,
-      "step": 16570
-    },
-    {
-      "epoch": 1.0830011110384943,
-      "grad_norm": 0.4168480634689331,
-      "learning_rate": 7.2986535021615565e-06,
-      "loss": 0.3042,
-      "step": 16571
-    },
-    {
-      "epoch": 1.0830664662440364,
-      "grad_norm": 0.4562520682811737,
-      "learning_rate": 7.298343395444982e-06,
-      "loss": 0.3362,
-      "step": 16572
-    },
-    {
-      "epoch": 1.0831318214495784,
-      "grad_norm": 0.48239827156066895,
-      "learning_rate": 7.298033277518592e-06,
-      "loss": 0.3679,
-      "step": 16573
-    },
-    {
-      "epoch": 1.0831971766551205,
-      "grad_norm": 0.44063931703567505,
-      "learning_rate": 7.2977231483838975e-06,
-      "loss": 0.3185,
-      "step": 16574
-    },
-    {
-      "epoch": 1.0832625318606628,
-      "grad_norm": 0.47119778394699097,
-      "learning_rate": 7.297413008042411e-06,
-      "loss": 0.3706,
-      "step": 16575
-    },
-    {
-      "epoch": 1.0833278870662049,
-      "grad_norm": 0.47423574328422546,
-      "learning_rate": 7.297102856495644e-06,
-      "loss": 0.3848,
-      "step": 16576
-    },
-    {
-      "epoch": 1.083393242271747,
-      "grad_norm": 0.4561114013195038,
-      "learning_rate": 7.296792693745113e-06,
-      "loss": 0.3536,
-      "step": 16577
-    },
-    {
-      "epoch": 1.083458597477289,
-      "grad_norm": 0.4443838894367218,
-      "learning_rate": 7.2964825197923275e-06,
-      "loss": 0.3422,
-      "step": 16578
-    },
-    {
-      "epoch": 1.0835239526828313,
-      "grad_norm": 0.5358433127403259,
-      "learning_rate": 7.2961723346387996e-06,
-      "loss": 0.3857,
-      "step": 16579
-    },
-    {
-      "epoch": 1.0835893078883734,
-      "grad_norm": 0.46897754073143005,
-      "learning_rate": 7.295862138286045e-06,
-      "loss": 0.3495,
-      "step": 16580
-    },
-    {
-      "epoch": 1.0836546630939154,
-      "grad_norm": 0.4311322569847107,
-      "learning_rate": 7.295551930735575e-06,
-      "loss": 0.3175,
-      "step": 16581
-    },
-    {
-      "epoch": 1.0837200182994575,
-      "grad_norm": 0.4757993221282959,
-      "learning_rate": 7.295241711988905e-06,
-      "loss": 0.3386,
-      "step": 16582
-    },
-    {
-      "epoch": 1.0837853735049996,
-      "grad_norm": 0.493779718875885,
-      "learning_rate": 7.294931482047544e-06,
-      "loss": 0.3759,
-      "step": 16583
-    },
-    {
-      "epoch": 1.0838507287105419,
-      "grad_norm": 0.459670752286911,
-      "learning_rate": 7.294621240913007e-06,
-      "loss": 0.3342,
-      "step": 16584
-    },
-    {
-      "epoch": 1.083916083916084,
-      "grad_norm": 0.405508816242218,
-      "learning_rate": 7.294310988586807e-06,
-      "loss": 0.2745,
-      "step": 16585
-    },
-    {
-      "epoch": 1.083981439121626,
-      "grad_norm": 0.4545801877975464,
-      "learning_rate": 7.294000725070458e-06,
-      "loss": 0.3489,
-      "step": 16586
-    },
-    {
-      "epoch": 1.084046794327168,
-      "grad_norm": 0.48119500279426575,
-      "learning_rate": 7.2936904503654736e-06,
-      "loss": 0.3517,
-      "step": 16587
-    },
-    {
-      "epoch": 1.0841121495327102,
-      "grad_norm": 0.4570417106151581,
-      "learning_rate": 7.293380164473365e-06,
-      "loss": 0.3557,
-      "step": 16588
-    },
-    {
-      "epoch": 1.0841775047382525,
-      "grad_norm": 0.47389963269233704,
-      "learning_rate": 7.293069867395648e-06,
-      "loss": 0.3664,
-      "step": 16589
-    },
-    {
-      "epoch": 1.0842428599437945,
-      "grad_norm": 0.46701309084892273,
-      "learning_rate": 7.292759559133834e-06,
-      "loss": 0.3657,
-      "step": 16590
-    },
-    {
-      "epoch": 1.0843082151493366,
-      "grad_norm": 0.45863738656044006,
-      "learning_rate": 7.2924492396894385e-06,
-      "loss": 0.3456,
-      "step": 16591
-    },
-    {
-      "epoch": 1.0843735703548787,
-      "grad_norm": 0.4857933223247528,
-      "learning_rate": 7.292138909063972e-06,
-      "loss": 0.3812,
-      "step": 16592
-    },
-    {
-      "epoch": 1.084438925560421,
-      "grad_norm": 0.45012250542640686,
-      "learning_rate": 7.291828567258952e-06,
-      "loss": 0.3387,
-      "step": 16593
-    },
-    {
-      "epoch": 1.084504280765963,
-      "grad_norm": 0.42833054065704346,
-      "learning_rate": 7.291518214275888e-06,
-      "loss": 0.3393,
-      "step": 16594
-    },
-    {
-      "epoch": 1.0845696359715051,
-      "grad_norm": 0.4619414806365967,
-      "learning_rate": 7.291207850116298e-06,
-      "loss": 0.3357,
-      "step": 16595
-    },
-    {
-      "epoch": 1.0846349911770472,
-      "grad_norm": 0.45225340127944946,
-      "learning_rate": 7.290897474781692e-06,
-      "loss": 0.3457,
-      "step": 16596
-    },
-    {
-      "epoch": 1.0847003463825893,
-      "grad_norm": 0.4553976356983185,
-      "learning_rate": 7.290587088273586e-06,
-      "loss": 0.3762,
-      "step": 16597
-    },
-    {
-      "epoch": 1.0847657015881316,
-      "grad_norm": 0.4105634093284607,
-      "learning_rate": 7.290276690593493e-06,
-      "loss": 0.3153,
-      "step": 16598
-    },
-    {
-      "epoch": 1.0848310567936736,
-      "grad_norm": 0.4671085476875305,
-      "learning_rate": 7.289966281742926e-06,
-      "loss": 0.3669,
-      "step": 16599
-    },
-    {
-      "epoch": 1.0848964119992157,
-      "grad_norm": 0.48248225450515747,
-      "learning_rate": 7.2896558617234014e-06,
-      "loss": 0.3773,
-      "step": 16600
-    },
-    {
-      "epoch": 1.0849617672047578,
-      "grad_norm": 0.42600542306900024,
-      "learning_rate": 7.2893454305364316e-06,
-      "loss": 0.3315,
-      "step": 16601
-    },
-    {
-      "epoch": 1.0850271224103,
-      "grad_norm": 0.42937231063842773,
-      "learning_rate": 7.289034988183532e-06,
-      "loss": 0.3507,
-      "step": 16602
-    },
-    {
-      "epoch": 1.0850924776158422,
-      "grad_norm": 0.441170871257782,
-      "learning_rate": 7.288724534666215e-06,
-      "loss": 0.3409,
-      "step": 16603
-    },
-    {
-      "epoch": 1.0851578328213842,
-      "grad_norm": 0.47401493787765503,
-      "learning_rate": 7.288414069985996e-06,
-      "loss": 0.3792,
-      "step": 16604
-    },
-    {
-      "epoch": 1.0852231880269263,
-      "grad_norm": 0.4646795690059662,
-      "learning_rate": 7.288103594144389e-06,
-      "loss": 0.3706,
-      "step": 16605
-    },
-    {
-      "epoch": 1.0852885432324684,
-      "grad_norm": 0.412864625453949,
-      "learning_rate": 7.287793107142908e-06,
-      "loss": 0.2999,
-      "step": 16606
-    },
-    {
-      "epoch": 1.0853538984380107,
-      "grad_norm": 0.4567086696624756,
-      "learning_rate": 7.287482608983066e-06,
-      "loss": 0.3645,
-      "step": 16607
-    },
-    {
-      "epoch": 1.0854192536435527,
-      "grad_norm": 0.4699552655220032,
-      "learning_rate": 7.2871720996663794e-06,
-      "loss": 0.3602,
-      "step": 16608
-    },
-    {
-      "epoch": 1.0854846088490948,
-      "grad_norm": 0.4367991089820862,
-      "learning_rate": 7.286861579194363e-06,
-      "loss": 0.321,
-      "step": 16609
-    },
-    {
-      "epoch": 1.0855499640546369,
-      "grad_norm": 0.4446961581707001,
-      "learning_rate": 7.28655104756853e-06,
-      "loss": 0.3254,
-      "step": 16610
-    },
-    {
-      "epoch": 1.085615319260179,
-      "grad_norm": 0.43319615721702576,
-      "learning_rate": 7.286240504790396e-06,
-      "loss": 0.3424,
-      "step": 16611
-    },
-    {
-      "epoch": 1.0856806744657213,
-      "grad_norm": 0.5113908648490906,
-      "learning_rate": 7.285929950861474e-06,
-      "loss": 0.3604,
-      "step": 16612
-    },
-    {
-      "epoch": 1.0857460296712633,
-      "grad_norm": 0.4660574793815613,
-      "learning_rate": 7.28561938578328e-06,
-      "loss": 0.3717,
-      "step": 16613
-    },
-    {
-      "epoch": 1.0858113848768054,
-      "grad_norm": 0.47809529304504395,
-      "learning_rate": 7.285308809557328e-06,
-      "loss": 0.3558,
-      "step": 16614
-    },
-    {
-      "epoch": 1.0858767400823475,
-      "grad_norm": 0.42160192131996155,
-      "learning_rate": 7.284998222185134e-06,
-      "loss": 0.332,
-      "step": 16615
-    },
-    {
-      "epoch": 1.0859420952878898,
-      "grad_norm": 0.7761409282684326,
-      "learning_rate": 7.284687623668212e-06,
-      "loss": 0.2734,
-      "step": 16616
-    },
-    {
-      "epoch": 1.0860074504934318,
-      "grad_norm": 0.4404222071170807,
-      "learning_rate": 7.284377014008077e-06,
-      "loss": 0.2978,
-      "step": 16617
-    },
-    {
-      "epoch": 1.086072805698974,
-      "grad_norm": 0.4651528000831604,
-      "learning_rate": 7.284066393206244e-06,
-      "loss": 0.3765,
-      "step": 16618
-    },
-    {
-      "epoch": 1.086138160904516,
-      "grad_norm": 0.44234445691108704,
-      "learning_rate": 7.283755761264226e-06,
-      "loss": 0.3143,
-      "step": 16619
-    },
-    {
-      "epoch": 1.086203516110058,
-      "grad_norm": 0.45051178336143494,
-      "learning_rate": 7.283445118183543e-06,
-      "loss": 0.3292,
-      "step": 16620
-    },
-    {
-      "epoch": 1.0862688713156003,
-      "grad_norm": 0.44953030347824097,
-      "learning_rate": 7.283134463965706e-06,
-      "loss": 0.3201,
-      "step": 16621
-    },
-    {
-      "epoch": 1.0863342265211424,
-      "grad_norm": 0.4718022048473358,
-      "learning_rate": 7.2828237986122304e-06,
-      "loss": 0.3647,
-      "step": 16622
-    },
-    {
-      "epoch": 1.0863995817266845,
-      "grad_norm": 0.46884533762931824,
-      "learning_rate": 7.2825131221246325e-06,
-      "loss": 0.323,
-      "step": 16623
-    },
-    {
-      "epoch": 1.0864649369322266,
-      "grad_norm": 0.460284560918808,
-      "learning_rate": 7.282202434504428e-06,
-      "loss": 0.3487,
-      "step": 16624
-    },
-    {
-      "epoch": 1.0865302921377689,
-      "grad_norm": 0.44646936655044556,
-      "learning_rate": 7.281891735753132e-06,
-      "loss": 0.3438,
-      "step": 16625
-    },
-    {
-      "epoch": 1.086595647343311,
-      "grad_norm": 0.43540501594543457,
-      "learning_rate": 7.281581025872258e-06,
-      "loss": 0.3481,
-      "step": 16626
-    },
-    {
-      "epoch": 1.086661002548853,
-      "grad_norm": 0.46796274185180664,
-      "learning_rate": 7.281270304863325e-06,
-      "loss": 0.3581,
-      "step": 16627
-    },
-    {
-      "epoch": 1.086726357754395,
-      "grad_norm": 0.471216082572937,
-      "learning_rate": 7.280959572727845e-06,
-      "loss": 0.3666,
-      "step": 16628
-    },
-    {
-      "epoch": 1.0867917129599372,
-      "grad_norm": 0.4512973427772522,
-      "learning_rate": 7.280648829467335e-06,
-      "loss": 0.3325,
-      "step": 16629
-    },
-    {
-      "epoch": 1.0868570681654794,
-      "grad_norm": 0.4285413324832916,
-      "learning_rate": 7.2803380750833105e-06,
-      "loss": 0.2966,
-      "step": 16630
-    },
-    {
-      "epoch": 1.0869224233710215,
-      "grad_norm": 0.46235746145248413,
-      "learning_rate": 7.280027309577288e-06,
-      "loss": 0.3445,
-      "step": 16631
-    },
-    {
-      "epoch": 1.0869877785765636,
-      "grad_norm": 0.4764554500579834,
-      "learning_rate": 7.279716532950781e-06,
-      "loss": 0.3852,
-      "step": 16632
-    },
-    {
-      "epoch": 1.0870531337821057,
-      "grad_norm": 0.4783000349998474,
-      "learning_rate": 7.279405745205308e-06,
-      "loss": 0.3554,
-      "step": 16633
-    },
-    {
-      "epoch": 1.087118488987648,
-      "grad_norm": 0.46820077300071716,
-      "learning_rate": 7.279094946342383e-06,
-      "loss": 0.3709,
-      "step": 16634
-    },
-    {
-      "epoch": 1.08718384419319,
-      "grad_norm": 0.4597938060760498,
-      "learning_rate": 7.2787841363635225e-06,
-      "loss": 0.3285,
-      "step": 16635
-    },
-    {
-      "epoch": 1.087249199398732,
-      "grad_norm": 0.4381828010082245,
-      "learning_rate": 7.278473315270242e-06,
-      "loss": 0.3536,
-      "step": 16636
-    },
-    {
-      "epoch": 1.0873145546042742,
-      "grad_norm": 0.4854438304901123,
-      "learning_rate": 7.278162483064057e-06,
-      "loss": 0.4449,
-      "step": 16637
-    },
-    {
-      "epoch": 1.0873799098098162,
-      "grad_norm": 0.45786261558532715,
-      "learning_rate": 7.277851639746487e-06,
-      "loss": 0.3612,
-      "step": 16638
-    },
-    {
-      "epoch": 1.0874452650153585,
-      "grad_norm": 0.4783628284931183,
-      "learning_rate": 7.277540785319044e-06,
-      "loss": 0.3537,
-      "step": 16639
-    },
-    {
-      "epoch": 1.0875106202209006,
-      "grad_norm": 0.44216424226760864,
-      "learning_rate": 7.277229919783246e-06,
-      "loss": 0.3429,
-      "step": 16640
-    },
-    {
-      "epoch": 1.0875759754264427,
-      "grad_norm": 0.4420805871486664,
-      "learning_rate": 7.276919043140607e-06,
-      "loss": 0.3023,
-      "step": 16641
-    },
-    {
-      "epoch": 1.0876413306319848,
-      "grad_norm": 0.4524914026260376,
-      "learning_rate": 7.276608155392646e-06,
-      "loss": 0.3175,
-      "step": 16642
-    },
-    {
-      "epoch": 1.087706685837527,
-      "grad_norm": 0.4534817039966583,
-      "learning_rate": 7.27629725654088e-06,
-      "loss": 0.3516,
-      "step": 16643
-    },
-    {
-      "epoch": 1.0877720410430691,
-      "grad_norm": 0.43316760659217834,
-      "learning_rate": 7.275986346586821e-06,
-      "loss": 0.3229,
-      "step": 16644
-    },
-    {
-      "epoch": 1.0878373962486112,
-      "grad_norm": 0.4456009566783905,
-      "learning_rate": 7.275675425531991e-06,
-      "loss": 0.3459,
-      "step": 16645
-    },
-    {
-      "epoch": 1.0879027514541533,
-      "grad_norm": 0.47747665643692017,
-      "learning_rate": 7.275364493377901e-06,
-      "loss": 0.3522,
-      "step": 16646
-    },
-    {
-      "epoch": 1.0879681066596953,
-      "grad_norm": 0.47131046652793884,
-      "learning_rate": 7.275053550126072e-06,
-      "loss": 0.3549,
-      "step": 16647
-    },
-    {
-      "epoch": 1.0880334618652376,
-      "grad_norm": 0.4717196524143219,
-      "learning_rate": 7.274742595778017e-06,
-      "loss": 0.2843,
-      "step": 16648
-    },
-    {
-      "epoch": 1.0880988170707797,
-      "grad_norm": 0.43104586005210876,
-      "learning_rate": 7.274431630335255e-06,
-      "loss": 0.3127,
-      "step": 16649
-    },
-    {
-      "epoch": 1.0881641722763218,
-      "grad_norm": 0.43192827701568604,
-      "learning_rate": 7.274120653799302e-06,
-      "loss": 0.3303,
-      "step": 16650
-    },
-    {
-      "epoch": 1.0882295274818639,
-      "grad_norm": 0.471676230430603,
-      "learning_rate": 7.273809666171675e-06,
-      "loss": 0.3577,
-      "step": 16651
-    },
-    {
-      "epoch": 1.0882948826874062,
-      "grad_norm": 0.44355347752571106,
-      "learning_rate": 7.27349866745389e-06,
-      "loss": 0.3378,
-      "step": 16652
-    },
-    {
-      "epoch": 1.0883602378929482,
-      "grad_norm": 0.4836813807487488,
-      "learning_rate": 7.273187657647465e-06,
-      "loss": 0.3551,
-      "step": 16653
-    },
-    {
-      "epoch": 1.0884255930984903,
-      "grad_norm": 0.46259772777557373,
-      "learning_rate": 7.272876636753916e-06,
-      "loss": 0.3915,
-      "step": 16654
-    },
-    {
-      "epoch": 1.0884909483040324,
-      "grad_norm": 0.468009352684021,
-      "learning_rate": 7.27256560477476e-06,
-      "loss": 0.3754,
-      "step": 16655
-    },
-    {
-      "epoch": 1.0885563035095744,
-      "grad_norm": 0.44383275508880615,
-      "learning_rate": 7.272254561711515e-06,
-      "loss": 0.3247,
-      "step": 16656
-    },
-    {
-      "epoch": 1.0886216587151167,
-      "grad_norm": 0.5151002407073975,
-      "learning_rate": 7.271943507565696e-06,
-      "loss": 0.3486,
-      "step": 16657
-    },
-    {
-      "epoch": 1.0886870139206588,
-      "grad_norm": 0.44885411858558655,
-      "learning_rate": 7.271632442338823e-06,
-      "loss": 0.3312,
-      "step": 16658
-    },
-    {
-      "epoch": 1.0887523691262009,
-      "grad_norm": 0.44856178760528564,
-      "learning_rate": 7.27132136603241e-06,
-      "loss": 0.3213,
-      "step": 16659
-    },
-    {
-      "epoch": 1.088817724331743,
-      "grad_norm": 0.44864511489868164,
-      "learning_rate": 7.2710102786479765e-06,
-      "loss": 0.3487,
-      "step": 16660
-    },
-    {
-      "epoch": 1.0888830795372852,
-      "grad_norm": 0.4368108808994293,
-      "learning_rate": 7.27069918018704e-06,
-      "loss": 0.2972,
-      "step": 16661
-    },
-    {
-      "epoch": 1.0889484347428273,
-      "grad_norm": 0.47293275594711304,
-      "learning_rate": 7.270388070651115e-06,
-      "loss": 0.3712,
-      "step": 16662
-    },
-    {
-      "epoch": 1.0890137899483694,
-      "grad_norm": 0.5326557159423828,
-      "learning_rate": 7.270076950041724e-06,
-      "loss": 0.3363,
-      "step": 16663
-    },
-    {
-      "epoch": 1.0890791451539115,
-      "grad_norm": 0.4424080550670624,
-      "learning_rate": 7.269765818360378e-06,
-      "loss": 0.3214,
-      "step": 16664
-    },
-    {
-      "epoch": 1.0891445003594535,
-      "grad_norm": 0.4424096643924713,
-      "learning_rate": 7.2694546756086e-06,
-      "loss": 0.3188,
-      "step": 16665
-    },
-    {
-      "epoch": 1.0892098555649958,
-      "grad_norm": 0.4487593173980713,
-      "learning_rate": 7.269143521787904e-06,
-      "loss": 0.3561,
-      "step": 16666
-    },
-    {
-      "epoch": 1.089275210770538,
-      "grad_norm": 0.4540695250034332,
-      "learning_rate": 7.2688323568998105e-06,
-      "loss": 0.3746,
-      "step": 16667
-    },
-    {
-      "epoch": 1.08934056597608,
-      "grad_norm": 0.4934118986129761,
-      "learning_rate": 7.2685211809458335e-06,
-      "loss": 0.3494,
-      "step": 16668
-    },
-    {
-      "epoch": 1.089405921181622,
-      "grad_norm": 0.4706835150718689,
-      "learning_rate": 7.2682099939274954e-06,
-      "loss": 0.3261,
-      "step": 16669
-    },
-    {
-      "epoch": 1.0894712763871643,
-      "grad_norm": 0.4687422215938568,
-      "learning_rate": 7.267898795846309e-06,
-      "loss": 0.3665,
-      "step": 16670
-    },
-    {
-      "epoch": 1.0895366315927064,
-      "grad_norm": 0.5007064342498779,
-      "learning_rate": 7.267587586703795e-06,
-      "loss": 0.4104,
-      "step": 16671
-    },
-    {
-      "epoch": 1.0896019867982485,
-      "grad_norm": 0.4735453128814697,
-      "learning_rate": 7.267276366501472e-06,
-      "loss": 0.3431,
-      "step": 16672
-    },
-    {
-      "epoch": 1.0896673420037906,
-      "grad_norm": 0.43810102343559265,
-      "learning_rate": 7.266965135240856e-06,
-      "loss": 0.3421,
-      "step": 16673
-    },
-    {
-      "epoch": 1.0897326972093326,
-      "grad_norm": 0.4781905710697174,
-      "learning_rate": 7.266653892923465e-06,
-      "loss": 0.3772,
-      "step": 16674
-    },
-    {
-      "epoch": 1.089798052414875,
-      "grad_norm": 0.4426269829273224,
-      "learning_rate": 7.2663426395508175e-06,
-      "loss": 0.3132,
-      "step": 16675
-    },
-    {
-      "epoch": 1.089863407620417,
-      "grad_norm": 0.44136500358581543,
-      "learning_rate": 7.266031375124433e-06,
-      "loss": 0.2931,
-      "step": 16676
-    },
-    {
-      "epoch": 1.089928762825959,
-      "grad_norm": 0.40366947650909424,
-      "learning_rate": 7.265720099645828e-06,
-      "loss": 0.2761,
-      "step": 16677
-    },
-    {
-      "epoch": 1.0899941180315011,
-      "grad_norm": 0.4639582633972168,
-      "learning_rate": 7.265408813116521e-06,
-      "loss": 0.3549,
-      "step": 16678
-    },
-    {
-      "epoch": 1.0900594732370434,
-      "grad_norm": 0.494266539812088,
-      "learning_rate": 7.26509751553803e-06,
-      "loss": 0.3884,
-      "step": 16679
-    },
-    {
-      "epoch": 1.0901248284425855,
-      "grad_norm": 0.4520329236984253,
-      "learning_rate": 7.264786206911875e-06,
-      "loss": 0.3515,
-      "step": 16680
-    },
-    {
-      "epoch": 1.0901901836481276,
-      "grad_norm": 0.44954192638397217,
-      "learning_rate": 7.264474887239572e-06,
-      "loss": 0.3645,
-      "step": 16681
-    },
-    {
-      "epoch": 1.0902555388536697,
-      "grad_norm": 0.431308388710022,
-      "learning_rate": 7.264163556522641e-06,
-      "loss": 0.3215,
-      "step": 16682
-    },
-    {
-      "epoch": 1.0903208940592117,
-      "grad_norm": 0.46347272396087646,
-      "learning_rate": 7.263852214762599e-06,
-      "loss": 0.3478,
-      "step": 16683
-    },
-    {
-      "epoch": 1.090386249264754,
-      "grad_norm": 0.4492810070514679,
-      "learning_rate": 7.263540861960966e-06,
-      "loss": 0.3225,
-      "step": 16684
-    },
-    {
-      "epoch": 1.090451604470296,
-      "grad_norm": 0.4365510940551758,
-      "learning_rate": 7.263229498119261e-06,
-      "loss": 0.2907,
-      "step": 16685
-    },
-    {
-      "epoch": 1.0905169596758382,
-      "grad_norm": 0.43292418122291565,
-      "learning_rate": 7.262918123239e-06,
-      "loss": 0.3079,
-      "step": 16686
-    },
-    {
-      "epoch": 1.0905823148813802,
-      "grad_norm": 0.4367837607860565,
-      "learning_rate": 7.2626067373217056e-06,
-      "loss": 0.3292,
-      "step": 16687
-    },
-    {
-      "epoch": 1.0906476700869225,
-      "grad_norm": 0.46882861852645874,
-      "learning_rate": 7.2622953403688925e-06,
-      "loss": 0.3513,
-      "step": 16688
-    },
-    {
-      "epoch": 1.0907130252924646,
-      "grad_norm": 0.4271145761013031,
-      "learning_rate": 7.261983932382083e-06,
-      "loss": 0.3155,
-      "step": 16689
-    },
-    {
-      "epoch": 1.0907783804980067,
-      "grad_norm": 0.4409774839878082,
-      "learning_rate": 7.261672513362792e-06,
-      "loss": 0.3328,
-      "step": 16690
-    },
-    {
-      "epoch": 1.0908437357035488,
-      "grad_norm": 0.44551536440849304,
-      "learning_rate": 7.261361083312542e-06,
-      "loss": 0.3674,
-      "step": 16691
-    },
-    {
-      "epoch": 1.0909090909090908,
-      "grad_norm": 0.4474148750305176,
-      "learning_rate": 7.261049642232851e-06,
-      "loss": 0.3619,
-      "step": 16692
-    },
-    {
-      "epoch": 1.0909744461146331,
-      "grad_norm": 0.4353700578212738,
-      "learning_rate": 7.260738190125236e-06,
-      "loss": 0.3308,
-      "step": 16693
-    },
-    {
-      "epoch": 1.0910398013201752,
-      "grad_norm": 0.4502025842666626,
-      "learning_rate": 7.260426726991219e-06,
-      "loss": 0.3334,
-      "step": 16694
-    },
-    {
-      "epoch": 1.0911051565257173,
-      "grad_norm": 0.4544939398765564,
-      "learning_rate": 7.260115252832316e-06,
-      "loss": 0.3559,
-      "step": 16695
-    },
-    {
-      "epoch": 1.0911705117312593,
-      "grad_norm": 0.4724854528903961,
-      "learning_rate": 7.25980376765005e-06,
-      "loss": 0.3604,
-      "step": 16696
-    },
-    {
-      "epoch": 1.0912358669368014,
-      "grad_norm": 0.492141991853714,
-      "learning_rate": 7.259492271445937e-06,
-      "loss": 0.4084,
-      "step": 16697
-    },
-    {
-      "epoch": 1.0913012221423437,
-      "grad_norm": 0.4570273756980896,
-      "learning_rate": 7.259180764221498e-06,
-      "loss": 0.3641,
-      "step": 16698
-    },
-    {
-      "epoch": 1.0913665773478858,
-      "grad_norm": 0.45201194286346436,
-      "learning_rate": 7.258869245978252e-06,
-      "loss": 0.3351,
-      "step": 16699
-    },
-    {
-      "epoch": 1.0914319325534279,
-      "grad_norm": 0.47722914814949036,
-      "learning_rate": 7.258557716717717e-06,
-      "loss": 0.3501,
-      "step": 16700
-    },
-    {
-      "epoch": 1.09149728775897,
-      "grad_norm": 0.43264567852020264,
-      "learning_rate": 7.258246176441414e-06,
-      "loss": 0.3307,
-      "step": 16701
-    },
-    {
-      "epoch": 1.0915626429645122,
-      "grad_norm": 0.4860178530216217,
-      "learning_rate": 7.257934625150862e-06,
-      "loss": 0.3704,
-      "step": 16702
-    },
-    {
-      "epoch": 1.0916279981700543,
-      "grad_norm": 0.4728662669658661,
-      "learning_rate": 7.257623062847582e-06,
-      "loss": 0.3698,
-      "step": 16703
-    },
-    {
-      "epoch": 1.0916933533755964,
-      "grad_norm": 0.47944486141204834,
-      "learning_rate": 7.25731148953309e-06,
-      "loss": 0.3921,
-      "step": 16704
-    },
-    {
-      "epoch": 1.0917587085811384,
-      "grad_norm": 0.4611533582210541,
-      "learning_rate": 7.25699990520891e-06,
-      "loss": 0.3652,
-      "step": 16705
-    },
-    {
-      "epoch": 1.0918240637866805,
-      "grad_norm": 0.4495225250720978,
-      "learning_rate": 7.256688309876558e-06,
-      "loss": 0.3677,
-      "step": 16706
-    },
-    {
-      "epoch": 1.0918894189922228,
-      "grad_norm": 0.40167364478111267,
-      "learning_rate": 7.256376703537556e-06,
-      "loss": 0.3059,
-      "step": 16707
-    },
-    {
-      "epoch": 1.0919547741977649,
-      "grad_norm": 0.445865273475647,
-      "learning_rate": 7.256065086193424e-06,
-      "loss": 0.3662,
-      "step": 16708
-    },
-    {
-      "epoch": 1.092020129403307,
-      "grad_norm": 0.4442897140979767,
-      "learning_rate": 7.25575345784568e-06,
-      "loss": 0.3276,
-      "step": 16709
-    },
-    {
-      "epoch": 1.092085484608849,
-      "grad_norm": 0.4639493525028229,
-      "learning_rate": 7.255441818495845e-06,
-      "loss": 0.3479,
-      "step": 16710
-    },
-    {
-      "epoch": 1.0921508398143913,
-      "grad_norm": 0.42112016677856445,
-      "learning_rate": 7.255130168145439e-06,
-      "loss": 0.3392,
-      "step": 16711
-    },
-    {
-      "epoch": 1.0922161950199334,
-      "grad_norm": 0.4336766302585602,
-      "learning_rate": 7.254818506795982e-06,
-      "loss": 0.3293,
-      "step": 16712
-    },
-    {
-      "epoch": 1.0922815502254755,
-      "grad_norm": 0.4338686764240265,
-      "learning_rate": 7.254506834448993e-06,
-      "loss": 0.3514,
-      "step": 16713
-    },
-    {
-      "epoch": 1.0923469054310175,
-      "grad_norm": 0.4458778500556946,
-      "learning_rate": 7.254195151105994e-06,
-      "loss": 0.3216,
-      "step": 16714
-    },
-    {
-      "epoch": 1.0924122606365596,
-      "grad_norm": 0.47227126359939575,
-      "learning_rate": 7.253883456768503e-06,
-      "loss": 0.3934,
-      "step": 16715
-    },
-    {
-      "epoch": 1.092477615842102,
-      "grad_norm": 0.4433448612689972,
-      "learning_rate": 7.253571751438045e-06,
-      "loss": 0.3563,
-      "step": 16716
-    },
-    {
-      "epoch": 1.092542971047644,
-      "grad_norm": 0.46793797612190247,
-      "learning_rate": 7.253260035116133e-06,
-      "loss": 0.3743,
-      "step": 16717
-    },
-    {
-      "epoch": 1.092608326253186,
-      "grad_norm": 0.4590999186038971,
-      "learning_rate": 7.252948307804293e-06,
-      "loss": 0.3688,
-      "step": 16718
-    },
-    {
-      "epoch": 1.0926736814587281,
-      "grad_norm": 0.4602227807044983,
-      "learning_rate": 7.252636569504044e-06,
-      "loss": 0.3868,
-      "step": 16719
-    },
-    {
-      "epoch": 1.0927390366642702,
-      "grad_norm": 0.4311557710170746,
-      "learning_rate": 7.252324820216905e-06,
-      "loss": 0.3202,
-      "step": 16720
-    },
-    {
-      "epoch": 1.0928043918698125,
-      "grad_norm": 0.41199976205825806,
-      "learning_rate": 7.252013059944398e-06,
-      "loss": 0.3134,
-      "step": 16721
-    },
-    {
-      "epoch": 1.0928697470753546,
-      "grad_norm": 0.4609593451023102,
-      "learning_rate": 7.251701288688042e-06,
-      "loss": 0.3304,
-      "step": 16722
-    },
-    {
-      "epoch": 1.0929351022808966,
-      "grad_norm": 0.4601035714149475,
-      "learning_rate": 7.251389506449361e-06,
-      "loss": 0.3679,
-      "step": 16723
-    },
-    {
-      "epoch": 1.0930004574864387,
-      "grad_norm": 0.445535808801651,
-      "learning_rate": 7.251077713229873e-06,
-      "loss": 0.3218,
-      "step": 16724
-    },
-    {
-      "epoch": 1.093065812691981,
-      "grad_norm": 0.4361458122730255,
-      "learning_rate": 7.250765909031098e-06,
-      "loss": 0.3401,
-      "step": 16725
-    },
-    {
-      "epoch": 1.093131167897523,
-      "grad_norm": 0.4398879110813141,
-      "learning_rate": 7.2504540938545585e-06,
-      "loss": 0.3459,
-      "step": 16726
-    },
-    {
-      "epoch": 1.0931965231030651,
-      "grad_norm": 0.4309949576854706,
-      "learning_rate": 7.250142267701774e-06,
-      "loss": 0.3259,
-      "step": 16727
-    },
-    {
-      "epoch": 1.0932618783086072,
-      "grad_norm": 0.43104955554008484,
-      "learning_rate": 7.249830430574267e-06,
-      "loss": 0.3568,
-      "step": 16728
-    },
-    {
-      "epoch": 1.0933272335141493,
-      "grad_norm": 0.45479997992515564,
-      "learning_rate": 7.249518582473558e-06,
-      "loss": 0.3697,
-      "step": 16729
-    },
-    {
-      "epoch": 1.0933925887196916,
-      "grad_norm": 0.4632102847099304,
-      "learning_rate": 7.249206723401167e-06,
-      "loss": 0.3612,
-      "step": 16730
-    },
-    {
-      "epoch": 1.0934579439252337,
-      "grad_norm": 0.429729700088501,
-      "learning_rate": 7.248894853358616e-06,
-      "loss": 0.3179,
-      "step": 16731
-    },
-    {
-      "epoch": 1.0935232991307757,
-      "grad_norm": 0.46751073002815247,
-      "learning_rate": 7.248582972347426e-06,
-      "loss": 0.3663,
-      "step": 16732
-    },
-    {
-      "epoch": 1.0935886543363178,
-      "grad_norm": 0.48656025528907776,
-      "learning_rate": 7.248271080369116e-06,
-      "loss": 0.3637,
-      "step": 16733
-    },
-    {
-      "epoch": 1.09365400954186,
-      "grad_norm": 0.4686853289604187,
-      "learning_rate": 7.24795917742521e-06,
-      "loss": 0.3496,
-      "step": 16734
-    },
-    {
-      "epoch": 1.0937193647474022,
-      "grad_norm": 0.4280194044113159,
-      "learning_rate": 7.247647263517228e-06,
-      "loss": 0.3324,
-      "step": 16735
-    },
-    {
-      "epoch": 1.0937847199529442,
-      "grad_norm": 0.4650658369064331,
-      "learning_rate": 7.247335338646693e-06,
-      "loss": 0.375,
-      "step": 16736
-    },
-    {
-      "epoch": 1.0938500751584863,
-      "grad_norm": 0.45358169078826904,
-      "learning_rate": 7.247023402815125e-06,
-      "loss": 0.3457,
-      "step": 16737
-    },
-    {
-      "epoch": 1.0939154303640284,
-      "grad_norm": 0.4725961685180664,
-      "learning_rate": 7.2467114560240435e-06,
-      "loss": 0.3802,
-      "step": 16738
-    },
-    {
-      "epoch": 1.0939807855695707,
-      "grad_norm": 0.45289045572280884,
-      "learning_rate": 7.246399498274974e-06,
-      "loss": 0.3774,
-      "step": 16739
-    },
-    {
-      "epoch": 1.0940461407751128,
-      "grad_norm": 0.5127090811729431,
-      "learning_rate": 7.246087529569435e-06,
-      "loss": 0.4047,
-      "step": 16740
-    },
-    {
-      "epoch": 1.0941114959806548,
-      "grad_norm": 0.43285638093948364,
-      "learning_rate": 7.245775549908948e-06,
-      "loss": 0.3237,
-      "step": 16741
-    },
-    {
-      "epoch": 1.094176851186197,
-      "grad_norm": 0.43867310881614685,
-      "learning_rate": 7.245463559295036e-06,
-      "loss": 0.2945,
-      "step": 16742
-    },
-    {
-      "epoch": 1.0942422063917392,
-      "grad_norm": 0.44885164499282837,
-      "learning_rate": 7.245151557729221e-06,
-      "loss": 0.3465,
-      "step": 16743
-    },
-    {
-      "epoch": 1.0943075615972813,
-      "grad_norm": 0.4602097272872925,
-      "learning_rate": 7.244839545213024e-06,
-      "loss": 0.3657,
-      "step": 16744
-    },
-    {
-      "epoch": 1.0943729168028233,
-      "grad_norm": 0.48046061396598816,
-      "learning_rate": 7.244527521747966e-06,
-      "loss": 0.3974,
-      "step": 16745
-    },
-    {
-      "epoch": 1.0944382720083654,
-      "grad_norm": 0.3908010423183441,
-      "learning_rate": 7.2442154873355716e-06,
-      "loss": 0.2774,
-      "step": 16746
-    },
-    {
-      "epoch": 1.0945036272139075,
-      "grad_norm": 0.4774419069290161,
-      "learning_rate": 7.2439034419773594e-06,
-      "loss": 0.347,
-      "step": 16747
-    },
-    {
-      "epoch": 1.0945689824194498,
-      "grad_norm": 0.46570348739624023,
-      "learning_rate": 7.243591385674853e-06,
-      "loss": 0.3568,
-      "step": 16748
-    },
-    {
-      "epoch": 1.0946343376249918,
-      "grad_norm": 0.4292398989200592,
-      "learning_rate": 7.243279318429574e-06,
-      "loss": 0.31,
-      "step": 16749
-    },
-    {
-      "epoch": 1.094699692830534,
-      "grad_norm": 0.43820425868034363,
-      "learning_rate": 7.2429672402430465e-06,
-      "loss": 0.3023,
-      "step": 16750
-    },
-    {
-      "epoch": 1.094765048036076,
-      "grad_norm": 0.44285324215888977,
-      "learning_rate": 7.242655151116788e-06,
-      "loss": 0.3122,
-      "step": 16751
-    },
-    {
-      "epoch": 1.0948304032416183,
-      "grad_norm": 0.46688154339790344,
-      "learning_rate": 7.242343051052325e-06,
-      "loss": 0.3507,
-      "step": 16752
-    },
-    {
-      "epoch": 1.0948957584471604,
-      "grad_norm": 0.4655936360359192,
-      "learning_rate": 7.242030940051177e-06,
-      "loss": 0.3698,
-      "step": 16753
-    },
-    {
-      "epoch": 1.0949611136527024,
-      "grad_norm": 0.4363328218460083,
-      "learning_rate": 7.241718818114868e-06,
-      "loss": 0.3251,
-      "step": 16754
-    },
-    {
-      "epoch": 1.0950264688582445,
-      "grad_norm": 0.4805351197719574,
-      "learning_rate": 7.241406685244918e-06,
-      "loss": 0.4087,
-      "step": 16755
-    },
-    {
-      "epoch": 1.0950918240637866,
-      "grad_norm": 0.44228002429008484,
-      "learning_rate": 7.241094541442854e-06,
-      "loss": 0.3326,
-      "step": 16756
-    },
-    {
-      "epoch": 1.0951571792693289,
-      "grad_norm": 0.44409671425819397,
-      "learning_rate": 7.240782386710194e-06,
-      "loss": 0.3334,
-      "step": 16757
-    },
-    {
-      "epoch": 1.095222534474871,
-      "grad_norm": 0.4377843141555786,
-      "learning_rate": 7.240470221048462e-06,
-      "loss": 0.335,
-      "step": 16758
-    },
-    {
-      "epoch": 1.095287889680413,
-      "grad_norm": 0.46416565775871277,
-      "learning_rate": 7.240158044459181e-06,
-      "loss": 0.3792,
-      "step": 16759
-    },
-    {
-      "epoch": 1.095353244885955,
-      "grad_norm": 0.422510027885437,
-      "learning_rate": 7.2398458569438726e-06,
-      "loss": 0.3248,
-      "step": 16760
-    },
-    {
-      "epoch": 1.0954186000914974,
-      "grad_norm": 0.440060019493103,
-      "learning_rate": 7.239533658504061e-06,
-      "loss": 0.3467,
-      "step": 16761
-    },
-    {
-      "epoch": 1.0954839552970395,
-      "grad_norm": 0.4493173658847809,
-      "learning_rate": 7.239221449141267e-06,
-      "loss": 0.3306,
-      "step": 16762
-    },
-    {
-      "epoch": 1.0955493105025815,
-      "grad_norm": 0.4283548891544342,
-      "learning_rate": 7.238909228857015e-06,
-      "loss": 0.3026,
-      "step": 16763
-    },
-    {
-      "epoch": 1.0956146657081236,
-      "grad_norm": 0.4168672263622284,
-      "learning_rate": 7.238596997652827e-06,
-      "loss": 0.2905,
-      "step": 16764
-    },
-    {
-      "epoch": 1.0956800209136657,
-      "grad_norm": 0.4448527693748474,
-      "learning_rate": 7.2382847555302245e-06,
-      "loss": 0.3385,
-      "step": 16765
-    },
-    {
-      "epoch": 1.095745376119208,
-      "grad_norm": 0.4585151672363281,
-      "learning_rate": 7.237972502490733e-06,
-      "loss": 0.3513,
-      "step": 16766
-    },
-    {
-      "epoch": 1.09581073132475,
-      "grad_norm": 0.4385329782962799,
-      "learning_rate": 7.237660238535874e-06,
-      "loss": 0.3327,
-      "step": 16767
-    },
-    {
-      "epoch": 1.0958760865302921,
-      "grad_norm": 0.43763411045074463,
-      "learning_rate": 7.237347963667172e-06,
-      "loss": 0.3275,
-      "step": 16768
-    },
-    {
-      "epoch": 1.0959414417358342,
-      "grad_norm": 0.43371498584747314,
-      "learning_rate": 7.2370356778861464e-06,
-      "loss": 0.3247,
-      "step": 16769
-    },
-    {
-      "epoch": 1.0960067969413765,
-      "grad_norm": 0.4241376519203186,
-      "learning_rate": 7.236723381194325e-06,
-      "loss": 0.3314,
-      "step": 16770
-    },
-    {
-      "epoch": 1.0960721521469186,
-      "grad_norm": 0.4330827295780182,
-      "learning_rate": 7.236411073593228e-06,
-      "loss": 0.3488,
-      "step": 16771
-    },
-    {
-      "epoch": 1.0961375073524606,
-      "grad_norm": 0.4511134922504425,
-      "learning_rate": 7.23609875508438e-06,
-      "loss": 0.3495,
-      "step": 16772
-    },
-    {
-      "epoch": 1.0962028625580027,
-      "grad_norm": 0.4721181392669678,
-      "learning_rate": 7.235786425669302e-06,
-      "loss": 0.3905,
-      "step": 16773
-    },
-    {
-      "epoch": 1.0962682177635448,
-      "grad_norm": 0.4329351484775543,
-      "learning_rate": 7.235474085349521e-06,
-      "loss": 0.3027,
-      "step": 16774
-    },
-    {
-      "epoch": 1.096333572969087,
-      "grad_norm": 0.46141180396080017,
-      "learning_rate": 7.235161734126558e-06,
-      "loss": 0.3889,
-      "step": 16775
-    },
-    {
-      "epoch": 1.0963989281746291,
-      "grad_norm": 0.44180065393447876,
-      "learning_rate": 7.234849372001936e-06,
-      "loss": 0.3602,
-      "step": 16776
-    },
-    {
-      "epoch": 1.0964642833801712,
-      "grad_norm": 0.480880469083786,
-      "learning_rate": 7.234536998977181e-06,
-      "loss": 0.3797,
-      "step": 16777
-    },
-    {
-      "epoch": 1.0965296385857133,
-      "grad_norm": 0.43409091234207153,
-      "learning_rate": 7.234224615053813e-06,
-      "loss": 0.3144,
-      "step": 16778
-    },
-    {
-      "epoch": 1.0965949937912556,
-      "grad_norm": 0.4776366651058197,
-      "learning_rate": 7.233912220233359e-06,
-      "loss": 0.3799,
-      "step": 16779
-    },
-    {
-      "epoch": 1.0966603489967977,
-      "grad_norm": 0.4754861891269684,
-      "learning_rate": 7.233599814517341e-06,
-      "loss": 0.3543,
-      "step": 16780
-    },
-    {
-      "epoch": 1.0967257042023397,
-      "grad_norm": 0.44954970479011536,
-      "learning_rate": 7.233287397907283e-06,
-      "loss": 0.3416,
-      "step": 16781
-    },
-    {
-      "epoch": 1.0967910594078818,
-      "grad_norm": 0.4645956754684448,
-      "learning_rate": 7.232974970404707e-06,
-      "loss": 0.3773,
-      "step": 16782
-    },
-    {
-      "epoch": 1.0968564146134239,
-      "grad_norm": 0.4409351050853729,
-      "learning_rate": 7.23266253201114e-06,
-      "loss": 0.3812,
-      "step": 16783
-    },
-    {
-      "epoch": 1.0969217698189662,
-      "grad_norm": 0.5002554059028625,
-      "learning_rate": 7.232350082728105e-06,
-      "loss": 0.3846,
-      "step": 16784
-    },
-    {
-      "epoch": 1.0969871250245082,
-      "grad_norm": 0.47278743982315063,
-      "learning_rate": 7.232037622557123e-06,
-      "loss": 0.3442,
-      "step": 16785
-    },
-    {
-      "epoch": 1.0970524802300503,
-      "grad_norm": 0.45567455887794495,
-      "learning_rate": 7.231725151499722e-06,
-      "loss": 0.3794,
-      "step": 16786
-    },
-    {
-      "epoch": 1.0971178354355924,
-      "grad_norm": 0.4341067969799042,
-      "learning_rate": 7.231412669557424e-06,
-      "loss": 0.3131,
-      "step": 16787
-    },
-    {
-      "epoch": 1.0971831906411347,
-      "grad_norm": 0.4420716166496277,
-      "learning_rate": 7.231100176731753e-06,
-      "loss": 0.3193,
-      "step": 16788
-    },
-    {
-      "epoch": 1.0972485458466767,
-      "grad_norm": 0.4943370223045349,
-      "learning_rate": 7.2307876730242336e-06,
-      "loss": 0.3535,
-      "step": 16789
-    },
-    {
-      "epoch": 1.0973139010522188,
-      "grad_norm": 0.4515933692455292,
-      "learning_rate": 7.23047515843639e-06,
-      "loss": 0.3519,
-      "step": 16790
-    },
-    {
-      "epoch": 1.097379256257761,
-      "grad_norm": 0.44382259249687195,
-      "learning_rate": 7.230162632969746e-06,
-      "loss": 0.2976,
-      "step": 16791
-    },
-    {
-      "epoch": 1.097444611463303,
-      "grad_norm": 0.4211956262588501,
-      "learning_rate": 7.229850096625828e-06,
-      "loss": 0.256,
-      "step": 16792
-    },
-    {
-      "epoch": 1.0975099666688453,
-      "grad_norm": 0.4627498984336853,
-      "learning_rate": 7.229537549406157e-06,
-      "loss": 0.3423,
-      "step": 16793
-    },
-    {
-      "epoch": 1.0975753218743873,
-      "grad_norm": 0.4721478521823883,
-      "learning_rate": 7.22922499131226e-06,
-      "loss": 0.3423,
-      "step": 16794
-    },
-    {
-      "epoch": 1.0976406770799294,
-      "grad_norm": 0.42021995782852173,
-      "learning_rate": 7.22891242234566e-06,
-      "loss": 0.2962,
-      "step": 16795
-    },
-    {
-      "epoch": 1.0977060322854715,
-      "grad_norm": 0.45924872159957886,
-      "learning_rate": 7.228599842507881e-06,
-      "loss": 0.3606,
-      "step": 16796
-    },
-    {
-      "epoch": 1.0977713874910138,
-      "grad_norm": 0.44723430275917053,
-      "learning_rate": 7.22828725180045e-06,
-      "loss": 0.3605,
-      "step": 16797
-    },
-    {
-      "epoch": 1.0978367426965558,
-      "grad_norm": 0.4550357162952423,
-      "learning_rate": 7.227974650224888e-06,
-      "loss": 0.3062,
-      "step": 16798
-    },
-    {
-      "epoch": 1.097902097902098,
-      "grad_norm": 0.4332388937473297,
-      "learning_rate": 7.227662037782723e-06,
-      "loss": 0.3046,
-      "step": 16799
-    },
-    {
-      "epoch": 1.09796745310764,
-      "grad_norm": 0.4315112233161926,
-      "learning_rate": 7.227349414475479e-06,
-      "loss": 0.3376,
-      "step": 16800
-    },
-    {
-      "epoch": 1.098032808313182,
-      "grad_norm": 0.43725505471229553,
-      "learning_rate": 7.227036780304679e-06,
-      "loss": 0.3364,
-      "step": 16801
-    },
-    {
-      "epoch": 1.0980981635187244,
-      "grad_norm": 0.44807305932044983,
-      "learning_rate": 7.22672413527185e-06,
-      "loss": 0.3466,
-      "step": 16802
-    },
-    {
-      "epoch": 1.0981635187242664,
-      "grad_norm": 0.49027779698371887,
-      "learning_rate": 7.226411479378517e-06,
-      "loss": 0.361,
-      "step": 16803
-    },
-    {
-      "epoch": 1.0982288739298085,
-      "grad_norm": 0.451984167098999,
-      "learning_rate": 7.2260988126262035e-06,
-      "loss": 0.3379,
-      "step": 16804
-    },
-    {
-      "epoch": 1.0982942291353506,
-      "grad_norm": 0.4550917446613312,
-      "learning_rate": 7.225786135016433e-06,
-      "loss": 0.3425,
-      "step": 16805
-    },
-    {
-      "epoch": 1.0983595843408929,
-      "grad_norm": 0.4698694944381714,
-      "learning_rate": 7.225473446550733e-06,
-      "loss": 0.3635,
-      "step": 16806
-    },
-    {
-      "epoch": 1.098424939546435,
-      "grad_norm": 0.461752712726593,
-      "learning_rate": 7.225160747230628e-06,
-      "loss": 0.3439,
-      "step": 16807
-    },
-    {
-      "epoch": 1.098490294751977,
-      "grad_norm": 0.4706531763076782,
-      "learning_rate": 7.224848037057646e-06,
-      "loss": 0.391,
-      "step": 16808
-    },
-    {
-      "epoch": 1.098555649957519,
-      "grad_norm": 0.47599413990974426,
-      "learning_rate": 7.224535316033305e-06,
-      "loss": 0.3778,
-      "step": 16809
-    },
-    {
-      "epoch": 1.0986210051630612,
-      "grad_norm": 0.44619935750961304,
-      "learning_rate": 7.224222584159139e-06,
-      "loss": 0.3472,
-      "step": 16810
-    },
-    {
-      "epoch": 1.0986863603686035,
-      "grad_norm": 0.4722409248352051,
-      "learning_rate": 7.223909841436666e-06,
-      "loss": 0.3464,
-      "step": 16811
-    },
-    {
-      "epoch": 1.0987517155741455,
-      "grad_norm": 0.4703991115093231,
-      "learning_rate": 7.223597087867414e-06,
-      "loss": 0.3813,
-      "step": 16812
-    },
-    {
-      "epoch": 1.0988170707796876,
-      "grad_norm": 0.4867233633995056,
-      "learning_rate": 7.22328432345291e-06,
-      "loss": 0.3903,
-      "step": 16813
-    },
-    {
-      "epoch": 1.0988824259852297,
-      "grad_norm": 0.46626606583595276,
-      "learning_rate": 7.222971548194679e-06,
-      "loss": 0.3176,
-      "step": 16814
-    },
-    {
-      "epoch": 1.0989477811907717,
-      "grad_norm": 0.45740047097206116,
-      "learning_rate": 7.2226587620942456e-06,
-      "loss": 0.3675,
-      "step": 16815
-    },
-    {
-      "epoch": 1.099013136396314,
-      "grad_norm": 0.47037217020988464,
-      "learning_rate": 7.222345965153133e-06,
-      "loss": 0.3611,
-      "step": 16816
-    },
-    {
-      "epoch": 1.0990784916018561,
-      "grad_norm": 0.4419795870780945,
-      "learning_rate": 7.222033157372871e-06,
-      "loss": 0.3487,
-      "step": 16817
-    },
-    {
-      "epoch": 1.0991438468073982,
-      "grad_norm": 0.42824700474739075,
-      "learning_rate": 7.221720338754983e-06,
-      "loss": 0.3176,
-      "step": 16818
-    },
-    {
-      "epoch": 1.0992092020129403,
-      "grad_norm": 0.44660231471061707,
-      "learning_rate": 7.221407509300995e-06,
-      "loss": 0.333,
-      "step": 16819
-    },
-    {
-      "epoch": 1.0992745572184826,
-      "grad_norm": 0.4343295991420746,
-      "learning_rate": 7.221094669012433e-06,
-      "loss": 0.2936,
-      "step": 16820
-    },
-    {
-      "epoch": 1.0993399124240246,
-      "grad_norm": 0.4637698531150818,
-      "learning_rate": 7.220781817890823e-06,
-      "loss": 0.3182,
-      "step": 16821
-    },
-    {
-      "epoch": 1.0994052676295667,
-      "grad_norm": 0.42962801456451416,
-      "learning_rate": 7.220468955937692e-06,
-      "loss": 0.3522,
-      "step": 16822
-    },
-    {
-      "epoch": 1.0994706228351088,
-      "grad_norm": 0.4624047875404358,
-      "learning_rate": 7.220156083154562e-06,
-      "loss": 0.3624,
-      "step": 16823
-    },
-    {
-      "epoch": 1.0995359780406508,
-      "grad_norm": 0.43798285722732544,
-      "learning_rate": 7.219843199542964e-06,
-      "loss": 0.3019,
-      "step": 16824
-    },
-    {
-      "epoch": 1.0996013332461931,
-      "grad_norm": 0.47579702734947205,
-      "learning_rate": 7.219530305104421e-06,
-      "loss": 0.3589,
-      "step": 16825
-    },
-    {
-      "epoch": 1.0996666884517352,
-      "grad_norm": 0.45312392711639404,
-      "learning_rate": 7.2192173998404595e-06,
-      "loss": 0.3251,
-      "step": 16826
-    },
-    {
-      "epoch": 1.0997320436572773,
-      "grad_norm": 0.4656205177307129,
-      "learning_rate": 7.218904483752605e-06,
-      "loss": 0.3657,
-      "step": 16827
-    },
-    {
-      "epoch": 1.0997973988628194,
-      "grad_norm": 0.44865405559539795,
-      "learning_rate": 7.218591556842386e-06,
-      "loss": 0.3208,
-      "step": 16828
-    },
-    {
-      "epoch": 1.0998627540683616,
-      "grad_norm": 0.4409961998462677,
-      "learning_rate": 7.218278619111326e-06,
-      "loss": 0.3416,
-      "step": 16829
-    },
-    {
-      "epoch": 1.0999281092739037,
-      "grad_norm": 0.49070578813552856,
-      "learning_rate": 7.217965670560955e-06,
-      "loss": 0.3824,
-      "step": 16830
-    },
-    {
-      "epoch": 1.0999934644794458,
-      "grad_norm": 0.44723668694496155,
-      "learning_rate": 7.217652711192796e-06,
-      "loss": 0.3456,
-      "step": 16831
-    },
-    {
-      "epoch": 1.1000588196849879,
-      "grad_norm": 0.4212040603160858,
-      "learning_rate": 7.217339741008376e-06,
-      "loss": 0.3011,
-      "step": 16832
-    },
-    {
-      "epoch": 1.10012417489053,
-      "grad_norm": 0.48938217759132385,
-      "learning_rate": 7.217026760009223e-06,
-      "loss": 0.3438,
-      "step": 16833
-    },
-    {
-      "epoch": 1.1001895300960722,
-      "grad_norm": 0.455473393201828,
-      "learning_rate": 7.216713768196861e-06,
-      "loss": 0.3657,
-      "step": 16834
-    },
-    {
-      "epoch": 1.1002548853016143,
-      "grad_norm": 0.4355417490005493,
-      "learning_rate": 7.21640076557282e-06,
-      "loss": 0.3286,
-      "step": 16835
-    },
-    {
-      "epoch": 1.1003202405071564,
-      "grad_norm": 0.4471430480480194,
-      "learning_rate": 7.216087752138622e-06,
-      "loss": 0.3384,
-      "step": 16836
-    },
-    {
-      "epoch": 1.1003855957126984,
-      "grad_norm": 0.496289998292923,
-      "learning_rate": 7.215774727895798e-06,
-      "loss": 0.3855,
-      "step": 16837
-    },
-    {
-      "epoch": 1.1004509509182405,
-      "grad_norm": 0.41548284888267517,
-      "learning_rate": 7.215461692845872e-06,
-      "loss": 0.3097,
-      "step": 16838
-    },
-    {
-      "epoch": 1.1005163061237828,
-      "grad_norm": 0.42794421315193176,
-      "learning_rate": 7.215148646990373e-06,
-      "loss": 0.2958,
-      "step": 16839
-    },
-    {
-      "epoch": 1.100581661329325,
-      "grad_norm": 0.4622229337692261,
-      "learning_rate": 7.214835590330825e-06,
-      "loss": 0.388,
-      "step": 16840
-    },
-    {
-      "epoch": 1.100647016534867,
-      "grad_norm": 0.4127064347267151,
-      "learning_rate": 7.214522522868758e-06,
-      "loss": 0.2861,
-      "step": 16841
-    },
-    {
-      "epoch": 1.100712371740409,
-      "grad_norm": 0.46275952458381653,
-      "learning_rate": 7.2142094446056974e-06,
-      "loss": 0.3678,
-      "step": 16842
-    },
-    {
-      "epoch": 1.1007777269459513,
-      "grad_norm": 0.4520907998085022,
-      "learning_rate": 7.213896355543169e-06,
-      "loss": 0.3444,
-      "step": 16843
-    },
-    {
-      "epoch": 1.1008430821514934,
-      "grad_norm": 0.43794530630111694,
-      "learning_rate": 7.213583255682702e-06,
-      "loss": 0.3608,
-      "step": 16844
-    },
-    {
-      "epoch": 1.1009084373570355,
-      "grad_norm": 0.4261733591556549,
-      "learning_rate": 7.213270145025822e-06,
-      "loss": 0.327,
-      "step": 16845
-    },
-    {
-      "epoch": 1.1009737925625775,
-      "grad_norm": 0.44544142484664917,
-      "learning_rate": 7.212957023574059e-06,
-      "loss": 0.324,
-      "step": 16846
-    },
-    {
-      "epoch": 1.1010391477681196,
-      "grad_norm": 0.44111916422843933,
-      "learning_rate": 7.212643891328935e-06,
-      "loss": 0.3333,
-      "step": 16847
-    },
-    {
-      "epoch": 1.101104502973662,
-      "grad_norm": 0.4675199091434479,
-      "learning_rate": 7.212330748291982e-06,
-      "loss": 0.3668,
-      "step": 16848
-    },
-    {
-      "epoch": 1.101169858179204,
-      "grad_norm": 0.4404946565628052,
-      "learning_rate": 7.212017594464725e-06,
-      "loss": 0.3765,
-      "step": 16849
-    },
-    {
-      "epoch": 1.101235213384746,
-      "grad_norm": 0.4739381670951843,
-      "learning_rate": 7.211704429848691e-06,
-      "loss": 0.3786,
-      "step": 16850
-    },
-    {
-      "epoch": 1.1013005685902881,
-      "grad_norm": 0.4321777820587158,
-      "learning_rate": 7.2113912544454105e-06,
-      "loss": 0.343,
-      "step": 16851
-    },
-    {
-      "epoch": 1.1013659237958304,
-      "grad_norm": 0.4226875305175781,
-      "learning_rate": 7.211078068256408e-06,
-      "loss": 0.3301,
-      "step": 16852
-    },
-    {
-      "epoch": 1.1014312790013725,
-      "grad_norm": 0.4376566410064697,
-      "learning_rate": 7.210764871283211e-06,
-      "loss": 0.3529,
-      "step": 16853
-    },
-    {
-      "epoch": 1.1014966342069146,
-      "grad_norm": 0.45876818895339966,
-      "learning_rate": 7.210451663527347e-06,
-      "loss": 0.3348,
-      "step": 16854
-    },
-    {
-      "epoch": 1.1015619894124566,
-      "grad_norm": 0.4542117118835449,
-      "learning_rate": 7.210138444990347e-06,
-      "loss": 0.3774,
-      "step": 16855
-    },
-    {
-      "epoch": 1.1016273446179987,
-      "grad_norm": 0.45462122559547424,
-      "learning_rate": 7.209825215673734e-06,
-      "loss": 0.3707,
-      "step": 16856
-    },
-    {
-      "epoch": 1.101692699823541,
-      "grad_norm": 0.4399607181549072,
-      "learning_rate": 7.209511975579039e-06,
-      "loss": 0.3169,
-      "step": 16857
-    },
-    {
-      "epoch": 1.101758055029083,
-      "grad_norm": 0.49060624837875366,
-      "learning_rate": 7.209198724707788e-06,
-      "loss": 0.4028,
-      "step": 16858
-    },
-    {
-      "epoch": 1.1018234102346252,
-      "grad_norm": 0.4618798792362213,
-      "learning_rate": 7.20888546306151e-06,
-      "loss": 0.3329,
-      "step": 16859
-    },
-    {
-      "epoch": 1.1018887654401672,
-      "grad_norm": 0.46027952432632446,
-      "learning_rate": 7.208572190641732e-06,
-      "loss": 0.34,
-      "step": 16860
-    },
-    {
-      "epoch": 1.1019541206457095,
-      "grad_norm": 0.45319730043411255,
-      "learning_rate": 7.208258907449982e-06,
-      "loss": 0.356,
-      "step": 16861
-    },
-    {
-      "epoch": 1.1020194758512516,
-      "grad_norm": 0.4011819362640381,
-      "learning_rate": 7.207945613487789e-06,
-      "loss": 0.2816,
-      "step": 16862
-    },
-    {
-      "epoch": 1.1020848310567937,
-      "grad_norm": 0.4350883662700653,
-      "learning_rate": 7.207632308756679e-06,
-      "loss": 0.3276,
-      "step": 16863
-    },
-    {
-      "epoch": 1.1021501862623357,
-      "grad_norm": 0.4412400424480438,
-      "learning_rate": 7.2073189932581835e-06,
-      "loss": 0.3586,
-      "step": 16864
-    },
-    {
-      "epoch": 1.1022155414678778,
-      "grad_norm": 0.43144404888153076,
-      "learning_rate": 7.207005666993827e-06,
-      "loss": 0.3026,
-      "step": 16865
-    },
-    {
-      "epoch": 1.10228089667342,
-      "grad_norm": 0.4390637278556824,
-      "learning_rate": 7.206692329965139e-06,
-      "loss": 0.3139,
-      "step": 16866
-    },
-    {
-      "epoch": 1.1023462518789622,
-      "grad_norm": 0.4494231343269348,
-      "learning_rate": 7.20637898217365e-06,
-      "loss": 0.3573,
-      "step": 16867
-    },
-    {
-      "epoch": 1.1024116070845043,
-      "grad_norm": 0.46788421273231506,
-      "learning_rate": 7.206065623620885e-06,
-      "loss": 0.4113,
-      "step": 16868
-    },
-    {
-      "epoch": 1.1024769622900463,
-      "grad_norm": 0.43642088770866394,
-      "learning_rate": 7.205752254308374e-06,
-      "loss": 0.3424,
-      "step": 16869
-    },
-    {
-      "epoch": 1.1025423174955886,
-      "grad_norm": 0.4403359293937683,
-      "learning_rate": 7.205438874237644e-06,
-      "loss": 0.341,
-      "step": 16870
-    },
-    {
-      "epoch": 1.1026076727011307,
-      "grad_norm": 0.43854424357414246,
-      "learning_rate": 7.205125483410226e-06,
-      "loss": 0.3129,
-      "step": 16871
-    },
-    {
-      "epoch": 1.1026730279066728,
-      "grad_norm": 0.4508354663848877,
-      "learning_rate": 7.204812081827645e-06,
-      "loss": 0.3557,
-      "step": 16872
-    },
-    {
-      "epoch": 1.1027383831122148,
-      "grad_norm": 0.47603222727775574,
-      "learning_rate": 7.204498669491435e-06,
-      "loss": 0.3784,
-      "step": 16873
-    },
-    {
-      "epoch": 1.102803738317757,
-      "grad_norm": 0.4385119676589966,
-      "learning_rate": 7.2041852464031195e-06,
-      "loss": 0.3386,
-      "step": 16874
-    },
-    {
-      "epoch": 1.1028690935232992,
-      "grad_norm": 0.4601020812988281,
-      "learning_rate": 7.203871812564229e-06,
-      "loss": 0.3329,
-      "step": 16875
-    },
-    {
-      "epoch": 1.1029344487288413,
-      "grad_norm": 0.47119858860969543,
-      "learning_rate": 7.203558367976292e-06,
-      "loss": 0.3425,
-      "step": 16876
-    },
-    {
-      "epoch": 1.1029998039343833,
-      "grad_norm": 0.5225222706794739,
-      "learning_rate": 7.203244912640839e-06,
-      "loss": 0.4024,
-      "step": 16877
-    },
-    {
-      "epoch": 1.1030651591399254,
-      "grad_norm": 0.4457537531852722,
-      "learning_rate": 7.202931446559395e-06,
-      "loss": 0.3248,
-      "step": 16878
-    },
-    {
-      "epoch": 1.1031305143454677,
-      "grad_norm": 0.4651030898094177,
-      "learning_rate": 7.202617969733492e-06,
-      "loss": 0.3476,
-      "step": 16879
-    },
-    {
-      "epoch": 1.1031958695510098,
-      "grad_norm": 0.5119554996490479,
-      "learning_rate": 7.202304482164659e-06,
-      "loss": 0.4074,
-      "step": 16880
-    },
-    {
-      "epoch": 1.1032612247565519,
-      "grad_norm": 0.43499282002449036,
-      "learning_rate": 7.201990983854422e-06,
-      "loss": 0.3081,
-      "step": 16881
-    },
-    {
-      "epoch": 1.103326579962094,
-      "grad_norm": 0.4564814269542694,
-      "learning_rate": 7.201677474804314e-06,
-      "loss": 0.3473,
-      "step": 16882
-    },
-    {
-      "epoch": 1.103391935167636,
-      "grad_norm": 0.5000506043434143,
-      "learning_rate": 7.201363955015861e-06,
-      "loss": 0.3481,
-      "step": 16883
-    },
-    {
-      "epoch": 1.1034572903731783,
-      "grad_norm": 0.44875526428222656,
-      "learning_rate": 7.201050424490594e-06,
-      "loss": 0.3359,
-      "step": 16884
-    },
-    {
-      "epoch": 1.1035226455787204,
-      "grad_norm": 0.47175514698028564,
-      "learning_rate": 7.200736883230042e-06,
-      "loss": 0.3684,
-      "step": 16885
-    },
-    {
-      "epoch": 1.1035880007842624,
-      "grad_norm": 0.46201175451278687,
-      "learning_rate": 7.200423331235733e-06,
-      "loss": 0.3293,
-      "step": 16886
-    },
-    {
-      "epoch": 1.1036533559898045,
-      "grad_norm": 0.4432675540447235,
-      "learning_rate": 7.200109768509198e-06,
-      "loss": 0.3247,
-      "step": 16887
-    },
-    {
-      "epoch": 1.1037187111953468,
-      "grad_norm": 0.43380874395370483,
-      "learning_rate": 7.1997961950519646e-06,
-      "loss": 0.3128,
-      "step": 16888
-    },
-    {
-      "epoch": 1.1037840664008889,
-      "grad_norm": 0.416278213262558,
-      "learning_rate": 7.199482610865563e-06,
-      "loss": 0.3319,
-      "step": 16889
-    },
-    {
-      "epoch": 1.103849421606431,
-      "grad_norm": 0.4704750180244446,
-      "learning_rate": 7.199169015951523e-06,
-      "loss": 0.3685,
-      "step": 16890
-    },
-    {
-      "epoch": 1.103914776811973,
-      "grad_norm": 0.432136207818985,
-      "learning_rate": 7.198855410311374e-06,
-      "loss": 0.3094,
-      "step": 16891
-    },
-    {
-      "epoch": 1.103980132017515,
-      "grad_norm": 0.44062837958335876,
-      "learning_rate": 7.198541793946645e-06,
-      "loss": 0.3376,
-      "step": 16892
-    },
-    {
-      "epoch": 1.1040454872230574,
-      "grad_norm": 0.4458533525466919,
-      "learning_rate": 7.1982281668588675e-06,
-      "loss": 0.3564,
-      "step": 16893
-    },
-    {
-      "epoch": 1.1041108424285995,
-      "grad_norm": 0.4284687042236328,
-      "learning_rate": 7.197914529049568e-06,
-      "loss": 0.3412,
-      "step": 16894
-    },
-    {
-      "epoch": 1.1041761976341415,
-      "grad_norm": 0.42662525177001953,
-      "learning_rate": 7.197600880520279e-06,
-      "loss": 0.3052,
-      "step": 16895
-    },
-    {
-      "epoch": 1.1042415528396836,
-      "grad_norm": 0.44712355732917786,
-      "learning_rate": 7.1972872212725284e-06,
-      "loss": 0.3426,
-      "step": 16896
-    },
-    {
-      "epoch": 1.104306908045226,
-      "grad_norm": 0.45476073026657104,
-      "learning_rate": 7.1969735513078475e-06,
-      "loss": 0.3327,
-      "step": 16897
-    },
-    {
-      "epoch": 1.104372263250768,
-      "grad_norm": 0.4307636022567749,
-      "learning_rate": 7.196659870627765e-06,
-      "loss": 0.317,
-      "step": 16898
-    },
-    {
-      "epoch": 1.10443761845631,
-      "grad_norm": 0.4835704267024994,
-      "learning_rate": 7.1963461792338115e-06,
-      "loss": 0.3859,
-      "step": 16899
-    },
-    {
-      "epoch": 1.1045029736618521,
-      "grad_norm": 0.4754636883735657,
-      "learning_rate": 7.196032477127517e-06,
-      "loss": 0.3721,
-      "step": 16900
-    },
-    {
-      "epoch": 1.1045683288673942,
-      "grad_norm": 0.4795001447200775,
-      "learning_rate": 7.195718764310411e-06,
-      "loss": 0.3767,
-      "step": 16901
-    },
-    {
-      "epoch": 1.1046336840729365,
-      "grad_norm": 0.4457133114337921,
-      "learning_rate": 7.195405040784025e-06,
-      "loss": 0.3298,
-      "step": 16902
-    },
-    {
-      "epoch": 1.1046990392784786,
-      "grad_norm": 0.454456090927124,
-      "learning_rate": 7.1950913065498865e-06,
-      "loss": 0.3312,
-      "step": 16903
-    },
-    {
-      "epoch": 1.1047643944840206,
-      "grad_norm": 0.4530438780784607,
-      "learning_rate": 7.194777561609527e-06,
-      "loss": 0.3398,
-      "step": 16904
-    },
-    {
-      "epoch": 1.1048297496895627,
-      "grad_norm": 0.44374170899391174,
-      "learning_rate": 7.194463805964478e-06,
-      "loss": 0.3344,
-      "step": 16905
-    },
-    {
-      "epoch": 1.104895104895105,
-      "grad_norm": 0.4460992217063904,
-      "learning_rate": 7.1941500396162675e-06,
-      "loss": 0.3491,
-      "step": 16906
-    },
-    {
-      "epoch": 1.104960460100647,
-      "grad_norm": 0.44110846519470215,
-      "learning_rate": 7.19383626256643e-06,
-      "loss": 0.3368,
-      "step": 16907
-    },
-    {
-      "epoch": 1.1050258153061892,
-      "grad_norm": 0.4690539836883545,
-      "learning_rate": 7.19352247481649e-06,
-      "loss": 0.3799,
-      "step": 16908
-    },
-    {
-      "epoch": 1.1050911705117312,
-      "grad_norm": 0.46475502848625183,
-      "learning_rate": 7.193208676367982e-06,
-      "loss": 0.3432,
-      "step": 16909
-    },
-    {
-      "epoch": 1.1051565257172733,
-      "grad_norm": 0.4278426468372345,
-      "learning_rate": 7.192894867222435e-06,
-      "loss": 0.3075,
-      "step": 16910
-    },
-    {
-      "epoch": 1.1052218809228156,
-      "grad_norm": 0.42875391244888306,
-      "learning_rate": 7.192581047381382e-06,
-      "loss": 0.3429,
-      "step": 16911
-    },
-    {
-      "epoch": 1.1052872361283577,
-      "grad_norm": 0.44750601053237915,
-      "learning_rate": 7.19226721684635e-06,
-      "loss": 0.367,
-      "step": 16912
-    },
-    {
-      "epoch": 1.1053525913338997,
-      "grad_norm": 0.4273897409439087,
-      "learning_rate": 7.191953375618872e-06,
-      "loss": 0.3022,
-      "step": 16913
-    },
-    {
-      "epoch": 1.1054179465394418,
-      "grad_norm": 0.4383806884288788,
-      "learning_rate": 7.191639523700478e-06,
-      "loss": 0.3435,
-      "step": 16914
-    },
-    {
-      "epoch": 1.105483301744984,
-      "grad_norm": 0.4066704213619232,
-      "learning_rate": 7.1913256610926975e-06,
-      "loss": 0.2842,
-      "step": 16915
-    },
-    {
-      "epoch": 1.1055486569505262,
-      "grad_norm": 0.4943954348564148,
-      "learning_rate": 7.191011787797064e-06,
-      "loss": 0.355,
-      "step": 16916
-    },
-    {
-      "epoch": 1.1056140121560682,
-      "grad_norm": 0.4474785625934601,
-      "learning_rate": 7.190697903815106e-06,
-      "loss": 0.3463,
-      "step": 16917
-    },
-    {
-      "epoch": 1.1056793673616103,
-      "grad_norm": 0.48214051127433777,
-      "learning_rate": 7.190384009148357e-06,
-      "loss": 0.38,
-      "step": 16918
-    },
-    {
-      "epoch": 1.1057447225671524,
-      "grad_norm": 0.4600478410720825,
-      "learning_rate": 7.190070103798346e-06,
-      "loss": 0.3504,
-      "step": 16919
-    },
-    {
-      "epoch": 1.1058100777726947,
-      "grad_norm": 0.4889717698097229,
-      "learning_rate": 7.1897561877666035e-06,
-      "loss": 0.3853,
-      "step": 16920
-    },
-    {
-      "epoch": 1.1058754329782368,
-      "grad_norm": 0.4611831307411194,
-      "learning_rate": 7.18944226105466e-06,
-      "loss": 0.3442,
-      "step": 16921
-    },
-    {
-      "epoch": 1.1059407881837788,
-      "grad_norm": 0.42291581630706787,
-      "learning_rate": 7.189128323664051e-06,
-      "loss": 0.3171,
-      "step": 16922
-    },
-    {
-      "epoch": 1.106006143389321,
-      "grad_norm": 0.46375003457069397,
-      "learning_rate": 7.1888143755963026e-06,
-      "loss": 0.3743,
-      "step": 16923
-    },
-    {
-      "epoch": 1.106071498594863,
-      "grad_norm": 0.4620855450630188,
-      "learning_rate": 7.188500416852949e-06,
-      "loss": 0.3471,
-      "step": 16924
-    },
-    {
-      "epoch": 1.1061368538004053,
-      "grad_norm": 0.41443929076194763,
-      "learning_rate": 7.188186447435521e-06,
-      "loss": 0.3073,
-      "step": 16925
-    },
-    {
-      "epoch": 1.1062022090059473,
-      "grad_norm": 0.4544248580932617,
-      "learning_rate": 7.187872467345549e-06,
-      "loss": 0.3464,
-      "step": 16926
-    },
-    {
-      "epoch": 1.1062675642114894,
-      "grad_norm": 0.4621868133544922,
-      "learning_rate": 7.187558476584566e-06,
-      "loss": 0.3777,
-      "step": 16927
-    },
-    {
-      "epoch": 1.1063329194170315,
-      "grad_norm": 0.4542226195335388,
-      "learning_rate": 7.1872444751541025e-06,
-      "loss": 0.3592,
-      "step": 16928
-    },
-    {
-      "epoch": 1.1063982746225738,
-      "grad_norm": 0.44520843029022217,
-      "learning_rate": 7.186930463055689e-06,
-      "loss": 0.3215,
-      "step": 16929
-    },
-    {
-      "epoch": 1.1064636298281159,
-      "grad_norm": 0.4341967701911926,
-      "learning_rate": 7.186616440290858e-06,
-      "loss": 0.3175,
-      "step": 16930
-    },
-    {
-      "epoch": 1.106528985033658,
-      "grad_norm": 0.44681474566459656,
-      "learning_rate": 7.186302406861142e-06,
-      "loss": 0.3556,
-      "step": 16931
-    },
-    {
-      "epoch": 1.1065943402392,
-      "grad_norm": 0.4931948781013489,
-      "learning_rate": 7.18598836276807e-06,
-      "loss": 0.3931,
-      "step": 16932
-    },
-    {
-      "epoch": 1.106659695444742,
-      "grad_norm": 0.45590752363204956,
-      "learning_rate": 7.185674308013177e-06,
-      "loss": 0.3807,
-      "step": 16933
-    },
-    {
-      "epoch": 1.1067250506502844,
-      "grad_norm": 0.41704118251800537,
-      "learning_rate": 7.185360242597994e-06,
-      "loss": 0.3275,
-      "step": 16934
-    },
-    {
-      "epoch": 1.1067904058558264,
-      "grad_norm": 0.47248926758766174,
-      "learning_rate": 7.18504616652405e-06,
-      "loss": 0.3537,
-      "step": 16935
-    },
-    {
-      "epoch": 1.1068557610613685,
-      "grad_norm": 0.4390184283256531,
-      "learning_rate": 7.184732079792881e-06,
-      "loss": 0.3107,
-      "step": 16936
-    },
-    {
-      "epoch": 1.1069211162669106,
-      "grad_norm": 0.4599098265171051,
-      "learning_rate": 7.1844179824060155e-06,
-      "loss": 0.3688,
-      "step": 16937
-    },
-    {
-      "epoch": 1.1069864714724529,
-      "grad_norm": 0.4395337402820587,
-      "learning_rate": 7.184103874364987e-06,
-      "loss": 0.3454,
-      "step": 16938
-    },
-    {
-      "epoch": 1.107051826677995,
-      "grad_norm": 0.44652825593948364,
-      "learning_rate": 7.183789755671328e-06,
-      "loss": 0.3321,
-      "step": 16939
-    },
-    {
-      "epoch": 1.107117181883537,
-      "grad_norm": 0.46515724062919617,
-      "learning_rate": 7.183475626326568e-06,
-      "loss": 0.3302,
-      "step": 16940
-    },
-    {
-      "epoch": 1.107182537089079,
-      "grad_norm": 0.45389124751091003,
-      "learning_rate": 7.183161486332242e-06,
-      "loss": 0.3465,
-      "step": 16941
-    },
-    {
-      "epoch": 1.1072478922946212,
-      "grad_norm": 0.5035362243652344,
-      "learning_rate": 7.182847335689882e-06,
-      "loss": 0.3836,
-      "step": 16942
-    },
-    {
-      "epoch": 1.1073132475001635,
-      "grad_norm": 0.4707864224910736,
-      "learning_rate": 7.182533174401017e-06,
-      "loss": 0.3422,
-      "step": 16943
-    },
-    {
-      "epoch": 1.1073786027057055,
-      "grad_norm": 0.4576711058616638,
-      "learning_rate": 7.182219002467183e-06,
-      "loss": 0.3255,
-      "step": 16944
-    },
-    {
-      "epoch": 1.1074439579112476,
-      "grad_norm": 0.44810256361961365,
-      "learning_rate": 7.181904819889912e-06,
-      "loss": 0.3555,
-      "step": 16945
-    },
-    {
-      "epoch": 1.1075093131167897,
-      "grad_norm": 0.4620898365974426,
-      "learning_rate": 7.181590626670734e-06,
-      "loss": 0.3527,
-      "step": 16946
-    },
-    {
-      "epoch": 1.1075746683223318,
-      "grad_norm": 0.4596816301345825,
-      "learning_rate": 7.181276422811183e-06,
-      "loss": 0.3592,
-      "step": 16947
-    },
-    {
-      "epoch": 1.107640023527874,
-      "grad_norm": 0.4484982490539551,
-      "learning_rate": 7.18096220831279e-06,
-      "loss": 0.3202,
-      "step": 16948
-    },
-    {
-      "epoch": 1.1077053787334161,
-      "grad_norm": 0.44571617245674133,
-      "learning_rate": 7.1806479831770905e-06,
-      "loss": 0.356,
-      "step": 16949
-    },
-    {
-      "epoch": 1.1077707339389582,
-      "grad_norm": 0.4459291398525238,
-      "learning_rate": 7.180333747405615e-06,
-      "loss": 0.3339,
-      "step": 16950
-    },
-    {
-      "epoch": 1.1078360891445003,
-      "grad_norm": 0.4683772921562195,
-      "learning_rate": 7.180019500999895e-06,
-      "loss": 0.3843,
-      "step": 16951
-    },
-    {
-      "epoch": 1.1079014443500426,
-      "grad_norm": 0.43534207344055176,
-      "learning_rate": 7.179705243961467e-06,
-      "loss": 0.3137,
-      "step": 16952
-    },
-    {
-      "epoch": 1.1079667995555846,
-      "grad_norm": 0.47841548919677734,
-      "learning_rate": 7.17939097629186e-06,
-      "loss": 0.3655,
-      "step": 16953
-    },
-    {
-      "epoch": 1.1080321547611267,
-      "grad_norm": 0.44662392139434814,
-      "learning_rate": 7.179076697992608e-06,
-      "loss": 0.3349,
-      "step": 16954
-    },
-    {
-      "epoch": 1.1080975099666688,
-      "grad_norm": 0.43585604429244995,
-      "learning_rate": 7.178762409065245e-06,
-      "loss": 0.3222,
-      "step": 16955
-    },
-    {
-      "epoch": 1.1081628651722109,
-      "grad_norm": 0.4386887550354004,
-      "learning_rate": 7.178448109511303e-06,
-      "loss": 0.3301,
-      "step": 16956
-    },
-    {
-      "epoch": 1.1082282203777531,
-      "grad_norm": 0.45249950885772705,
-      "learning_rate": 7.178133799332313e-06,
-      "loss": 0.374,
-      "step": 16957
-    },
-    {
-      "epoch": 1.1082935755832952,
-      "grad_norm": 0.44922906160354614,
-      "learning_rate": 7.177819478529811e-06,
-      "loss": 0.3431,
-      "step": 16958
-    },
-    {
-      "epoch": 1.1083589307888373,
-      "grad_norm": 0.4794429838657379,
-      "learning_rate": 7.177505147105329e-06,
-      "loss": 0.3538,
-      "step": 16959
-    },
-    {
-      "epoch": 1.1084242859943794,
-      "grad_norm": 0.4555109739303589,
-      "learning_rate": 7.177190805060402e-06,
-      "loss": 0.3371,
-      "step": 16960
-    },
-    {
-      "epoch": 1.1084896411999217,
-      "grad_norm": 0.45904722809791565,
-      "learning_rate": 7.176876452396558e-06,
-      "loss": 0.375,
-      "step": 16961
-    },
-    {
-      "epoch": 1.1085549964054637,
-      "grad_norm": 0.4988830089569092,
-      "learning_rate": 7.1765620891153354e-06,
-      "loss": 0.4157,
-      "step": 16962
-    },
-    {
-      "epoch": 1.1086203516110058,
-      "grad_norm": 0.462028831243515,
-      "learning_rate": 7.1762477152182655e-06,
-      "loss": 0.342,
-      "step": 16963
-    },
-    {
-      "epoch": 1.1086857068165479,
-      "grad_norm": 0.4333418011665344,
-      "learning_rate": 7.17593333070688e-06,
-      "loss": 0.3029,
-      "step": 16964
-    },
-    {
-      "epoch": 1.10875106202209,
-      "grad_norm": 0.4603612422943115,
-      "learning_rate": 7.175618935582716e-06,
-      "loss": 0.3413,
-      "step": 16965
-    },
-    {
-      "epoch": 1.1088164172276322,
-      "grad_norm": 0.43920230865478516,
-      "learning_rate": 7.175304529847303e-06,
-      "loss": 0.3438,
-      "step": 16966
-    },
-    {
-      "epoch": 1.1088817724331743,
-      "grad_norm": 0.45885127782821655,
-      "learning_rate": 7.174990113502176e-06,
-      "loss": 0.3498,
-      "step": 16967
-    },
-    {
-      "epoch": 1.1089471276387164,
-      "grad_norm": 0.44494903087615967,
-      "learning_rate": 7.17467568654887e-06,
-      "loss": 0.3211,
-      "step": 16968
-    },
-    {
-      "epoch": 1.1090124828442585,
-      "grad_norm": 0.44241607189178467,
-      "learning_rate": 7.174361248988917e-06,
-      "loss": 0.3431,
-      "step": 16969
-    },
-    {
-      "epoch": 1.1090778380498008,
-      "grad_norm": 0.4518430531024933,
-      "learning_rate": 7.1740468008238494e-06,
-      "loss": 0.3164,
-      "step": 16970
-    },
-    {
-      "epoch": 1.1091431932553428,
-      "grad_norm": 0.44299453496932983,
-      "learning_rate": 7.173732342055204e-06,
-      "loss": 0.3187,
-      "step": 16971
-    },
-    {
-      "epoch": 1.109208548460885,
-      "grad_norm": 0.4495779573917389,
-      "learning_rate": 7.173417872684513e-06,
-      "loss": 0.338,
-      "step": 16972
-    },
-    {
-      "epoch": 1.109273903666427,
-      "grad_norm": 0.42455625534057617,
-      "learning_rate": 7.17310339271331e-06,
-      "loss": 0.3367,
-      "step": 16973
-    },
-    {
-      "epoch": 1.109339258871969,
-      "grad_norm": 0.4374285042285919,
-      "learning_rate": 7.172788902143128e-06,
-      "loss": 0.3414,
-      "step": 16974
-    },
-    {
-      "epoch": 1.1094046140775113,
-      "grad_norm": 0.4373040199279785,
-      "learning_rate": 7.172474400975502e-06,
-      "loss": 0.3114,
-      "step": 16975
-    },
-    {
-      "epoch": 1.1094699692830534,
-      "grad_norm": 0.5009270906448364,
-      "learning_rate": 7.172159889211966e-06,
-      "loss": 0.3965,
-      "step": 16976
-    },
-    {
-      "epoch": 1.1095353244885955,
-      "grad_norm": 0.4263479709625244,
-      "learning_rate": 7.171845366854053e-06,
-      "loss": 0.3364,
-      "step": 16977
-    },
-    {
-      "epoch": 1.1096006796941376,
-      "grad_norm": 0.44614484906196594,
-      "learning_rate": 7.1715308339033e-06,
-      "loss": 0.3529,
-      "step": 16978
-    },
-    {
-      "epoch": 1.1096660348996799,
-      "grad_norm": 0.44420912861824036,
-      "learning_rate": 7.171216290361237e-06,
-      "loss": 0.3363,
-      "step": 16979
-    },
-    {
-      "epoch": 1.109731390105222,
-      "grad_norm": 0.4217231571674347,
-      "learning_rate": 7.1709017362294e-06,
-      "loss": 0.3295,
-      "step": 16980
-    },
-    {
-      "epoch": 1.109796745310764,
-      "grad_norm": 0.4209829270839691,
-      "learning_rate": 7.170587171509325e-06,
-      "loss": 0.3025,
-      "step": 16981
-    },
-    {
-      "epoch": 1.109862100516306,
-      "grad_norm": 0.485849529504776,
-      "learning_rate": 7.170272596202542e-06,
-      "loss": 0.3501,
-      "step": 16982
-    },
-    {
-      "epoch": 1.1099274557218481,
-      "grad_norm": 0.4528864920139313,
-      "learning_rate": 7.169958010310589e-06,
-      "loss": 0.3316,
-      "step": 16983
-    },
-    {
-      "epoch": 1.1099928109273904,
-      "grad_norm": 0.4277852475643158,
-      "learning_rate": 7.169643413834998e-06,
-      "loss": 0.3273,
-      "step": 16984
-    },
-    {
-      "epoch": 1.1100581661329325,
-      "grad_norm": 0.44532978534698486,
-      "learning_rate": 7.169328806777306e-06,
-      "loss": 0.3384,
-      "step": 16985
-    },
-    {
-      "epoch": 1.1101235213384746,
-      "grad_norm": 0.42431533336639404,
-      "learning_rate": 7.169014189139044e-06,
-      "loss": 0.3247,
-      "step": 16986
-    },
-    {
-      "epoch": 1.1101888765440167,
-      "grad_norm": 0.4513096213340759,
-      "learning_rate": 7.16869956092175e-06,
-      "loss": 0.3473,
-      "step": 16987
-    },
-    {
-      "epoch": 1.110254231749559,
-      "grad_norm": 0.42926570773124695,
-      "learning_rate": 7.168384922126955e-06,
-      "loss": 0.328,
-      "step": 16988
-    },
-    {
-      "epoch": 1.110319586955101,
-      "grad_norm": 0.4437406659126282,
-      "learning_rate": 7.168070272756198e-06,
-      "loss": 0.3527,
-      "step": 16989
-    },
-    {
-      "epoch": 1.110384942160643,
-      "grad_norm": 0.4713248312473297,
-      "learning_rate": 7.167755612811009e-06,
-      "loss": 0.3474,
-      "step": 16990
-    },
-    {
-      "epoch": 1.1104502973661852,
-      "grad_norm": 0.42764395475387573,
-      "learning_rate": 7.167440942292926e-06,
-      "loss": 0.3138,
-      "step": 16991
-    },
-    {
-      "epoch": 1.1105156525717272,
-      "grad_norm": 0.44529637694358826,
-      "learning_rate": 7.167126261203483e-06,
-      "loss": 0.326,
-      "step": 16992
-    },
-    {
-      "epoch": 1.1105810077772695,
-      "grad_norm": 0.4588848948478699,
-      "learning_rate": 7.166811569544213e-06,
-      "loss": 0.3536,
-      "step": 16993
-    },
-    {
-      "epoch": 1.1106463629828116,
-      "grad_norm": 0.4248858094215393,
-      "learning_rate": 7.1664968673166545e-06,
-      "loss": 0.3144,
-      "step": 16994
-    },
-    {
-      "epoch": 1.1107117181883537,
-      "grad_norm": 0.4718970060348511,
-      "learning_rate": 7.1661821545223385e-06,
-      "loss": 0.3595,
-      "step": 16995
-    },
-    {
-      "epoch": 1.1107770733938958,
-      "grad_norm": 0.44681426882743835,
-      "learning_rate": 7.165867431162802e-06,
-      "loss": 0.316,
-      "step": 16996
-    },
-    {
-      "epoch": 1.110842428599438,
-      "grad_norm": 0.4472370743751526,
-      "learning_rate": 7.165552697239579e-06,
-      "loss": 0.3217,
-      "step": 16997
-    },
-    {
-      "epoch": 1.1109077838049801,
-      "grad_norm": 0.5172298550605774,
-      "learning_rate": 7.1652379527542075e-06,
-      "loss": 0.3161,
-      "step": 16998
-    },
-    {
-      "epoch": 1.1109731390105222,
-      "grad_norm": 0.46733778715133667,
-      "learning_rate": 7.164923197708219e-06,
-      "loss": 0.3647,
-      "step": 16999
-    },
-    {
-      "epoch": 1.1110384942160643,
-      "grad_norm": 0.44942665100097656,
-      "learning_rate": 7.164608432103149e-06,
-      "loss": 0.3586,
-      "step": 17000
-    },
-    {
-      "epoch": 1.1111038494216063,
-      "grad_norm": 0.4257781505584717,
-      "learning_rate": 7.164293655940537e-06,
-      "loss": 0.3365,
-      "step": 17001
-    },
-    {
-      "epoch": 1.1111692046271486,
-      "grad_norm": 0.42540618777275085,
-      "learning_rate": 7.163978869221912e-06,
-      "loss": 0.2947,
-      "step": 17002
-    },
-    {
-      "epoch": 1.1112345598326907,
-      "grad_norm": 0.44917404651641846,
-      "learning_rate": 7.1636640719488145e-06,
-      "loss": 0.3513,
-      "step": 17003
-    },
-    {
-      "epoch": 1.1112999150382328,
-      "grad_norm": 0.45079484581947327,
-      "learning_rate": 7.163349264122776e-06,
-      "loss": 0.353,
-      "step": 17004
-    },
-    {
-      "epoch": 1.1113652702437748,
-      "grad_norm": 0.44623708724975586,
-      "learning_rate": 7.163034445745335e-06,
-      "loss": 0.362,
-      "step": 17005
-    },
-    {
-      "epoch": 1.1114306254493171,
-      "grad_norm": 0.4129381477832794,
-      "learning_rate": 7.162719616818024e-06,
-      "loss": 0.3163,
-      "step": 17006
-    },
-    {
-      "epoch": 1.1114959806548592,
-      "grad_norm": 0.4586268365383148,
-      "learning_rate": 7.162404777342382e-06,
-      "loss": 0.3562,
-      "step": 17007
-    },
-    {
-      "epoch": 1.1115613358604013,
-      "grad_norm": 0.4528006911277771,
-      "learning_rate": 7.1620899273199404e-06,
-      "loss": 0.3402,
-      "step": 17008
-    },
-    {
-      "epoch": 1.1116266910659434,
-      "grad_norm": 0.4343746602535248,
-      "learning_rate": 7.161775066752239e-06,
-      "loss": 0.3353,
-      "step": 17009
-    },
-    {
-      "epoch": 1.1116920462714854,
-      "grad_norm": 0.45098209381103516,
-      "learning_rate": 7.161460195640812e-06,
-      "loss": 0.3769,
-      "step": 17010
-    },
-    {
-      "epoch": 1.1117574014770277,
-      "grad_norm": 0.45243769884109497,
-      "learning_rate": 7.161145313987194e-06,
-      "loss": 0.3783,
-      "step": 17011
-    },
-    {
-      "epoch": 1.1118227566825698,
-      "grad_norm": 0.4434148073196411,
-      "learning_rate": 7.160830421792922e-06,
-      "loss": 0.3198,
-      "step": 17012
-    },
-    {
-      "epoch": 1.1118881118881119,
-      "grad_norm": 0.44666770100593567,
-      "learning_rate": 7.160515519059531e-06,
-      "loss": 0.3365,
-      "step": 17013
-    },
-    {
-      "epoch": 1.111953467093654,
-      "grad_norm": 0.45525074005126953,
-      "learning_rate": 7.160200605788559e-06,
-      "loss": 0.311,
-      "step": 17014
-    },
-    {
-      "epoch": 1.1120188222991962,
-      "grad_norm": 0.450960636138916,
-      "learning_rate": 7.1598856819815374e-06,
-      "loss": 0.3373,
-      "step": 17015
-    },
-    {
-      "epoch": 1.1120841775047383,
-      "grad_norm": 0.4486139118671417,
-      "learning_rate": 7.159570747640008e-06,
-      "loss": 0.3087,
-      "step": 17016
-    },
-    {
-      "epoch": 1.1121495327102804,
-      "grad_norm": 0.5001394152641296,
-      "learning_rate": 7.1592558027655016e-06,
-      "loss": 0.3286,
-      "step": 17017
-    },
-    {
-      "epoch": 1.1122148879158225,
-      "grad_norm": 0.4579133689403534,
-      "learning_rate": 7.158940847359558e-06,
-      "loss": 0.3631,
-      "step": 17018
-    },
-    {
-      "epoch": 1.1122802431213645,
-      "grad_norm": 0.46382418274879456,
-      "learning_rate": 7.158625881423711e-06,
-      "loss": 0.365,
-      "step": 17019
-    },
-    {
-      "epoch": 1.1123455983269068,
-      "grad_norm": 0.443668395280838,
-      "learning_rate": 7.158310904959498e-06,
-      "loss": 0.3347,
-      "step": 17020
-    },
-    {
-      "epoch": 1.112410953532449,
-      "grad_norm": 0.4584289491176605,
-      "learning_rate": 7.157995917968455e-06,
-      "loss": 0.3545,
-      "step": 17021
-    },
-    {
-      "epoch": 1.112476308737991,
-      "grad_norm": 0.5357624292373657,
-      "learning_rate": 7.157680920452119e-06,
-      "loss": 0.3404,
-      "step": 17022
-    },
-    {
-      "epoch": 1.112541663943533,
-      "grad_norm": 0.4695124626159668,
-      "learning_rate": 7.157365912412027e-06,
-      "loss": 0.3458,
-      "step": 17023
-    },
-    {
-      "epoch": 1.1126070191490753,
-      "grad_norm": 0.4648081064224243,
-      "learning_rate": 7.1570508938497105e-06,
-      "loss": 0.3763,
-      "step": 17024
-    },
-    {
-      "epoch": 1.1126723743546174,
-      "grad_norm": 0.4514486491680145,
-      "learning_rate": 7.156735864766713e-06,
-      "loss": 0.3454,
-      "step": 17025
-    },
-    {
-      "epoch": 1.1127377295601595,
-      "grad_norm": 0.4472137689590454,
-      "learning_rate": 7.156420825164565e-06,
-      "loss": 0.3281,
-      "step": 17026
-    },
-    {
-      "epoch": 1.1128030847657016,
-      "grad_norm": 0.49321311712265015,
-      "learning_rate": 7.156105775044806e-06,
-      "loss": 0.3666,
-      "step": 17027
-    },
-    {
-      "epoch": 1.1128684399712436,
-      "grad_norm": 0.42386817932128906,
-      "learning_rate": 7.155790714408972e-06,
-      "loss": 0.3051,
-      "step": 17028
-    },
-    {
-      "epoch": 1.112933795176786,
-      "grad_norm": 0.42108550667762756,
-      "learning_rate": 7.155475643258601e-06,
-      "loss": 0.2892,
-      "step": 17029
-    },
-    {
-      "epoch": 1.112999150382328,
-      "grad_norm": 0.44198665022850037,
-      "learning_rate": 7.155160561595229e-06,
-      "loss": 0.343,
-      "step": 17030
-    },
-    {
-      "epoch": 1.11306450558787,
-      "grad_norm": 0.49344179034233093,
-      "learning_rate": 7.15484546942039e-06,
-      "loss": 0.4028,
-      "step": 17031
-    },
-    {
-      "epoch": 1.1131298607934121,
-      "grad_norm": 0.4633345603942871,
-      "learning_rate": 7.154530366735626e-06,
-      "loss": 0.331,
-      "step": 17032
-    },
-    {
-      "epoch": 1.1131952159989542,
-      "grad_norm": 0.4551130533218384,
-      "learning_rate": 7.154215253542468e-06,
-      "loss": 0.3422,
-      "step": 17033
-    },
-    {
-      "epoch": 1.1132605712044965,
-      "grad_norm": 0.41660186648368835,
-      "learning_rate": 7.153900129842458e-06,
-      "loss": 0.2963,
-      "step": 17034
-    },
-    {
-      "epoch": 1.1133259264100386,
-      "grad_norm": 0.45533299446105957,
-      "learning_rate": 7.153584995637129e-06,
-      "loss": 0.3208,
-      "step": 17035
-    },
-    {
-      "epoch": 1.1133912816155807,
-      "grad_norm": 0.461054265499115,
-      "learning_rate": 7.1532698509280215e-06,
-      "loss": 0.3527,
-      "step": 17036
-    },
-    {
-      "epoch": 1.1134566368211227,
-      "grad_norm": 0.49841558933258057,
-      "learning_rate": 7.152954695716672e-06,
-      "loss": 0.367,
-      "step": 17037
-    },
-    {
-      "epoch": 1.113521992026665,
-      "grad_norm": 0.45055150985717773,
-      "learning_rate": 7.152639530004615e-06,
-      "loss": 0.3344,
-      "step": 17038
-    },
-    {
-      "epoch": 1.113587347232207,
-      "grad_norm": 0.43217653036117554,
-      "learning_rate": 7.152324353793389e-06,
-      "loss": 0.32,
-      "step": 17039
-    },
-    {
-      "epoch": 1.1136527024377492,
-      "grad_norm": 0.46337515115737915,
-      "learning_rate": 7.1520091670845316e-06,
-      "loss": 0.3885,
-      "step": 17040
-    },
-    {
-      "epoch": 1.1137180576432912,
-      "grad_norm": 0.39758139848709106,
-      "learning_rate": 7.15169396987958e-06,
-      "loss": 0.2717,
-      "step": 17041
-    },
-    {
-      "epoch": 1.1137834128488333,
-      "grad_norm": 0.48315921425819397,
-      "learning_rate": 7.151378762180072e-06,
-      "loss": 0.381,
-      "step": 17042
-    },
-    {
-      "epoch": 1.1138487680543756,
-      "grad_norm": 0.4399372637271881,
-      "learning_rate": 7.151063543987544e-06,
-      "loss": 0.3466,
-      "step": 17043
-    },
-    {
-      "epoch": 1.1139141232599177,
-      "grad_norm": 0.45378926396369934,
-      "learning_rate": 7.150748315303535e-06,
-      "loss": 0.3627,
-      "step": 17044
-    },
-    {
-      "epoch": 1.1139794784654597,
-      "grad_norm": 0.45435047149658203,
-      "learning_rate": 7.150433076129581e-06,
-      "loss": 0.3663,
-      "step": 17045
-    },
-    {
-      "epoch": 1.1140448336710018,
-      "grad_norm": 0.4683096706867218,
-      "learning_rate": 7.15011782646722e-06,
-      "loss": 0.3699,
-      "step": 17046
-    },
-    {
-      "epoch": 1.1141101888765441,
-      "grad_norm": 0.4378737807273865,
-      "learning_rate": 7.14980256631799e-06,
-      "loss": 0.3424,
-      "step": 17047
-    },
-    {
-      "epoch": 1.1141755440820862,
-      "grad_norm": 0.44605475664138794,
-      "learning_rate": 7.149487295683426e-06,
-      "loss": 0.359,
-      "step": 17048
-    },
-    {
-      "epoch": 1.1142408992876283,
-      "grad_norm": 0.46032944321632385,
-      "learning_rate": 7.149172014565069e-06,
-      "loss": 0.3676,
-      "step": 17049
-    },
-    {
-      "epoch": 1.1143062544931703,
-      "grad_norm": 0.47950488328933716,
-      "learning_rate": 7.148856722964456e-06,
-      "loss": 0.3446,
-      "step": 17050
-    },
-    {
-      "epoch": 1.1143716096987124,
-      "grad_norm": 0.438961923122406,
-      "learning_rate": 7.148541420883123e-06,
-      "loss": 0.3495,
-      "step": 17051
-    },
-    {
-      "epoch": 1.1144369649042547,
-      "grad_norm": 0.4428153932094574,
-      "learning_rate": 7.14822610832261e-06,
-      "loss": 0.368,
-      "step": 17052
-    },
-    {
-      "epoch": 1.1145023201097968,
-      "grad_norm": 0.4724746644496918,
-      "learning_rate": 7.147910785284453e-06,
-      "loss": 0.2795,
-      "step": 17053
-    },
-    {
-      "epoch": 1.1145676753153388,
-      "grad_norm": 0.4283941984176636,
-      "learning_rate": 7.147595451770193e-06,
-      "loss": 0.3386,
-      "step": 17054
-    },
-    {
-      "epoch": 1.114633030520881,
-      "grad_norm": 0.4613955318927765,
-      "learning_rate": 7.147280107781365e-06,
-      "loss": 0.3307,
-      "step": 17055
-    },
-    {
-      "epoch": 1.1146983857264232,
-      "grad_norm": 0.4663427174091339,
-      "learning_rate": 7.1469647533195075e-06,
-      "loss": 0.3536,
-      "step": 17056
-    },
-    {
-      "epoch": 1.1147637409319653,
-      "grad_norm": 0.4404975473880768,
-      "learning_rate": 7.14664938838616e-06,
-      "loss": 0.3439,
-      "step": 17057
-    },
-    {
-      "epoch": 1.1148290961375074,
-      "grad_norm": 0.4159106910228729,
-      "learning_rate": 7.146334012982859e-06,
-      "loss": 0.2972,
-      "step": 17058
-    },
-    {
-      "epoch": 1.1148944513430494,
-      "grad_norm": 0.45903849601745605,
-      "learning_rate": 7.146018627111144e-06,
-      "loss": 0.3559,
-      "step": 17059
-    },
-    {
-      "epoch": 1.1149598065485915,
-      "grad_norm": 0.4485066533088684,
-      "learning_rate": 7.145703230772552e-06,
-      "loss": 0.3582,
-      "step": 17060
-    },
-    {
-      "epoch": 1.1150251617541338,
-      "grad_norm": 0.4486883580684662,
-      "learning_rate": 7.1453878239686235e-06,
-      "loss": 0.3249,
-      "step": 17061
-    },
-    {
-      "epoch": 1.1150905169596759,
-      "grad_norm": 0.5061477422714233,
-      "learning_rate": 7.145072406700894e-06,
-      "loss": 0.3625,
-      "step": 17062
-    },
-    {
-      "epoch": 1.115155872165218,
-      "grad_norm": 0.45438152551651,
-      "learning_rate": 7.144756978970904e-06,
-      "loss": 0.3392,
-      "step": 17063
-    },
-    {
-      "epoch": 1.11522122737076,
-      "grad_norm": 0.4226972162723541,
-      "learning_rate": 7.14444154078019e-06,
-      "loss": 0.3355,
-      "step": 17064
-    },
-    {
-      "epoch": 1.115286582576302,
-      "grad_norm": 0.4342862069606781,
-      "learning_rate": 7.144126092130293e-06,
-      "loss": 0.3066,
-      "step": 17065
-    },
-    {
-      "epoch": 1.1153519377818444,
-      "grad_norm": 0.47154200077056885,
-      "learning_rate": 7.14381063302275e-06,
-      "loss": 0.4142,
-      "step": 17066
-    },
-    {
-      "epoch": 1.1154172929873865,
-      "grad_norm": 0.40317896008491516,
-      "learning_rate": 7.1434951634591e-06,
-      "loss": 0.2885,
-      "step": 17067
-    },
-    {
-      "epoch": 1.1154826481929285,
-      "grad_norm": 0.40597960352897644,
-      "learning_rate": 7.143179683440882e-06,
-      "loss": 0.2789,
-      "step": 17068
-    },
-    {
-      "epoch": 1.1155480033984706,
-      "grad_norm": 0.4608341157436371,
-      "learning_rate": 7.142864192969632e-06,
-      "loss": 0.3355,
-      "step": 17069
-    },
-    {
-      "epoch": 1.115613358604013,
-      "grad_norm": 0.39357125759124756,
-      "learning_rate": 7.142548692046893e-06,
-      "loss": 0.2598,
-      "step": 17070
-    },
-    {
-      "epoch": 1.115678713809555,
-      "grad_norm": 0.46614694595336914,
-      "learning_rate": 7.142233180674201e-06,
-      "loss": 0.3533,
-      "step": 17071
-    },
-    {
-      "epoch": 1.115744069015097,
-      "grad_norm": 0.5154941082000732,
-      "learning_rate": 7.141917658853095e-06,
-      "loss": 0.3515,
-      "step": 17072
-    },
-    {
-      "epoch": 1.1158094242206391,
-      "grad_norm": 0.422382652759552,
-      "learning_rate": 7.141602126585116e-06,
-      "loss": 0.3078,
-      "step": 17073
-    },
-    {
-      "epoch": 1.1158747794261812,
-      "grad_norm": 0.4533899426460266,
-      "learning_rate": 7.1412865838718e-06,
-      "loss": 0.3355,
-      "step": 17074
-    },
-    {
-      "epoch": 1.1159401346317235,
-      "grad_norm": 0.41989266872406006,
-      "learning_rate": 7.140971030714689e-06,
-      "loss": 0.3067,
-      "step": 17075
-    },
-    {
-      "epoch": 1.1160054898372656,
-      "grad_norm": 0.4281536936759949,
-      "learning_rate": 7.14065546711532e-06,
-      "loss": 0.3241,
-      "step": 17076
-    },
-    {
-      "epoch": 1.1160708450428076,
-      "grad_norm": 0.4291760325431824,
-      "learning_rate": 7.1403398930752324e-06,
-      "loss": 0.3083,
-      "step": 17077
-    },
-    {
-      "epoch": 1.1161362002483497,
-      "grad_norm": 0.4882870614528656,
-      "learning_rate": 7.1400243085959655e-06,
-      "loss": 0.3699,
-      "step": 17078
-    },
-    {
-      "epoch": 1.116201555453892,
-      "grad_norm": 0.4481711685657501,
-      "learning_rate": 7.139708713679059e-06,
-      "loss": 0.3397,
-      "step": 17079
-    },
-    {
-      "epoch": 1.116266910659434,
-      "grad_norm": 0.42289987206459045,
-      "learning_rate": 7.1393931083260515e-06,
-      "loss": 0.3328,
-      "step": 17080
-    },
-    {
-      "epoch": 1.1163322658649761,
-      "grad_norm": 0.44162896275520325,
-      "learning_rate": 7.1390774925384835e-06,
-      "loss": 0.3447,
-      "step": 17081
-    },
-    {
-      "epoch": 1.1163976210705182,
-      "grad_norm": 0.44941946864128113,
-      "learning_rate": 7.138761866317893e-06,
-      "loss": 0.3567,
-      "step": 17082
-    },
-    {
-      "epoch": 1.1164629762760603,
-      "grad_norm": 0.4282679855823517,
-      "learning_rate": 7.1384462296658196e-06,
-      "loss": 0.3486,
-      "step": 17083
-    },
-    {
-      "epoch": 1.1165283314816026,
-      "grad_norm": 0.46869492530822754,
-      "learning_rate": 7.138130582583804e-06,
-      "loss": 0.3442,
-      "step": 17084
-    },
-    {
-      "epoch": 1.1165936866871446,
-      "grad_norm": 0.4324144423007965,
-      "learning_rate": 7.137814925073383e-06,
-      "loss": 0.3361,
-      "step": 17085
-    },
-    {
-      "epoch": 1.1166590418926867,
-      "grad_norm": 0.42913392186164856,
-      "learning_rate": 7.137499257136101e-06,
-      "loss": 0.3381,
-      "step": 17086
-    },
-    {
-      "epoch": 1.1167243970982288,
-      "grad_norm": 0.429119735956192,
-      "learning_rate": 7.137183578773492e-06,
-      "loss": 0.3093,
-      "step": 17087
-    },
-    {
-      "epoch": 1.116789752303771,
-      "grad_norm": 0.4642656743526459,
-      "learning_rate": 7.1368678899871e-06,
-      "loss": 0.3809,
-      "step": 17088
-    },
-    {
-      "epoch": 1.1168551075093132,
-      "grad_norm": 0.4095107614994049,
-      "learning_rate": 7.136552190778462e-06,
-      "loss": 0.3191,
-      "step": 17089
-    },
-    {
-      "epoch": 1.1169204627148552,
-      "grad_norm": 0.48212072253227234,
-      "learning_rate": 7.136236481149119e-06,
-      "loss": 0.3946,
-      "step": 17090
-    },
-    {
-      "epoch": 1.1169858179203973,
-      "grad_norm": 0.44342443346977234,
-      "learning_rate": 7.13592076110061e-06,
-      "loss": 0.3605,
-      "step": 17091
-    },
-    {
-      "epoch": 1.1170511731259394,
-      "grad_norm": 0.43573877215385437,
-      "learning_rate": 7.135605030634477e-06,
-      "loss": 0.3141,
-      "step": 17092
-    },
-    {
-      "epoch": 1.1171165283314817,
-      "grad_norm": 0.4341411292552948,
-      "learning_rate": 7.1352892897522564e-06,
-      "loss": 0.338,
-      "step": 17093
-    },
-    {
-      "epoch": 1.1171818835370237,
-      "grad_norm": 0.44523581862449646,
-      "learning_rate": 7.134973538455492e-06,
-      "loss": 0.3067,
-      "step": 17094
-    },
-    {
-      "epoch": 1.1172472387425658,
-      "grad_norm": 0.4501059651374817,
-      "learning_rate": 7.1346577767457225e-06,
-      "loss": 0.3807,
-      "step": 17095
-    },
-    {
-      "epoch": 1.117312593948108,
-      "grad_norm": 0.45240458846092224,
-      "learning_rate": 7.134342004624486e-06,
-      "loss": 0.3493,
-      "step": 17096
-    },
-    {
-      "epoch": 1.1173779491536502,
-      "grad_norm": 0.4435292184352875,
-      "learning_rate": 7.134026222093325e-06,
-      "loss": 0.3455,
-      "step": 17097
-    },
-    {
-      "epoch": 1.1174433043591923,
-      "grad_norm": 0.47913676500320435,
-      "learning_rate": 7.133710429153778e-06,
-      "loss": 0.3685,
-      "step": 17098
-    },
-    {
-      "epoch": 1.1175086595647343,
-      "grad_norm": 0.4573393166065216,
-      "learning_rate": 7.133394625807386e-06,
-      "loss": 0.3274,
-      "step": 17099
-    },
-    {
-      "epoch": 1.1175740147702764,
-      "grad_norm": 0.49698054790496826,
-      "learning_rate": 7.133078812055689e-06,
-      "loss": 0.3822,
-      "step": 17100
-    },
-    {
-      "epoch": 1.1176393699758185,
-      "grad_norm": 0.4798371195793152,
-      "learning_rate": 7.132762987900229e-06,
-      "loss": 0.3962,
-      "step": 17101
-    },
-    {
-      "epoch": 1.1177047251813608,
-      "grad_norm": 0.4546820819377899,
-      "learning_rate": 7.132447153342545e-06,
-      "loss": 0.3626,
-      "step": 17102
-    },
-    {
-      "epoch": 1.1177700803869028,
-      "grad_norm": 0.42338061332702637,
-      "learning_rate": 7.1321313083841755e-06,
-      "loss": 0.3095,
-      "step": 17103
-    },
-    {
-      "epoch": 1.117835435592445,
-      "grad_norm": 0.4631916284561157,
-      "learning_rate": 7.131815453026665e-06,
-      "loss": 0.3072,
-      "step": 17104
-    },
-    {
-      "epoch": 1.117900790797987,
-      "grad_norm": 0.45935285091400146,
-      "learning_rate": 7.131499587271551e-06,
-      "loss": 0.3538,
-      "step": 17105
-    },
-    {
-      "epoch": 1.1179661460035293,
-      "grad_norm": 0.5046459436416626,
-      "learning_rate": 7.131183711120376e-06,
-      "loss": 0.4196,
-      "step": 17106
-    },
-    {
-      "epoch": 1.1180315012090714,
-      "grad_norm": 0.44703638553619385,
-      "learning_rate": 7.13086782457468e-06,
-      "loss": 0.3624,
-      "step": 17107
-    },
-    {
-      "epoch": 1.1180968564146134,
-      "grad_norm": 0.4890722632408142,
-      "learning_rate": 7.130551927636002e-06,
-      "loss": 0.3827,
-      "step": 17108
-    },
-    {
-      "epoch": 1.1181622116201555,
-      "grad_norm": 0.4590572118759155,
-      "learning_rate": 7.130236020305885e-06,
-      "loss": 0.3573,
-      "step": 17109
-    },
-    {
-      "epoch": 1.1182275668256976,
-      "grad_norm": 0.4598666727542877,
-      "learning_rate": 7.1299201025858685e-06,
-      "loss": 0.3667,
-      "step": 17110
-    },
-    {
-      "epoch": 1.1182929220312399,
-      "grad_norm": 0.4720495343208313,
-      "learning_rate": 7.129604174477493e-06,
-      "loss": 0.3939,
-      "step": 17111
-    },
-    {
-      "epoch": 1.118358277236782,
-      "grad_norm": 0.5097851753234863,
-      "learning_rate": 7.129288235982303e-06,
-      "loss": 0.4218,
-      "step": 17112
-    },
-    {
-      "epoch": 1.118423632442324,
-      "grad_norm": 0.4786634147167206,
-      "learning_rate": 7.128972287101835e-06,
-      "loss": 0.3638,
-      "step": 17113
-    },
-    {
-      "epoch": 1.118488987647866,
-      "grad_norm": 0.44928136467933655,
-      "learning_rate": 7.1286563278376306e-06,
-      "loss": 0.3444,
-      "step": 17114
-    },
-    {
-      "epoch": 1.1185543428534084,
-      "grad_norm": 0.44733908772468567,
-      "learning_rate": 7.128340358191234e-06,
-      "loss": 0.3408,
-      "step": 17115
-    },
-    {
-      "epoch": 1.1186196980589505,
-      "grad_norm": 0.4302018880844116,
-      "learning_rate": 7.128024378164181e-06,
-      "loss": 0.3191,
-      "step": 17116
-    },
-    {
-      "epoch": 1.1186850532644925,
-      "grad_norm": 0.48240160942077637,
-      "learning_rate": 7.127708387758019e-06,
-      "loss": 0.3985,
-      "step": 17117
-    },
-    {
-      "epoch": 1.1187504084700346,
-      "grad_norm": 0.4628893733024597,
-      "learning_rate": 7.127392386974283e-06,
-      "loss": 0.3424,
-      "step": 17118
-    },
-    {
-      "epoch": 1.1188157636755767,
-      "grad_norm": 0.47859904170036316,
-      "learning_rate": 7.12707637581452e-06,
-      "loss": 0.3783,
-      "step": 17119
-    },
-    {
-      "epoch": 1.118881118881119,
-      "grad_norm": 0.535497784614563,
-      "learning_rate": 7.126760354280268e-06,
-      "loss": 0.4356,
-      "step": 17120
-    },
-    {
-      "epoch": 1.118946474086661,
-      "grad_norm": 0.47680404782295227,
-      "learning_rate": 7.126444322373069e-06,
-      "loss": 0.3919,
-      "step": 17121
-    },
-    {
-      "epoch": 1.119011829292203,
-      "grad_norm": 0.4746123254299164,
-      "learning_rate": 7.126128280094464e-06,
-      "loss": 0.3583,
-      "step": 17122
-    },
-    {
-      "epoch": 1.1190771844977452,
-      "grad_norm": 0.44779065251350403,
-      "learning_rate": 7.125812227445994e-06,
-      "loss": 0.3289,
-      "step": 17123
-    },
-    {
-      "epoch": 1.1191425397032875,
-      "grad_norm": 0.4273865818977356,
-      "learning_rate": 7.125496164429203e-06,
-      "loss": 0.3279,
-      "step": 17124
-    },
-    {
-      "epoch": 1.1192078949088295,
-      "grad_norm": 0.46545201539993286,
-      "learning_rate": 7.1251800910456295e-06,
-      "loss": 0.3697,
-      "step": 17125
-    },
-    {
-      "epoch": 1.1192732501143716,
-      "grad_norm": 0.43651264905929565,
-      "learning_rate": 7.124864007296818e-06,
-      "loss": 0.3335,
-      "step": 17126
-    },
-    {
-      "epoch": 1.1193386053199137,
-      "grad_norm": 0.439040869474411,
-      "learning_rate": 7.1245479131843065e-06,
-      "loss": 0.3244,
-      "step": 17127
-    },
-    {
-      "epoch": 1.1194039605254558,
-      "grad_norm": 0.46707066893577576,
-      "learning_rate": 7.124231808709642e-06,
-      "loss": 0.3564,
-      "step": 17128
-    },
-    {
-      "epoch": 1.119469315730998,
-      "grad_norm": 0.4685092568397522,
-      "learning_rate": 7.123915693874359e-06,
-      "loss": 0.3397,
-      "step": 17129
-    },
-    {
-      "epoch": 1.1195346709365401,
-      "grad_norm": 0.5096257925033569,
-      "learning_rate": 7.1235995686800065e-06,
-      "loss": 0.414,
-      "step": 17130
-    },
-    {
-      "epoch": 1.1196000261420822,
-      "grad_norm": 0.48026302456855774,
-      "learning_rate": 7.123283433128122e-06,
-      "loss": 0.4028,
-      "step": 17131
-    },
-    {
-      "epoch": 1.1196653813476243,
-      "grad_norm": 0.4236884117126465,
-      "learning_rate": 7.122967287220248e-06,
-      "loss": 0.2925,
-      "step": 17132
-    },
-    {
-      "epoch": 1.1197307365531666,
-      "grad_norm": 0.4994780719280243,
-      "learning_rate": 7.122651130957929e-06,
-      "loss": 0.3995,
-      "step": 17133
-    },
-    {
-      "epoch": 1.1197960917587086,
-      "grad_norm": 0.44404107332229614,
-      "learning_rate": 7.122334964342703e-06,
-      "loss": 0.3332,
-      "step": 17134
-    },
-    {
-      "epoch": 1.1198614469642507,
-      "grad_norm": 0.44905710220336914,
-      "learning_rate": 7.122018787376116e-06,
-      "loss": 0.3577,
-      "step": 17135
-    },
-    {
-      "epoch": 1.1199268021697928,
-      "grad_norm": 0.45197802782058716,
-      "learning_rate": 7.1217026000597066e-06,
-      "loss": 0.3721,
-      "step": 17136
-    },
-    {
-      "epoch": 1.1199921573753349,
-      "grad_norm": 0.4590807259082794,
-      "learning_rate": 7.1213864023950195e-06,
-      "loss": 0.3588,
-      "step": 17137
-    },
-    {
-      "epoch": 1.1200575125808772,
-      "grad_norm": 0.43286678194999695,
-      "learning_rate": 7.1210701943835945e-06,
-      "loss": 0.3369,
-      "step": 17138
-    },
-    {
-      "epoch": 1.1201228677864192,
-      "grad_norm": 0.4448758065700531,
-      "learning_rate": 7.1207539760269776e-06,
-      "loss": 0.3432,
-      "step": 17139
-    },
-    {
-      "epoch": 1.1201882229919613,
-      "grad_norm": 0.47225460410118103,
-      "learning_rate": 7.1204377473267085e-06,
-      "loss": 0.3713,
-      "step": 17140
-    },
-    {
-      "epoch": 1.1202535781975034,
-      "grad_norm": 0.46399086713790894,
-      "learning_rate": 7.120121508284329e-06,
-      "loss": 0.3534,
-      "step": 17141
-    },
-    {
-      "epoch": 1.1203189334030457,
-      "grad_norm": 0.4282212555408478,
-      "learning_rate": 7.119805258901382e-06,
-      "loss": 0.3155,
-      "step": 17142
-    },
-    {
-      "epoch": 1.1203842886085877,
-      "grad_norm": 0.4380335509777069,
-      "learning_rate": 7.1194889991794115e-06,
-      "loss": 0.3339,
-      "step": 17143
-    },
-    {
-      "epoch": 1.1204496438141298,
-      "grad_norm": 0.41226276755332947,
-      "learning_rate": 7.1191727291199585e-06,
-      "loss": 0.2779,
-      "step": 17144
-    },
-    {
-      "epoch": 1.1205149990196719,
-      "grad_norm": 0.4880550503730774,
-      "learning_rate": 7.118856448724565e-06,
-      "loss": 0.3772,
-      "step": 17145
-    },
-    {
-      "epoch": 1.120580354225214,
-      "grad_norm": 0.4138718545436859,
-      "learning_rate": 7.118540157994775e-06,
-      "loss": 0.3123,
-      "step": 17146
-    },
-    {
-      "epoch": 1.1206457094307563,
-      "grad_norm": 0.4755130410194397,
-      "learning_rate": 7.118223856932132e-06,
-      "loss": 0.3422,
-      "step": 17147
-    },
-    {
-      "epoch": 1.1207110646362983,
-      "grad_norm": 0.475567489862442,
-      "learning_rate": 7.117907545538177e-06,
-      "loss": 0.3583,
-      "step": 17148
-    },
-    {
-      "epoch": 1.1207764198418404,
-      "grad_norm": 0.44259950518608093,
-      "learning_rate": 7.117591223814453e-06,
-      "loss": 0.3487,
-      "step": 17149
-    },
-    {
-      "epoch": 1.1208417750473825,
-      "grad_norm": 0.4870612621307373,
-      "learning_rate": 7.117274891762503e-06,
-      "loss": 0.3889,
-      "step": 17150
-    },
-    {
-      "epoch": 1.1209071302529245,
-      "grad_norm": 0.4553016722202301,
-      "learning_rate": 7.116958549383869e-06,
-      "loss": 0.3538,
-      "step": 17151
-    },
-    {
-      "epoch": 1.1209724854584668,
-      "grad_norm": 0.4354605972766876,
-      "learning_rate": 7.116642196680095e-06,
-      "loss": 0.3306,
-      "step": 17152
-    },
-    {
-      "epoch": 1.121037840664009,
-      "grad_norm": 0.43308666348457336,
-      "learning_rate": 7.116325833652726e-06,
-      "loss": 0.3248,
-      "step": 17153
-    },
-    {
-      "epoch": 1.121103195869551,
-      "grad_norm": 0.4298359453678131,
-      "learning_rate": 7.116009460303301e-06,
-      "loss": 0.3055,
-      "step": 17154
-    },
-    {
-      "epoch": 1.121168551075093,
-      "grad_norm": 0.45559221506118774,
-      "learning_rate": 7.115693076633364e-06,
-      "loss": 0.354,
-      "step": 17155
-    },
-    {
-      "epoch": 1.1212339062806354,
-      "grad_norm": 0.49907585978507996,
-      "learning_rate": 7.1153766826444595e-06,
-      "loss": 0.4032,
-      "step": 17156
-    },
-    {
-      "epoch": 1.1212992614861774,
-      "grad_norm": 0.43977731466293335,
-      "learning_rate": 7.1150602783381304e-06,
-      "loss": 0.3404,
-      "step": 17157
-    },
-    {
-      "epoch": 1.1213646166917195,
-      "grad_norm": 0.43656906485557556,
-      "learning_rate": 7.114743863715918e-06,
-      "loss": 0.3361,
-      "step": 17158
-    },
-    {
-      "epoch": 1.1214299718972616,
-      "grad_norm": 0.4312306344509125,
-      "learning_rate": 7.11442743877937e-06,
-      "loss": 0.3148,
-      "step": 17159
-    },
-    {
-      "epoch": 1.1214953271028036,
-      "grad_norm": 0.49295681715011597,
-      "learning_rate": 7.114111003530025e-06,
-      "loss": 0.3862,
-      "step": 17160
-    },
-    {
-      "epoch": 1.121560682308346,
-      "grad_norm": 0.45185586810112,
-      "learning_rate": 7.113794557969429e-06,
-      "loss": 0.3599,
-      "step": 17161
-    },
-    {
-      "epoch": 1.121626037513888,
-      "grad_norm": 0.46818238496780396,
-      "learning_rate": 7.113478102099124e-06,
-      "loss": 0.3434,
-      "step": 17162
-    },
-    {
-      "epoch": 1.12169139271943,
-      "grad_norm": 0.4316209852695465,
-      "learning_rate": 7.113161635920654e-06,
-      "loss": 0.3291,
-      "step": 17163
-    },
-    {
-      "epoch": 1.1217567479249722,
-      "grad_norm": 0.44057488441467285,
-      "learning_rate": 7.112845159435564e-06,
-      "loss": 0.3314,
-      "step": 17164
-    },
-    {
-      "epoch": 1.1218221031305144,
-      "grad_norm": 0.4619652032852173,
-      "learning_rate": 7.112528672645395e-06,
-      "loss": 0.3739,
-      "step": 17165
-    },
-    {
-      "epoch": 1.1218874583360565,
-      "grad_norm": 0.4332674741744995,
-      "learning_rate": 7.112212175551691e-06,
-      "loss": 0.3151,
-      "step": 17166
-    },
-    {
-      "epoch": 1.1219528135415986,
-      "grad_norm": 0.40701258182525635,
-      "learning_rate": 7.111895668155999e-06,
-      "loss": 0.2896,
-      "step": 17167
-    },
-    {
-      "epoch": 1.1220181687471407,
-      "grad_norm": 0.4592593312263489,
-      "learning_rate": 7.111579150459857e-06,
-      "loss": 0.3421,
-      "step": 17168
-    },
-    {
-      "epoch": 1.1220835239526827,
-      "grad_norm": 0.45556318759918213,
-      "learning_rate": 7.111262622464815e-06,
-      "loss": 0.3604,
-      "step": 17169
-    },
-    {
-      "epoch": 1.122148879158225,
-      "grad_norm": 0.4613608121871948,
-      "learning_rate": 7.110946084172413e-06,
-      "loss": 0.3299,
-      "step": 17170
-    },
-    {
-      "epoch": 1.122214234363767,
-      "grad_norm": 0.48709607124328613,
-      "learning_rate": 7.1106295355841955e-06,
-      "loss": 0.3556,
-      "step": 17171
-    },
-    {
-      "epoch": 1.1222795895693092,
-      "grad_norm": 0.48541393876075745,
-      "learning_rate": 7.110312976701706e-06,
-      "loss": 0.3564,
-      "step": 17172
-    },
-    {
-      "epoch": 1.1223449447748512,
-      "grad_norm": 0.4681394696235657,
-      "learning_rate": 7.109996407526489e-06,
-      "loss": 0.3664,
-      "step": 17173
-    },
-    {
-      "epoch": 1.1224102999803933,
-      "grad_norm": 0.46374887228012085,
-      "learning_rate": 7.1096798280600885e-06,
-      "loss": 0.3517,
-      "step": 17174
-    },
-    {
-      "epoch": 1.1224756551859356,
-      "grad_norm": 0.4465112090110779,
-      "learning_rate": 7.10936323830405e-06,
-      "loss": 0.3474,
-      "step": 17175
-    },
-    {
-      "epoch": 1.1225410103914777,
-      "grad_norm": 0.46198412775993347,
-      "learning_rate": 7.109046638259913e-06,
-      "loss": 0.3666,
-      "step": 17176
-    },
-    {
-      "epoch": 1.1226063655970198,
-      "grad_norm": 0.4331059455871582,
-      "learning_rate": 7.108730027929228e-06,
-      "loss": 0.3321,
-      "step": 17177
-    },
-    {
-      "epoch": 1.1226717208025618,
-      "grad_norm": 0.4709119200706482,
-      "learning_rate": 7.108413407313535e-06,
-      "loss": 0.3912,
-      "step": 17178
-    },
-    {
-      "epoch": 1.1227370760081041,
-      "grad_norm": 0.44842803478240967,
-      "learning_rate": 7.1080967764143795e-06,
-      "loss": 0.3671,
-      "step": 17179
-    },
-    {
-      "epoch": 1.1228024312136462,
-      "grad_norm": 0.435075044631958,
-      "learning_rate": 7.107780135233306e-06,
-      "loss": 0.3467,
-      "step": 17180
-    },
-    {
-      "epoch": 1.1228677864191883,
-      "grad_norm": 0.4641771912574768,
-      "learning_rate": 7.107463483771858e-06,
-      "loss": 0.3622,
-      "step": 17181
-    },
-    {
-      "epoch": 1.1229331416247303,
-      "grad_norm": 0.44011834263801575,
-      "learning_rate": 7.107146822031581e-06,
-      "loss": 0.3295,
-      "step": 17182
-    },
-    {
-      "epoch": 1.1229984968302724,
-      "grad_norm": 0.4464653730392456,
-      "learning_rate": 7.106830150014019e-06,
-      "loss": 0.3349,
-      "step": 17183
-    },
-    {
-      "epoch": 1.1230638520358147,
-      "grad_norm": 0.4518473446369171,
-      "learning_rate": 7.106513467720717e-06,
-      "loss": 0.3593,
-      "step": 17184
-    },
-    {
-      "epoch": 1.1231292072413568,
-      "grad_norm": 0.46642863750457764,
-      "learning_rate": 7.106196775153217e-06,
-      "loss": 0.3619,
-      "step": 17185
-    },
-    {
-      "epoch": 1.1231945624468989,
-      "grad_norm": 0.42322593927383423,
-      "learning_rate": 7.105880072313067e-06,
-      "loss": 0.3039,
-      "step": 17186
-    },
-    {
-      "epoch": 1.123259917652441,
-      "grad_norm": 0.475157231092453,
-      "learning_rate": 7.1055633592018115e-06,
-      "loss": 0.3536,
-      "step": 17187
-    },
-    {
-      "epoch": 1.1233252728579832,
-      "grad_norm": 0.4709039330482483,
-      "learning_rate": 7.105246635820993e-06,
-      "loss": 0.3579,
-      "step": 17188
-    },
-    {
-      "epoch": 1.1233906280635253,
-      "grad_norm": 0.44005364179611206,
-      "learning_rate": 7.1049299021721575e-06,
-      "loss": 0.3304,
-      "step": 17189
-    },
-    {
-      "epoch": 1.1234559832690674,
-      "grad_norm": 0.44961556792259216,
-      "learning_rate": 7.104613158256848e-06,
-      "loss": 0.3742,
-      "step": 17190
-    },
-    {
-      "epoch": 1.1235213384746094,
-      "grad_norm": 0.4559229612350464,
-      "learning_rate": 7.104296404076614e-06,
-      "loss": 0.319,
-      "step": 17191
-    },
-    {
-      "epoch": 1.1235866936801515,
-      "grad_norm": 0.4463317096233368,
-      "learning_rate": 7.103979639632995e-06,
-      "loss": 0.3539,
-      "step": 17192
-    },
-    {
-      "epoch": 1.1236520488856938,
-      "grad_norm": 0.4417072534561157,
-      "learning_rate": 7.1036628649275394e-06,
-      "loss": 0.3378,
-      "step": 17193
-    },
-    {
-      "epoch": 1.1237174040912359,
-      "grad_norm": 0.4529339373111725,
-      "learning_rate": 7.10334607996179e-06,
-      "loss": 0.3673,
-      "step": 17194
-    },
-    {
-      "epoch": 1.123782759296778,
-      "grad_norm": 0.4649065136909485,
-      "learning_rate": 7.103029284737295e-06,
-      "loss": 0.3341,
-      "step": 17195
-    },
-    {
-      "epoch": 1.12384811450232,
-      "grad_norm": 0.44754651188850403,
-      "learning_rate": 7.102712479255597e-06,
-      "loss": 0.3473,
-      "step": 17196
-    },
-    {
-      "epoch": 1.1239134697078623,
-      "grad_norm": 0.48458895087242126,
-      "learning_rate": 7.102395663518241e-06,
-      "loss": 0.3475,
-      "step": 17197
-    },
-    {
-      "epoch": 1.1239788249134044,
-      "grad_norm": 0.437276154756546,
-      "learning_rate": 7.102078837526774e-06,
-      "loss": 0.3279,
-      "step": 17198
-    },
-    {
-      "epoch": 1.1240441801189465,
-      "grad_norm": 0.43904247879981995,
-      "learning_rate": 7.10176200128274e-06,
-      "loss": 0.349,
-      "step": 17199
-    },
-    {
-      "epoch": 1.1241095353244885,
-      "grad_norm": 0.4686586260795593,
-      "learning_rate": 7.101445154787685e-06,
-      "loss": 0.3532,
-      "step": 17200
-    },
-    {
-      "epoch": 1.1241748905300306,
-      "grad_norm": 0.46144312620162964,
-      "learning_rate": 7.1011282980431525e-06,
-      "loss": 0.3368,
-      "step": 17201
-    },
-    {
-      "epoch": 1.124240245735573,
-      "grad_norm": 0.4716927707195282,
-      "learning_rate": 7.10081143105069e-06,
-      "loss": 0.3552,
-      "step": 17202
-    },
-    {
-      "epoch": 1.124305600941115,
-      "grad_norm": 0.4556477963924408,
-      "learning_rate": 7.100494553811843e-06,
-      "loss": 0.3556,
-      "step": 17203
-    },
-    {
-      "epoch": 1.124370956146657,
-      "grad_norm": 0.4434954524040222,
-      "learning_rate": 7.100177666328156e-06,
-      "loss": 0.3342,
-      "step": 17204
-    },
-    {
-      "epoch": 1.1244363113521991,
-      "grad_norm": 0.48426884412765503,
-      "learning_rate": 7.099860768601174e-06,
-      "loss": 0.3437,
-      "step": 17205
-    },
-    {
-      "epoch": 1.1245016665577414,
-      "grad_norm": 0.4554997384548187,
-      "learning_rate": 7.0995438606324454e-06,
-      "loss": 0.3594,
-      "step": 17206
-    },
-    {
-      "epoch": 1.1245670217632835,
-      "grad_norm": 0.4420515298843384,
-      "learning_rate": 7.099226942423514e-06,
-      "loss": 0.3036,
-      "step": 17207
-    },
-    {
-      "epoch": 1.1246323769688256,
-      "grad_norm": 0.4349117875099182,
-      "learning_rate": 7.098910013975924e-06,
-      "loss": 0.3294,
-      "step": 17208
-    },
-    {
-      "epoch": 1.1246977321743676,
-      "grad_norm": 0.41185781359672546,
-      "learning_rate": 7.098593075291225e-06,
-      "loss": 0.3016,
-      "step": 17209
-    },
-    {
-      "epoch": 1.1247630873799097,
-      "grad_norm": 0.44127434492111206,
-      "learning_rate": 7.0982761263709575e-06,
-      "loss": 0.3238,
-      "step": 17210
-    },
-    {
-      "epoch": 1.124828442585452,
-      "grad_norm": 0.4465019106864929,
-      "learning_rate": 7.097959167216672e-06,
-      "loss": 0.3065,
-      "step": 17211
-    },
-    {
-      "epoch": 1.124893797790994,
-      "grad_norm": 0.448458194732666,
-      "learning_rate": 7.097642197829914e-06,
-      "loss": 0.333,
-      "step": 17212
-    },
-    {
-      "epoch": 1.1249591529965361,
-      "grad_norm": 0.40496647357940674,
-      "learning_rate": 7.097325218212228e-06,
-      "loss": 0.2952,
-      "step": 17213
-    },
-    {
-      "epoch": 1.1250245082020782,
-      "grad_norm": 0.4957025349140167,
-      "learning_rate": 7.097008228365157e-06,
-      "loss": 0.4294,
-      "step": 17214
-    },
-    {
-      "epoch": 1.1250898634076205,
-      "grad_norm": 0.41872408986091614,
-      "learning_rate": 7.096691228290255e-06,
-      "loss": 0.3298,
-      "step": 17215
-    },
-    {
-      "epoch": 1.1251552186131626,
-      "grad_norm": 0.41255950927734375,
-      "learning_rate": 7.0963742179890595e-06,
-      "loss": 0.2982,
-      "step": 17216
-    },
-    {
-      "epoch": 1.1252205738187047,
-      "grad_norm": 0.4755578637123108,
-      "learning_rate": 7.096057197463123e-06,
-      "loss": 0.3683,
-      "step": 17217
-    },
-    {
-      "epoch": 1.1252859290242467,
-      "grad_norm": 0.46129918098449707,
-      "learning_rate": 7.095740166713989e-06,
-      "loss": 0.3584,
-      "step": 17218
-    },
-    {
-      "epoch": 1.1253512842297888,
-      "grad_norm": 0.4215393364429474,
-      "learning_rate": 7.0954231257432034e-06,
-      "loss": 0.3159,
-      "step": 17219
-    },
-    {
-      "epoch": 1.125416639435331,
-      "grad_norm": 0.4933900833129883,
-      "learning_rate": 7.095106074552316e-06,
-      "loss": 0.3759,
-      "step": 17220
-    },
-    {
-      "epoch": 1.1254819946408732,
-      "grad_norm": 0.4249468147754669,
-      "learning_rate": 7.0947890131428674e-06,
-      "loss": 0.2888,
-      "step": 17221
-    },
-    {
-      "epoch": 1.1255473498464152,
-      "grad_norm": 0.46237921714782715,
-      "learning_rate": 7.094471941516409e-06,
-      "loss": 0.377,
-      "step": 17222
-    },
-    {
-      "epoch": 1.1256127050519573,
-      "grad_norm": 0.4628616273403168,
-      "learning_rate": 7.0941548596744835e-06,
-      "loss": 0.3554,
-      "step": 17223
-    },
-    {
-      "epoch": 1.1256780602574996,
-      "grad_norm": 0.4421902894973755,
-      "learning_rate": 7.093837767618641e-06,
-      "loss": 0.3316,
-      "step": 17224
-    },
-    {
-      "epoch": 1.1257434154630417,
-      "grad_norm": 0.44672462344169617,
-      "learning_rate": 7.093520665350425e-06,
-      "loss": 0.3217,
-      "step": 17225
-    },
-    {
-      "epoch": 1.1258087706685838,
-      "grad_norm": 0.4612988531589508,
-      "learning_rate": 7.093203552871384e-06,
-      "loss": 0.3672,
-      "step": 17226
-    },
-    {
-      "epoch": 1.1258741258741258,
-      "grad_norm": 0.47889474034309387,
-      "learning_rate": 7.0928864301830646e-06,
-      "loss": 0.3712,
-      "step": 17227
-    },
-    {
-      "epoch": 1.125939481079668,
-      "grad_norm": 0.5105122923851013,
-      "learning_rate": 7.092569297287012e-06,
-      "loss": 0.3045,
-      "step": 17228
-    },
-    {
-      "epoch": 1.1260048362852102,
-      "grad_norm": 0.4600527882575989,
-      "learning_rate": 7.092252154184776e-06,
-      "loss": 0.3523,
-      "step": 17229
-    },
-    {
-      "epoch": 1.1260701914907523,
-      "grad_norm": 0.43037718534469604,
-      "learning_rate": 7.0919350008778985e-06,
-      "loss": 0.3056,
-      "step": 17230
-    },
-    {
-      "epoch": 1.1261355466962943,
-      "grad_norm": 0.45579734444618225,
-      "learning_rate": 7.091617837367931e-06,
-      "loss": 0.3468,
-      "step": 17231
-    },
-    {
-      "epoch": 1.1262009019018364,
-      "grad_norm": 0.455057829618454,
-      "learning_rate": 7.09130066365642e-06,
-      "loss": 0.3353,
-      "step": 17232
-    },
-    {
-      "epoch": 1.1262662571073787,
-      "grad_norm": 0.46316277980804443,
-      "learning_rate": 7.09098347974491e-06,
-      "loss": 0.3341,
-      "step": 17233
-    },
-    {
-      "epoch": 1.1263316123129208,
-      "grad_norm": 0.467692494392395,
-      "learning_rate": 7.090666285634947e-06,
-      "loss": 0.3758,
-      "step": 17234
-    },
-    {
-      "epoch": 1.1263969675184629,
-      "grad_norm": 0.4547955393791199,
-      "learning_rate": 7.090349081328083e-06,
-      "loss": 0.3544,
-      "step": 17235
-    },
-    {
-      "epoch": 1.126462322724005,
-      "grad_norm": 0.4579767882823944,
-      "learning_rate": 7.0900318668258614e-06,
-      "loss": 0.3571,
-      "step": 17236
-    },
-    {
-      "epoch": 1.126527677929547,
-      "grad_norm": 0.46709659695625305,
-      "learning_rate": 7.08971464212983e-06,
-      "loss": 0.3283,
-      "step": 17237
-    },
-    {
-      "epoch": 1.1265930331350893,
-      "grad_norm": 0.4610109031200409,
-      "learning_rate": 7.089397407241537e-06,
-      "loss": 0.3211,
-      "step": 17238
-    },
-    {
-      "epoch": 1.1266583883406314,
-      "grad_norm": 0.4366426169872284,
-      "learning_rate": 7.089080162162528e-06,
-      "loss": 0.3692,
-      "step": 17239
-    },
-    {
-      "epoch": 1.1267237435461734,
-      "grad_norm": 0.4222938120365143,
-      "learning_rate": 7.088762906894353e-06,
-      "loss": 0.2909,
-      "step": 17240
-    },
-    {
-      "epoch": 1.1267890987517155,
-      "grad_norm": 0.436970591545105,
-      "learning_rate": 7.088445641438556e-06,
-      "loss": 0.3162,
-      "step": 17241
-    },
-    {
-      "epoch": 1.1268544539572578,
-      "grad_norm": 0.4476236402988434,
-      "learning_rate": 7.088128365796687e-06,
-      "loss": 0.3716,
-      "step": 17242
-    },
-    {
-      "epoch": 1.1269198091627999,
-      "grad_norm": 0.4152679443359375,
-      "learning_rate": 7.0878110799702925e-06,
-      "loss": 0.2782,
-      "step": 17243
-    },
-    {
-      "epoch": 1.126985164368342,
-      "grad_norm": 0.4600754976272583,
-      "learning_rate": 7.08749378396092e-06,
-      "loss": 0.3583,
-      "step": 17244
-    },
-    {
-      "epoch": 1.127050519573884,
-      "grad_norm": 0.44209912419319153,
-      "learning_rate": 7.087176477770118e-06,
-      "loss": 0.3286,
-      "step": 17245
-    },
-    {
-      "epoch": 1.127115874779426,
-      "grad_norm": 0.46539053320884705,
-      "learning_rate": 7.0868591613994305e-06,
-      "loss": 0.3895,
-      "step": 17246
-    },
-    {
-      "epoch": 1.1271812299849684,
-      "grad_norm": 0.42643725872039795,
-      "learning_rate": 7.0865418348504115e-06,
-      "loss": 0.3279,
-      "step": 17247
-    },
-    {
-      "epoch": 1.1272465851905105,
-      "grad_norm": 0.4708240330219269,
-      "learning_rate": 7.086224498124603e-06,
-      "loss": 0.3709,
-      "step": 17248
-    },
-    {
-      "epoch": 1.1273119403960525,
-      "grad_norm": 0.4962260127067566,
-      "learning_rate": 7.085907151223555e-06,
-      "loss": 0.3836,
-      "step": 17249
-    },
-    {
-      "epoch": 1.1273772956015946,
-      "grad_norm": 0.4395652711391449,
-      "learning_rate": 7.085589794148815e-06,
-      "loss": 0.3448,
-      "step": 17250
-    },
-    {
-      "epoch": 1.127442650807137,
-      "grad_norm": 0.4021154046058655,
-      "learning_rate": 7.085272426901932e-06,
-      "loss": 0.2926,
-      "step": 17251
-    },
-    {
-      "epoch": 1.127508006012679,
-      "grad_norm": 0.4648852050304413,
-      "learning_rate": 7.0849550494844535e-06,
-      "loss": 0.3571,
-      "step": 17252
-    },
-    {
-      "epoch": 1.127573361218221,
-      "grad_norm": 0.4426382780075073,
-      "learning_rate": 7.084637661897926e-06,
-      "loss": 0.3423,
-      "step": 17253
-    },
-    {
-      "epoch": 1.1276387164237631,
-      "grad_norm": 0.4879113733768463,
-      "learning_rate": 7.084320264143898e-06,
-      "loss": 0.404,
-      "step": 17254
-    },
-    {
-      "epoch": 1.1277040716293052,
-      "grad_norm": 0.4888441562652588,
-      "learning_rate": 7.084002856223918e-06,
-      "loss": 0.3869,
-      "step": 17255
-    },
-    {
-      "epoch": 1.1277694268348475,
-      "grad_norm": 0.4416881799697876,
-      "learning_rate": 7.0836854381395355e-06,
-      "loss": 0.3111,
-      "step": 17256
-    },
-    {
-      "epoch": 1.1278347820403896,
-      "grad_norm": 0.4856817126274109,
-      "learning_rate": 7.083368009892295e-06,
-      "loss": 0.3985,
-      "step": 17257
-    },
-    {
-      "epoch": 1.1279001372459316,
-      "grad_norm": 0.459825724363327,
-      "learning_rate": 7.083050571483749e-06,
-      "loss": 0.3739,
-      "step": 17258
-    },
-    {
-      "epoch": 1.1279654924514737,
-      "grad_norm": 0.45025455951690674,
-      "learning_rate": 7.082733122915443e-06,
-      "loss": 0.3269,
-      "step": 17259
-    },
-    {
-      "epoch": 1.128030847657016,
-      "grad_norm": 0.42813920974731445,
-      "learning_rate": 7.082415664188927e-06,
-      "loss": 0.3193,
-      "step": 17260
-    },
-    {
-      "epoch": 1.128096202862558,
-      "grad_norm": 0.4929443597793579,
-      "learning_rate": 7.082098195305746e-06,
-      "loss": 0.3561,
-      "step": 17261
-    },
-    {
-      "epoch": 1.1281615580681001,
-      "grad_norm": 0.4523185193538666,
-      "learning_rate": 7.081780716267452e-06,
-      "loss": 0.3523,
-      "step": 17262
-    },
-    {
-      "epoch": 1.1282269132736422,
-      "grad_norm": 0.4614277184009552,
-      "learning_rate": 7.081463227075593e-06,
-      "loss": 0.3751,
-      "step": 17263
-    },
-    {
-      "epoch": 1.1282922684791843,
-      "grad_norm": 0.42780762910842896,
-      "learning_rate": 7.0811457277317156e-06,
-      "loss": 0.3102,
-      "step": 17264
-    },
-    {
-      "epoch": 1.1283576236847266,
-      "grad_norm": 0.44840002059936523,
-      "learning_rate": 7.08082821823737e-06,
-      "loss": 0.3278,
-      "step": 17265
-    },
-    {
-      "epoch": 1.1284229788902687,
-      "grad_norm": 0.4721134305000305,
-      "learning_rate": 7.0805106985941045e-06,
-      "loss": 0.3775,
-      "step": 17266
-    },
-    {
-      "epoch": 1.1284883340958107,
-      "grad_norm": 0.44914865493774414,
-      "learning_rate": 7.080193168803468e-06,
-      "loss": 0.3528,
-      "step": 17267
-    },
-    {
-      "epoch": 1.1285536893013528,
-      "grad_norm": 0.44426921010017395,
-      "learning_rate": 7.0798756288670075e-06,
-      "loss": 0.3063,
-      "step": 17268
-    },
-    {
-      "epoch": 1.128619044506895,
-      "grad_norm": 0.4684821665287018,
-      "learning_rate": 7.079558078786274e-06,
-      "loss": 0.3765,
-      "step": 17269
-    },
-    {
-      "epoch": 1.1286843997124372,
-      "grad_norm": 0.4223394989967346,
-      "learning_rate": 7.079240518562814e-06,
-      "loss": 0.3101,
-      "step": 17270
-    },
-    {
-      "epoch": 1.1287497549179792,
-      "grad_norm": 0.47070014476776123,
-      "learning_rate": 7.078922948198179e-06,
-      "loss": 0.4017,
-      "step": 17271
-    },
-    {
-      "epoch": 1.1288151101235213,
-      "grad_norm": 0.4264320731163025,
-      "learning_rate": 7.078605367693917e-06,
-      "loss": 0.3007,
-      "step": 17272
-    },
-    {
-      "epoch": 1.1288804653290634,
-      "grad_norm": 0.40820813179016113,
-      "learning_rate": 7.0782877770515755e-06,
-      "loss": 0.2684,
-      "step": 17273
-    },
-    {
-      "epoch": 1.1289458205346055,
-      "grad_norm": 0.42854416370391846,
-      "learning_rate": 7.077970176272706e-06,
-      "loss": 0.3291,
-      "step": 17274
-    },
-    {
-      "epoch": 1.1290111757401478,
-      "grad_norm": 0.4528523087501526,
-      "learning_rate": 7.077652565358855e-06,
-      "loss": 0.3345,
-      "step": 17275
-    },
-    {
-      "epoch": 1.1290765309456898,
-      "grad_norm": 0.5198642015457153,
-      "learning_rate": 7.077334944311572e-06,
-      "loss": 0.3885,
-      "step": 17276
-    },
-    {
-      "epoch": 1.129141886151232,
-      "grad_norm": 0.4706581234931946,
-      "learning_rate": 7.077017313132407e-06,
-      "loss": 0.3607,
-      "step": 17277
-    },
-    {
-      "epoch": 1.129207241356774,
-      "grad_norm": 0.4525388479232788,
-      "learning_rate": 7.07669967182291e-06,
-      "loss": 0.3389,
-      "step": 17278
-    },
-    {
-      "epoch": 1.1292725965623163,
-      "grad_norm": 0.4727247655391693,
-      "learning_rate": 7.076382020384628e-06,
-      "loss": 0.3645,
-      "step": 17279
-    },
-    {
-      "epoch": 1.1293379517678583,
-      "grad_norm": 0.483717679977417,
-      "learning_rate": 7.076064358819113e-06,
-      "loss": 0.3592,
-      "step": 17280
-    },
-    {
-      "epoch": 1.1294033069734004,
-      "grad_norm": 0.49619531631469727,
-      "learning_rate": 7.075746687127912e-06,
-      "loss": 0.391,
-      "step": 17281
-    },
-    {
-      "epoch": 1.1294686621789425,
-      "grad_norm": 0.4986744523048401,
-      "learning_rate": 7.075429005312575e-06,
-      "loss": 0.4127,
-      "step": 17282
-    },
-    {
-      "epoch": 1.1295340173844846,
-      "grad_norm": 0.4431566894054413,
-      "learning_rate": 7.075111313374653e-06,
-      "loss": 0.3448,
-      "step": 17283
-    },
-    {
-      "epoch": 1.1295993725900269,
-      "grad_norm": 0.45887160301208496,
-      "learning_rate": 7.074793611315692e-06,
-      "loss": 0.3302,
-      "step": 17284
-    },
-    {
-      "epoch": 1.129664727795569,
-      "grad_norm": 0.7908359169960022,
-      "learning_rate": 7.074475899137246e-06,
-      "loss": 0.3337,
-      "step": 17285
-    },
-    {
-      "epoch": 1.129730083001111,
-      "grad_norm": 0.46458959579467773,
-      "learning_rate": 7.074158176840862e-06,
-      "loss": 0.3844,
-      "step": 17286
-    },
-    {
-      "epoch": 1.129795438206653,
-      "grad_norm": 0.4635846018791199,
-      "learning_rate": 7.073840444428089e-06,
-      "loss": 0.3298,
-      "step": 17287
-    },
-    {
-      "epoch": 1.1298607934121954,
-      "grad_norm": 0.4312458038330078,
-      "learning_rate": 7.073522701900478e-06,
-      "loss": 0.3172,
-      "step": 17288
-    },
-    {
-      "epoch": 1.1299261486177374,
-      "grad_norm": 0.4425434172153473,
-      "learning_rate": 7.073204949259579e-06,
-      "loss": 0.3209,
-      "step": 17289
-    },
-    {
-      "epoch": 1.1299915038232795,
-      "grad_norm": 0.47515788674354553,
-      "learning_rate": 7.072887186506941e-06,
-      "loss": 0.3655,
-      "step": 17290
-    },
-    {
-      "epoch": 1.1300568590288216,
-      "grad_norm": 0.4508166015148163,
-      "learning_rate": 7.072569413644113e-06,
-      "loss": 0.3149,
-      "step": 17291
-    },
-    {
-      "epoch": 1.1301222142343637,
-      "grad_norm": 0.4490228295326233,
-      "learning_rate": 7.0722516306726485e-06,
-      "loss": 0.3145,
-      "step": 17292
-    },
-    {
-      "epoch": 1.130187569439906,
-      "grad_norm": 0.4573231637477875,
-      "learning_rate": 7.071933837594093e-06,
-      "loss": 0.3391,
-      "step": 17293
-    },
-    {
-      "epoch": 1.130252924645448,
-      "grad_norm": 0.4532839357852936,
-      "learning_rate": 7.071616034409998e-06,
-      "loss": 0.3115,
-      "step": 17294
-    },
-    {
-      "epoch": 1.13031827985099,
-      "grad_norm": 0.4641956090927124,
-      "learning_rate": 7.071298221121916e-06,
-      "loss": 0.3409,
-      "step": 17295
-    },
-    {
-      "epoch": 1.1303836350565322,
-      "grad_norm": 0.47702524065971375,
-      "learning_rate": 7.070980397731394e-06,
-      "loss": 0.373,
-      "step": 17296
-    },
-    {
-      "epoch": 1.1304489902620745,
-      "grad_norm": 0.46357297897338867,
-      "learning_rate": 7.070662564239983e-06,
-      "loss": 0.3628,
-      "step": 17297
-    },
-    {
-      "epoch": 1.1305143454676165,
-      "grad_norm": 0.4431051015853882,
-      "learning_rate": 7.070344720649234e-06,
-      "loss": 0.3188,
-      "step": 17298
-    },
-    {
-      "epoch": 1.1305797006731586,
-      "grad_norm": 0.475641667842865,
-      "learning_rate": 7.0700268669606945e-06,
-      "loss": 0.3754,
-      "step": 17299
-    },
-    {
-      "epoch": 1.1306450558787007,
-      "grad_norm": 0.45539844036102295,
-      "learning_rate": 7.0697090031759186e-06,
-      "loss": 0.3549,
-      "step": 17300
-    },
-    {
-      "epoch": 1.1307104110842428,
-      "grad_norm": 0.454843133687973,
-      "learning_rate": 7.069391129296455e-06,
-      "loss": 0.3482,
-      "step": 17301
-    },
-    {
-      "epoch": 1.130775766289785,
-      "grad_norm": 0.4210955500602722,
-      "learning_rate": 7.069073245323852e-06,
-      "loss": 0.3487,
-      "step": 17302
-    },
-    {
-      "epoch": 1.1308411214953271,
-      "grad_norm": 0.45039451122283936,
-      "learning_rate": 7.068755351259664e-06,
-      "loss": 0.3586,
-      "step": 17303
-    },
-    {
-      "epoch": 1.1309064767008692,
-      "grad_norm": 0.45915690064430237,
-      "learning_rate": 7.068437447105439e-06,
-      "loss": 0.3412,
-      "step": 17304
-    },
-    {
-      "epoch": 1.1309718319064113,
-      "grad_norm": 0.4373253881931305,
-      "learning_rate": 7.068119532862727e-06,
-      "loss": 0.3269,
-      "step": 17305
-    },
-    {
-      "epoch": 1.1310371871119536,
-      "grad_norm": 0.4582827389240265,
-      "learning_rate": 7.06780160853308e-06,
-      "loss": 0.3389,
-      "step": 17306
-    },
-    {
-      "epoch": 1.1311025423174956,
-      "grad_norm": 0.4619799554347992,
-      "learning_rate": 7.06748367411805e-06,
-      "loss": 0.3718,
-      "step": 17307
-    },
-    {
-      "epoch": 1.1311678975230377,
-      "grad_norm": 0.43546101450920105,
-      "learning_rate": 7.067165729619183e-06,
-      "loss": 0.3282,
-      "step": 17308
-    },
-    {
-      "epoch": 1.1312332527285798,
-      "grad_norm": 0.43418750166893005,
-      "learning_rate": 7.066847775038035e-06,
-      "loss": 0.3286,
-      "step": 17309
-    },
-    {
-      "epoch": 1.1312986079341218,
-      "grad_norm": 0.4487304389476776,
-      "learning_rate": 7.0665298103761525e-06,
-      "loss": 0.3483,
-      "step": 17310
-    },
-    {
-      "epoch": 1.1313639631396641,
-      "grad_norm": 0.45989903807640076,
-      "learning_rate": 7.066211835635089e-06,
-      "loss": 0.3212,
-      "step": 17311
-    },
-    {
-      "epoch": 1.1314293183452062,
-      "grad_norm": 0.5298823118209839,
-      "learning_rate": 7.065893850816396e-06,
-      "loss": 0.436,
-      "step": 17312
-    },
-    {
-      "epoch": 1.1314946735507483,
-      "grad_norm": 0.48152822256088257,
-      "learning_rate": 7.06557585592162e-06,
-      "loss": 0.3857,
-      "step": 17313
-    },
-    {
-      "epoch": 1.1315600287562904,
-      "grad_norm": 0.4999312162399292,
-      "learning_rate": 7.065257850952318e-06,
-      "loss": 0.3871,
-      "step": 17314
-    },
-    {
-      "epoch": 1.1316253839618327,
-      "grad_norm": 0.42304015159606934,
-      "learning_rate": 7.064939835910035e-06,
-      "loss": 0.3182,
-      "step": 17315
-    },
-    {
-      "epoch": 1.1316907391673747,
-      "grad_norm": 0.438496857881546,
-      "learning_rate": 7.064621810796328e-06,
-      "loss": 0.3471,
-      "step": 17316
-    },
-    {
-      "epoch": 1.1317560943729168,
-      "grad_norm": 0.4253050684928894,
-      "learning_rate": 7.064303775612745e-06,
-      "loss": 0.2895,
-      "step": 17317
-    },
-    {
-      "epoch": 1.1318214495784589,
-      "grad_norm": 0.46796098351478577,
-      "learning_rate": 7.063985730360836e-06,
-      "loss": 0.3597,
-      "step": 17318
-    },
-    {
-      "epoch": 1.131886804784001,
-      "grad_norm": 0.43548649549484253,
-      "learning_rate": 7.063667675042153e-06,
-      "loss": 0.316,
-      "step": 17319
-    },
-    {
-      "epoch": 1.1319521599895432,
-      "grad_norm": 0.4427529573440552,
-      "learning_rate": 7.063349609658248e-06,
-      "loss": 0.3383,
-      "step": 17320
-    },
-    {
-      "epoch": 1.1320175151950853,
-      "grad_norm": 0.42355749011039734,
-      "learning_rate": 7.063031534210673e-06,
-      "loss": 0.3267,
-      "step": 17321
-    },
-    {
-      "epoch": 1.1320828704006274,
-      "grad_norm": 0.4459403455257416,
-      "learning_rate": 7.062713448700979e-06,
-      "loss": 0.3042,
-      "step": 17322
-    },
-    {
-      "epoch": 1.1321482256061695,
-      "grad_norm": 0.49038663506507874,
-      "learning_rate": 7.062395353130716e-06,
-      "loss": 0.366,
-      "step": 17323
-    },
-    {
-      "epoch": 1.1322135808117118,
-      "grad_norm": 0.4366703927516937,
-      "learning_rate": 7.062077247501436e-06,
-      "loss": 0.3511,
-      "step": 17324
-    },
-    {
-      "epoch": 1.1322789360172538,
-      "grad_norm": 0.43770113587379456,
-      "learning_rate": 7.0617591318146915e-06,
-      "loss": 0.3072,
-      "step": 17325
-    },
-    {
-      "epoch": 1.132344291222796,
-      "grad_norm": 0.4596775770187378,
-      "learning_rate": 7.061441006072033e-06,
-      "loss": 0.3467,
-      "step": 17326
-    },
-    {
-      "epoch": 1.132409646428338,
-      "grad_norm": 0.44386380910873413,
-      "learning_rate": 7.061122870275013e-06,
-      "loss": 0.3455,
-      "step": 17327
-    },
-    {
-      "epoch": 1.13247500163388,
-      "grad_norm": 0.4150366187095642,
-      "learning_rate": 7.0608047244251834e-06,
-      "loss": 0.3035,
-      "step": 17328
-    },
-    {
-      "epoch": 1.1325403568394223,
-      "grad_norm": 0.4691945016384125,
-      "learning_rate": 7.060486568524094e-06,
-      "loss": 0.3742,
-      "step": 17329
-    },
-    {
-      "epoch": 1.1326057120449644,
-      "grad_norm": 0.47214385867118835,
-      "learning_rate": 7.060168402573298e-06,
-      "loss": 0.3936,
-      "step": 17330
-    },
-    {
-      "epoch": 1.1326710672505065,
-      "grad_norm": 0.4524897634983063,
-      "learning_rate": 7.059850226574347e-06,
-      "loss": 0.3194,
-      "step": 17331
-    },
-    {
-      "epoch": 1.1327364224560486,
-      "grad_norm": 0.4278281033039093,
-      "learning_rate": 7.0595320405287935e-06,
-      "loss": 0.3272,
-      "step": 17332
-    },
-    {
-      "epoch": 1.1328017776615908,
-      "grad_norm": 0.44247061014175415,
-      "learning_rate": 7.059213844438189e-06,
-      "loss": 0.3389,
-      "step": 17333
-    },
-    {
-      "epoch": 1.132867132867133,
-      "grad_norm": 0.42941904067993164,
-      "learning_rate": 7.058895638304084e-06,
-      "loss": 0.3351,
-      "step": 17334
-    },
-    {
-      "epoch": 1.132932488072675,
-      "grad_norm": 0.4680541753768921,
-      "learning_rate": 7.0585774221280325e-06,
-      "loss": 0.3967,
-      "step": 17335
-    },
-    {
-      "epoch": 1.132997843278217,
-      "grad_norm": 0.42064833641052246,
-      "learning_rate": 7.0582591959115855e-06,
-      "loss": 0.3048,
-      "step": 17336
-    },
-    {
-      "epoch": 1.1330631984837591,
-      "grad_norm": 0.4507811665534973,
-      "learning_rate": 7.057940959656295e-06,
-      "loss": 0.3747,
-      "step": 17337
-    },
-    {
-      "epoch": 1.1331285536893014,
-      "grad_norm": 0.4998791515827179,
-      "learning_rate": 7.057622713363714e-06,
-      "loss": 0.3948,
-      "step": 17338
-    },
-    {
-      "epoch": 1.1331939088948435,
-      "grad_norm": 0.4143909215927124,
-      "learning_rate": 7.057304457035395e-06,
-      "loss": 0.2958,
-      "step": 17339
-    },
-    {
-      "epoch": 1.1332592641003856,
-      "grad_norm": 0.43302014470100403,
-      "learning_rate": 7.0569861906728885e-06,
-      "loss": 0.3005,
-      "step": 17340
-    },
-    {
-      "epoch": 1.1333246193059276,
-      "grad_norm": 0.4498705565929413,
-      "learning_rate": 7.056667914277748e-06,
-      "loss": 0.3151,
-      "step": 17341
-    },
-    {
-      "epoch": 1.13338997451147,
-      "grad_norm": 0.43818414211273193,
-      "learning_rate": 7.056349627851524e-06,
-      "loss": 0.3361,
-      "step": 17342
-    },
-    {
-      "epoch": 1.133455329717012,
-      "grad_norm": 0.45202022790908813,
-      "learning_rate": 7.056031331395773e-06,
-      "loss": 0.3702,
-      "step": 17343
-    },
-    {
-      "epoch": 1.133520684922554,
-      "grad_norm": 0.4247361421585083,
-      "learning_rate": 7.0557130249120434e-06,
-      "loss": 0.3453,
-      "step": 17344
-    },
-    {
-      "epoch": 1.1335860401280962,
-      "grad_norm": 0.45013856887817383,
-      "learning_rate": 7.05539470840189e-06,
-      "loss": 0.3532,
-      "step": 17345
-    },
-    {
-      "epoch": 1.1336513953336382,
-      "grad_norm": 0.4905974864959717,
-      "learning_rate": 7.055076381866863e-06,
-      "loss": 0.3757,
-      "step": 17346
-    },
-    {
-      "epoch": 1.1337167505391805,
-      "grad_norm": 0.4609545171260834,
-      "learning_rate": 7.054758045308518e-06,
-      "loss": 0.3376,
-      "step": 17347
-    },
-    {
-      "epoch": 1.1337821057447226,
-      "grad_norm": 0.5002155303955078,
-      "learning_rate": 7.0544396987284055e-06,
-      "loss": 0.3553,
-      "step": 17348
-    },
-    {
-      "epoch": 1.1338474609502647,
-      "grad_norm": 0.42377957701683044,
-      "learning_rate": 7.0541213421280795e-06,
-      "loss": 0.3133,
-      "step": 17349
-    },
-    {
-      "epoch": 1.1339128161558067,
-      "grad_norm": 0.4309867024421692,
-      "learning_rate": 7.053802975509092e-06,
-      "loss": 0.3233,
-      "step": 17350
-    },
-    {
-      "epoch": 1.133978171361349,
-      "grad_norm": 0.44097164273262024,
-      "learning_rate": 7.053484598872995e-06,
-      "loss": 0.3527,
-      "step": 17351
-    },
-    {
-      "epoch": 1.1340435265668911,
-      "grad_norm": 0.43234068155288696,
-      "learning_rate": 7.053166212221343e-06,
-      "loss": 0.3388,
-      "step": 17352
-    },
-    {
-      "epoch": 1.1341088817724332,
-      "grad_norm": 0.502611517906189,
-      "learning_rate": 7.0528478155556875e-06,
-      "loss": 0.3413,
-      "step": 17353
-    },
-    {
-      "epoch": 1.1341742369779753,
-      "grad_norm": 0.4061679542064667,
-      "learning_rate": 7.0525294088775835e-06,
-      "loss": 0.28,
-      "step": 17354
-    },
-    {
-      "epoch": 1.1342395921835173,
-      "grad_norm": 0.4602612257003784,
-      "learning_rate": 7.052210992188582e-06,
-      "loss": 0.3532,
-      "step": 17355
-    },
-    {
-      "epoch": 1.1343049473890596,
-      "grad_norm": 0.4474022090435028,
-      "learning_rate": 7.051892565490235e-06,
-      "loss": 0.3531,
-      "step": 17356
-    },
-    {
-      "epoch": 1.1343703025946017,
-      "grad_norm": 0.4401814043521881,
-      "learning_rate": 7.051574128784099e-06,
-      "loss": 0.3305,
-      "step": 17357
-    },
-    {
-      "epoch": 1.1344356578001438,
-      "grad_norm": 0.4765278995037079,
-      "learning_rate": 7.051255682071725e-06,
-      "loss": 0.3655,
-      "step": 17358
-    },
-    {
-      "epoch": 1.1345010130056858,
-      "grad_norm": 0.4542257487773895,
-      "learning_rate": 7.050937225354666e-06,
-      "loss": 0.3807,
-      "step": 17359
-    },
-    {
-      "epoch": 1.1345663682112281,
-      "grad_norm": 0.48853549361228943,
-      "learning_rate": 7.050618758634475e-06,
-      "loss": 0.3486,
-      "step": 17360
-    },
-    {
-      "epoch": 1.1346317234167702,
-      "grad_norm": 0.4309377372264862,
-      "learning_rate": 7.0503002819127065e-06,
-      "loss": 0.339,
-      "step": 17361
-    },
-    {
-      "epoch": 1.1346970786223123,
-      "grad_norm": 0.43339797854423523,
-      "learning_rate": 7.049981795190912e-06,
-      "loss": 0.328,
-      "step": 17362
-    },
-    {
-      "epoch": 1.1347624338278544,
-      "grad_norm": 0.4359287619590759,
-      "learning_rate": 7.049663298470648e-06,
-      "loss": 0.3394,
-      "step": 17363
-    },
-    {
-      "epoch": 1.1348277890333964,
-      "grad_norm": 0.4393298625946045,
-      "learning_rate": 7.049344791753465e-06,
-      "loss": 0.344,
-      "step": 17364
-    },
-    {
-      "epoch": 1.1348931442389387,
-      "grad_norm": 0.44869673252105713,
-      "learning_rate": 7.049026275040918e-06,
-      "loss": 0.3265,
-      "step": 17365
-    },
-    {
-      "epoch": 1.1349584994444808,
-      "grad_norm": 0.44936755299568176,
-      "learning_rate": 7.048707748334559e-06,
-      "loss": 0.3354,
-      "step": 17366
-    },
-    {
-      "epoch": 1.1350238546500229,
-      "grad_norm": 0.44537898898124695,
-      "learning_rate": 7.048389211635943e-06,
-      "loss": 0.3392,
-      "step": 17367
-    },
-    {
-      "epoch": 1.135089209855565,
-      "grad_norm": 0.4435463547706604,
-      "learning_rate": 7.048070664946624e-06,
-      "loss": 0.3456,
-      "step": 17368
-    },
-    {
-      "epoch": 1.1351545650611072,
-      "grad_norm": 0.47491025924682617,
-      "learning_rate": 7.0477521082681545e-06,
-      "loss": 0.3888,
-      "step": 17369
-    },
-    {
-      "epoch": 1.1352199202666493,
-      "grad_norm": 0.45296916365623474,
-      "learning_rate": 7.047433541602089e-06,
-      "loss": 0.3213,
-      "step": 17370
-    },
-    {
-      "epoch": 1.1352852754721914,
-      "grad_norm": 0.47255194187164307,
-      "learning_rate": 7.047114964949979e-06,
-      "loss": 0.3834,
-      "step": 17371
-    },
-    {
-      "epoch": 1.1353506306777335,
-      "grad_norm": 0.43570056557655334,
-      "learning_rate": 7.046796378313382e-06,
-      "loss": 0.3601,
-      "step": 17372
-    },
-    {
-      "epoch": 1.1354159858832755,
-      "grad_norm": 0.45375707745552063,
-      "learning_rate": 7.046477781693848e-06,
-      "loss": 0.3781,
-      "step": 17373
-    },
-    {
-      "epoch": 1.1354813410888178,
-      "grad_norm": 0.4503547251224518,
-      "learning_rate": 7.046159175092935e-06,
-      "loss": 0.3285,
-      "step": 17374
-    },
-    {
-      "epoch": 1.13554669629436,
-      "grad_norm": 0.44876909255981445,
-      "learning_rate": 7.045840558512194e-06,
-      "loss": 0.312,
-      "step": 17375
-    },
-    {
-      "epoch": 1.135612051499902,
-      "grad_norm": 0.4256334602832794,
-      "learning_rate": 7.0455219319531795e-06,
-      "loss": 0.3399,
-      "step": 17376
-    },
-    {
-      "epoch": 1.135677406705444,
-      "grad_norm": 0.4213603138923645,
-      "learning_rate": 7.045203295417446e-06,
-      "loss": 0.3154,
-      "step": 17377
-    },
-    {
-      "epoch": 1.1357427619109863,
-      "grad_norm": 0.45095837116241455,
-      "learning_rate": 7.044884648906548e-06,
-      "loss": 0.3509,
-      "step": 17378
-    },
-    {
-      "epoch": 1.1358081171165284,
-      "grad_norm": 0.5239996910095215,
-      "learning_rate": 7.044565992422041e-06,
-      "loss": 0.4397,
-      "step": 17379
-    },
-    {
-      "epoch": 1.1358734723220705,
-      "grad_norm": 0.47133052349090576,
-      "learning_rate": 7.044247325965474e-06,
-      "loss": 0.3864,
-      "step": 17380
-    },
-    {
-      "epoch": 1.1359388275276125,
-      "grad_norm": 0.4539269804954529,
-      "learning_rate": 7.043928649538406e-06,
-      "loss": 0.3475,
-      "step": 17381
-    },
-    {
-      "epoch": 1.1360041827331546,
-      "grad_norm": 0.48296546936035156,
-      "learning_rate": 7.04360996314239e-06,
-      "loss": 0.3258,
-      "step": 17382
-    },
-    {
-      "epoch": 1.1360695379386967,
-      "grad_norm": 0.4233575761318207,
-      "learning_rate": 7.04329126677898e-06,
-      "loss": 0.3677,
-      "step": 17383
-    },
-    {
-      "epoch": 1.136134893144239,
-      "grad_norm": 0.42596665024757385,
-      "learning_rate": 7.04297256044973e-06,
-      "loss": 0.3294,
-      "step": 17384
-    },
-    {
-      "epoch": 1.136200248349781,
-      "grad_norm": 0.4421609342098236,
-      "learning_rate": 7.0426538441561965e-06,
-      "loss": 0.3398,
-      "step": 17385
-    },
-    {
-      "epoch": 1.1362656035553231,
-      "grad_norm": 0.45286962389945984,
-      "learning_rate": 7.042335117899932e-06,
-      "loss": 0.353,
-      "step": 17386
-    },
-    {
-      "epoch": 1.1363309587608652,
-      "grad_norm": 0.42848026752471924,
-      "learning_rate": 7.0420163816824906e-06,
-      "loss": 0.3356,
-      "step": 17387
-    },
-    {
-      "epoch": 1.1363963139664075,
-      "grad_norm": 0.4370197057723999,
-      "learning_rate": 7.041697635505429e-06,
-      "loss": 0.3385,
-      "step": 17388
-    },
-    {
-      "epoch": 1.1364616691719496,
-      "grad_norm": 0.46252763271331787,
-      "learning_rate": 7.0413788793703e-06,
-      "loss": 0.37,
-      "step": 17389
-    },
-    {
-      "epoch": 1.1365270243774916,
-      "grad_norm": 0.5076103806495667,
-      "learning_rate": 7.041060113278659e-06,
-      "loss": 0.4124,
-      "step": 17390
-    },
-    {
-      "epoch": 1.1365923795830337,
-      "grad_norm": 0.45653036236763,
-      "learning_rate": 7.040741337232061e-06,
-      "loss": 0.3391,
-      "step": 17391
-    },
-    {
-      "epoch": 1.1366577347885758,
-      "grad_norm": 0.4983507990837097,
-      "learning_rate": 7.040422551232061e-06,
-      "loss": 0.3108,
-      "step": 17392
-    },
-    {
-      "epoch": 1.136723089994118,
-      "grad_norm": 0.49780896306037903,
-      "learning_rate": 7.040103755280213e-06,
-      "loss": 0.401,
-      "step": 17393
-    },
-    {
-      "epoch": 1.1367884451996602,
-      "grad_norm": 0.4493112564086914,
-      "learning_rate": 7.039784949378073e-06,
-      "loss": 0.3412,
-      "step": 17394
-    },
-    {
-      "epoch": 1.1368538004052022,
-      "grad_norm": 0.43972402811050415,
-      "learning_rate": 7.039466133527194e-06,
-      "loss": 0.3299,
-      "step": 17395
-    },
-    {
-      "epoch": 1.1369191556107443,
-      "grad_norm": 0.4225434362888336,
-      "learning_rate": 7.039147307729133e-06,
-      "loss": 0.2964,
-      "step": 17396
-    },
-    {
-      "epoch": 1.1369845108162866,
-      "grad_norm": 0.45975571870803833,
-      "learning_rate": 7.038828471985444e-06,
-      "loss": 0.3665,
-      "step": 17397
-    },
-    {
-      "epoch": 1.1370498660218287,
-      "grad_norm": 0.4576628506183624,
-      "learning_rate": 7.03850962629768e-06,
-      "loss": 0.3618,
-      "step": 17398
-    },
-    {
-      "epoch": 1.1371152212273707,
-      "grad_norm": 0.4470856785774231,
-      "learning_rate": 7.038190770667401e-06,
-      "loss": 0.3628,
-      "step": 17399
-    },
-    {
-      "epoch": 1.1371805764329128,
-      "grad_norm": 0.4435643255710602,
-      "learning_rate": 7.037871905096159e-06,
-      "loss": 0.2982,
-      "step": 17400
-    },
-    {
-      "epoch": 1.1372459316384549,
-      "grad_norm": 0.46543169021606445,
-      "learning_rate": 7.03755302958551e-06,
-      "loss": 0.3415,
-      "step": 17401
-    },
-    {
-      "epoch": 1.1373112868439972,
-      "grad_norm": 0.45349591970443726,
-      "learning_rate": 7.037234144137009e-06,
-      "loss": 0.3329,
-      "step": 17402
-    },
-    {
-      "epoch": 1.1373766420495393,
-      "grad_norm": 0.4510424733161926,
-      "learning_rate": 7.03691524875221e-06,
-      "loss": 0.3286,
-      "step": 17403
-    },
-    {
-      "epoch": 1.1374419972550813,
-      "grad_norm": 0.42036014795303345,
-      "learning_rate": 7.03659634343267e-06,
-      "loss": 0.2916,
-      "step": 17404
-    },
-    {
-      "epoch": 1.1375073524606234,
-      "grad_norm": 0.48616576194763184,
-      "learning_rate": 7.036277428179945e-06,
-      "loss": 0.3563,
-      "step": 17405
-    },
-    {
-      "epoch": 1.1375727076661657,
-      "grad_norm": 0.45995670557022095,
-      "learning_rate": 7.035958502995589e-06,
-      "loss": 0.3807,
-      "step": 17406
-    },
-    {
-      "epoch": 1.1376380628717078,
-      "grad_norm": 0.4322569966316223,
-      "learning_rate": 7.035639567881158e-06,
-      "loss": 0.3034,
-      "step": 17407
-    },
-    {
-      "epoch": 1.1377034180772498,
-      "grad_norm": 0.4015660881996155,
-      "learning_rate": 7.035320622838208e-06,
-      "loss": 0.294,
-      "step": 17408
-    },
-    {
-      "epoch": 1.137768773282792,
-      "grad_norm": 0.44619235396385193,
-      "learning_rate": 7.035001667868293e-06,
-      "loss": 0.3385,
-      "step": 17409
-    },
-    {
-      "epoch": 1.137834128488334,
-      "grad_norm": 0.4452032446861267,
-      "learning_rate": 7.034682702972971e-06,
-      "loss": 0.3633,
-      "step": 17410
-    },
-    {
-      "epoch": 1.1378994836938763,
-      "grad_norm": 0.4686889052391052,
-      "learning_rate": 7.034363728153797e-06,
-      "loss": 0.3536,
-      "step": 17411
-    },
-    {
-      "epoch": 1.1379648388994184,
-      "grad_norm": 0.4439161419868469,
-      "learning_rate": 7.0340447434123264e-06,
-      "loss": 0.3422,
-      "step": 17412
-    },
-    {
-      "epoch": 1.1380301941049604,
-      "grad_norm": 0.4715063273906708,
-      "learning_rate": 7.0337257487501145e-06,
-      "loss": 0.3479,
-      "step": 17413
-    },
-    {
-      "epoch": 1.1380955493105025,
-      "grad_norm": 0.4390163719654083,
-      "learning_rate": 7.033406744168716e-06,
-      "loss": 0.323,
-      "step": 17414
-    },
-    {
-      "epoch": 1.1381609045160448,
-      "grad_norm": 0.4483895003795624,
-      "learning_rate": 7.03308772966969e-06,
-      "loss": 0.356,
-      "step": 17415
-    },
-    {
-      "epoch": 1.1382262597215869,
-      "grad_norm": 0.4739121198654175,
-      "learning_rate": 7.03276870525459e-06,
-      "loss": 0.3489,
-      "step": 17416
-    },
-    {
-      "epoch": 1.138291614927129,
-      "grad_norm": 0.43798214197158813,
-      "learning_rate": 7.032449670924974e-06,
-      "loss": 0.3043,
-      "step": 17417
-    },
-    {
-      "epoch": 1.138356970132671,
-      "grad_norm": 0.4182601273059845,
-      "learning_rate": 7.032130626682395e-06,
-      "loss": 0.3046,
-      "step": 17418
-    },
-    {
-      "epoch": 1.138422325338213,
-      "grad_norm": 0.4340640604496002,
-      "learning_rate": 7.031811572528413e-06,
-      "loss": 0.3359,
-      "step": 17419
-    },
-    {
-      "epoch": 1.1384876805437554,
-      "grad_norm": 0.4664618968963623,
-      "learning_rate": 7.031492508464581e-06,
-      "loss": 0.3772,
-      "step": 17420
-    },
-    {
-      "epoch": 1.1385530357492974,
-      "grad_norm": 0.4334782660007477,
-      "learning_rate": 7.031173434492458e-06,
-      "loss": 0.2973,
-      "step": 17421
-    },
-    {
-      "epoch": 1.1386183909548395,
-      "grad_norm": 0.4512031376361847,
-      "learning_rate": 7.030854350613596e-06,
-      "loss": 0.3546,
-      "step": 17422
-    },
-    {
-      "epoch": 1.1386837461603816,
-      "grad_norm": 0.4125944674015045,
-      "learning_rate": 7.030535256829555e-06,
-      "loss": 0.3129,
-      "step": 17423
-    },
-    {
-      "epoch": 1.1387491013659239,
-      "grad_norm": 0.45096346735954285,
-      "learning_rate": 7.03021615314189e-06,
-      "loss": 0.3389,
-      "step": 17424
-    },
-    {
-      "epoch": 1.138814456571466,
-      "grad_norm": 0.41682204604148865,
-      "learning_rate": 7.029897039552157e-06,
-      "loss": 0.3124,
-      "step": 17425
-    },
-    {
-      "epoch": 1.138879811777008,
-      "grad_norm": 0.42640236020088196,
-      "learning_rate": 7.029577916061913e-06,
-      "loss": 0.3147,
-      "step": 17426
-    },
-    {
-      "epoch": 1.13894516698255,
-      "grad_norm": 0.4052651524543762,
-      "learning_rate": 7.029258782672714e-06,
-      "loss": 0.3047,
-      "step": 17427
-    },
-    {
-      "epoch": 1.1390105221880922,
-      "grad_norm": 0.4783850312232971,
-      "learning_rate": 7.028939639386118e-06,
-      "loss": 0.3281,
-      "step": 17428
-    },
-    {
-      "epoch": 1.1390758773936345,
-      "grad_norm": 0.4420865774154663,
-      "learning_rate": 7.02862048620368e-06,
-      "loss": 0.3389,
-      "step": 17429
-    },
-    {
-      "epoch": 1.1391412325991765,
-      "grad_norm": 0.4368607997894287,
-      "learning_rate": 7.028301323126958e-06,
-      "loss": 0.3255,
-      "step": 17430
-    },
-    {
-      "epoch": 1.1392065878047186,
-      "grad_norm": 0.4040074646472931,
-      "learning_rate": 7.027982150157506e-06,
-      "loss": 0.3003,
-      "step": 17431
-    },
-    {
-      "epoch": 1.1392719430102607,
-      "grad_norm": 0.477368026971817,
-      "learning_rate": 7.027662967296885e-06,
-      "loss": 0.4153,
-      "step": 17432
-    },
-    {
-      "epoch": 1.139337298215803,
-      "grad_norm": 0.4319467544555664,
-      "learning_rate": 7.0273437745466465e-06,
-      "loss": 0.3039,
-      "step": 17433
-    },
-    {
-      "epoch": 1.139402653421345,
-      "grad_norm": 0.45744410157203674,
-      "learning_rate": 7.027024571908351e-06,
-      "loss": 0.3536,
-      "step": 17434
-    },
-    {
-      "epoch": 1.1394680086268871,
-      "grad_norm": 0.49372386932373047,
-      "learning_rate": 7.026705359383554e-06,
-      "loss": 0.3724,
-      "step": 17435
-    },
-    {
-      "epoch": 1.1395333638324292,
-      "grad_norm": 0.45175907015800476,
-      "learning_rate": 7.026386136973814e-06,
-      "loss": 0.3877,
-      "step": 17436
-    },
-    {
-      "epoch": 1.1395987190379713,
-      "grad_norm": 0.4616450071334839,
-      "learning_rate": 7.0260669046806864e-06,
-      "loss": 0.4162,
-      "step": 17437
-    },
-    {
-      "epoch": 1.1396640742435136,
-      "grad_norm": 0.4368988573551178,
-      "learning_rate": 7.0257476625057285e-06,
-      "loss": 0.3153,
-      "step": 17438
-    },
-    {
-      "epoch": 1.1397294294490556,
-      "grad_norm": 0.4661909341812134,
-      "learning_rate": 7.025428410450498e-06,
-      "loss": 0.3211,
-      "step": 17439
-    },
-    {
-      "epoch": 1.1397947846545977,
-      "grad_norm": 0.45357945561408997,
-      "learning_rate": 7.025109148516552e-06,
-      "loss": 0.345,
-      "step": 17440
-    },
-    {
-      "epoch": 1.1398601398601398,
-      "grad_norm": 0.44125261902809143,
-      "learning_rate": 7.024789876705445e-06,
-      "loss": 0.3397,
-      "step": 17441
-    },
-    {
-      "epoch": 1.139925495065682,
-      "grad_norm": 0.4618300795555115,
-      "learning_rate": 7.024470595018738e-06,
-      "loss": 0.3772,
-      "step": 17442
-    },
-    {
-      "epoch": 1.1399908502712242,
-      "grad_norm": 0.478115051984787,
-      "learning_rate": 7.024151303457985e-06,
-      "loss": 0.3532,
-      "step": 17443
-    },
-    {
-      "epoch": 1.1400562054767662,
-      "grad_norm": 0.5167980194091797,
-      "learning_rate": 7.023832002024746e-06,
-      "loss": 0.4005,
-      "step": 17444
-    },
-    {
-      "epoch": 1.1401215606823083,
-      "grad_norm": 0.45658424496650696,
-      "learning_rate": 7.0235126907205775e-06,
-      "loss": 0.36,
-      "step": 17445
-    },
-    {
-      "epoch": 1.1401869158878504,
-      "grad_norm": 0.44158434867858887,
-      "learning_rate": 7.023193369547037e-06,
-      "loss": 0.3505,
-      "step": 17446
-    },
-    {
-      "epoch": 1.1402522710933927,
-      "grad_norm": 0.45263805985450745,
-      "learning_rate": 7.022874038505679e-06,
-      "loss": 0.3631,
-      "step": 17447
-    },
-    {
-      "epoch": 1.1403176262989347,
-      "grad_norm": 0.4402000606060028,
-      "learning_rate": 7.022554697598065e-06,
-      "loss": 0.3399,
-      "step": 17448
-    },
-    {
-      "epoch": 1.1403829815044768,
-      "grad_norm": 0.4918239712715149,
-      "learning_rate": 7.022235346825751e-06,
-      "loss": 0.3941,
-      "step": 17449
-    },
-    {
-      "epoch": 1.1404483367100189,
-      "grad_norm": 0.42927759885787964,
-      "learning_rate": 7.021915986190295e-06,
-      "loss": 0.2929,
-      "step": 17450
-    },
-    {
-      "epoch": 1.1405136919155612,
-      "grad_norm": 0.5000864863395691,
-      "learning_rate": 7.021596615693256e-06,
-      "loss": 0.3882,
-      "step": 17451
-    },
-    {
-      "epoch": 1.1405790471211033,
-      "grad_norm": 0.41612523794174194,
-      "learning_rate": 7.021277235336187e-06,
-      "loss": 0.3019,
-      "step": 17452
-    },
-    {
-      "epoch": 1.1406444023266453,
-      "grad_norm": 0.46470025181770325,
-      "learning_rate": 7.020957845120649e-06,
-      "loss": 0.343,
-      "step": 17453
-    },
-    {
-      "epoch": 1.1407097575321874,
-      "grad_norm": 0.43162819743156433,
-      "learning_rate": 7.0206384450482e-06,
-      "loss": 0.3456,
-      "step": 17454
-    },
-    {
-      "epoch": 1.1407751127377295,
-      "grad_norm": 0.45312026143074036,
-      "learning_rate": 7.020319035120399e-06,
-      "loss": 0.3533,
-      "step": 17455
-    },
-    {
-      "epoch": 1.1408404679432718,
-      "grad_norm": 0.4765598177909851,
-      "learning_rate": 7.019999615338799e-06,
-      "loss": 0.4242,
-      "step": 17456
-    },
-    {
-      "epoch": 1.1409058231488138,
-      "grad_norm": 0.41143378615379333,
-      "learning_rate": 7.019680185704964e-06,
-      "loss": 0.2977,
-      "step": 17457
-    },
-    {
-      "epoch": 1.140971178354356,
-      "grad_norm": 0.44259515404701233,
-      "learning_rate": 7.019360746220447e-06,
-      "loss": 0.3014,
-      "step": 17458
-    },
-    {
-      "epoch": 1.141036533559898,
-      "grad_norm": 0.4750520586967468,
-      "learning_rate": 7.0190412968868095e-06,
-      "loss": 0.3738,
-      "step": 17459
-    },
-    {
-      "epoch": 1.1411018887654403,
-      "grad_norm": 0.5006691217422485,
-      "learning_rate": 7.018721837705608e-06,
-      "loss": 0.4195,
-      "step": 17460
-    },
-    {
-      "epoch": 1.1411672439709823,
-      "grad_norm": 0.42368045449256897,
-      "learning_rate": 7.018402368678399e-06,
-      "loss": 0.3209,
-      "step": 17461
-    },
-    {
-      "epoch": 1.1412325991765244,
-      "grad_norm": 0.47609299421310425,
-      "learning_rate": 7.0180828898067445e-06,
-      "loss": 0.3686,
-      "step": 17462
-    },
-    {
-      "epoch": 1.1412979543820665,
-      "grad_norm": 0.5127665996551514,
-      "learning_rate": 7.0177634010922e-06,
-      "loss": 0.3663,
-      "step": 17463
-    },
-    {
-      "epoch": 1.1413633095876086,
-      "grad_norm": 0.4121999740600586,
-      "learning_rate": 7.017443902536325e-06,
-      "loss": 0.2799,
-      "step": 17464
-    },
-    {
-      "epoch": 1.1414286647931509,
-      "grad_norm": 0.4780321717262268,
-      "learning_rate": 7.0171243941406755e-06,
-      "loss": 0.3486,
-      "step": 17465
-    },
-    {
-      "epoch": 1.141494019998693,
-      "grad_norm": 0.482979953289032,
-      "learning_rate": 7.016804875906813e-06,
-      "loss": 0.3814,
-      "step": 17466
-    },
-    {
-      "epoch": 1.141559375204235,
-      "grad_norm": 0.43570512533187866,
-      "learning_rate": 7.0164853478362925e-06,
-      "loss": 0.3134,
-      "step": 17467
-    },
-    {
-      "epoch": 1.141624730409777,
-      "grad_norm": 0.4267442524433136,
-      "learning_rate": 7.016165809930676e-06,
-      "loss": 0.3265,
-      "step": 17468
-    },
-    {
-      "epoch": 1.1416900856153194,
-      "grad_norm": 0.45084404945373535,
-      "learning_rate": 7.01584626219152e-06,
-      "loss": 0.3261,
-      "step": 17469
-    },
-    {
-      "epoch": 1.1417554408208614,
-      "grad_norm": 0.4325253367424011,
-      "learning_rate": 7.015526704620383e-06,
-      "loss": 0.3033,
-      "step": 17470
-    },
-    {
-      "epoch": 1.1418207960264035,
-      "grad_norm": 0.4355321228504181,
-      "learning_rate": 7.015207137218826e-06,
-      "loss": 0.2788,
-      "step": 17471
-    },
-    {
-      "epoch": 1.1418861512319456,
-      "grad_norm": 0.4479791224002838,
-      "learning_rate": 7.014887559988403e-06,
-      "loss": 0.3348,
-      "step": 17472
-    },
-    {
-      "epoch": 1.1419515064374877,
-      "grad_norm": 0.4449215531349182,
-      "learning_rate": 7.0145679729306775e-06,
-      "loss": 0.3448,
-      "step": 17473
-    },
-    {
-      "epoch": 1.14201686164303,
-      "grad_norm": 0.4846709072589874,
-      "learning_rate": 7.014248376047205e-06,
-      "loss": 0.3845,
-      "step": 17474
-    },
-    {
-      "epoch": 1.142082216848572,
-      "grad_norm": 0.4538358449935913,
-      "learning_rate": 7.013928769339545e-06,
-      "loss": 0.3541,
-      "step": 17475
-    },
-    {
-      "epoch": 1.142147572054114,
-      "grad_norm": 0.4530636668205261,
-      "learning_rate": 7.013609152809256e-06,
-      "loss": 0.3285,
-      "step": 17476
-    },
-    {
-      "epoch": 1.1422129272596562,
-      "grad_norm": 0.4355792701244354,
-      "learning_rate": 7.013289526457901e-06,
-      "loss": 0.3523,
-      "step": 17477
-    },
-    {
-      "epoch": 1.1422782824651985,
-      "grad_norm": 0.47689175605773926,
-      "learning_rate": 7.012969890287033e-06,
-      "loss": 0.3395,
-      "step": 17478
-    },
-    {
-      "epoch": 1.1423436376707405,
-      "grad_norm": 0.44496792554855347,
-      "learning_rate": 7.0126502442982125e-06,
-      "loss": 0.3261,
-      "step": 17479
-    },
-    {
-      "epoch": 1.1424089928762826,
-      "grad_norm": 0.44975581765174866,
-      "learning_rate": 7.012330588493001e-06,
-      "loss": 0.3583,
-      "step": 17480
-    },
-    {
-      "epoch": 1.1424743480818247,
-      "grad_norm": 0.42740964889526367,
-      "learning_rate": 7.012010922872956e-06,
-      "loss": 0.3192,
-      "step": 17481
-    },
-    {
-      "epoch": 1.1425397032873668,
-      "grad_norm": 0.4553546607494354,
-      "learning_rate": 7.011691247439636e-06,
-      "loss": 0.3381,
-      "step": 17482
-    },
-    {
-      "epoch": 1.142605058492909,
-      "grad_norm": 0.4539209306240082,
-      "learning_rate": 7.011371562194601e-06,
-      "loss": 0.3636,
-      "step": 17483
-    },
-    {
-      "epoch": 1.1426704136984511,
-      "grad_norm": 0.47124984860420227,
-      "learning_rate": 7.011051867139412e-06,
-      "loss": 0.3436,
-      "step": 17484
-    },
-    {
-      "epoch": 1.1427357689039932,
-      "grad_norm": 0.4807920753955841,
-      "learning_rate": 7.010732162275624e-06,
-      "loss": 0.3838,
-      "step": 17485
-    },
-    {
-      "epoch": 1.1428011241095353,
-      "grad_norm": 0.5294554829597473,
-      "learning_rate": 7.0104124476048e-06,
-      "loss": 0.3826,
-      "step": 17486
-    },
-    {
-      "epoch": 1.1428664793150776,
-      "grad_norm": 0.4663105010986328,
-      "learning_rate": 7.010092723128497e-06,
-      "loss": 0.3824,
-      "step": 17487
-    },
-    {
-      "epoch": 1.1429318345206196,
-      "grad_norm": 0.4112255573272705,
-      "learning_rate": 7.0097729888482764e-06,
-      "loss": 0.2874,
-      "step": 17488
-    },
-    {
-      "epoch": 1.1429971897261617,
-      "grad_norm": 0.47517916560173035,
-      "learning_rate": 7.009453244765697e-06,
-      "loss": 0.3853,
-      "step": 17489
-    },
-    {
-      "epoch": 1.1430625449317038,
-      "grad_norm": 0.4611404240131378,
-      "learning_rate": 7.009133490882316e-06,
-      "loss": 0.3286,
-      "step": 17490
-    },
-    {
-      "epoch": 1.1431279001372459,
-      "grad_norm": 0.4590357840061188,
-      "learning_rate": 7.008813727199697e-06,
-      "loss": 0.3412,
-      "step": 17491
-    },
-    {
-      "epoch": 1.143193255342788,
-      "grad_norm": 0.45553115010261536,
-      "learning_rate": 7.008493953719396e-06,
-      "loss": 0.3621,
-      "step": 17492
-    },
-    {
-      "epoch": 1.1432586105483302,
-      "grad_norm": 0.43632274866104126,
-      "learning_rate": 7.008174170442975e-06,
-      "loss": 0.3244,
-      "step": 17493
-    },
-    {
-      "epoch": 1.1433239657538723,
-      "grad_norm": 0.42470139265060425,
-      "learning_rate": 7.007854377371992e-06,
-      "loss": 0.3033,
-      "step": 17494
-    },
-    {
-      "epoch": 1.1433893209594144,
-      "grad_norm": 0.4470202922821045,
-      "learning_rate": 7.007534574508009e-06,
-      "loss": 0.3364,
-      "step": 17495
-    },
-    {
-      "epoch": 1.1434546761649567,
-      "grad_norm": 0.45505478978157043,
-      "learning_rate": 7.007214761852583e-06,
-      "loss": 0.3493,
-      "step": 17496
-    },
-    {
-      "epoch": 1.1435200313704987,
-      "grad_norm": 0.4418489933013916,
-      "learning_rate": 7.0068949394072765e-06,
-      "loss": 0.3341,
-      "step": 17497
-    },
-    {
-      "epoch": 1.1435853865760408,
-      "grad_norm": 0.4500032067298889,
-      "learning_rate": 7.006575107173647e-06,
-      "loss": 0.3466,
-      "step": 17498
-    },
-    {
-      "epoch": 1.1436507417815829,
-      "grad_norm": 0.48150408267974854,
-      "learning_rate": 7.006255265153257e-06,
-      "loss": 0.3992,
-      "step": 17499
-    },
-    {
-      "epoch": 1.143716096987125,
-      "grad_norm": 0.45815691351890564,
-      "learning_rate": 7.005935413347664e-06,
-      "loss": 0.3628,
-      "step": 17500
-    },
-    {
-      "epoch": 1.143781452192667,
-      "grad_norm": 0.4734034836292267,
-      "learning_rate": 7.005615551758429e-06,
-      "loss": 0.3366,
-      "step": 17501
-    },
-    {
-      "epoch": 1.1438468073982093,
-      "grad_norm": 0.44199973344802856,
-      "learning_rate": 7.005295680387113e-06,
-      "loss": 0.3496,
-      "step": 17502
-    },
-    {
-      "epoch": 1.1439121626037514,
-      "grad_norm": 0.4660091698169708,
-      "learning_rate": 7.004975799235274e-06,
-      "loss": 0.3755,
-      "step": 17503
-    },
-    {
-      "epoch": 1.1439775178092935,
-      "grad_norm": 0.45670491456985474,
-      "learning_rate": 7.004655908304474e-06,
-      "loss": 0.3547,
-      "step": 17504
-    },
-    {
-      "epoch": 1.1440428730148355,
-      "grad_norm": 0.45676979422569275,
-      "learning_rate": 7.004336007596273e-06,
-      "loss": 0.3471,
-      "step": 17505
-    },
-    {
-      "epoch": 1.1441082282203778,
-      "grad_norm": 0.42680084705352783,
-      "learning_rate": 7.0040160971122315e-06,
-      "loss": 0.3421,
-      "step": 17506
-    },
-    {
-      "epoch": 1.14417358342592,
-      "grad_norm": 0.4537406861782074,
-      "learning_rate": 7.003696176853908e-06,
-      "loss": 0.3673,
-      "step": 17507
-    },
-    {
-      "epoch": 1.144238938631462,
-      "grad_norm": 0.4100183844566345,
-      "learning_rate": 7.003376246822865e-06,
-      "loss": 0.3022,
-      "step": 17508
-    },
-    {
-      "epoch": 1.144304293837004,
-      "grad_norm": 0.46144744753837585,
-      "learning_rate": 7.003056307020662e-06,
-      "loss": 0.3927,
-      "step": 17509
-    },
-    {
-      "epoch": 1.1443696490425461,
-      "grad_norm": 0.4399982690811157,
-      "learning_rate": 7.002736357448858e-06,
-      "loss": 0.3438,
-      "step": 17510
-    },
-    {
-      "epoch": 1.1444350042480884,
-      "grad_norm": 0.4397696852684021,
-      "learning_rate": 7.002416398109016e-06,
-      "loss": 0.3508,
-      "step": 17511
-    },
-    {
-      "epoch": 1.1445003594536305,
-      "grad_norm": 0.4235154986381531,
-      "learning_rate": 7.002096429002696e-06,
-      "loss": 0.3124,
-      "step": 17512
-    },
-    {
-      "epoch": 1.1445657146591726,
-      "grad_norm": 0.4556167721748352,
-      "learning_rate": 7.001776450131458e-06,
-      "loss": 0.3625,
-      "step": 17513
-    },
-    {
-      "epoch": 1.1446310698647146,
-      "grad_norm": 0.4215102195739746,
-      "learning_rate": 7.001456461496862e-06,
-      "loss": 0.3155,
-      "step": 17514
-    },
-    {
-      "epoch": 1.144696425070257,
-      "grad_norm": 0.44292616844177246,
-      "learning_rate": 7.0011364631004695e-06,
-      "loss": 0.3468,
-      "step": 17515
-    },
-    {
-      "epoch": 1.144761780275799,
-      "grad_norm": 0.49077779054641724,
-      "learning_rate": 7.000816454943842e-06,
-      "loss": 0.3644,
-      "step": 17516
-    },
-    {
-      "epoch": 1.144827135481341,
-      "grad_norm": 0.42309874296188354,
-      "learning_rate": 7.000496437028539e-06,
-      "loss": 0.3288,
-      "step": 17517
-    },
-    {
-      "epoch": 1.1448924906868831,
-      "grad_norm": 0.419172078371048,
-      "learning_rate": 7.000176409356122e-06,
-      "loss": 0.2896,
-      "step": 17518
-    },
-    {
-      "epoch": 1.1449578458924252,
-      "grad_norm": 0.5341691970825195,
-      "learning_rate": 6.999856371928151e-06,
-      "loss": 0.4543,
-      "step": 17519
-    },
-    {
-      "epoch": 1.1450232010979675,
-      "grad_norm": 0.4835638403892517,
-      "learning_rate": 6.9995363247461874e-06,
-      "loss": 0.3563,
-      "step": 17520
-    },
-    {
-      "epoch": 1.1450885563035096,
-      "grad_norm": 0.43735870718955994,
-      "learning_rate": 6.9992162678117935e-06,
-      "loss": 0.3101,
-      "step": 17521
-    },
-    {
-      "epoch": 1.1451539115090517,
-      "grad_norm": 0.477425754070282,
-      "learning_rate": 6.998896201126529e-06,
-      "loss": 0.3656,
-      "step": 17522
-    },
-    {
-      "epoch": 1.1452192667145937,
-      "grad_norm": 0.4202739894390106,
-      "learning_rate": 6.998576124691956e-06,
-      "loss": 0.292,
-      "step": 17523
-    },
-    {
-      "epoch": 1.145284621920136,
-      "grad_norm": 0.47121620178222656,
-      "learning_rate": 6.998256038509635e-06,
-      "loss": 0.3313,
-      "step": 17524
-    },
-    {
-      "epoch": 1.145349977125678,
-      "grad_norm": 0.44480106234550476,
-      "learning_rate": 6.997935942581125e-06,
-      "loss": 0.3318,
-      "step": 17525
-    },
-    {
-      "epoch": 1.1454153323312202,
-      "grad_norm": 0.4418734312057495,
-      "learning_rate": 6.997615836907991e-06,
-      "loss": 0.3407,
-      "step": 17526
-    },
-    {
-      "epoch": 1.1454806875367622,
-      "grad_norm": 0.47333475947380066,
-      "learning_rate": 6.997295721491792e-06,
-      "loss": 0.3563,
-      "step": 17527
-    },
-    {
-      "epoch": 1.1455460427423043,
-      "grad_norm": 0.4485928416252136,
-      "learning_rate": 6.99697559633409e-06,
-      "loss": 0.3343,
-      "step": 17528
-    },
-    {
-      "epoch": 1.1456113979478466,
-      "grad_norm": 0.4543781876564026,
-      "learning_rate": 6.996655461436447e-06,
-      "loss": 0.3431,
-      "step": 17529
-    },
-    {
-      "epoch": 1.1456767531533887,
-      "grad_norm": 0.42821744084358215,
-      "learning_rate": 6.9963353168004225e-06,
-      "loss": 0.3175,
-      "step": 17530
-    },
-    {
-      "epoch": 1.1457421083589308,
-      "grad_norm": 0.43268415331840515,
-      "learning_rate": 6.996015162427578e-06,
-      "loss": 0.325,
-      "step": 17531
-    },
-    {
-      "epoch": 1.1458074635644728,
-      "grad_norm": 0.4574171304702759,
-      "learning_rate": 6.995694998319478e-06,
-      "loss": 0.3478,
-      "step": 17532
-    },
-    {
-      "epoch": 1.1458728187700151,
-      "grad_norm": 0.5189958810806274,
-      "learning_rate": 6.995374824477681e-06,
-      "loss": 0.4125,
-      "step": 17533
-    },
-    {
-      "epoch": 1.1459381739755572,
-      "grad_norm": 0.468641996383667,
-      "learning_rate": 6.99505464090375e-06,
-      "loss": 0.3532,
-      "step": 17534
-    },
-    {
-      "epoch": 1.1460035291810993,
-      "grad_norm": 0.4337131381034851,
-      "learning_rate": 6.994734447599247e-06,
-      "loss": 0.3247,
-      "step": 17535
-    },
-    {
-      "epoch": 1.1460688843866413,
-      "grad_norm": 0.45070910453796387,
-      "learning_rate": 6.994414244565732e-06,
-      "loss": 0.338,
-      "step": 17536
-    },
-    {
-      "epoch": 1.1461342395921834,
-      "grad_norm": 0.46367672085762024,
-      "learning_rate": 6.994094031804768e-06,
-      "loss": 0.3399,
-      "step": 17537
-    },
-    {
-      "epoch": 1.1461995947977257,
-      "grad_norm": 0.426435261964798,
-      "learning_rate": 6.993773809317918e-06,
-      "loss": 0.3239,
-      "step": 17538
-    },
-    {
-      "epoch": 1.1462649500032678,
-      "grad_norm": 0.4395165741443634,
-      "learning_rate": 6.99345357710674e-06,
-      "loss": 0.3019,
-      "step": 17539
-    },
-    {
-      "epoch": 1.1463303052088099,
-      "grad_norm": 0.4694999158382416,
-      "learning_rate": 6.9931333351728e-06,
-      "loss": 0.3633,
-      "step": 17540
-    },
-    {
-      "epoch": 1.146395660414352,
-      "grad_norm": 0.47705039381980896,
-      "learning_rate": 6.992813083517658e-06,
-      "loss": 0.3617,
-      "step": 17541
-    },
-    {
-      "epoch": 1.1464610156198942,
-      "grad_norm": 0.47072869539260864,
-      "learning_rate": 6.992492822142877e-06,
-      "loss": 0.3725,
-      "step": 17542
-    },
-    {
-      "epoch": 1.1465263708254363,
-      "grad_norm": 0.44895902276039124,
-      "learning_rate": 6.992172551050017e-06,
-      "loss": 0.3579,
-      "step": 17543
-    },
-    {
-      "epoch": 1.1465917260309784,
-      "grad_norm": 0.4714765250682831,
-      "learning_rate": 6.991852270240641e-06,
-      "loss": 0.3867,
-      "step": 17544
-    },
-    {
-      "epoch": 1.1466570812365204,
-      "grad_norm": 0.4682939350605011,
-      "learning_rate": 6.991531979716313e-06,
-      "loss": 0.3784,
-      "step": 17545
-    },
-    {
-      "epoch": 1.1467224364420625,
-      "grad_norm": 0.46957314014434814,
-      "learning_rate": 6.991211679478591e-06,
-      "loss": 0.3572,
-      "step": 17546
-    },
-    {
-      "epoch": 1.1467877916476048,
-      "grad_norm": 0.4267326593399048,
-      "learning_rate": 6.990891369529044e-06,
-      "loss": 0.3181,
-      "step": 17547
-    },
-    {
-      "epoch": 1.1468531468531469,
-      "grad_norm": 0.5047838091850281,
-      "learning_rate": 6.990571049869227e-06,
-      "loss": 0.4355,
-      "step": 17548
-    },
-    {
-      "epoch": 1.146918502058689,
-      "grad_norm": 0.4337526857852936,
-      "learning_rate": 6.990250720500706e-06,
-      "loss": 0.3486,
-      "step": 17549
-    },
-    {
-      "epoch": 1.146983857264231,
-      "grad_norm": 0.4063011407852173,
-      "learning_rate": 6.989930381425042e-06,
-      "loss": 0.2792,
-      "step": 17550
-    },
-    {
-      "epoch": 1.1470492124697733,
-      "grad_norm": 0.4726828634738922,
-      "learning_rate": 6.989610032643799e-06,
-      "loss": 0.3203,
-      "step": 17551
-    },
-    {
-      "epoch": 1.1471145676753154,
-      "grad_norm": 0.46124228835105896,
-      "learning_rate": 6.989289674158538e-06,
-      "loss": 0.3609,
-      "step": 17552
-    },
-    {
-      "epoch": 1.1471799228808575,
-      "grad_norm": 0.4671699106693268,
-      "learning_rate": 6.988969305970823e-06,
-      "loss": 0.3576,
-      "step": 17553
-    },
-    {
-      "epoch": 1.1472452780863995,
-      "grad_norm": 0.4478949010372162,
-      "learning_rate": 6.9886489280822155e-06,
-      "loss": 0.3381,
-      "step": 17554
-    },
-    {
-      "epoch": 1.1473106332919416,
-      "grad_norm": 0.4265245795249939,
-      "learning_rate": 6.988328540494278e-06,
-      "loss": 0.3315,
-      "step": 17555
-    },
-    {
-      "epoch": 1.147375988497484,
-      "grad_norm": 0.46849754452705383,
-      "learning_rate": 6.988008143208574e-06,
-      "loss": 0.3821,
-      "step": 17556
-    },
-    {
-      "epoch": 1.147441343703026,
-      "grad_norm": 0.43850085139274597,
-      "learning_rate": 6.987687736226664e-06,
-      "loss": 0.3496,
-      "step": 17557
-    },
-    {
-      "epoch": 1.147506698908568,
-      "grad_norm": 0.41968628764152527,
-      "learning_rate": 6.987367319550113e-06,
-      "loss": 0.2851,
-      "step": 17558
-    },
-    {
-      "epoch": 1.1475720541141101,
-      "grad_norm": 0.4520379602909088,
-      "learning_rate": 6.987046893180483e-06,
-      "loss": 0.3583,
-      "step": 17559
-    },
-    {
-      "epoch": 1.1476374093196524,
-      "grad_norm": 0.4553259611129761,
-      "learning_rate": 6.986726457119339e-06,
-      "loss": 0.3364,
-      "step": 17560
-    },
-    {
-      "epoch": 1.1477027645251945,
-      "grad_norm": 0.48487669229507446,
-      "learning_rate": 6.986406011368239e-06,
-      "loss": 0.3601,
-      "step": 17561
-    },
-    {
-      "epoch": 1.1477681197307366,
-      "grad_norm": 0.42817962169647217,
-      "learning_rate": 6.98608555592875e-06,
-      "loss": 0.3134,
-      "step": 17562
-    },
-    {
-      "epoch": 1.1478334749362786,
-      "grad_norm": 0.4315100610256195,
-      "learning_rate": 6.985765090802434e-06,
-      "loss": 0.3078,
-      "step": 17563
-    },
-    {
-      "epoch": 1.1478988301418207,
-      "grad_norm": 0.4427310824394226,
-      "learning_rate": 6.985444615990852e-06,
-      "loss": 0.3253,
-      "step": 17564
-    },
-    {
-      "epoch": 1.147964185347363,
-      "grad_norm": 0.43725985288619995,
-      "learning_rate": 6.985124131495571e-06,
-      "loss": 0.3281,
-      "step": 17565
-    },
-    {
-      "epoch": 1.148029540552905,
-      "grad_norm": 0.4995577037334442,
-      "learning_rate": 6.984803637318149e-06,
-      "loss": 0.4056,
-      "step": 17566
-    },
-    {
-      "epoch": 1.1480948957584471,
-      "grad_norm": 0.43872538208961487,
-      "learning_rate": 6.984483133460155e-06,
-      "loss": 0.3165,
-      "step": 17567
-    },
-    {
-      "epoch": 1.1481602509639892,
-      "grad_norm": 0.4488801062107086,
-      "learning_rate": 6.984162619923149e-06,
-      "loss": 0.3374,
-      "step": 17568
-    },
-    {
-      "epoch": 1.1482256061695315,
-      "grad_norm": 0.49378132820129395,
-      "learning_rate": 6.983842096708694e-06,
-      "loss": 0.3707,
-      "step": 17569
-    },
-    {
-      "epoch": 1.1482909613750736,
-      "grad_norm": 0.4640345275402069,
-      "learning_rate": 6.983521563818353e-06,
-      "loss": 0.3884,
-      "step": 17570
-    },
-    {
-      "epoch": 1.1483563165806157,
-      "grad_norm": 0.43276870250701904,
-      "learning_rate": 6.983201021253692e-06,
-      "loss": 0.3334,
-      "step": 17571
-    },
-    {
-      "epoch": 1.1484216717861577,
-      "grad_norm": 0.4319801330566406,
-      "learning_rate": 6.98288046901627e-06,
-      "loss": 0.3075,
-      "step": 17572
-    },
-    {
-      "epoch": 1.1484870269916998,
-      "grad_norm": 0.456879198551178,
-      "learning_rate": 6.982559907107655e-06,
-      "loss": 0.3089,
-      "step": 17573
-    },
-    {
-      "epoch": 1.148552382197242,
-      "grad_norm": 0.4286223351955414,
-      "learning_rate": 6.982239335529408e-06,
-      "loss": 0.3112,
-      "step": 17574
-    },
-    {
-      "epoch": 1.1486177374027842,
-      "grad_norm": 0.4080653786659241,
-      "learning_rate": 6.981918754283092e-06,
-      "loss": 0.2859,
-      "step": 17575
-    },
-    {
-      "epoch": 1.1486830926083262,
-      "grad_norm": 0.46638861298561096,
-      "learning_rate": 6.981598163370273e-06,
-      "loss": 0.3636,
-      "step": 17576
-    },
-    {
-      "epoch": 1.1487484478138683,
-      "grad_norm": 0.4485006332397461,
-      "learning_rate": 6.981277562792512e-06,
-      "loss": 0.3506,
-      "step": 17577
-    },
-    {
-      "epoch": 1.1488138030194106,
-      "grad_norm": 0.4528709352016449,
-      "learning_rate": 6.9809569525513755e-06,
-      "loss": 0.3436,
-      "step": 17578
-    },
-    {
-      "epoch": 1.1488791582249527,
-      "grad_norm": 0.43235430121421814,
-      "learning_rate": 6.980636332648424e-06,
-      "loss": 0.3247,
-      "step": 17579
-    },
-    {
-      "epoch": 1.1489445134304948,
-      "grad_norm": 0.4458158016204834,
-      "learning_rate": 6.980315703085224e-06,
-      "loss": 0.3461,
-      "step": 17580
-    },
-    {
-      "epoch": 1.1490098686360368,
-      "grad_norm": 0.4641415476799011,
-      "learning_rate": 6.979995063863339e-06,
-      "loss": 0.3714,
-      "step": 17581
-    },
-    {
-      "epoch": 1.149075223841579,
-      "grad_norm": 0.44364693760871887,
-      "learning_rate": 6.97967441498433e-06,
-      "loss": 0.3324,
-      "step": 17582
-    },
-    {
-      "epoch": 1.1491405790471212,
-      "grad_norm": 0.45736822485923767,
-      "learning_rate": 6.979353756449765e-06,
-      "loss": 0.3605,
-      "step": 17583
-    },
-    {
-      "epoch": 1.1492059342526633,
-      "grad_norm": 0.4471518397331238,
-      "learning_rate": 6.979033088261205e-06,
-      "loss": 0.3289,
-      "step": 17584
-    },
-    {
-      "epoch": 1.1492712894582053,
-      "grad_norm": 0.45127806067466736,
-      "learning_rate": 6.978712410420215e-06,
-      "loss": 0.331,
-      "step": 17585
-    },
-    {
-      "epoch": 1.1493366446637474,
-      "grad_norm": 0.4386953115463257,
-      "learning_rate": 6.978391722928359e-06,
-      "loss": 0.3444,
-      "step": 17586
-    },
-    {
-      "epoch": 1.1494019998692897,
-      "grad_norm": 0.4590171277523041,
-      "learning_rate": 6.978071025787202e-06,
-      "loss": 0.3367,
-      "step": 17587
-    },
-    {
-      "epoch": 1.1494673550748318,
-      "grad_norm": 0.44228118658065796,
-      "learning_rate": 6.977750318998306e-06,
-      "loss": 0.34,
-      "step": 17588
-    },
-    {
-      "epoch": 1.1495327102803738,
-      "grad_norm": 0.4404328167438507,
-      "learning_rate": 6.9774296025632396e-06,
-      "loss": 0.3181,
-      "step": 17589
-    },
-    {
-      "epoch": 1.149598065485916,
-      "grad_norm": 0.4597437381744385,
-      "learning_rate": 6.977108876483562e-06,
-      "loss": 0.3594,
-      "step": 17590
-    },
-    {
-      "epoch": 1.149663420691458,
-      "grad_norm": 0.4291709065437317,
-      "learning_rate": 6.976788140760839e-06,
-      "loss": 0.3273,
-      "step": 17591
-    },
-    {
-      "epoch": 1.1497287758970003,
-      "grad_norm": 0.44610679149627686,
-      "learning_rate": 6.9764673953966355e-06,
-      "loss": 0.3387,
-      "step": 17592
-    },
-    {
-      "epoch": 1.1497941311025424,
-      "grad_norm": 0.44380804896354675,
-      "learning_rate": 6.976146640392516e-06,
-      "loss": 0.3483,
-      "step": 17593
-    },
-    {
-      "epoch": 1.1498594863080844,
-      "grad_norm": 0.4359813332557678,
-      "learning_rate": 6.975825875750045e-06,
-      "loss": 0.3295,
-      "step": 17594
-    },
-    {
-      "epoch": 1.1499248415136265,
-      "grad_norm": 0.4144361615180969,
-      "learning_rate": 6.975505101470786e-06,
-      "loss": 0.3056,
-      "step": 17595
-    },
-    {
-      "epoch": 1.1499901967191688,
-      "grad_norm": 0.4788469970226288,
-      "learning_rate": 6.975184317556305e-06,
-      "loss": 0.3747,
-      "step": 17596
-    },
-    {
-      "epoch": 1.1500555519247109,
-      "grad_norm": 0.41650545597076416,
-      "learning_rate": 6.9748635240081656e-06,
-      "loss": 0.3002,
-      "step": 17597
-    },
-    {
-      "epoch": 1.150120907130253,
-      "grad_norm": 0.5181236863136292,
-      "learning_rate": 6.974542720827932e-06,
-      "loss": 0.3124,
-      "step": 17598
-    },
-    {
-      "epoch": 1.150186262335795,
-      "grad_norm": 0.42471063137054443,
-      "learning_rate": 6.9742219080171706e-06,
-      "loss": 0.2872,
-      "step": 17599
-    },
-    {
-      "epoch": 1.150251617541337,
-      "grad_norm": 0.43877092003822327,
-      "learning_rate": 6.973901085577445e-06,
-      "loss": 0.3382,
-      "step": 17600
-    },
-    {
-      "epoch": 1.1503169727468794,
-      "grad_norm": 0.4526621699333191,
-      "learning_rate": 6.9735802535103194e-06,
-      "loss": 0.321,
-      "step": 17601
-    },
-    {
-      "epoch": 1.1503823279524215,
-      "grad_norm": 0.4256632328033447,
-      "learning_rate": 6.973259411817359e-06,
-      "loss": 0.323,
-      "step": 17602
-    },
-    {
-      "epoch": 1.1504476831579635,
-      "grad_norm": 0.4291374683380127,
-      "learning_rate": 6.972938560500129e-06,
-      "loss": 0.3168,
-      "step": 17603
-    },
-    {
-      "epoch": 1.1505130383635056,
-      "grad_norm": 0.43796679377555847,
-      "learning_rate": 6.972617699560194e-06,
-      "loss": 0.3158,
-      "step": 17604
-    },
-    {
-      "epoch": 1.150578393569048,
-      "grad_norm": 0.47402623295783997,
-      "learning_rate": 6.972296828999119e-06,
-      "loss": 0.3625,
-      "step": 17605
-    },
-    {
-      "epoch": 1.15064374877459,
-      "grad_norm": 0.4425138831138611,
-      "learning_rate": 6.971975948818469e-06,
-      "loss": 0.3465,
-      "step": 17606
-    },
-    {
-      "epoch": 1.150709103980132,
-      "grad_norm": 0.4444705545902252,
-      "learning_rate": 6.971655059019811e-06,
-      "loss": 0.342,
-      "step": 17607
-    },
-    {
-      "epoch": 1.1507744591856741,
-      "grad_norm": 0.4593660235404968,
-      "learning_rate": 6.9713341596047066e-06,
-      "loss": 0.3692,
-      "step": 17608
-    },
-    {
-      "epoch": 1.1508398143912162,
-      "grad_norm": 0.4351654052734375,
-      "learning_rate": 6.971013250574722e-06,
-      "loss": 0.3101,
-      "step": 17609
-    },
-    {
-      "epoch": 1.1509051695967583,
-      "grad_norm": 0.5078103542327881,
-      "learning_rate": 6.970692331931425e-06,
-      "loss": 0.4376,
-      "step": 17610
-    },
-    {
-      "epoch": 1.1509705248023006,
-      "grad_norm": 0.4377695918083191,
-      "learning_rate": 6.970371403676377e-06,
-      "loss": 0.3433,
-      "step": 17611
-    },
-    {
-      "epoch": 1.1510358800078426,
-      "grad_norm": 0.4759107232093811,
-      "learning_rate": 6.9700504658111465e-06,
-      "loss": 0.3577,
-      "step": 17612
-    },
-    {
-      "epoch": 1.1511012352133847,
-      "grad_norm": 0.4467790126800537,
-      "learning_rate": 6.969729518337296e-06,
-      "loss": 0.3493,
-      "step": 17613
-    },
-    {
-      "epoch": 1.1511665904189268,
-      "grad_norm": 0.5905129909515381,
-      "learning_rate": 6.969408561256393e-06,
-      "loss": 0.3877,
-      "step": 17614
-    },
-    {
-      "epoch": 1.151231945624469,
-      "grad_norm": 0.4773085117340088,
-      "learning_rate": 6.969087594570001e-06,
-      "loss": 0.3467,
-      "step": 17615
-    },
-    {
-      "epoch": 1.1512973008300111,
-      "grad_norm": 0.4675079584121704,
-      "learning_rate": 6.968766618279688e-06,
-      "loss": 0.3375,
-      "step": 17616
-    },
-    {
-      "epoch": 1.1513626560355532,
-      "grad_norm": 0.39735788106918335,
-      "learning_rate": 6.968445632387017e-06,
-      "loss": 0.2697,
-      "step": 17617
-    },
-    {
-      "epoch": 1.1514280112410953,
-      "grad_norm": 0.4492117464542389,
-      "learning_rate": 6.9681246368935565e-06,
-      "loss": 0.3555,
-      "step": 17618
-    },
-    {
-      "epoch": 1.1514933664466374,
-      "grad_norm": 0.42149829864501953,
-      "learning_rate": 6.967803631800868e-06,
-      "loss": 0.3139,
-      "step": 17619
-    },
-    {
-      "epoch": 1.1515587216521797,
-      "grad_norm": 0.4348292648792267,
-      "learning_rate": 6.9674826171105214e-06,
-      "loss": 0.347,
-      "step": 17620
-    },
-    {
-      "epoch": 1.1516240768577217,
-      "grad_norm": 0.4530472457408905,
-      "learning_rate": 6.96716159282408e-06,
-      "loss": 0.3629,
-      "step": 17621
-    },
-    {
-      "epoch": 1.1516894320632638,
-      "grad_norm": 0.48482373356819153,
-      "learning_rate": 6.96684055894311e-06,
-      "loss": 0.3436,
-      "step": 17622
-    },
-    {
-      "epoch": 1.1517547872688059,
-      "grad_norm": 0.4720245599746704,
-      "learning_rate": 6.9665195154691765e-06,
-      "loss": 0.3515,
-      "step": 17623
-    },
-    {
-      "epoch": 1.1518201424743482,
-      "grad_norm": 0.46872296929359436,
-      "learning_rate": 6.966198462403846e-06,
-      "loss": 0.3815,
-      "step": 17624
-    },
-    {
-      "epoch": 1.1518854976798902,
-      "grad_norm": 0.4645110070705414,
-      "learning_rate": 6.965877399748686e-06,
-      "loss": 0.353,
-      "step": 17625
-    },
-    {
-      "epoch": 1.1519508528854323,
-      "grad_norm": 0.4311627745628357,
-      "learning_rate": 6.96555632750526e-06,
-      "loss": 0.3039,
-      "step": 17626
-    },
-    {
-      "epoch": 1.1520162080909744,
-      "grad_norm": 0.4705878496170044,
-      "learning_rate": 6.965235245675135e-06,
-      "loss": 0.362,
-      "step": 17627
-    },
-    {
-      "epoch": 1.1520815632965165,
-      "grad_norm": 0.45533472299575806,
-      "learning_rate": 6.964914154259876e-06,
-      "loss": 0.3374,
-      "step": 17628
-    },
-    {
-      "epoch": 1.1521469185020587,
-      "grad_norm": 0.42912060022354126,
-      "learning_rate": 6.964593053261051e-06,
-      "loss": 0.3403,
-      "step": 17629
-    },
-    {
-      "epoch": 1.1522122737076008,
-      "grad_norm": 0.458856463432312,
-      "learning_rate": 6.9642719426802255e-06,
-      "loss": 0.344,
-      "step": 17630
-    },
-    {
-      "epoch": 1.152277628913143,
-      "grad_norm": 0.41162678599357605,
-      "learning_rate": 6.963950822518963e-06,
-      "loss": 0.2972,
-      "step": 17631
-    },
-    {
-      "epoch": 1.152342984118685,
-      "grad_norm": 0.46424978971481323,
-      "learning_rate": 6.963629692778835e-06,
-      "loss": 0.3678,
-      "step": 17632
-    },
-    {
-      "epoch": 1.1524083393242273,
-      "grad_norm": 0.46035653352737427,
-      "learning_rate": 6.963308553461402e-06,
-      "loss": 0.3763,
-      "step": 17633
-    },
-    {
-      "epoch": 1.1524736945297693,
-      "grad_norm": 0.45069998502731323,
-      "learning_rate": 6.962987404568235e-06,
-      "loss": 0.3582,
-      "step": 17634
-    },
-    {
-      "epoch": 1.1525390497353114,
-      "grad_norm": 0.4333679676055908,
-      "learning_rate": 6.962666246100897e-06,
-      "loss": 0.33,
-      "step": 17635
-    },
-    {
-      "epoch": 1.1526044049408535,
-      "grad_norm": 0.4406760036945343,
-      "learning_rate": 6.962345078060957e-06,
-      "loss": 0.3489,
-      "step": 17636
-    },
-    {
-      "epoch": 1.1526697601463956,
-      "grad_norm": 0.47684451937675476,
-      "learning_rate": 6.962023900449979e-06,
-      "loss": 0.33,
-      "step": 17637
-    },
-    {
-      "epoch": 1.1527351153519378,
-      "grad_norm": 0.4457132816314697,
-      "learning_rate": 6.961702713269532e-06,
-      "loss": 0.3284,
-      "step": 17638
-    },
-    {
-      "epoch": 1.15280047055748,
-      "grad_norm": 0.46327537298202515,
-      "learning_rate": 6.96138151652118e-06,
-      "loss": 0.3782,
-      "step": 17639
-    },
-    {
-      "epoch": 1.152865825763022,
-      "grad_norm": 0.4420397877693176,
-      "learning_rate": 6.961060310206491e-06,
-      "loss": 0.3444,
-      "step": 17640
-    },
-    {
-      "epoch": 1.152931180968564,
-      "grad_norm": 0.4100053608417511,
-      "learning_rate": 6.9607390943270325e-06,
-      "loss": 0.3062,
-      "step": 17641
-    },
-    {
-      "epoch": 1.1529965361741064,
-      "grad_norm": 0.49853360652923584,
-      "learning_rate": 6.960417868884368e-06,
-      "loss": 0.3897,
-      "step": 17642
-    },
-    {
-      "epoch": 1.1530618913796484,
-      "grad_norm": 0.49394646286964417,
-      "learning_rate": 6.960096633880069e-06,
-      "loss": 0.3565,
-      "step": 17643
-    },
-    {
-      "epoch": 1.1531272465851905,
-      "grad_norm": 0.46234485507011414,
-      "learning_rate": 6.959775389315698e-06,
-      "loss": 0.2861,
-      "step": 17644
-    },
-    {
-      "epoch": 1.1531926017907326,
-      "grad_norm": 0.4874555766582489,
-      "learning_rate": 6.959454135192825e-06,
-      "loss": 0.376,
-      "step": 17645
-    },
-    {
-      "epoch": 1.1532579569962746,
-      "grad_norm": 0.491465300321579,
-      "learning_rate": 6.9591328715130135e-06,
-      "loss": 0.3935,
-      "step": 17646
-    },
-    {
-      "epoch": 1.153323312201817,
-      "grad_norm": 0.46655192971229553,
-      "learning_rate": 6.958811598277834e-06,
-      "loss": 0.3553,
-      "step": 17647
-    },
-    {
-      "epoch": 1.153388667407359,
-      "grad_norm": 0.4627162516117096,
-      "learning_rate": 6.958490315488851e-06,
-      "loss": 0.3721,
-      "step": 17648
-    },
-    {
-      "epoch": 1.153454022612901,
-      "grad_norm": 0.43710124492645264,
-      "learning_rate": 6.958169023147632e-06,
-      "loss": 0.3179,
-      "step": 17649
-    },
-    {
-      "epoch": 1.1535193778184432,
-      "grad_norm": 0.4593852162361145,
-      "learning_rate": 6.957847721255745e-06,
-      "loss": 0.3253,
-      "step": 17650
-    },
-    {
-      "epoch": 1.1535847330239855,
-      "grad_norm": 0.49218297004699707,
-      "learning_rate": 6.957526409814755e-06,
-      "loss": 0.3889,
-      "step": 17651
-    },
-    {
-      "epoch": 1.1536500882295275,
-      "grad_norm": 0.46324622631073,
-      "learning_rate": 6.957205088826233e-06,
-      "loss": 0.3722,
-      "step": 17652
-    },
-    {
-      "epoch": 1.1537154434350696,
-      "grad_norm": 0.45687490701675415,
-      "learning_rate": 6.956883758291742e-06,
-      "loss": 0.3406,
-      "step": 17653
-    },
-    {
-      "epoch": 1.1537807986406117,
-      "grad_norm": 0.46883320808410645,
-      "learning_rate": 6.956562418212851e-06,
-      "loss": 0.3465,
-      "step": 17654
-    },
-    {
-      "epoch": 1.1538461538461537,
-      "grad_norm": 0.47016748785972595,
-      "learning_rate": 6.956241068591127e-06,
-      "loss": 0.36,
-      "step": 17655
-    },
-    {
-      "epoch": 1.153911509051696,
-      "grad_norm": 0.43132883310317993,
-      "learning_rate": 6.955919709428139e-06,
-      "loss": 0.323,
-      "step": 17656
-    },
-    {
-      "epoch": 1.153976864257238,
-      "grad_norm": 0.4467172920703888,
-      "learning_rate": 6.955598340725451e-06,
-      "loss": 0.3083,
-      "step": 17657
-    },
-    {
-      "epoch": 1.1540422194627802,
-      "grad_norm": 0.4533108174800873,
-      "learning_rate": 6.9552769624846335e-06,
-      "loss": 0.3499,
-      "step": 17658
-    },
-    {
-      "epoch": 1.1541075746683223,
-      "grad_norm": 0.4377674162387848,
-      "learning_rate": 6.954955574707254e-06,
-      "loss": 0.3076,
-      "step": 17659
-    },
-    {
-      "epoch": 1.1541729298738646,
-      "grad_norm": 0.45000824332237244,
-      "learning_rate": 6.954634177394877e-06,
-      "loss": 0.3548,
-      "step": 17660
-    },
-    {
-      "epoch": 1.1542382850794066,
-      "grad_norm": 0.45199882984161377,
-      "learning_rate": 6.954312770549073e-06,
-      "loss": 0.3378,
-      "step": 17661
-    },
-    {
-      "epoch": 1.1543036402849487,
-      "grad_norm": 0.4631260335445404,
-      "learning_rate": 6.953991354171408e-06,
-      "loss": 0.3984,
-      "step": 17662
-    },
-    {
-      "epoch": 1.1543689954904908,
-      "grad_norm": 0.44825875759124756,
-      "learning_rate": 6.953669928263451e-06,
-      "loss": 0.3303,
-      "step": 17663
-    },
-    {
-      "epoch": 1.1544343506960328,
-      "grad_norm": 0.45595115423202515,
-      "learning_rate": 6.953348492826768e-06,
-      "loss": 0.3467,
-      "step": 17664
-    },
-    {
-      "epoch": 1.1544997059015751,
-      "grad_norm": 0.43641722202301025,
-      "learning_rate": 6.953027047862928e-06,
-      "loss": 0.3045,
-      "step": 17665
-    },
-    {
-      "epoch": 1.1545650611071172,
-      "grad_norm": 0.6549420356750488,
-      "learning_rate": 6.9527055933735e-06,
-      "loss": 0.3434,
-      "step": 17666
-    },
-    {
-      "epoch": 1.1546304163126593,
-      "grad_norm": 0.43091461062431335,
-      "learning_rate": 6.952384129360048e-06,
-      "loss": 0.3141,
-      "step": 17667
-    },
-    {
-      "epoch": 1.1546957715182014,
-      "grad_norm": 0.4651185870170593,
-      "learning_rate": 6.952062655824143e-06,
-      "loss": 0.3909,
-      "step": 17668
-    },
-    {
-      "epoch": 1.1547611267237436,
-      "grad_norm": 0.47757580876350403,
-      "learning_rate": 6.951741172767352e-06,
-      "loss": 0.3397,
-      "step": 17669
-    },
-    {
-      "epoch": 1.1548264819292857,
-      "grad_norm": 0.4739300310611725,
-      "learning_rate": 6.951419680191245e-06,
-      "loss": 0.3436,
-      "step": 17670
-    },
-    {
-      "epoch": 1.1548918371348278,
-      "grad_norm": 0.4194020926952362,
-      "learning_rate": 6.951098178097386e-06,
-      "loss": 0.295,
-      "step": 17671
-    },
-    {
-      "epoch": 1.1549571923403699,
-      "grad_norm": 0.5356296896934509,
-      "learning_rate": 6.950776666487347e-06,
-      "loss": 0.419,
-      "step": 17672
-    },
-    {
-      "epoch": 1.155022547545912,
-      "grad_norm": 0.4514598548412323,
-      "learning_rate": 6.950455145362694e-06,
-      "loss": 0.3575,
-      "step": 17673
-    },
-    {
-      "epoch": 1.1550879027514542,
-      "grad_norm": 0.457670658826828,
-      "learning_rate": 6.950133614724994e-06,
-      "loss": 0.3522,
-      "step": 17674
-    },
-    {
-      "epoch": 1.1551532579569963,
-      "grad_norm": 0.4478275179862976,
-      "learning_rate": 6.949812074575817e-06,
-      "loss": 0.3332,
-      "step": 17675
-    },
-    {
-      "epoch": 1.1552186131625384,
-      "grad_norm": 0.44643643498420715,
-      "learning_rate": 6.949490524916734e-06,
-      "loss": 0.3601,
-      "step": 17676
-    },
-    {
-      "epoch": 1.1552839683680804,
-      "grad_norm": 0.3907199800014496,
-      "learning_rate": 6.949168965749307e-06,
-      "loss": 0.2759,
-      "step": 17677
-    },
-    {
-      "epoch": 1.1553493235736227,
-      "grad_norm": 0.4266630709171295,
-      "learning_rate": 6.948847397075108e-06,
-      "loss": 0.3327,
-      "step": 17678
-    },
-    {
-      "epoch": 1.1554146787791648,
-      "grad_norm": 0.4935249090194702,
-      "learning_rate": 6.948525818895706e-06,
-      "loss": 0.3744,
-      "step": 17679
-    },
-    {
-      "epoch": 1.155480033984707,
-      "grad_norm": 0.46187731623649597,
-      "learning_rate": 6.948204231212668e-06,
-      "loss": 0.3264,
-      "step": 17680
-    },
-    {
-      "epoch": 1.155545389190249,
-      "grad_norm": 0.4567898213863373,
-      "learning_rate": 6.947882634027564e-06,
-      "loss": 0.3579,
-      "step": 17681
-    },
-    {
-      "epoch": 1.155610744395791,
-      "grad_norm": 0.4668227732181549,
-      "learning_rate": 6.94756102734196e-06,
-      "loss": 0.3983,
-      "step": 17682
-    },
-    {
-      "epoch": 1.1556760996013333,
-      "grad_norm": 0.4595906436443329,
-      "learning_rate": 6.947239411157428e-06,
-      "loss": 0.3598,
-      "step": 17683
-    },
-    {
-      "epoch": 1.1557414548068754,
-      "grad_norm": 0.44565173983573914,
-      "learning_rate": 6.9469177854755344e-06,
-      "loss": 0.3388,
-      "step": 17684
-    },
-    {
-      "epoch": 1.1558068100124175,
-      "grad_norm": 0.409270703792572,
-      "learning_rate": 6.9465961502978485e-06,
-      "loss": 0.3103,
-      "step": 17685
-    },
-    {
-      "epoch": 1.1558721652179595,
-      "grad_norm": 0.44563353061676025,
-      "learning_rate": 6.946274505625939e-06,
-      "loss": 0.3063,
-      "step": 17686
-    },
-    {
-      "epoch": 1.1559375204235018,
-      "grad_norm": 0.4690755605697632,
-      "learning_rate": 6.945952851461373e-06,
-      "loss": 0.4142,
-      "step": 17687
-    },
-    {
-      "epoch": 1.156002875629044,
-      "grad_norm": 0.6353440880775452,
-      "learning_rate": 6.945631187805721e-06,
-      "loss": 0.3421,
-      "step": 17688
-    },
-    {
-      "epoch": 1.156068230834586,
-      "grad_norm": 0.4302791953086853,
-      "learning_rate": 6.945309514660552e-06,
-      "loss": 0.3397,
-      "step": 17689
-    },
-    {
-      "epoch": 1.156133586040128,
-      "grad_norm": 0.4376680254936218,
-      "learning_rate": 6.944987832027434e-06,
-      "loss": 0.3278,
-      "step": 17690
-    },
-    {
-      "epoch": 1.1561989412456701,
-      "grad_norm": 0.4576703906059265,
-      "learning_rate": 6.944666139907937e-06,
-      "loss": 0.3536,
-      "step": 17691
-    },
-    {
-      "epoch": 1.1562642964512124,
-      "grad_norm": 0.4490030109882355,
-      "learning_rate": 6.944344438303631e-06,
-      "loss": 0.3558,
-      "step": 17692
-    },
-    {
-      "epoch": 1.1563296516567545,
-      "grad_norm": 0.45046737790107727,
-      "learning_rate": 6.944022727216082e-06,
-      "loss": 0.3403,
-      "step": 17693
-    },
-    {
-      "epoch": 1.1563950068622966,
-      "grad_norm": 0.44379544258117676,
-      "learning_rate": 6.943701006646863e-06,
-      "loss": 0.3344,
-      "step": 17694
-    },
-    {
-      "epoch": 1.1564603620678386,
-      "grad_norm": 0.4549551010131836,
-      "learning_rate": 6.943379276597537e-06,
-      "loss": 0.3387,
-      "step": 17695
-    },
-    {
-      "epoch": 1.156525717273381,
-      "grad_norm": 0.4415263831615448,
-      "learning_rate": 6.9430575370696796e-06,
-      "loss": 0.3227,
-      "step": 17696
-    },
-    {
-      "epoch": 1.156591072478923,
-      "grad_norm": 0.4730100631713867,
-      "learning_rate": 6.9427357880648576e-06,
-      "loss": 0.3809,
-      "step": 17697
-    },
-    {
-      "epoch": 1.156656427684465,
-      "grad_norm": 0.42756032943725586,
-      "learning_rate": 6.942414029584639e-06,
-      "loss": 0.2957,
-      "step": 17698
-    },
-    {
-      "epoch": 1.1567217828900072,
-      "grad_norm": 0.4337129592895508,
-      "learning_rate": 6.942092261630594e-06,
-      "loss": 0.3325,
-      "step": 17699
-    },
-    {
-      "epoch": 1.1567871380955492,
-      "grad_norm": 0.4741310775279999,
-      "learning_rate": 6.941770484204294e-06,
-      "loss": 0.4079,
-      "step": 17700
-    },
-    {
-      "epoch": 1.1568524933010915,
-      "grad_norm": 0.42564648389816284,
-      "learning_rate": 6.9414486973073045e-06,
-      "loss": 0.3486,
-      "step": 17701
-    },
-    {
-      "epoch": 1.1569178485066336,
-      "grad_norm": 0.42298024892807007,
-      "learning_rate": 6.941126900941198e-06,
-      "loss": 0.3225,
-      "step": 17702
-    },
-    {
-      "epoch": 1.1569832037121757,
-      "grad_norm": 0.4448823630809784,
-      "learning_rate": 6.940805095107544e-06,
-      "loss": 0.3463,
-      "step": 17703
-    },
-    {
-      "epoch": 1.1570485589177177,
-      "grad_norm": 0.42664510011672974,
-      "learning_rate": 6.94048327980791e-06,
-      "loss": 0.3466,
-      "step": 17704
-    },
-    {
-      "epoch": 1.15711391412326,
-      "grad_norm": 0.4197862446308136,
-      "learning_rate": 6.940161455043867e-06,
-      "loss": 0.3309,
-      "step": 17705
-    },
-    {
-      "epoch": 1.157179269328802,
-      "grad_norm": 0.44740229845046997,
-      "learning_rate": 6.939839620816985e-06,
-      "loss": 0.3104,
-      "step": 17706
-    },
-    {
-      "epoch": 1.1572446245343442,
-      "grad_norm": 0.4196939170360565,
-      "learning_rate": 6.939517777128833e-06,
-      "loss": 0.2841,
-      "step": 17707
-    },
-    {
-      "epoch": 1.1573099797398863,
-      "grad_norm": 0.46484148502349854,
-      "learning_rate": 6.93919592398098e-06,
-      "loss": 0.3176,
-      "step": 17708
-    },
-    {
-      "epoch": 1.1573753349454283,
-      "grad_norm": 0.41649261116981506,
-      "learning_rate": 6.938874061374997e-06,
-      "loss": 0.3064,
-      "step": 17709
-    },
-    {
-      "epoch": 1.1574406901509706,
-      "grad_norm": 0.427366703748703,
-      "learning_rate": 6.938552189312454e-06,
-      "loss": 0.3121,
-      "step": 17710
-    },
-    {
-      "epoch": 1.1575060453565127,
-      "grad_norm": 0.4428960382938385,
-      "learning_rate": 6.93823030779492e-06,
-      "loss": 0.3787,
-      "step": 17711
-    },
-    {
-      "epoch": 1.1575714005620548,
-      "grad_norm": 0.45263683795928955,
-      "learning_rate": 6.937908416823967e-06,
-      "loss": 0.3844,
-      "step": 17712
-    },
-    {
-      "epoch": 1.1576367557675968,
-      "grad_norm": 0.44688180088996887,
-      "learning_rate": 6.937586516401162e-06,
-      "loss": 0.3575,
-      "step": 17713
-    },
-    {
-      "epoch": 1.1577021109731391,
-      "grad_norm": 0.47628867626190186,
-      "learning_rate": 6.937264606528074e-06,
-      "loss": 0.3529,
-      "step": 17714
-    },
-    {
-      "epoch": 1.1577674661786812,
-      "grad_norm": 0.4480690360069275,
-      "learning_rate": 6.9369426872062804e-06,
-      "loss": 0.3399,
-      "step": 17715
-    },
-    {
-      "epoch": 1.1578328213842233,
-      "grad_norm": 0.47651416063308716,
-      "learning_rate": 6.9366207584373425e-06,
-      "loss": 0.3867,
-      "step": 17716
-    },
-    {
-      "epoch": 1.1578981765897653,
-      "grad_norm": 0.4467816948890686,
-      "learning_rate": 6.936298820222835e-06,
-      "loss": 0.3522,
-      "step": 17717
-    },
-    {
-      "epoch": 1.1579635317953074,
-      "grad_norm": 0.44865909218788147,
-      "learning_rate": 6.935976872564327e-06,
-      "loss": 0.3101,
-      "step": 17718
-    },
-    {
-      "epoch": 1.1580288870008495,
-      "grad_norm": 0.4765969216823578,
-      "learning_rate": 6.93565491546339e-06,
-      "loss": 0.3837,
-      "step": 17719
-    },
-    {
-      "epoch": 1.1580942422063918,
-      "grad_norm": 0.42105284333229065,
-      "learning_rate": 6.935332948921594e-06,
-      "loss": 0.3226,
-      "step": 17720
-    },
-    {
-      "epoch": 1.1581595974119339,
-      "grad_norm": 0.5028206706047058,
-      "learning_rate": 6.935010972940508e-06,
-      "loss": 0.4379,
-      "step": 17721
-    },
-    {
-      "epoch": 1.158224952617476,
-      "grad_norm": 0.4653327763080597,
-      "learning_rate": 6.934688987521702e-06,
-      "loss": 0.3671,
-      "step": 17722
-    },
-    {
-      "epoch": 1.1582903078230182,
-      "grad_norm": 0.4500982463359833,
-      "learning_rate": 6.934366992666749e-06,
-      "loss": 0.3709,
-      "step": 17723
-    },
-    {
-      "epoch": 1.1583556630285603,
-      "grad_norm": 0.44609150290489197,
-      "learning_rate": 6.934044988377218e-06,
-      "loss": 0.3462,
-      "step": 17724
-    },
-    {
-      "epoch": 1.1584210182341024,
-      "grad_norm": 0.4506785273551941,
-      "learning_rate": 6.9337229746546784e-06,
-      "loss": 0.3338,
-      "step": 17725
-    },
-    {
-      "epoch": 1.1584863734396444,
-      "grad_norm": 0.48381128907203674,
-      "learning_rate": 6.933400951500704e-06,
-      "loss": 0.3838,
-      "step": 17726
-    },
-    {
-      "epoch": 1.1585517286451865,
-      "grad_norm": 0.4423137605190277,
-      "learning_rate": 6.93307891891686e-06,
-      "loss": 0.3434,
-      "step": 17727
-    },
-    {
-      "epoch": 1.1586170838507286,
-      "grad_norm": 0.47422897815704346,
-      "learning_rate": 6.932756876904724e-06,
-      "loss": 0.385,
-      "step": 17728
-    },
-    {
-      "epoch": 1.1586824390562709,
-      "grad_norm": 0.40563276410102844,
-      "learning_rate": 6.932434825465862e-06,
-      "loss": 0.2791,
-      "step": 17729
-    },
-    {
-      "epoch": 1.158747794261813,
-      "grad_norm": 0.4454939365386963,
-      "learning_rate": 6.932112764601845e-06,
-      "loss": 0.3452,
-      "step": 17730
-    },
-    {
-      "epoch": 1.158813149467355,
-      "grad_norm": 0.4607152044773102,
-      "learning_rate": 6.931790694314246e-06,
-      "loss": 0.3982,
-      "step": 17731
-    },
-    {
-      "epoch": 1.158878504672897,
-      "grad_norm": 0.4172993004322052,
-      "learning_rate": 6.931468614604633e-06,
-      "loss": 0.3101,
-      "step": 17732
-    },
-    {
-      "epoch": 1.1589438598784394,
-      "grad_norm": 0.42265626788139343,
-      "learning_rate": 6.931146525474579e-06,
-      "loss": 0.3181,
-      "step": 17733
-    },
-    {
-      "epoch": 1.1590092150839815,
-      "grad_norm": 0.45781445503234863,
-      "learning_rate": 6.930824426925654e-06,
-      "loss": 0.3529,
-      "step": 17734
-    },
-    {
-      "epoch": 1.1590745702895235,
-      "grad_norm": 0.4498709738254547,
-      "learning_rate": 6.930502318959431e-06,
-      "loss": 0.3337,
-      "step": 17735
-    },
-    {
-      "epoch": 1.1591399254950656,
-      "grad_norm": 0.4513479173183441,
-      "learning_rate": 6.930180201577479e-06,
-      "loss": 0.3381,
-      "step": 17736
-    },
-    {
-      "epoch": 1.1592052807006077,
-      "grad_norm": 0.4964551627635956,
-      "learning_rate": 6.929858074781368e-06,
-      "loss": 0.4226,
-      "step": 17737
-    },
-    {
-      "epoch": 1.15927063590615,
-      "grad_norm": 0.47749781608581543,
-      "learning_rate": 6.929535938572671e-06,
-      "loss": 0.3853,
-      "step": 17738
-    },
-    {
-      "epoch": 1.159335991111692,
-      "grad_norm": 0.44127658009529114,
-      "learning_rate": 6.92921379295296e-06,
-      "loss": 0.3188,
-      "step": 17739
-    },
-    {
-      "epoch": 1.1594013463172341,
-      "grad_norm": 0.47328656911849976,
-      "learning_rate": 6.928891637923804e-06,
-      "loss": 0.3156,
-      "step": 17740
-    },
-    {
-      "epoch": 1.1594667015227762,
-      "grad_norm": 0.48997533321380615,
-      "learning_rate": 6.928569473486776e-06,
-      "loss": 0.359,
-      "step": 17741
-    },
-    {
-      "epoch": 1.1595320567283185,
-      "grad_norm": 0.41671857237815857,
-      "learning_rate": 6.928247299643447e-06,
-      "loss": 0.2964,
-      "step": 17742
-    },
-    {
-      "epoch": 1.1595974119338606,
-      "grad_norm": 0.4364735186100006,
-      "learning_rate": 6.927925116395385e-06,
-      "loss": 0.3275,
-      "step": 17743
-    },
-    {
-      "epoch": 1.1596627671394026,
-      "grad_norm": 0.4537159502506256,
-      "learning_rate": 6.927602923744168e-06,
-      "loss": 0.3778,
-      "step": 17744
-    },
-    {
-      "epoch": 1.1597281223449447,
-      "grad_norm": 0.44723019003868103,
-      "learning_rate": 6.9272807216913606e-06,
-      "loss": 0.3503,
-      "step": 17745
-    },
-    {
-      "epoch": 1.1597934775504868,
-      "grad_norm": 0.43190619349479675,
-      "learning_rate": 6.926958510238539e-06,
-      "loss": 0.3023,
-      "step": 17746
-    },
-    {
-      "epoch": 1.159858832756029,
-      "grad_norm": 0.46130454540252686,
-      "learning_rate": 6.926636289387273e-06,
-      "loss": 0.3464,
-      "step": 17747
-    },
-    {
-      "epoch": 1.1599241879615712,
-      "grad_norm": 0.504646897315979,
-      "learning_rate": 6.926314059139134e-06,
-      "loss": 0.4161,
-      "step": 17748
-    },
-    {
-      "epoch": 1.1599895431671132,
-      "grad_norm": 0.4602302014827728,
-      "learning_rate": 6.925991819495694e-06,
-      "loss": 0.3713,
-      "step": 17749
-    },
-    {
-      "epoch": 1.1600548983726553,
-      "grad_norm": 0.46173661947250366,
-      "learning_rate": 6.925669570458526e-06,
-      "loss": 0.3683,
-      "step": 17750
-    },
-    {
-      "epoch": 1.1601202535781976,
-      "grad_norm": 0.46331456303596497,
-      "learning_rate": 6.925347312029201e-06,
-      "loss": 0.3431,
-      "step": 17751
-    },
-    {
-      "epoch": 1.1601856087837397,
-      "grad_norm": 0.4749436676502228,
-      "learning_rate": 6.925025044209287e-06,
-      "loss": 0.3864,
-      "step": 17752
-    },
-    {
-      "epoch": 1.1602509639892817,
-      "grad_norm": 0.4597143530845642,
-      "learning_rate": 6.924702767000361e-06,
-      "loss": 0.3766,
-      "step": 17753
-    },
-    {
-      "epoch": 1.1603163191948238,
-      "grad_norm": 0.4547629952430725,
-      "learning_rate": 6.924380480403991e-06,
-      "loss": 0.3411,
-      "step": 17754
-    },
-    {
-      "epoch": 1.1603816744003659,
-      "grad_norm": 0.5494674444198608,
-      "learning_rate": 6.924058184421753e-06,
-      "loss": 0.3224,
-      "step": 17755
-    },
-    {
-      "epoch": 1.1604470296059082,
-      "grad_norm": 0.49714577198028564,
-      "learning_rate": 6.923735879055215e-06,
-      "loss": 0.4199,
-      "step": 17756
-    },
-    {
-      "epoch": 1.1605123848114502,
-      "grad_norm": 0.43896931409835815,
-      "learning_rate": 6.923413564305953e-06,
-      "loss": 0.3384,
-      "step": 17757
-    },
-    {
-      "epoch": 1.1605777400169923,
-      "grad_norm": 0.49413934350013733,
-      "learning_rate": 6.923091240175534e-06,
-      "loss": 0.3911,
-      "step": 17758
-    },
-    {
-      "epoch": 1.1606430952225344,
-      "grad_norm": 0.4775960445404053,
-      "learning_rate": 6.9227689066655355e-06,
-      "loss": 0.3785,
-      "step": 17759
-    },
-    {
-      "epoch": 1.1607084504280767,
-      "grad_norm": 0.4625619947910309,
-      "learning_rate": 6.922446563777525e-06,
-      "loss": 0.372,
-      "step": 17760
-    },
-    {
-      "epoch": 1.1607738056336188,
-      "grad_norm": 0.45002758502960205,
-      "learning_rate": 6.9221242115130775e-06,
-      "loss": 0.3351,
-      "step": 17761
-    },
-    {
-      "epoch": 1.1608391608391608,
-      "grad_norm": 0.5079486966133118,
-      "learning_rate": 6.921801849873763e-06,
-      "loss": 0.3899,
-      "step": 17762
-    },
-    {
-      "epoch": 1.160904516044703,
-      "grad_norm": 0.4468036890029907,
-      "learning_rate": 6.921479478861156e-06,
-      "loss": 0.3339,
-      "step": 17763
-    },
-    {
-      "epoch": 1.160969871250245,
-      "grad_norm": 0.3972158133983612,
-      "learning_rate": 6.921157098476829e-06,
-      "loss": 0.2574,
-      "step": 17764
-    },
-    {
-      "epoch": 1.1610352264557873,
-      "grad_norm": 0.42045748233795166,
-      "learning_rate": 6.9208347087223505e-06,
-      "loss": 0.3088,
-      "step": 17765
-    },
-    {
-      "epoch": 1.1611005816613293,
-      "grad_norm": 0.48316705226898193,
-      "learning_rate": 6.920512309599298e-06,
-      "loss": 0.3765,
-      "step": 17766
-    },
-    {
-      "epoch": 1.1611659368668714,
-      "grad_norm": 0.4461493194103241,
-      "learning_rate": 6.920189901109241e-06,
-      "loss": 0.3493,
-      "step": 17767
-    },
-    {
-      "epoch": 1.1612312920724135,
-      "grad_norm": 0.4933953583240509,
-      "learning_rate": 6.919867483253753e-06,
-      "loss": 0.349,
-      "step": 17768
-    },
-    {
-      "epoch": 1.1612966472779558,
-      "grad_norm": 0.47226646542549133,
-      "learning_rate": 6.919545056034406e-06,
-      "loss": 0.3645,
-      "step": 17769
-    },
-    {
-      "epoch": 1.1613620024834979,
-      "grad_norm": 0.4418177008628845,
-      "learning_rate": 6.919222619452772e-06,
-      "loss": 0.3337,
-      "step": 17770
-    },
-    {
-      "epoch": 1.16142735768904,
-      "grad_norm": 0.43331435322761536,
-      "learning_rate": 6.918900173510426e-06,
-      "loss": 0.3225,
-      "step": 17771
-    },
-    {
-      "epoch": 1.161492712894582,
-      "grad_norm": 0.49668073654174805,
-      "learning_rate": 6.918577718208937e-06,
-      "loss": 0.3043,
-      "step": 17772
-    },
-    {
-      "epoch": 1.161558068100124,
-      "grad_norm": 0.44126200675964355,
-      "learning_rate": 6.918255253549882e-06,
-      "loss": 0.327,
-      "step": 17773
-    },
-    {
-      "epoch": 1.1616234233056664,
-      "grad_norm": 0.42060813307762146,
-      "learning_rate": 6.917932779534828e-06,
-      "loss": 0.3096,
-      "step": 17774
-    },
-    {
-      "epoch": 1.1616887785112084,
-      "grad_norm": 0.44270995259284973,
-      "learning_rate": 6.917610296165356e-06,
-      "loss": 0.3379,
-      "step": 17775
-    },
-    {
-      "epoch": 1.1617541337167505,
-      "grad_norm": 0.47732046246528625,
-      "learning_rate": 6.917287803443031e-06,
-      "loss": 0.3892,
-      "step": 17776
-    },
-    {
-      "epoch": 1.1618194889222926,
-      "grad_norm": 0.45870134234428406,
-      "learning_rate": 6.91696530136943e-06,
-      "loss": 0.342,
-      "step": 17777
-    },
-    {
-      "epoch": 1.1618848441278349,
-      "grad_norm": 0.4363591969013214,
-      "learning_rate": 6.9166427899461266e-06,
-      "loss": 0.3613,
-      "step": 17778
-    },
-    {
-      "epoch": 1.161950199333377,
-      "grad_norm": 0.4601621627807617,
-      "learning_rate": 6.916320269174691e-06,
-      "loss": 0.3556,
-      "step": 17779
-    },
-    {
-      "epoch": 1.162015554538919,
-      "grad_norm": 0.4446568489074707,
-      "learning_rate": 6.915997739056697e-06,
-      "loss": 0.3062,
-      "step": 17780
-    },
-    {
-      "epoch": 1.162080909744461,
-      "grad_norm": 0.5495322346687317,
-      "learning_rate": 6.91567519959372e-06,
-      "loss": 0.4479,
-      "step": 17781
-    },
-    {
-      "epoch": 1.1621462649500032,
-      "grad_norm": 0.482977032661438,
-      "learning_rate": 6.9153526507873305e-06,
-      "loss": 0.3716,
-      "step": 17782
-    },
-    {
-      "epoch": 1.1622116201555455,
-      "grad_norm": 0.45155516266822815,
-      "learning_rate": 6.915030092639101e-06,
-      "loss": 0.3546,
-      "step": 17783
-    },
-    {
-      "epoch": 1.1622769753610875,
-      "grad_norm": 0.4688279628753662,
-      "learning_rate": 6.914707525150609e-06,
-      "loss": 0.3787,
-      "step": 17784
-    },
-    {
-      "epoch": 1.1623423305666296,
-      "grad_norm": 0.4659804403781891,
-      "learning_rate": 6.914384948323423e-06,
-      "loss": 0.3616,
-      "step": 17785
-    },
-    {
-      "epoch": 1.1624076857721717,
-      "grad_norm": 0.4333311915397644,
-      "learning_rate": 6.9140623621591194e-06,
-      "loss": 0.3351,
-      "step": 17786
-    },
-    {
-      "epoch": 1.162473040977714,
-      "grad_norm": 0.45172226428985596,
-      "learning_rate": 6.9137397666592695e-06,
-      "loss": 0.338,
-      "step": 17787
-    },
-    {
-      "epoch": 1.162538396183256,
-      "grad_norm": 0.45358651876449585,
-      "learning_rate": 6.913417161825449e-06,
-      "loss": 0.3667,
-      "step": 17788
-    },
-    {
-      "epoch": 1.1626037513887981,
-      "grad_norm": 0.45765697956085205,
-      "learning_rate": 6.913094547659231e-06,
-      "loss": 0.3433,
-      "step": 17789
-    },
-    {
-      "epoch": 1.1626691065943402,
-      "grad_norm": 0.44429054856300354,
-      "learning_rate": 6.912771924162186e-06,
-      "loss": 0.3509,
-      "step": 17790
-    },
-    {
-      "epoch": 1.1627344617998823,
-      "grad_norm": 0.4193040728569031,
-      "learning_rate": 6.912449291335891e-06,
-      "loss": 0.3035,
-      "step": 17791
-    },
-    {
-      "epoch": 1.1627998170054246,
-      "grad_norm": 0.4288153648376465,
-      "learning_rate": 6.912126649181917e-06,
-      "loss": 0.289,
-      "step": 17792
-    },
-    {
-      "epoch": 1.1628651722109666,
-      "grad_norm": 0.4442983865737915,
-      "learning_rate": 6.91180399770184e-06,
-      "loss": 0.3481,
-      "step": 17793
-    },
-    {
-      "epoch": 1.1629305274165087,
-      "grad_norm": 0.4443356990814209,
-      "learning_rate": 6.911481336897232e-06,
-      "loss": 0.3515,
-      "step": 17794
-    },
-    {
-      "epoch": 1.1629958826220508,
-      "grad_norm": 0.45085409283638,
-      "learning_rate": 6.911158666769668e-06,
-      "loss": 0.3724,
-      "step": 17795
-    },
-    {
-      "epoch": 1.163061237827593,
-      "grad_norm": 0.4919372797012329,
-      "learning_rate": 6.9108359873207224e-06,
-      "loss": 0.381,
-      "step": 17796
-    },
-    {
-      "epoch": 1.1631265930331351,
-      "grad_norm": 0.46046745777130127,
-      "learning_rate": 6.9105132985519655e-06,
-      "loss": 0.3274,
-      "step": 17797
-    },
-    {
-      "epoch": 1.1631919482386772,
-      "grad_norm": 0.46758633852005005,
-      "learning_rate": 6.910190600464975e-06,
-      "loss": 0.3695,
-      "step": 17798
-    },
-    {
-      "epoch": 1.1632573034442193,
-      "grad_norm": 0.4499673545360565,
-      "learning_rate": 6.909867893061322e-06,
-      "loss": 0.3334,
-      "step": 17799
-    },
-    {
-      "epoch": 1.1633226586497614,
-      "grad_norm": 0.44051435589790344,
-      "learning_rate": 6.909545176342582e-06,
-      "loss": 0.3454,
-      "step": 17800
-    },
-    {
-      "epoch": 1.1633880138553037,
-      "grad_norm": 0.47183161973953247,
-      "learning_rate": 6.909222450310326e-06,
-      "loss": 0.3527,
-      "step": 17801
-    },
-    {
-      "epoch": 1.1634533690608457,
-      "grad_norm": 0.45104748010635376,
-      "learning_rate": 6.908899714966136e-06,
-      "loss": 0.3158,
-      "step": 17802
-    },
-    {
-      "epoch": 1.1635187242663878,
-      "grad_norm": 0.4534842073917389,
-      "learning_rate": 6.9085769703115755e-06,
-      "loss": 0.3206,
-      "step": 17803
-    },
-    {
-      "epoch": 1.1635840794719299,
-      "grad_norm": 0.49478787183761597,
-      "learning_rate": 6.908254216348227e-06,
-      "loss": 0.38,
-      "step": 17804
-    },
-    {
-      "epoch": 1.1636494346774722,
-      "grad_norm": 0.43771791458129883,
-      "learning_rate": 6.907931453077661e-06,
-      "loss": 0.3526,
-      "step": 17805
-    },
-    {
-      "epoch": 1.1637147898830142,
-      "grad_norm": 0.42836201190948486,
-      "learning_rate": 6.907608680501452e-06,
-      "loss": 0.3058,
-      "step": 17806
-    },
-    {
-      "epoch": 1.1637801450885563,
-      "grad_norm": 0.46062058210372925,
-      "learning_rate": 6.907285898621174e-06,
-      "loss": 0.3535,
-      "step": 17807
-    },
-    {
-      "epoch": 1.1638455002940984,
-      "grad_norm": 0.47398918867111206,
-      "learning_rate": 6.906963107438405e-06,
-      "loss": 0.3896,
-      "step": 17808
-    },
-    {
-      "epoch": 1.1639108554996405,
-      "grad_norm": 0.4412325918674469,
-      "learning_rate": 6.906640306954714e-06,
-      "loss": 0.3461,
-      "step": 17809
-    },
-    {
-      "epoch": 1.1639762107051828,
-      "grad_norm": 0.4329341948032379,
-      "learning_rate": 6.906317497171677e-06,
-      "loss": 0.3211,
-      "step": 17810
-    },
-    {
-      "epoch": 1.1640415659107248,
-      "grad_norm": 0.44721320271492004,
-      "learning_rate": 6.9059946780908705e-06,
-      "loss": 0.3465,
-      "step": 17811
-    },
-    {
-      "epoch": 1.164106921116267,
-      "grad_norm": 0.485296368598938,
-      "learning_rate": 6.905671849713866e-06,
-      "loss": 0.4212,
-      "step": 17812
-    },
-    {
-      "epoch": 1.164172276321809,
-      "grad_norm": 0.4662768840789795,
-      "learning_rate": 6.905349012042242e-06,
-      "loss": 0.3689,
-      "step": 17813
-    },
-    {
-      "epoch": 1.1642376315273513,
-      "grad_norm": 0.44688907265663147,
-      "learning_rate": 6.905026165077569e-06,
-      "loss": 0.3446,
-      "step": 17814
-    },
-    {
-      "epoch": 1.1643029867328933,
-      "grad_norm": 0.44490015506744385,
-      "learning_rate": 6.904703308821424e-06,
-      "loss": 0.3175,
-      "step": 17815
-    },
-    {
-      "epoch": 1.1643683419384354,
-      "grad_norm": 0.4581475853919983,
-      "learning_rate": 6.904380443275381e-06,
-      "loss": 0.3194,
-      "step": 17816
-    },
-    {
-      "epoch": 1.1644336971439775,
-      "grad_norm": 0.42191195487976074,
-      "learning_rate": 6.904057568441015e-06,
-      "loss": 0.3052,
-      "step": 17817
-    },
-    {
-      "epoch": 1.1644990523495196,
-      "grad_norm": 0.4517256021499634,
-      "learning_rate": 6.9037346843199e-06,
-      "loss": 0.3487,
-      "step": 17818
-    },
-    {
-      "epoch": 1.1645644075550619,
-      "grad_norm": 0.4611721634864807,
-      "learning_rate": 6.903411790913612e-06,
-      "loss": 0.3529,
-      "step": 17819
-    },
-    {
-      "epoch": 1.164629762760604,
-      "grad_norm": 0.46251773834228516,
-      "learning_rate": 6.903088888223727e-06,
-      "loss": 0.3625,
-      "step": 17820
-    },
-    {
-      "epoch": 1.164695117966146,
-      "grad_norm": 0.47848039865493774,
-      "learning_rate": 6.902765976251817e-06,
-      "loss": 0.3975,
-      "step": 17821
-    },
-    {
-      "epoch": 1.164760473171688,
-      "grad_norm": 0.4365377426147461,
-      "learning_rate": 6.902443054999457e-06,
-      "loss": 0.3437,
-      "step": 17822
-    },
-    {
-      "epoch": 1.1648258283772304,
-      "grad_norm": 0.46792539954185486,
-      "learning_rate": 6.902120124468224e-06,
-      "loss": 0.3529,
-      "step": 17823
-    },
-    {
-      "epoch": 1.1648911835827724,
-      "grad_norm": 0.4476098120212555,
-      "learning_rate": 6.901797184659691e-06,
-      "loss": 0.3185,
-      "step": 17824
-    },
-    {
-      "epoch": 1.1649565387883145,
-      "grad_norm": 0.49423664808273315,
-      "learning_rate": 6.901474235575435e-06,
-      "loss": 0.343,
-      "step": 17825
-    },
-    {
-      "epoch": 1.1650218939938566,
-      "grad_norm": 0.42680567502975464,
-      "learning_rate": 6.9011512772170304e-06,
-      "loss": 0.3039,
-      "step": 17826
-    },
-    {
-      "epoch": 1.1650872491993987,
-      "grad_norm": 0.47794899344444275,
-      "learning_rate": 6.900828309586054e-06,
-      "loss": 0.3699,
-      "step": 17827
-    },
-    {
-      "epoch": 1.165152604404941,
-      "grad_norm": 0.43290624022483826,
-      "learning_rate": 6.900505332684078e-06,
-      "loss": 0.3418,
-      "step": 17828
-    },
-    {
-      "epoch": 1.165217959610483,
-      "grad_norm": 0.4656946659088135,
-      "learning_rate": 6.900182346512679e-06,
-      "loss": 0.3703,
-      "step": 17829
-    },
-    {
-      "epoch": 1.165283314816025,
-      "grad_norm": 0.5058578848838806,
-      "learning_rate": 6.899859351073432e-06,
-      "loss": 0.411,
-      "step": 17830
-    },
-    {
-      "epoch": 1.1653486700215672,
-      "grad_norm": 0.46284496784210205,
-      "learning_rate": 6.899536346367914e-06,
-      "loss": 0.377,
-      "step": 17831
-    },
-    {
-      "epoch": 1.1654140252271095,
-      "grad_norm": 0.47973090410232544,
-      "learning_rate": 6.8992133323976985e-06,
-      "loss": 0.3759,
-      "step": 17832
-    },
-    {
-      "epoch": 1.1654793804326515,
-      "grad_norm": 0.43616053462028503,
-      "learning_rate": 6.898890309164362e-06,
-      "loss": 0.3073,
-      "step": 17833
-    },
-    {
-      "epoch": 1.1655447356381936,
-      "grad_norm": 0.44696760177612305,
-      "learning_rate": 6.8985672766694785e-06,
-      "loss": 0.3433,
-      "step": 17834
-    },
-    {
-      "epoch": 1.1656100908437357,
-      "grad_norm": 0.44514200091362,
-      "learning_rate": 6.898244234914626e-06,
-      "loss": 0.3291,
-      "step": 17835
-    },
-    {
-      "epoch": 1.1656754460492778,
-      "grad_norm": 0.4564495086669922,
-      "learning_rate": 6.8979211839013785e-06,
-      "loss": 0.3639,
-      "step": 17836
-    },
-    {
-      "epoch": 1.1657408012548198,
-      "grad_norm": 0.4443361163139343,
-      "learning_rate": 6.8975981236313105e-06,
-      "loss": 0.3421,
-      "step": 17837
-    },
-    {
-      "epoch": 1.1658061564603621,
-      "grad_norm": 0.4628848433494568,
-      "learning_rate": 6.897275054106001e-06,
-      "loss": 0.3543,
-      "step": 17838
-    },
-    {
-      "epoch": 1.1658715116659042,
-      "grad_norm": 0.5016350150108337,
-      "learning_rate": 6.896951975327022e-06,
-      "loss": 0.4043,
-      "step": 17839
-    },
-    {
-      "epoch": 1.1659368668714463,
-      "grad_norm": 0.46526437997817993,
-      "learning_rate": 6.896628887295953e-06,
-      "loss": 0.3563,
-      "step": 17840
-    },
-    {
-      "epoch": 1.1660022220769883,
-      "grad_norm": 0.4498814642429352,
-      "learning_rate": 6.896305790014367e-06,
-      "loss": 0.3353,
-      "step": 17841
-    },
-    {
-      "epoch": 1.1660675772825306,
-      "grad_norm": 0.44351500272750854,
-      "learning_rate": 6.895982683483842e-06,
-      "loss": 0.3213,
-      "step": 17842
-    },
-    {
-      "epoch": 1.1661329324880727,
-      "grad_norm": 0.4809412658214569,
-      "learning_rate": 6.89565956770595e-06,
-      "loss": 0.3811,
-      "step": 17843
-    },
-    {
-      "epoch": 1.1661982876936148,
-      "grad_norm": 0.469224750995636,
-      "learning_rate": 6.895336442682271e-06,
-      "loss": 0.3414,
-      "step": 17844
-    },
-    {
-      "epoch": 1.1662636428991568,
-      "grad_norm": 0.4733920097351074,
-      "learning_rate": 6.895013308414379e-06,
-      "loss": 0.3381,
-      "step": 17845
-    },
-    {
-      "epoch": 1.166328998104699,
-      "grad_norm": 0.4229116141796112,
-      "learning_rate": 6.894690164903851e-06,
-      "loss": 0.3322,
-      "step": 17846
-    },
-    {
-      "epoch": 1.1663943533102412,
-      "grad_norm": 0.442176878452301,
-      "learning_rate": 6.894367012152263e-06,
-      "loss": 0.3645,
-      "step": 17847
-    },
-    {
-      "epoch": 1.1664597085157833,
-      "grad_norm": 0.4134009778499603,
-      "learning_rate": 6.89404385016119e-06,
-      "loss": 0.3302,
-      "step": 17848
-    },
-    {
-      "epoch": 1.1665250637213254,
-      "grad_norm": 0.47841525077819824,
-      "learning_rate": 6.89372067893221e-06,
-      "loss": 0.3731,
-      "step": 17849
-    },
-    {
-      "epoch": 1.1665904189268674,
-      "grad_norm": 0.42091986536979675,
-      "learning_rate": 6.893397498466897e-06,
-      "loss": 0.2816,
-      "step": 17850
-    },
-    {
-      "epoch": 1.1666557741324097,
-      "grad_norm": 0.4541703760623932,
-      "learning_rate": 6.89307430876683e-06,
-      "loss": 0.3566,
-      "step": 17851
-    },
-    {
-      "epoch": 1.1667211293379518,
-      "grad_norm": 0.4824099540710449,
-      "learning_rate": 6.892751109833582e-06,
-      "loss": 0.4082,
-      "step": 17852
-    },
-    {
-      "epoch": 1.1667864845434939,
-      "grad_norm": 0.4634397029876709,
-      "learning_rate": 6.892427901668732e-06,
-      "loss": 0.4,
-      "step": 17853
-    },
-    {
-      "epoch": 1.166851839749036,
-      "grad_norm": 0.4780362844467163,
-      "learning_rate": 6.892104684273856e-06,
-      "loss": 0.387,
-      "step": 17854
-    },
-    {
-      "epoch": 1.166917194954578,
-      "grad_norm": 0.4563106596469879,
-      "learning_rate": 6.8917814576505296e-06,
-      "loss": 0.3391,
-      "step": 17855
-    },
-    {
-      "epoch": 1.1669825501601203,
-      "grad_norm": 0.4266104996204376,
-      "learning_rate": 6.891458221800329e-06,
-      "loss": 0.308,
-      "step": 17856
-    },
-    {
-      "epoch": 1.1670479053656624,
-      "grad_norm": 0.46188053488731384,
-      "learning_rate": 6.891134976724831e-06,
-      "loss": 0.3884,
-      "step": 17857
-    },
-    {
-      "epoch": 1.1671132605712045,
-      "grad_norm": 0.45284342765808105,
-      "learning_rate": 6.890811722425615e-06,
-      "loss": 0.3515,
-      "step": 17858
-    },
-    {
-      "epoch": 1.1671786157767465,
-      "grad_norm": 0.46304458379745483,
-      "learning_rate": 6.890488458904253e-06,
-      "loss": 0.3471,
-      "step": 17859
-    },
-    {
-      "epoch": 1.1672439709822888,
-      "grad_norm": 0.5158681869506836,
-      "learning_rate": 6.890165186162326e-06,
-      "loss": 0.4391,
-      "step": 17860
-    },
-    {
-      "epoch": 1.167309326187831,
-      "grad_norm": 0.41425713896751404,
-      "learning_rate": 6.889841904201405e-06,
-      "loss": 0.2848,
-      "step": 17861
-    },
-    {
-      "epoch": 1.167374681393373,
-      "grad_norm": 0.4457348883152008,
-      "learning_rate": 6.889518613023074e-06,
-      "loss": 0.3369,
-      "step": 17862
-    },
-    {
-      "epoch": 1.167440036598915,
-      "grad_norm": 0.43998560309410095,
-      "learning_rate": 6.889195312628905e-06,
-      "loss": 0.3146,
-      "step": 17863
-    },
-    {
-      "epoch": 1.1675053918044571,
-      "grad_norm": 0.47874224185943604,
-      "learning_rate": 6.888872003020475e-06,
-      "loss": 0.3399,
-      "step": 17864
-    },
-    {
-      "epoch": 1.1675707470099994,
-      "grad_norm": 0.4386598765850067,
-      "learning_rate": 6.888548684199363e-06,
-      "loss": 0.3483,
-      "step": 17865
-    },
-    {
-      "epoch": 1.1676361022155415,
-      "grad_norm": 0.48109519481658936,
-      "learning_rate": 6.888225356167144e-06,
-      "loss": 0.3436,
-      "step": 17866
-    },
-    {
-      "epoch": 1.1677014574210836,
-      "grad_norm": 0.47091230750083923,
-      "learning_rate": 6.887902018925396e-06,
-      "loss": 0.3733,
-      "step": 17867
-    },
-    {
-      "epoch": 1.1677668126266256,
-      "grad_norm": 0.43233412504196167,
-      "learning_rate": 6.887578672475695e-06,
-      "loss": 0.322,
-      "step": 17868
-    },
-    {
-      "epoch": 1.167832167832168,
-      "grad_norm": 0.42750242352485657,
-      "learning_rate": 6.887255316819621e-06,
-      "loss": 0.3278,
-      "step": 17869
-    },
-    {
-      "epoch": 1.16789752303771,
-      "grad_norm": 0.4594009816646576,
-      "learning_rate": 6.886931951958746e-06,
-      "loss": 0.3568,
-      "step": 17870
-    },
-    {
-      "epoch": 1.167962878243252,
-      "grad_norm": 0.47360721230506897,
-      "learning_rate": 6.8866085778946535e-06,
-      "loss": 0.351,
-      "step": 17871
-    },
-    {
-      "epoch": 1.1680282334487941,
-      "grad_norm": 0.4647512137889862,
-      "learning_rate": 6.886285194628914e-06,
-      "loss": 0.3348,
-      "step": 17872
-    },
-    {
-      "epoch": 1.1680935886543362,
-      "grad_norm": 0.4592376947402954,
-      "learning_rate": 6.885961802163111e-06,
-      "loss": 0.3706,
-      "step": 17873
-    },
-    {
-      "epoch": 1.1681589438598785,
-      "grad_norm": 0.4431360363960266,
-      "learning_rate": 6.885638400498819e-06,
-      "loss": 0.3496,
-      "step": 17874
-    },
-    {
-      "epoch": 1.1682242990654206,
-      "grad_norm": 0.4597468972206116,
-      "learning_rate": 6.8853149896376125e-06,
-      "loss": 0.3638,
-      "step": 17875
-    },
-    {
-      "epoch": 1.1682896542709627,
-      "grad_norm": 0.4695626497268677,
-      "learning_rate": 6.884991569581074e-06,
-      "loss": 0.3899,
-      "step": 17876
-    },
-    {
-      "epoch": 1.1683550094765047,
-      "grad_norm": 0.4574628174304962,
-      "learning_rate": 6.884668140330777e-06,
-      "loss": 0.3438,
-      "step": 17877
-    },
-    {
-      "epoch": 1.168420364682047,
-      "grad_norm": 0.44924554228782654,
-      "learning_rate": 6.884344701888303e-06,
-      "loss": 0.3581,
-      "step": 17878
-    },
-    {
-      "epoch": 1.168485719887589,
-      "grad_norm": 0.39787736535072327,
-      "learning_rate": 6.8840212542552245e-06,
-      "loss": 0.2588,
-      "step": 17879
-    },
-    {
-      "epoch": 1.1685510750931312,
-      "grad_norm": 0.43013879656791687,
-      "learning_rate": 6.883697797433123e-06,
-      "loss": 0.3417,
-      "step": 17880
-    },
-    {
-      "epoch": 1.1686164302986732,
-      "grad_norm": 0.4616085886955261,
-      "learning_rate": 6.883374331423574e-06,
-      "loss": 0.3781,
-      "step": 17881
-    },
-    {
-      "epoch": 1.1686817855042153,
-      "grad_norm": 0.4305798411369324,
-      "learning_rate": 6.8830508562281575e-06,
-      "loss": 0.2914,
-      "step": 17882
-    },
-    {
-      "epoch": 1.1687471407097576,
-      "grad_norm": 0.4259510934352875,
-      "learning_rate": 6.882727371848448e-06,
-      "loss": 0.326,
-      "step": 17883
-    },
-    {
-      "epoch": 1.1688124959152997,
-      "grad_norm": 0.44786539673805237,
-      "learning_rate": 6.8824038782860255e-06,
-      "loss": 0.3315,
-      "step": 17884
-    },
-    {
-      "epoch": 1.1688778511208417,
-      "grad_norm": 0.44690102338790894,
-      "learning_rate": 6.882080375542468e-06,
-      "loss": 0.3348,
-      "step": 17885
-    },
-    {
-      "epoch": 1.1689432063263838,
-      "grad_norm": 0.4234074652194977,
-      "learning_rate": 6.881756863619351e-06,
-      "loss": 0.3269,
-      "step": 17886
-    },
-    {
-      "epoch": 1.1690085615319261,
-      "grad_norm": 0.46854862570762634,
-      "learning_rate": 6.8814333425182545e-06,
-      "loss": 0.3723,
-      "step": 17887
-    },
-    {
-      "epoch": 1.1690739167374682,
-      "grad_norm": 0.450252890586853,
-      "learning_rate": 6.881109812240755e-06,
-      "loss": 0.3484,
-      "step": 17888
-    },
-    {
-      "epoch": 1.1691392719430103,
-      "grad_norm": 0.4471893608570099,
-      "learning_rate": 6.880786272788433e-06,
-      "loss": 0.3578,
-      "step": 17889
-    },
-    {
-      "epoch": 1.1692046271485523,
-      "grad_norm": 0.471813440322876,
-      "learning_rate": 6.880462724162863e-06,
-      "loss": 0.3707,
-      "step": 17890
-    },
-    {
-      "epoch": 1.1692699823540944,
-      "grad_norm": 0.44897350668907166,
-      "learning_rate": 6.8801391663656256e-06,
-      "loss": 0.3486,
-      "step": 17891
-    },
-    {
-      "epoch": 1.1693353375596367,
-      "grad_norm": 0.4620920717716217,
-      "learning_rate": 6.879815599398299e-06,
-      "loss": 0.3596,
-      "step": 17892
-    },
-    {
-      "epoch": 1.1694006927651788,
-      "grad_norm": 0.44977056980133057,
-      "learning_rate": 6.8794920232624594e-06,
-      "loss": 0.3241,
-      "step": 17893
-    },
-    {
-      "epoch": 1.1694660479707208,
-      "grad_norm": 0.40761831402778625,
-      "learning_rate": 6.8791684379596865e-06,
-      "loss": 0.3117,
-      "step": 17894
-    },
-    {
-      "epoch": 1.169531403176263,
-      "grad_norm": 0.45136305689811707,
-      "learning_rate": 6.878844843491556e-06,
-      "loss": 0.3411,
-      "step": 17895
-    },
-    {
-      "epoch": 1.1695967583818052,
-      "grad_norm": 0.48416197299957275,
-      "learning_rate": 6.878521239859652e-06,
-      "loss": 0.3472,
-      "step": 17896
-    },
-    {
-      "epoch": 1.1696621135873473,
-      "grad_norm": 0.4517776668071747,
-      "learning_rate": 6.8781976270655456e-06,
-      "loss": 0.3441,
-      "step": 17897
-    },
-    {
-      "epoch": 1.1697274687928894,
-      "grad_norm": 0.4336404800415039,
-      "learning_rate": 6.877874005110822e-06,
-      "loss": 0.3115,
-      "step": 17898
-    },
-    {
-      "epoch": 1.1697928239984314,
-      "grad_norm": 0.47181832790374756,
-      "learning_rate": 6.877550373997054e-06,
-      "loss": 0.3448,
-      "step": 17899
-    },
-    {
-      "epoch": 1.1698581792039735,
-      "grad_norm": 0.43525490164756775,
-      "learning_rate": 6.877226733725824e-06,
-      "loss": 0.3438,
-      "step": 17900
-    },
-    {
-      "epoch": 1.1699235344095158,
-      "grad_norm": 0.4715517461299896,
-      "learning_rate": 6.876903084298709e-06,
-      "loss": 0.3539,
-      "step": 17901
-    },
-    {
-      "epoch": 1.1699888896150579,
-      "grad_norm": 0.45197275280952454,
-      "learning_rate": 6.876579425717285e-06,
-      "loss": 0.3148,
-      "step": 17902
-    },
-    {
-      "epoch": 1.1700542448206,
-      "grad_norm": 0.4133222997188568,
-      "learning_rate": 6.876255757983134e-06,
-      "loss": 0.2975,
-      "step": 17903
-    },
-    {
-      "epoch": 1.170119600026142,
-      "grad_norm": 0.4909352660179138,
-      "learning_rate": 6.875932081097836e-06,
-      "loss": 0.3903,
-      "step": 17904
-    },
-    {
-      "epoch": 1.1701849552316843,
-      "grad_norm": 0.4491754472255707,
-      "learning_rate": 6.875608395062966e-06,
-      "loss": 0.3523,
-      "step": 17905
-    },
-    {
-      "epoch": 1.1702503104372264,
-      "grad_norm": 0.47093915939331055,
-      "learning_rate": 6.8752846998801025e-06,
-      "loss": 0.3744,
-      "step": 17906
-    },
-    {
-      "epoch": 1.1703156656427685,
-      "grad_norm": 0.46209442615509033,
-      "learning_rate": 6.8749609955508275e-06,
-      "loss": 0.3467,
-      "step": 17907
-    },
-    {
-      "epoch": 1.1703810208483105,
-      "grad_norm": 0.4198269844055176,
-      "learning_rate": 6.874637282076717e-06,
-      "loss": 0.3002,
-      "step": 17908
-    },
-    {
-      "epoch": 1.1704463760538526,
-      "grad_norm": 0.4138481616973877,
-      "learning_rate": 6.874313559459352e-06,
-      "loss": 0.2938,
-      "step": 17909
-    },
-    {
-      "epoch": 1.170511731259395,
-      "grad_norm": 0.4861946403980255,
-      "learning_rate": 6.873989827700309e-06,
-      "loss": 0.3591,
-      "step": 17910
-    },
-    {
-      "epoch": 1.170577086464937,
-      "grad_norm": 0.4596599340438843,
-      "learning_rate": 6.87366608680117e-06,
-      "loss": 0.3359,
-      "step": 17911
-    },
-    {
-      "epoch": 1.170642441670479,
-      "grad_norm": 0.44050946831703186,
-      "learning_rate": 6.873342336763513e-06,
-      "loss": 0.3522,
-      "step": 17912
-    },
-    {
-      "epoch": 1.1707077968760211,
-      "grad_norm": 0.4483802914619446,
-      "learning_rate": 6.873018577588915e-06,
-      "loss": 0.3739,
-      "step": 17913
-    },
-    {
-      "epoch": 1.1707731520815634,
-      "grad_norm": 0.4418737590312958,
-      "learning_rate": 6.872694809278957e-06,
-      "loss": 0.356,
-      "step": 17914
-    },
-    {
-      "epoch": 1.1708385072871055,
-      "grad_norm": 0.4524979591369629,
-      "learning_rate": 6.872371031835217e-06,
-      "loss": 0.3578,
-      "step": 17915
-    },
-    {
-      "epoch": 1.1709038624926476,
-      "grad_norm": 0.4703204333782196,
-      "learning_rate": 6.872047245259276e-06,
-      "loss": 0.361,
-      "step": 17916
-    },
-    {
-      "epoch": 1.1709692176981896,
-      "grad_norm": 0.45783576369285583,
-      "learning_rate": 6.871723449552711e-06,
-      "loss": 0.3662,
-      "step": 17917
-    },
-    {
-      "epoch": 1.1710345729037317,
-      "grad_norm": 0.42472904920578003,
-      "learning_rate": 6.871399644717103e-06,
-      "loss": 0.3266,
-      "step": 17918
-    },
-    {
-      "epoch": 1.171099928109274,
-      "grad_norm": 0.49030977487564087,
-      "learning_rate": 6.87107583075403e-06,
-      "loss": 0.3957,
-      "step": 17919
-    },
-    {
-      "epoch": 1.171165283314816,
-      "grad_norm": 0.5081116557121277,
-      "learning_rate": 6.870752007665072e-06,
-      "loss": 0.3791,
-      "step": 17920
-    },
-    {
-      "epoch": 1.1712306385203581,
-      "grad_norm": 0.5148254036903381,
-      "learning_rate": 6.87042817545181e-06,
-      "loss": 0.3364,
-      "step": 17921
-    },
-    {
-      "epoch": 1.1712959937259002,
-      "grad_norm": 0.4610372483730316,
-      "learning_rate": 6.870104334115819e-06,
-      "loss": 0.3724,
-      "step": 17922
-    },
-    {
-      "epoch": 1.1713613489314425,
-      "grad_norm": 0.4673974812030792,
-      "learning_rate": 6.869780483658684e-06,
-      "loss": 0.3504,
-      "step": 17923
-    },
-    {
-      "epoch": 1.1714267041369846,
-      "grad_norm": 0.4357917904853821,
-      "learning_rate": 6.86945662408198e-06,
-      "loss": 0.3364,
-      "step": 17924
-    },
-    {
-      "epoch": 1.1714920593425266,
-      "grad_norm": 0.42845776677131653,
-      "learning_rate": 6.86913275538729e-06,
-      "loss": 0.313,
-      "step": 17925
-    },
-    {
-      "epoch": 1.1715574145480687,
-      "grad_norm": 0.424049973487854,
-      "learning_rate": 6.868808877576191e-06,
-      "loss": 0.3193,
-      "step": 17926
-    },
-    {
-      "epoch": 1.1716227697536108,
-      "grad_norm": 0.46391093730926514,
-      "learning_rate": 6.868484990650264e-06,
-      "loss": 0.3976,
-      "step": 17927
-    },
-    {
-      "epoch": 1.171688124959153,
-      "grad_norm": 0.4479975402355194,
-      "learning_rate": 6.868161094611088e-06,
-      "loss": 0.3427,
-      "step": 17928
-    },
-    {
-      "epoch": 1.1717534801646952,
-      "grad_norm": 0.44597792625427246,
-      "learning_rate": 6.867837189460244e-06,
-      "loss": 0.3146,
-      "step": 17929
-    },
-    {
-      "epoch": 1.1718188353702372,
-      "grad_norm": 0.4174831211566925,
-      "learning_rate": 6.86751327519931e-06,
-      "loss": 0.2994,
-      "step": 17930
-    },
-    {
-      "epoch": 1.1718841905757793,
-      "grad_norm": 0.4617291986942291,
-      "learning_rate": 6.867189351829866e-06,
-      "loss": 0.3786,
-      "step": 17931
-    },
-    {
-      "epoch": 1.1719495457813216,
-      "grad_norm": 0.43605837225914,
-      "learning_rate": 6.866865419353494e-06,
-      "loss": 0.3054,
-      "step": 17932
-    },
-    {
-      "epoch": 1.1720149009868637,
-      "grad_norm": 0.46060532331466675,
-      "learning_rate": 6.866541477771772e-06,
-      "loss": 0.3479,
-      "step": 17933
-    },
-    {
-      "epoch": 1.1720802561924057,
-      "grad_norm": 0.45015233755111694,
-      "learning_rate": 6.866217527086281e-06,
-      "loss": 0.3334,
-      "step": 17934
-    },
-    {
-      "epoch": 1.1721456113979478,
-      "grad_norm": 0.43359237909317017,
-      "learning_rate": 6.8658935672986e-06,
-      "loss": 0.3339,
-      "step": 17935
-    },
-    {
-      "epoch": 1.17221096660349,
-      "grad_norm": 0.4737197756767273,
-      "learning_rate": 6.865569598410311e-06,
-      "loss": 0.362,
-      "step": 17936
-    },
-    {
-      "epoch": 1.1722763218090322,
-      "grad_norm": 0.44698429107666016,
-      "learning_rate": 6.865245620422991e-06,
-      "loss": 0.3612,
-      "step": 17937
-    },
-    {
-      "epoch": 1.1723416770145743,
-      "grad_norm": 0.4733181893825531,
-      "learning_rate": 6.864921633338224e-06,
-      "loss": 0.3675,
-      "step": 17938
-    },
-    {
-      "epoch": 1.1724070322201163,
-      "grad_norm": 0.44927340745925903,
-      "learning_rate": 6.864597637157586e-06,
-      "loss": 0.3319,
-      "step": 17939
-    },
-    {
-      "epoch": 1.1724723874256584,
-      "grad_norm": 0.45221009850502014,
-      "learning_rate": 6.864273631882661e-06,
-      "loss": 0.3829,
-      "step": 17940
-    },
-    {
-      "epoch": 1.1725377426312007,
-      "grad_norm": 0.446685791015625,
-      "learning_rate": 6.863949617515027e-06,
-      "loss": 0.316,
-      "step": 17941
-    },
-    {
-      "epoch": 1.1726030978367428,
-      "grad_norm": 0.4642782211303711,
-      "learning_rate": 6.863625594056264e-06,
-      "loss": 0.378,
-      "step": 17942
-    },
-    {
-      "epoch": 1.1726684530422848,
-      "grad_norm": 0.49931108951568604,
-      "learning_rate": 6.8633015615079555e-06,
-      "loss": 0.3358,
-      "step": 17943
-    },
-    {
-      "epoch": 1.172733808247827,
-      "grad_norm": 0.4535532295703888,
-      "learning_rate": 6.862977519871678e-06,
-      "loss": 0.3257,
-      "step": 17944
-    },
-    {
-      "epoch": 1.172799163453369,
-      "grad_norm": 0.45970287919044495,
-      "learning_rate": 6.862653469149014e-06,
-      "loss": 0.3246,
-      "step": 17945
-    },
-    {
-      "epoch": 1.172864518658911,
-      "grad_norm": 0.509086012840271,
-      "learning_rate": 6.862329409341545e-06,
-      "loss": 0.3155,
-      "step": 17946
-    },
-    {
-      "epoch": 1.1729298738644534,
-      "grad_norm": 0.4492672085762024,
-      "learning_rate": 6.86200534045085e-06,
-      "loss": 0.3446,
-      "step": 17947
-    },
-    {
-      "epoch": 1.1729952290699954,
-      "grad_norm": 0.48400983214378357,
-      "learning_rate": 6.861681262478508e-06,
-      "loss": 0.4013,
-      "step": 17948
-    },
-    {
-      "epoch": 1.1730605842755375,
-      "grad_norm": 0.43667343258857727,
-      "learning_rate": 6.8613571754261036e-06,
-      "loss": 0.3211,
-      "step": 17949
-    },
-    {
-      "epoch": 1.1731259394810796,
-      "grad_norm": 0.46238502860069275,
-      "learning_rate": 6.861033079295215e-06,
-      "loss": 0.3235,
-      "step": 17950
-    },
-    {
-      "epoch": 1.1731912946866219,
-      "grad_norm": 0.4288751482963562,
-      "learning_rate": 6.860708974087422e-06,
-      "loss": 0.3117,
-      "step": 17951
-    },
-    {
-      "epoch": 1.173256649892164,
-      "grad_norm": 0.43703600764274597,
-      "learning_rate": 6.860384859804308e-06,
-      "loss": 0.3282,
-      "step": 17952
-    },
-    {
-      "epoch": 1.173322005097706,
-      "grad_norm": 0.43666911125183105,
-      "learning_rate": 6.860060736447452e-06,
-      "loss": 0.3464,
-      "step": 17953
-    },
-    {
-      "epoch": 1.173387360303248,
-      "grad_norm": 0.467313677072525,
-      "learning_rate": 6.8597366040184365e-06,
-      "loss": 0.3535,
-      "step": 17954
-    },
-    {
-      "epoch": 1.1734527155087902,
-      "grad_norm": 0.4325222074985504,
-      "learning_rate": 6.8594124625188395e-06,
-      "loss": 0.2938,
-      "step": 17955
-    },
-    {
-      "epoch": 1.1735180707143325,
-      "grad_norm": 0.45261213183403015,
-      "learning_rate": 6.859088311950245e-06,
-      "loss": 0.3602,
-      "step": 17956
-    },
-    {
-      "epoch": 1.1735834259198745,
-      "grad_norm": 0.48950883746147156,
-      "learning_rate": 6.858764152314234e-06,
-      "loss": 0.3792,
-      "step": 17957
-    },
-    {
-      "epoch": 1.1736487811254166,
-      "grad_norm": 0.4489250183105469,
-      "learning_rate": 6.858439983612384e-06,
-      "loss": 0.3597,
-      "step": 17958
-    },
-    {
-      "epoch": 1.1737141363309587,
-      "grad_norm": 0.40376684069633484,
-      "learning_rate": 6.858115805846279e-06,
-      "loss": 0.2981,
-      "step": 17959
-    },
-    {
-      "epoch": 1.173779491536501,
-      "grad_norm": 0.4631209373474121,
-      "learning_rate": 6.857791619017499e-06,
-      "loss": 0.39,
-      "step": 17960
-    },
-    {
-      "epoch": 1.173844846742043,
-      "grad_norm": 0.4509110450744629,
-      "learning_rate": 6.857467423127626e-06,
-      "loss": 0.3629,
-      "step": 17961
-    },
-    {
-      "epoch": 1.173910201947585,
-      "grad_norm": 0.4790554642677307,
-      "learning_rate": 6.857143218178242e-06,
-      "loss": 0.3856,
-      "step": 17962
-    },
-    {
-      "epoch": 1.1739755571531272,
-      "grad_norm": 0.4161204397678375,
-      "learning_rate": 6.856819004170926e-06,
-      "loss": 0.3239,
-      "step": 17963
-    },
-    {
-      "epoch": 1.1740409123586693,
-      "grad_norm": 0.4591652750968933,
-      "learning_rate": 6.8564947811072606e-06,
-      "loss": 0.3707,
-      "step": 17964
-    },
-    {
-      "epoch": 1.1741062675642115,
-      "grad_norm": 0.44288069009780884,
-      "learning_rate": 6.856170548988827e-06,
-      "loss": 0.3542,
-      "step": 17965
-    },
-    {
-      "epoch": 1.1741716227697536,
-      "grad_norm": 0.42953959107398987,
-      "learning_rate": 6.855846307817206e-06,
-      "loss": 0.3028,
-      "step": 17966
-    },
-    {
-      "epoch": 1.1742369779752957,
-      "grad_norm": 0.45490196347236633,
-      "learning_rate": 6.8555220575939816e-06,
-      "loss": 0.3233,
-      "step": 17967
-    },
-    {
-      "epoch": 1.1743023331808378,
-      "grad_norm": 0.4385647475719452,
-      "learning_rate": 6.8551977983207314e-06,
-      "loss": 0.3477,
-      "step": 17968
-    },
-    {
-      "epoch": 1.17436768838638,
-      "grad_norm": 0.41461601853370667,
-      "learning_rate": 6.85487352999904e-06,
-      "loss": 0.3006,
-      "step": 17969
-    },
-    {
-      "epoch": 1.1744330435919221,
-      "grad_norm": 0.42973747849464417,
-      "learning_rate": 6.854549252630488e-06,
-      "loss": 0.3175,
-      "step": 17970
-    },
-    {
-      "epoch": 1.1744983987974642,
-      "grad_norm": 0.501099169254303,
-      "learning_rate": 6.854224966216656e-06,
-      "loss": 0.4013,
-      "step": 17971
-    },
-    {
-      "epoch": 1.1745637540030063,
-      "grad_norm": 0.4343818724155426,
-      "learning_rate": 6.853900670759127e-06,
-      "loss": 0.3505,
-      "step": 17972
-    },
-    {
-      "epoch": 1.1746291092085483,
-      "grad_norm": 0.4484075605869293,
-      "learning_rate": 6.853576366259481e-06,
-      "loss": 0.331,
-      "step": 17973
-    },
-    {
-      "epoch": 1.1746944644140906,
-      "grad_norm": 0.47266674041748047,
-      "learning_rate": 6.853252052719302e-06,
-      "loss": 0.3887,
-      "step": 17974
-    },
-    {
-      "epoch": 1.1747598196196327,
-      "grad_norm": 0.45336630940437317,
-      "learning_rate": 6.852927730140171e-06,
-      "loss": 0.3515,
-      "step": 17975
-    },
-    {
-      "epoch": 1.1748251748251748,
-      "grad_norm": 0.42022061347961426,
-      "learning_rate": 6.852603398523668e-06,
-      "loss": 0.3281,
-      "step": 17976
-    },
-    {
-      "epoch": 1.1748905300307169,
-      "grad_norm": 0.41096192598342896,
-      "learning_rate": 6.8522790578713785e-06,
-      "loss": 0.3085,
-      "step": 17977
-    },
-    {
-      "epoch": 1.1749558852362592,
-      "grad_norm": 0.4493367373943329,
-      "learning_rate": 6.8519547081848804e-06,
-      "loss": 0.3501,
-      "step": 17978
-    },
-    {
-      "epoch": 1.1750212404418012,
-      "grad_norm": 0.4425660967826843,
-      "learning_rate": 6.8516303494657585e-06,
-      "loss": 0.3773,
-      "step": 17979
-    },
-    {
-      "epoch": 1.1750865956473433,
-      "grad_norm": 0.4296548366546631,
-      "learning_rate": 6.851305981715595e-06,
-      "loss": 0.3014,
-      "step": 17980
-    },
-    {
-      "epoch": 1.1751519508528854,
-      "grad_norm": 0.43769317865371704,
-      "learning_rate": 6.850981604935969e-06,
-      "loss": 0.3179,
-      "step": 17981
-    },
-    {
-      "epoch": 1.1752173060584274,
-      "grad_norm": 0.40472865104675293,
-      "learning_rate": 6.850657219128465e-06,
-      "loss": 0.2798,
-      "step": 17982
-    },
-    {
-      "epoch": 1.1752826612639697,
-      "grad_norm": 0.4564395546913147,
-      "learning_rate": 6.850332824294666e-06,
-      "loss": 0.3953,
-      "step": 17983
-    },
-    {
-      "epoch": 1.1753480164695118,
-      "grad_norm": 0.5194770693778992,
-      "learning_rate": 6.850008420436152e-06,
-      "loss": 0.3887,
-      "step": 17984
-    },
-    {
-      "epoch": 1.1754133716750539,
-      "grad_norm": 0.4259186387062073,
-      "learning_rate": 6.849684007554505e-06,
-      "loss": 0.3249,
-      "step": 17985
-    },
-    {
-      "epoch": 1.175478726880596,
-      "grad_norm": 0.45520758628845215,
-      "learning_rate": 6.8493595856513085e-06,
-      "loss": 0.3679,
-      "step": 17986
-    },
-    {
-      "epoch": 1.1755440820861383,
-      "grad_norm": 0.4616486132144928,
-      "learning_rate": 6.849035154728145e-06,
-      "loss": 0.3656,
-      "step": 17987
-    },
-    {
-      "epoch": 1.1756094372916803,
-      "grad_norm": 0.4529689848423004,
-      "learning_rate": 6.848710714786597e-06,
-      "loss": 0.3262,
-      "step": 17988
-    },
-    {
-      "epoch": 1.1756747924972224,
-      "grad_norm": 0.4659520089626312,
-      "learning_rate": 6.848386265828247e-06,
-      "loss": 0.3768,
-      "step": 17989
-    },
-    {
-      "epoch": 1.1757401477027645,
-      "grad_norm": 0.4518166780471802,
-      "learning_rate": 6.8480618078546755e-06,
-      "loss": 0.3296,
-      "step": 17990
-    },
-    {
-      "epoch": 1.1758055029083065,
-      "grad_norm": 0.42663031816482544,
-      "learning_rate": 6.847737340867466e-06,
-      "loss": 0.3271,
-      "step": 17991
-    },
-    {
-      "epoch": 1.1758708581138488,
-      "grad_norm": 0.44441017508506775,
-      "learning_rate": 6.847412864868203e-06,
-      "loss": 0.3583,
-      "step": 17992
-    },
-    {
-      "epoch": 1.175936213319391,
-      "grad_norm": 0.4520832300186157,
-      "learning_rate": 6.847088379858466e-06,
-      "loss": 0.3435,
-      "step": 17993
-    },
-    {
-      "epoch": 1.176001568524933,
-      "grad_norm": 0.4381174147129059,
-      "learning_rate": 6.846763885839839e-06,
-      "loss": 0.3236,
-      "step": 17994
-    },
-    {
-      "epoch": 1.176066923730475,
-      "grad_norm": 0.46784117817878723,
-      "learning_rate": 6.846439382813906e-06,
-      "loss": 0.398,
-      "step": 17995
-    },
-    {
-      "epoch": 1.1761322789360174,
-      "grad_norm": 0.4486067295074463,
-      "learning_rate": 6.846114870782248e-06,
-      "loss": 0.339,
-      "step": 17996
-    },
-    {
-      "epoch": 1.1761976341415594,
-      "grad_norm": 0.4041271209716797,
-      "learning_rate": 6.845790349746447e-06,
-      "loss": 0.2812,
-      "step": 17997
-    },
-    {
-      "epoch": 1.1762629893471015,
-      "grad_norm": 0.4424710273742676,
-      "learning_rate": 6.845465819708088e-06,
-      "loss": 0.3598,
-      "step": 17998
-    },
-    {
-      "epoch": 1.1763283445526436,
-      "grad_norm": 0.42119044065475464,
-      "learning_rate": 6.845141280668753e-06,
-      "loss": 0.2893,
-      "step": 17999
-    },
-    {
-      "epoch": 1.1763936997581856,
-      "grad_norm": 0.4513300955295563,
-      "learning_rate": 6.844816732630024e-06,
-      "loss": 0.3302,
-      "step": 18000
-    },
-    {
-      "epoch": 1.176459054963728,
-      "grad_norm": 0.4212856888771057,
-      "learning_rate": 6.844492175593486e-06,
-      "loss": 0.3114,
-      "step": 18001
-    },
-    {
-      "epoch": 1.17652441016927,
-      "grad_norm": 0.46957045793533325,
-      "learning_rate": 6.844167609560719e-06,
-      "loss": 0.3836,
-      "step": 18002
-    },
-    {
-      "epoch": 1.176589765374812,
-      "grad_norm": 0.4354605972766876,
-      "learning_rate": 6.843843034533309e-06,
-      "loss": 0.3219,
-      "step": 18003
-    },
-    {
-      "epoch": 1.1766551205803542,
-      "grad_norm": 0.47783634066581726,
-      "learning_rate": 6.843518450512838e-06,
-      "loss": 0.3954,
-      "step": 18004
-    },
-    {
-      "epoch": 1.1767204757858964,
-      "grad_norm": 0.4695514738559723,
-      "learning_rate": 6.843193857500888e-06,
-      "loss": 0.3413,
-      "step": 18005
-    },
-    {
-      "epoch": 1.1767858309914385,
-      "grad_norm": 0.41028517484664917,
-      "learning_rate": 6.842869255499044e-06,
-      "loss": 0.2983,
-      "step": 18006
-    },
-    {
-      "epoch": 1.1768511861969806,
-      "grad_norm": 0.4204944670200348,
-      "learning_rate": 6.842544644508886e-06,
-      "loss": 0.2998,
-      "step": 18007
-    },
-    {
-      "epoch": 1.1769165414025227,
-      "grad_norm": 0.4137457013130188,
-      "learning_rate": 6.842220024532003e-06,
-      "loss": 0.295,
-      "step": 18008
-    },
-    {
-      "epoch": 1.1769818966080647,
-      "grad_norm": 0.4356216490268707,
-      "learning_rate": 6.841895395569972e-06,
-      "loss": 0.2984,
-      "step": 18009
-    },
-    {
-      "epoch": 1.177047251813607,
-      "grad_norm": 0.4260989725589752,
-      "learning_rate": 6.8415707576243806e-06,
-      "loss": 0.3023,
-      "step": 18010
-    },
-    {
-      "epoch": 1.177112607019149,
-      "grad_norm": 0.4351992607116699,
-      "learning_rate": 6.841246110696809e-06,
-      "loss": 0.3071,
-      "step": 18011
-    },
-    {
-      "epoch": 1.1771779622246912,
-      "grad_norm": 0.44766145944595337,
-      "learning_rate": 6.840921454788844e-06,
-      "loss": 0.3562,
-      "step": 18012
-    },
-    {
-      "epoch": 1.1772433174302332,
-      "grad_norm": 0.43967947363853455,
-      "learning_rate": 6.840596789902065e-06,
-      "loss": 0.3328,
-      "step": 18013
-    },
-    {
-      "epoch": 1.1773086726357755,
-      "grad_norm": 0.46116381883621216,
-      "learning_rate": 6.84027211603806e-06,
-      "loss": 0.374,
-      "step": 18014
-    },
-    {
-      "epoch": 1.1773740278413176,
-      "grad_norm": 0.41465020179748535,
-      "learning_rate": 6.83994743319841e-06,
-      "loss": 0.3381,
-      "step": 18015
-    },
-    {
-      "epoch": 1.1774393830468597,
-      "grad_norm": 0.47038406133651733,
-      "learning_rate": 6.839622741384697e-06,
-      "loss": 0.3456,
-      "step": 18016
-    },
-    {
-      "epoch": 1.1775047382524018,
-      "grad_norm": 0.45181000232696533,
-      "learning_rate": 6.839298040598509e-06,
-      "loss": 0.371,
-      "step": 18017
-    },
-    {
-      "epoch": 1.1775700934579438,
-      "grad_norm": 0.4977002739906311,
-      "learning_rate": 6.838973330841425e-06,
-      "loss": 0.4463,
-      "step": 18018
-    },
-    {
-      "epoch": 1.1776354486634861,
-      "grad_norm": 0.4479779601097107,
-      "learning_rate": 6.838648612115033e-06,
-      "loss": 0.359,
-      "step": 18019
-    },
-    {
-      "epoch": 1.1777008038690282,
-      "grad_norm": 0.4374707341194153,
-      "learning_rate": 6.8383238844209144e-06,
-      "loss": 0.3255,
-      "step": 18020
-    },
-    {
-      "epoch": 1.1777661590745703,
-      "grad_norm": 0.4624062180519104,
-      "learning_rate": 6.837999147760653e-06,
-      "loss": 0.3627,
-      "step": 18021
-    },
-    {
-      "epoch": 1.1778315142801123,
-      "grad_norm": 0.4592170715332031,
-      "learning_rate": 6.837674402135832e-06,
-      "loss": 0.3425,
-      "step": 18022
-    },
-    {
-      "epoch": 1.1778968694856546,
-      "grad_norm": 0.4088749885559082,
-      "learning_rate": 6.837349647548039e-06,
-      "loss": 0.2792,
-      "step": 18023
-    },
-    {
-      "epoch": 1.1779622246911967,
-      "grad_norm": 0.481981486082077,
-      "learning_rate": 6.837024883998853e-06,
-      "loss": 0.382,
-      "step": 18024
-    },
-    {
-      "epoch": 1.1780275798967388,
-      "grad_norm": 0.46223366260528564,
-      "learning_rate": 6.8367001114898605e-06,
-      "loss": 0.3844,
-      "step": 18025
-    },
-    {
-      "epoch": 1.1780929351022809,
-      "grad_norm": 0.4582575857639313,
-      "learning_rate": 6.836375330022646e-06,
-      "loss": 0.3676,
-      "step": 18026
-    },
-    {
-      "epoch": 1.178158290307823,
-      "grad_norm": 0.47594618797302246,
-      "learning_rate": 6.836050539598792e-06,
-      "loss": 0.3668,
-      "step": 18027
-    },
-    {
-      "epoch": 1.1782236455133652,
-      "grad_norm": 0.45693692564964294,
-      "learning_rate": 6.835725740219884e-06,
-      "loss": 0.3543,
-      "step": 18028
-    },
-    {
-      "epoch": 1.1782890007189073,
-      "grad_norm": 0.4364302456378937,
-      "learning_rate": 6.835400931887505e-06,
-      "loss": 0.3492,
-      "step": 18029
-    },
-    {
-      "epoch": 1.1783543559244494,
-      "grad_norm": 0.5103545188903809,
-      "learning_rate": 6.835076114603242e-06,
-      "loss": 0.3329,
-      "step": 18030
-    },
-    {
-      "epoch": 1.1784197111299914,
-      "grad_norm": 0.42652902007102966,
-      "learning_rate": 6.834751288368674e-06,
-      "loss": 0.3129,
-      "step": 18031
-    },
-    {
-      "epoch": 1.1784850663355337,
-      "grad_norm": 0.4542108178138733,
-      "learning_rate": 6.8344264531853896e-06,
-      "loss": 0.3731,
-      "step": 18032
-    },
-    {
-      "epoch": 1.1785504215410758,
-      "grad_norm": 0.5060190558433533,
-      "learning_rate": 6.834101609054973e-06,
-      "loss": 0.3602,
-      "step": 18033
-    },
-    {
-      "epoch": 1.1786157767466179,
-      "grad_norm": 0.45472753047943115,
-      "learning_rate": 6.833776755979006e-06,
-      "loss": 0.3628,
-      "step": 18034
-    },
-    {
-      "epoch": 1.17868113195216,
-      "grad_norm": 0.4651648998260498,
-      "learning_rate": 6.833451893959076e-06,
-      "loss": 0.3496,
-      "step": 18035
-    },
-    {
-      "epoch": 1.178746487157702,
-      "grad_norm": 0.4575902223587036,
-      "learning_rate": 6.833127022996764e-06,
-      "loss": 0.3791,
-      "step": 18036
-    },
-    {
-      "epoch": 1.1788118423632443,
-      "grad_norm": 0.4694966971874237,
-      "learning_rate": 6.832802143093657e-06,
-      "loss": 0.3645,
-      "step": 18037
-    },
-    {
-      "epoch": 1.1788771975687864,
-      "grad_norm": 0.4156460464000702,
-      "learning_rate": 6.832477254251339e-06,
-      "loss": 0.2651,
-      "step": 18038
-    },
-    {
-      "epoch": 1.1789425527743285,
-      "grad_norm": 0.46168753504753113,
-      "learning_rate": 6.832152356471396e-06,
-      "loss": 0.3498,
-      "step": 18039
-    },
-    {
-      "epoch": 1.1790079079798705,
-      "grad_norm": 0.5003201365470886,
-      "learning_rate": 6.831827449755408e-06,
-      "loss": 0.4062,
-      "step": 18040
-    },
-    {
-      "epoch": 1.1790732631854128,
-      "grad_norm": 0.4539341330528259,
-      "learning_rate": 6.831502534104966e-06,
-      "loss": 0.3593,
-      "step": 18041
-    },
-    {
-      "epoch": 1.179138618390955,
-      "grad_norm": 0.4290333092212677,
-      "learning_rate": 6.831177609521651e-06,
-      "loss": 0.3138,
-      "step": 18042
-    },
-    {
-      "epoch": 1.179203973596497,
-      "grad_norm": 0.4450630247592926,
-      "learning_rate": 6.830852676007048e-06,
-      "loss": 0.3692,
-      "step": 18043
-    },
-    {
-      "epoch": 1.179269328802039,
-      "grad_norm": 0.4453296363353729,
-      "learning_rate": 6.830527733562743e-06,
-      "loss": 0.3635,
-      "step": 18044
-    },
-    {
-      "epoch": 1.1793346840075811,
-      "grad_norm": 0.44329598546028137,
-      "learning_rate": 6.8302027821903185e-06,
-      "loss": 0.3469,
-      "step": 18045
-    },
-    {
-      "epoch": 1.1794000392131234,
-      "grad_norm": 0.4198266863822937,
-      "learning_rate": 6.829877821891362e-06,
-      "loss": 0.3051,
-      "step": 18046
-    },
-    {
-      "epoch": 1.1794653944186655,
-      "grad_norm": 0.440769761800766,
-      "learning_rate": 6.829552852667457e-06,
-      "loss": 0.3198,
-      "step": 18047
-    },
-    {
-      "epoch": 1.1795307496242076,
-      "grad_norm": 0.4311501383781433,
-      "learning_rate": 6.82922787452019e-06,
-      "loss": 0.2923,
-      "step": 18048
-    },
-    {
-      "epoch": 1.1795961048297496,
-      "grad_norm": 0.42883557081222534,
-      "learning_rate": 6.828902887451143e-06,
-      "loss": 0.3423,
-      "step": 18049
-    },
-    {
-      "epoch": 1.179661460035292,
-      "grad_norm": 0.409929096698761,
-      "learning_rate": 6.828577891461905e-06,
-      "loss": 0.2918,
-      "step": 18050
-    },
-    {
-      "epoch": 1.179726815240834,
-      "grad_norm": 0.4326501786708832,
-      "learning_rate": 6.8282528865540585e-06,
-      "loss": 0.328,
-      "step": 18051
-    },
-    {
-      "epoch": 1.179792170446376,
-      "grad_norm": 0.44737958908081055,
-      "learning_rate": 6.82792787272919e-06,
-      "loss": 0.3356,
-      "step": 18052
-    },
-    {
-      "epoch": 1.1798575256519181,
-      "grad_norm": 0.4553202986717224,
-      "learning_rate": 6.827602849988883e-06,
-      "loss": 0.3593,
-      "step": 18053
-    },
-    {
-      "epoch": 1.1799228808574602,
-      "grad_norm": 0.4543243646621704,
-      "learning_rate": 6.827277818334724e-06,
-      "loss": 0.3824,
-      "step": 18054
-    },
-    {
-      "epoch": 1.1799882360630023,
-      "grad_norm": 0.4133303463459015,
-      "learning_rate": 6.826952777768299e-06,
-      "loss": 0.2872,
-      "step": 18055
-    },
-    {
-      "epoch": 1.1800535912685446,
-      "grad_norm": 0.43816137313842773,
-      "learning_rate": 6.826627728291191e-06,
-      "loss": 0.3177,
-      "step": 18056
-    },
-    {
-      "epoch": 1.1801189464740867,
-      "grad_norm": 0.41716015338897705,
-      "learning_rate": 6.826302669904987e-06,
-      "loss": 0.312,
-      "step": 18057
-    },
-    {
-      "epoch": 1.1801843016796287,
-      "grad_norm": 0.4646686911582947,
-      "learning_rate": 6.825977602611271e-06,
-      "loss": 0.3586,
-      "step": 18058
-    },
-    {
-      "epoch": 1.180249656885171,
-      "grad_norm": 0.4383331835269928,
-      "learning_rate": 6.825652526411632e-06,
-      "loss": 0.3362,
-      "step": 18059
-    },
-    {
-      "epoch": 1.180315012090713,
-      "grad_norm": 0.4383421838283539,
-      "learning_rate": 6.825327441307652e-06,
-      "loss": 0.3199,
-      "step": 18060
-    },
-    {
-      "epoch": 1.1803803672962552,
-      "grad_norm": 0.45525622367858887,
-      "learning_rate": 6.825002347300919e-06,
-      "loss": 0.3394,
-      "step": 18061
-    },
-    {
-      "epoch": 1.1804457225017972,
-      "grad_norm": 0.4475131034851074,
-      "learning_rate": 6.824677244393017e-06,
-      "loss": 0.322,
-      "step": 18062
-    },
-    {
-      "epoch": 1.1805110777073393,
-      "grad_norm": 0.44492107629776,
-      "learning_rate": 6.82435213258553e-06,
-      "loss": 0.3685,
-      "step": 18063
-    },
-    {
-      "epoch": 1.1805764329128814,
-      "grad_norm": 0.44649577140808105,
-      "learning_rate": 6.824027011880047e-06,
-      "loss": 0.3119,
-      "step": 18064
-    },
-    {
-      "epoch": 1.1806417881184237,
-      "grad_norm": 0.44526177644729614,
-      "learning_rate": 6.823701882278151e-06,
-      "loss": 0.3457,
-      "step": 18065
-    },
-    {
-      "epoch": 1.1807071433239658,
-      "grad_norm": 0.4460541009902954,
-      "learning_rate": 6.823376743781432e-06,
-      "loss": 0.3343,
-      "step": 18066
-    },
-    {
-      "epoch": 1.1807724985295078,
-      "grad_norm": 0.4453355073928833,
-      "learning_rate": 6.82305159639147e-06,
-      "loss": 0.3244,
-      "step": 18067
-    },
-    {
-      "epoch": 1.18083785373505,
-      "grad_norm": 0.49602919816970825,
-      "learning_rate": 6.822726440109854e-06,
-      "loss": 0.4142,
-      "step": 18068
-    },
-    {
-      "epoch": 1.1809032089405922,
-      "grad_norm": 0.44526922702789307,
-      "learning_rate": 6.8224012749381714e-06,
-      "loss": 0.3234,
-      "step": 18069
-    },
-    {
-      "epoch": 1.1809685641461343,
-      "grad_norm": 0.4330725371837616,
-      "learning_rate": 6.822076100878006e-06,
-      "loss": 0.3019,
-      "step": 18070
-    },
-    {
-      "epoch": 1.1810339193516763,
-      "grad_norm": 0.4495363235473633,
-      "learning_rate": 6.821750917930945e-06,
-      "loss": 0.3354,
-      "step": 18071
-    },
-    {
-      "epoch": 1.1810992745572184,
-      "grad_norm": 0.4604150056838989,
-      "learning_rate": 6.821425726098572e-06,
-      "loss": 0.3837,
-      "step": 18072
-    },
-    {
-      "epoch": 1.1811646297627605,
-      "grad_norm": 0.44327229261398315,
-      "learning_rate": 6.821100525382476e-06,
-      "loss": 0.3367,
-      "step": 18073
-    },
-    {
-      "epoch": 1.1812299849683028,
-      "grad_norm": 0.4184994697570801,
-      "learning_rate": 6.82077531578424e-06,
-      "loss": 0.3098,
-      "step": 18074
-    },
-    {
-      "epoch": 1.1812953401738449,
-      "grad_norm": 0.4676985740661621,
-      "learning_rate": 6.820450097305454e-06,
-      "loss": 0.3401,
-      "step": 18075
-    },
-    {
-      "epoch": 1.181360695379387,
-      "grad_norm": 0.47087225317955017,
-      "learning_rate": 6.820124869947702e-06,
-      "loss": 0.3685,
-      "step": 18076
-    },
-    {
-      "epoch": 1.181426050584929,
-      "grad_norm": 0.42198848724365234,
-      "learning_rate": 6.819799633712569e-06,
-      "loss": 0.3365,
-      "step": 18077
-    },
-    {
-      "epoch": 1.1814914057904713,
-      "grad_norm": 0.45783135294914246,
-      "learning_rate": 6.819474388601644e-06,
-      "loss": 0.3503,
-      "step": 18078
-    },
-    {
-      "epoch": 1.1815567609960134,
-      "grad_norm": 0.44989100098609924,
-      "learning_rate": 6.8191491346165114e-06,
-      "loss": 0.3586,
-      "step": 18079
-    },
-    {
-      "epoch": 1.1816221162015554,
-      "grad_norm": 0.4347403645515442,
-      "learning_rate": 6.81882387175876e-06,
-      "loss": 0.335,
-      "step": 18080
-    },
-    {
-      "epoch": 1.1816874714070975,
-      "grad_norm": 0.48094576597213745,
-      "learning_rate": 6.818498600029972e-06,
-      "loss": 0.3587,
-      "step": 18081
-    },
-    {
-      "epoch": 1.1817528266126396,
-      "grad_norm": 0.4099532663822174,
-      "learning_rate": 6.818173319431738e-06,
-      "loss": 0.3026,
-      "step": 18082
-    },
-    {
-      "epoch": 1.1818181818181819,
-      "grad_norm": 0.47157734632492065,
-      "learning_rate": 6.817848029965641e-06,
-      "loss": 0.3631,
-      "step": 18083
-    },
-    {
-      "epoch": 1.181883537023724,
-      "grad_norm": 0.4975356161594391,
-      "learning_rate": 6.817522731633271e-06,
-      "loss": 0.3912,
-      "step": 18084
-    },
-    {
-      "epoch": 1.181948892229266,
-      "grad_norm": 0.4343315362930298,
-      "learning_rate": 6.817197424436212e-06,
-      "loss": 0.321,
-      "step": 18085
-    },
-    {
-      "epoch": 1.182014247434808,
-      "grad_norm": 0.46990543603897095,
-      "learning_rate": 6.816872108376054e-06,
-      "loss": 0.37,
-      "step": 18086
-    },
-    {
-      "epoch": 1.1820796026403504,
-      "grad_norm": 0.4753810465335846,
-      "learning_rate": 6.816546783454379e-06,
-      "loss": 0.39,
-      "step": 18087
-    },
-    {
-      "epoch": 1.1821449578458925,
-      "grad_norm": 0.45715683698654175,
-      "learning_rate": 6.8162214496727765e-06,
-      "loss": 0.3603,
-      "step": 18088
-    },
-    {
-      "epoch": 1.1822103130514345,
-      "grad_norm": 0.45637282729148865,
-      "learning_rate": 6.815896107032833e-06,
-      "loss": 0.3615,
-      "step": 18089
-    },
-    {
-      "epoch": 1.1822756682569766,
-      "grad_norm": 0.4541541635990143,
-      "learning_rate": 6.815570755536134e-06,
-      "loss": 0.3329,
-      "step": 18090
-    },
-    {
-      "epoch": 1.1823410234625187,
-      "grad_norm": 0.460896760225296,
-      "learning_rate": 6.815245395184269e-06,
-      "loss": 0.3797,
-      "step": 18091
-    },
-    {
-      "epoch": 1.182406378668061,
-      "grad_norm": 0.41321203112602234,
-      "learning_rate": 6.814920025978822e-06,
-      "loss": 0.2847,
-      "step": 18092
-    },
-    {
-      "epoch": 1.182471733873603,
-      "grad_norm": 0.44384047389030457,
-      "learning_rate": 6.814594647921384e-06,
-      "loss": 0.322,
-      "step": 18093
-    },
-    {
-      "epoch": 1.1825370890791451,
-      "grad_norm": 0.5093187689781189,
-      "learning_rate": 6.814269261013537e-06,
-      "loss": 0.407,
-      "step": 18094
-    },
-    {
-      "epoch": 1.1826024442846872,
-      "grad_norm": 0.502673327922821,
-      "learning_rate": 6.81394386525687e-06,
-      "loss": 0.4091,
-      "step": 18095
-    },
-    {
-      "epoch": 1.1826677994902295,
-      "grad_norm": 0.4660140573978424,
-      "learning_rate": 6.813618460652971e-06,
-      "loss": 0.3839,
-      "step": 18096
-    },
-    {
-      "epoch": 1.1827331546957716,
-      "grad_norm": 0.44856759905815125,
-      "learning_rate": 6.813293047203426e-06,
-      "loss": 0.3727,
-      "step": 18097
-    },
-    {
-      "epoch": 1.1827985099013136,
-      "grad_norm": 0.4533741772174835,
-      "learning_rate": 6.812967624909823e-06,
-      "loss": 0.3414,
-      "step": 18098
-    },
-    {
-      "epoch": 1.1828638651068557,
-      "grad_norm": 0.47776374220848083,
-      "learning_rate": 6.81264219377375e-06,
-      "loss": 0.383,
-      "step": 18099
-    },
-    {
-      "epoch": 1.1829292203123978,
-      "grad_norm": 0.4365442991256714,
-      "learning_rate": 6.812316753796791e-06,
-      "loss": 0.3531,
-      "step": 18100
-    },
-    {
-      "epoch": 1.18299457551794,
-      "grad_norm": 0.48415282368659973,
-      "learning_rate": 6.811991304980536e-06,
-      "loss": 0.3852,
-      "step": 18101
-    },
-    {
-      "epoch": 1.1830599307234821,
-      "grad_norm": 0.45601290464401245,
-      "learning_rate": 6.8116658473265725e-06,
-      "loss": 0.3581,
-      "step": 18102
-    },
-    {
-      "epoch": 1.1831252859290242,
-      "grad_norm": 0.43471112847328186,
-      "learning_rate": 6.811340380836486e-06,
-      "loss": 0.3346,
-      "step": 18103
-    },
-    {
-      "epoch": 1.1831906411345663,
-      "grad_norm": 0.4638686776161194,
-      "learning_rate": 6.811014905511866e-06,
-      "loss": 0.3616,
-      "step": 18104
-    },
-    {
-      "epoch": 1.1832559963401086,
-      "grad_norm": 0.46512529253959656,
-      "learning_rate": 6.810689421354297e-06,
-      "loss": 0.3722,
-      "step": 18105
-    },
-    {
-      "epoch": 1.1833213515456507,
-      "grad_norm": 0.4197812080383301,
-      "learning_rate": 6.810363928365371e-06,
-      "loss": 0.3259,
-      "step": 18106
-    },
-    {
-      "epoch": 1.1833867067511927,
-      "grad_norm": 0.4628385603427887,
-      "learning_rate": 6.810038426546672e-06,
-      "loss": 0.3653,
-      "step": 18107
-    },
-    {
-      "epoch": 1.1834520619567348,
-      "grad_norm": 0.43266236782073975,
-      "learning_rate": 6.809712915899788e-06,
-      "loss": 0.2942,
-      "step": 18108
-    },
-    {
-      "epoch": 1.1835174171622769,
-      "grad_norm": 0.4563673734664917,
-      "learning_rate": 6.809387396426308e-06,
-      "loss": 0.3504,
-      "step": 18109
-    },
-    {
-      "epoch": 1.1835827723678192,
-      "grad_norm": 0.4399195611476898,
-      "learning_rate": 6.809061868127817e-06,
-      "loss": 0.3247,
-      "step": 18110
-    },
-    {
-      "epoch": 1.1836481275733612,
-      "grad_norm": 0.45540115237236023,
-      "learning_rate": 6.8087363310059075e-06,
-      "loss": 0.3551,
-      "step": 18111
-    },
-    {
-      "epoch": 1.1837134827789033,
-      "grad_norm": 0.45073649287223816,
-      "learning_rate": 6.808410785062161e-06,
-      "loss": 0.3349,
-      "step": 18112
-    },
-    {
-      "epoch": 1.1837788379844454,
-      "grad_norm": 0.4552570581436157,
-      "learning_rate": 6.808085230298172e-06,
-      "loss": 0.3421,
-      "step": 18113
-    },
-    {
-      "epoch": 1.1838441931899877,
-      "grad_norm": 0.4731093943119049,
-      "learning_rate": 6.807759666715522e-06,
-      "loss": 0.3883,
-      "step": 18114
-    },
-    {
-      "epoch": 1.1839095483955298,
-      "grad_norm": 0.44675278663635254,
-      "learning_rate": 6.807434094315803e-06,
-      "loss": 0.3526,
-      "step": 18115
-    },
-    {
-      "epoch": 1.1839749036010718,
-      "grad_norm": 0.47140228748321533,
-      "learning_rate": 6.807108513100602e-06,
-      "loss": 0.3963,
-      "step": 18116
-    },
-    {
-      "epoch": 1.184040258806614,
-      "grad_norm": 0.4452797472476959,
-      "learning_rate": 6.806782923071506e-06,
-      "loss": 0.3123,
-      "step": 18117
-    },
-    {
-      "epoch": 1.184105614012156,
-      "grad_norm": 0.4607033133506775,
-      "learning_rate": 6.8064573242301056e-06,
-      "loss": 0.3494,
-      "step": 18118
-    },
-    {
-      "epoch": 1.1841709692176983,
-      "grad_norm": 0.43921059370040894,
-      "learning_rate": 6.806131716577985e-06,
-      "loss": 0.3067,
-      "step": 18119
-    },
-    {
-      "epoch": 1.1842363244232403,
-      "grad_norm": 0.46310117840766907,
-      "learning_rate": 6.805806100116735e-06,
-      "loss": 0.3681,
-      "step": 18120
-    },
-    {
-      "epoch": 1.1843016796287824,
-      "grad_norm": 0.4808109402656555,
-      "learning_rate": 6.805480474847943e-06,
-      "loss": 0.3667,
-      "step": 18121
-    },
-    {
-      "epoch": 1.1843670348343245,
-      "grad_norm": 0.4988486170768738,
-      "learning_rate": 6.805154840773198e-06,
-      "loss": 0.3889,
-      "step": 18122
-    },
-    {
-      "epoch": 1.1844323900398668,
-      "grad_norm": 0.46786004304885864,
-      "learning_rate": 6.804829197894086e-06,
-      "loss": 0.3852,
-      "step": 18123
-    },
-    {
-      "epoch": 1.1844977452454089,
-      "grad_norm": 0.46828898787498474,
-      "learning_rate": 6.804503546212198e-06,
-      "loss": 0.3394,
-      "step": 18124
-    },
-    {
-      "epoch": 1.184563100450951,
-      "grad_norm": 0.4303341507911682,
-      "learning_rate": 6.804177885729119e-06,
-      "loss": 0.3183,
-      "step": 18125
-    },
-    {
-      "epoch": 1.184628455656493,
-      "grad_norm": 0.46578550338745117,
-      "learning_rate": 6.803852216446443e-06,
-      "loss": 0.3197,
-      "step": 18126
-    },
-    {
-      "epoch": 1.184693810862035,
-      "grad_norm": 0.45559433102607727,
-      "learning_rate": 6.803526538365752e-06,
-      "loss": 0.3527,
-      "step": 18127
-    },
-    {
-      "epoch": 1.1847591660675774,
-      "grad_norm": 0.4716496467590332,
-      "learning_rate": 6.803200851488638e-06,
-      "loss": 0.3583,
-      "step": 18128
-    },
-    {
-      "epoch": 1.1848245212731194,
-      "grad_norm": 0.4363313317298889,
-      "learning_rate": 6.802875155816689e-06,
-      "loss": 0.3159,
-      "step": 18129
-    },
-    {
-      "epoch": 1.1848898764786615,
-      "grad_norm": 0.43479371070861816,
-      "learning_rate": 6.802549451351494e-06,
-      "loss": 0.3507,
-      "step": 18130
-    },
-    {
-      "epoch": 1.1849552316842036,
-      "grad_norm": 0.4546174705028534,
-      "learning_rate": 6.80222373809464e-06,
-      "loss": 0.3647,
-      "step": 18131
-    },
-    {
-      "epoch": 1.1850205868897459,
-      "grad_norm": 0.4394910931587219,
-      "learning_rate": 6.8018980160477155e-06,
-      "loss": 0.3369,
-      "step": 18132
-    },
-    {
-      "epoch": 1.185085942095288,
-      "grad_norm": 0.4569244980812073,
-      "learning_rate": 6.801572285212311e-06,
-      "loss": 0.3609,
-      "step": 18133
-    },
-    {
-      "epoch": 1.18515129730083,
-      "grad_norm": 0.4328986406326294,
-      "learning_rate": 6.801246545590016e-06,
-      "loss": 0.3182,
-      "step": 18134
-    },
-    {
-      "epoch": 1.185216652506372,
-      "grad_norm": 0.4514637291431427,
-      "learning_rate": 6.800920797182416e-06,
-      "loss": 0.3076,
-      "step": 18135
-    },
-    {
-      "epoch": 1.1852820077119142,
-      "grad_norm": 0.4802858233451843,
-      "learning_rate": 6.800595039991101e-06,
-      "loss": 0.3824,
-      "step": 18136
-    },
-    {
-      "epoch": 1.1853473629174565,
-      "grad_norm": 0.4062332510948181,
-      "learning_rate": 6.8002692740176615e-06,
-      "loss": 0.2999,
-      "step": 18137
-    },
-    {
-      "epoch": 1.1854127181229985,
-      "grad_norm": 0.44033023715019226,
-      "learning_rate": 6.799943499263683e-06,
-      "loss": 0.3488,
-      "step": 18138
-    },
-    {
-      "epoch": 1.1854780733285406,
-      "grad_norm": 0.4408622086048126,
-      "learning_rate": 6.7996177157307574e-06,
-      "loss": 0.3293,
-      "step": 18139
-    },
-    {
-      "epoch": 1.1855434285340827,
-      "grad_norm": 0.4373001456260681,
-      "learning_rate": 6.799291923420475e-06,
-      "loss": 0.3306,
-      "step": 18140
-    },
-    {
-      "epoch": 1.185608783739625,
-      "grad_norm": 0.44007745385169983,
-      "learning_rate": 6.79896612233442e-06,
-      "loss": 0.3139,
-      "step": 18141
-    },
-    {
-      "epoch": 1.185674138945167,
-      "grad_norm": 0.48035043478012085,
-      "learning_rate": 6.7986403124741836e-06,
-      "loss": 0.39,
-      "step": 18142
-    },
-    {
-      "epoch": 1.1857394941507091,
-      "grad_norm": 0.43871375918388367,
-      "learning_rate": 6.798314493841356e-06,
-      "loss": 0.3506,
-      "step": 18143
-    },
-    {
-      "epoch": 1.1858048493562512,
-      "grad_norm": 0.49691376090049744,
-      "learning_rate": 6.797988666437527e-06,
-      "loss": 0.4343,
-      "step": 18144
-    },
-    {
-      "epoch": 1.1858702045617933,
-      "grad_norm": 0.46149203181266785,
-      "learning_rate": 6.797662830264283e-06,
-      "loss": 0.3351,
-      "step": 18145
-    },
-    {
-      "epoch": 1.1859355597673356,
-      "grad_norm": 0.4465920329093933,
-      "learning_rate": 6.797336985323215e-06,
-      "loss": 0.3617,
-      "step": 18146
-    },
-    {
-      "epoch": 1.1860009149728776,
-      "grad_norm": 0.5167786478996277,
-      "learning_rate": 6.797011131615912e-06,
-      "loss": 0.3092,
-      "step": 18147
-    },
-    {
-      "epoch": 1.1860662701784197,
-      "grad_norm": 0.4532272517681122,
-      "learning_rate": 6.796685269143962e-06,
-      "loss": 0.3655,
-      "step": 18148
-    },
-    {
-      "epoch": 1.1861316253839618,
-      "grad_norm": 0.4581167995929718,
-      "learning_rate": 6.796359397908957e-06,
-      "loss": 0.356,
-      "step": 18149
-    },
-    {
-      "epoch": 1.186196980589504,
-      "grad_norm": 0.46039530634880066,
-      "learning_rate": 6.796033517912483e-06,
-      "loss": 0.3574,
-      "step": 18150
-    },
-    {
-      "epoch": 1.1862623357950461,
-      "grad_norm": 0.4544249176979065,
-      "learning_rate": 6.795707629156134e-06,
-      "loss": 0.3652,
-      "step": 18151
-    },
-    {
-      "epoch": 1.1863276910005882,
-      "grad_norm": 0.43098723888397217,
-      "learning_rate": 6.7953817316414946e-06,
-      "loss": 0.31,
-      "step": 18152
-    },
-    {
-      "epoch": 1.1863930462061303,
-      "grad_norm": 0.44438937306404114,
-      "learning_rate": 6.795055825370158e-06,
-      "loss": 0.3508,
-      "step": 18153
-    },
-    {
-      "epoch": 1.1864584014116724,
-      "grad_norm": 0.4400290250778198,
-      "learning_rate": 6.794729910343712e-06,
-      "loss": 0.3403,
-      "step": 18154
-    },
-    {
-      "epoch": 1.1865237566172147,
-      "grad_norm": 0.4446077048778534,
-      "learning_rate": 6.794403986563746e-06,
-      "loss": 0.3293,
-      "step": 18155
-    },
-    {
-      "epoch": 1.1865891118227567,
-      "grad_norm": 0.44412150979042053,
-      "learning_rate": 6.794078054031852e-06,
-      "loss": 0.3233,
-      "step": 18156
-    },
-    {
-      "epoch": 1.1866544670282988,
-      "grad_norm": 0.45930489897727966,
-      "learning_rate": 6.793752112749616e-06,
-      "loss": 0.3671,
-      "step": 18157
-    },
-    {
-      "epoch": 1.1867198222338409,
-      "grad_norm": 0.46669676899909973,
-      "learning_rate": 6.793426162718629e-06,
-      "loss": 0.312,
-      "step": 18158
-    },
-    {
-      "epoch": 1.1867851774393832,
-      "grad_norm": 0.43126240372657776,
-      "learning_rate": 6.793100203940481e-06,
-      "loss": 0.3312,
-      "step": 18159
-    },
-    {
-      "epoch": 1.1868505326449252,
-      "grad_norm": 0.46476152539253235,
-      "learning_rate": 6.792774236416764e-06,
-      "loss": 0.3863,
-      "step": 18160
-    },
-    {
-      "epoch": 1.1869158878504673,
-      "grad_norm": 0.43273457884788513,
-      "learning_rate": 6.792448260149065e-06,
-      "loss": 0.3454,
-      "step": 18161
-    },
-    {
-      "epoch": 1.1869812430560094,
-      "grad_norm": 0.43777090311050415,
-      "learning_rate": 6.7921222751389746e-06,
-      "loss": 0.3272,
-      "step": 18162
-    },
-    {
-      "epoch": 1.1870465982615515,
-      "grad_norm": 0.436638742685318,
-      "learning_rate": 6.791796281388084e-06,
-      "loss": 0.3388,
-      "step": 18163
-    },
-    {
-      "epoch": 1.1871119534670938,
-      "grad_norm": 0.46859070658683777,
-      "learning_rate": 6.79147027889798e-06,
-      "loss": 0.3713,
-      "step": 18164
-    },
-    {
-      "epoch": 1.1871773086726358,
-      "grad_norm": 0.4711884558200836,
-      "learning_rate": 6.791144267670258e-06,
-      "loss": 0.3519,
-      "step": 18165
-    },
-    {
-      "epoch": 1.187242663878178,
-      "grad_norm": 0.4488896131515503,
-      "learning_rate": 6.790818247706502e-06,
-      "loss": 0.3804,
-      "step": 18166
-    },
-    {
-      "epoch": 1.18730801908372,
-      "grad_norm": 0.48901477456092834,
-      "learning_rate": 6.790492219008306e-06,
-      "loss": 0.4211,
-      "step": 18167
-    },
-    {
-      "epoch": 1.1873733742892623,
-      "grad_norm": 0.4810936450958252,
-      "learning_rate": 6.790166181577259e-06,
-      "loss": 0.3855,
-      "step": 18168
-    },
-    {
-      "epoch": 1.1874387294948043,
-      "grad_norm": 0.4303014278411865,
-      "learning_rate": 6.789840135414952e-06,
-      "loss": 0.308,
-      "step": 18169
-    },
-    {
-      "epoch": 1.1875040847003464,
-      "grad_norm": 0.4382808804512024,
-      "learning_rate": 6.7895140805229745e-06,
-      "loss": 0.3377,
-      "step": 18170
-    },
-    {
-      "epoch": 1.1875694399058885,
-      "grad_norm": 0.4406227171421051,
-      "learning_rate": 6.789188016902917e-06,
-      "loss": 0.3416,
-      "step": 18171
-    },
-    {
-      "epoch": 1.1876347951114306,
-      "grad_norm": 0.4523710608482361,
-      "learning_rate": 6.78886194455637e-06,
-      "loss": 0.327,
-      "step": 18172
-    },
-    {
-      "epoch": 1.1877001503169726,
-      "grad_norm": 0.47915711998939514,
-      "learning_rate": 6.788535863484922e-06,
-      "loss": 0.3772,
-      "step": 18173
-    },
-    {
-      "epoch": 1.187765505522515,
-      "grad_norm": 0.5097954273223877,
-      "learning_rate": 6.788209773690166e-06,
-      "loss": 0.4274,
-      "step": 18174
-    },
-    {
-      "epoch": 1.187830860728057,
-      "grad_norm": 0.46755391359329224,
-      "learning_rate": 6.787883675173691e-06,
-      "loss": 0.3384,
-      "step": 18175
-    },
-    {
-      "epoch": 1.187896215933599,
-      "grad_norm": 0.45391738414764404,
-      "learning_rate": 6.787557567937089e-06,
-      "loss": 0.361,
-      "step": 18176
-    },
-    {
-      "epoch": 1.1879615711391411,
-      "grad_norm": 0.45209866762161255,
-      "learning_rate": 6.787231451981949e-06,
-      "loss": 0.3293,
-      "step": 18177
-    },
-    {
-      "epoch": 1.1880269263446834,
-      "grad_norm": 0.43817824125289917,
-      "learning_rate": 6.786905327309863e-06,
-      "loss": 0.3186,
-      "step": 18178
-    },
-    {
-      "epoch": 1.1880922815502255,
-      "grad_norm": 0.6073752045631409,
-      "learning_rate": 6.786579193922418e-06,
-      "loss": 0.3617,
-      "step": 18179
-    },
-    {
-      "epoch": 1.1881576367557676,
-      "grad_norm": 0.4198361337184906,
-      "learning_rate": 6.78625305182121e-06,
-      "loss": 0.3138,
-      "step": 18180
-    },
-    {
-      "epoch": 1.1882229919613096,
-      "grad_norm": 0.46060195565223694,
-      "learning_rate": 6.7859269010078255e-06,
-      "loss": 0.3766,
-      "step": 18181
-    },
-    {
-      "epoch": 1.1882883471668517,
-      "grad_norm": 0.4528558552265167,
-      "learning_rate": 6.785600741483857e-06,
-      "loss": 0.3605,
-      "step": 18182
-    },
-    {
-      "epoch": 1.188353702372394,
-      "grad_norm": 0.4450226426124573,
-      "learning_rate": 6.785274573250896e-06,
-      "loss": 0.3322,
-      "step": 18183
-    },
-    {
-      "epoch": 1.188419057577936,
-      "grad_norm": 0.4433911144733429,
-      "learning_rate": 6.7849483963105314e-06,
-      "loss": 0.3423,
-      "step": 18184
-    },
-    {
-      "epoch": 1.1884844127834782,
-      "grad_norm": 0.44697684049606323,
-      "learning_rate": 6.784622210664355e-06,
-      "loss": 0.2936,
-      "step": 18185
-    },
-    {
-      "epoch": 1.1885497679890202,
-      "grad_norm": 0.4662095308303833,
-      "learning_rate": 6.784296016313958e-06,
-      "loss": 0.3624,
-      "step": 18186
-    },
-    {
-      "epoch": 1.1886151231945625,
-      "grad_norm": 0.4551587402820587,
-      "learning_rate": 6.783969813260932e-06,
-      "loss": 0.3524,
-      "step": 18187
-    },
-    {
-      "epoch": 1.1886804784001046,
-      "grad_norm": 0.46663618087768555,
-      "learning_rate": 6.783643601506866e-06,
-      "loss": 0.3618,
-      "step": 18188
-    },
-    {
-      "epoch": 1.1887458336056467,
-      "grad_norm": 0.4618913531303406,
-      "learning_rate": 6.783317381053354e-06,
-      "loss": 0.3654,
-      "step": 18189
-    },
-    {
-      "epoch": 1.1888111888111887,
-      "grad_norm": 0.44946882128715515,
-      "learning_rate": 6.782991151901983e-06,
-      "loss": 0.3071,
-      "step": 18190
-    },
-    {
-      "epoch": 1.1888765440167308,
-      "grad_norm": 0.5025569200515747,
-      "learning_rate": 6.782664914054349e-06,
-      "loss": 0.3888,
-      "step": 18191
-    },
-    {
-      "epoch": 1.1889418992222731,
-      "grad_norm": 0.4740449786186218,
-      "learning_rate": 6.78233866751204e-06,
-      "loss": 0.3712,
-      "step": 18192
-    },
-    {
-      "epoch": 1.1890072544278152,
-      "grad_norm": 0.43006962537765503,
-      "learning_rate": 6.782012412276646e-06,
-      "loss": 0.3298,
-      "step": 18193
-    },
-    {
-      "epoch": 1.1890726096333573,
-      "grad_norm": 0.44759437441825867,
-      "learning_rate": 6.781686148349762e-06,
-      "loss": 0.3543,
-      "step": 18194
-    },
-    {
-      "epoch": 1.1891379648388993,
-      "grad_norm": 0.5013605356216431,
-      "learning_rate": 6.781359875732976e-06,
-      "loss": 0.3386,
-      "step": 18195
-    },
-    {
-      "epoch": 1.1892033200444416,
-      "grad_norm": 0.438453733921051,
-      "learning_rate": 6.781033594427882e-06,
-      "loss": 0.3291,
-      "step": 18196
-    },
-    {
-      "epoch": 1.1892686752499837,
-      "grad_norm": 0.48102906346321106,
-      "learning_rate": 6.780707304436069e-06,
-      "loss": 0.4035,
-      "step": 18197
-    },
-    {
-      "epoch": 1.1893340304555258,
-      "grad_norm": 0.4649040699005127,
-      "learning_rate": 6.780381005759131e-06,
-      "loss": 0.3428,
-      "step": 18198
-    },
-    {
-      "epoch": 1.1893993856610678,
-      "grad_norm": 0.450388640165329,
-      "learning_rate": 6.780054698398657e-06,
-      "loss": 0.359,
-      "step": 18199
-    },
-    {
-      "epoch": 1.18946474086661,
-      "grad_norm": 0.5376363396644592,
-      "learning_rate": 6.779728382356241e-06,
-      "loss": 0.387,
-      "step": 18200
-    },
-    {
-      "epoch": 1.1895300960721522,
-      "grad_norm": 0.486173152923584,
-      "learning_rate": 6.7794020576334705e-06,
-      "loss": 0.3701,
-      "step": 18201
-    },
-    {
-      "epoch": 1.1895954512776943,
-      "grad_norm": 0.44787952303886414,
-      "learning_rate": 6.779075724231942e-06,
-      "loss": 0.3636,
-      "step": 18202
-    },
-    {
-      "epoch": 1.1896608064832364,
-      "grad_norm": 0.4356648623943329,
-      "learning_rate": 6.778749382153245e-06,
-      "loss": 0.333,
-      "step": 18203
-    },
-    {
-      "epoch": 1.1897261616887784,
-      "grad_norm": 0.42434027791023254,
-      "learning_rate": 6.778423031398968e-06,
-      "loss": 0.2784,
-      "step": 18204
-    },
-    {
-      "epoch": 1.1897915168943207,
-      "grad_norm": 0.45641762018203735,
-      "learning_rate": 6.778096671970709e-06,
-      "loss": 0.3339,
-      "step": 18205
-    },
-    {
-      "epoch": 1.1898568720998628,
-      "grad_norm": 0.43157657980918884,
-      "learning_rate": 6.7777703038700546e-06,
-      "loss": 0.3295,
-      "step": 18206
-    },
-    {
-      "epoch": 1.1899222273054049,
-      "grad_norm": 0.4396790564060211,
-      "learning_rate": 6.7774439270986e-06,
-      "loss": 0.3312,
-      "step": 18207
-    },
-    {
-      "epoch": 1.189987582510947,
-      "grad_norm": 0.447013258934021,
-      "learning_rate": 6.777117541657935e-06,
-      "loss": 0.3442,
-      "step": 18208
-    },
-    {
-      "epoch": 1.190052937716489,
-      "grad_norm": 0.4444758892059326,
-      "learning_rate": 6.776791147549652e-06,
-      "loss": 0.3545,
-      "step": 18209
-    },
-    {
-      "epoch": 1.1901182929220313,
-      "grad_norm": 0.4434734284877777,
-      "learning_rate": 6.776464744775344e-06,
-      "loss": 0.3232,
-      "step": 18210
-    },
-    {
-      "epoch": 1.1901836481275734,
-      "grad_norm": 0.4431239366531372,
-      "learning_rate": 6.7761383333366e-06,
-      "loss": 0.3371,
-      "step": 18211
-    },
-    {
-      "epoch": 1.1902490033331155,
-      "grad_norm": 0.48813825845718384,
-      "learning_rate": 6.775811913235015e-06,
-      "loss": 0.3763,
-      "step": 18212
-    },
-    {
-      "epoch": 1.1903143585386575,
-      "grad_norm": 0.43185481429100037,
-      "learning_rate": 6.775485484472181e-06,
-      "loss": 0.3367,
-      "step": 18213
-    },
-    {
-      "epoch": 1.1903797137441998,
-      "grad_norm": 0.4722045063972473,
-      "learning_rate": 6.775159047049689e-06,
-      "loss": 0.3731,
-      "step": 18214
-    },
-    {
-      "epoch": 1.190445068949742,
-      "grad_norm": 0.4516633450984955,
-      "learning_rate": 6.77483260096913e-06,
-      "loss": 0.3773,
-      "step": 18215
-    },
-    {
-      "epoch": 1.190510424155284,
-      "grad_norm": 0.420488566160202,
-      "learning_rate": 6.774506146232098e-06,
-      "loss": 0.3009,
-      "step": 18216
-    },
-    {
-      "epoch": 1.190575779360826,
-      "grad_norm": 0.46615272760391235,
-      "learning_rate": 6.774179682840185e-06,
-      "loss": 0.3495,
-      "step": 18217
-    },
-    {
-      "epoch": 1.190641134566368,
-      "grad_norm": 0.4303334951400757,
-      "learning_rate": 6.773853210794983e-06,
-      "loss": 0.3512,
-      "step": 18218
-    },
-    {
-      "epoch": 1.1907064897719104,
-      "grad_norm": 0.45234352350234985,
-      "learning_rate": 6.773526730098085e-06,
-      "loss": 0.3491,
-      "step": 18219
-    },
-    {
-      "epoch": 1.1907718449774525,
-      "grad_norm": 0.4492032527923584,
-      "learning_rate": 6.773200240751083e-06,
-      "loss": 0.35,
-      "step": 18220
-    },
-    {
-      "epoch": 1.1908372001829945,
-      "grad_norm": 0.46623244881629944,
-      "learning_rate": 6.772873742755568e-06,
-      "loss": 0.3798,
-      "step": 18221
-    },
-    {
-      "epoch": 1.1909025553885366,
-      "grad_norm": 0.4667704403400421,
-      "learning_rate": 6.772547236113134e-06,
-      "loss": 0.3679,
-      "step": 18222
-    },
-    {
-      "epoch": 1.190967910594079,
-      "grad_norm": 0.4145137071609497,
-      "learning_rate": 6.772220720825373e-06,
-      "loss": 0.3064,
-      "step": 18223
-    },
-    {
-      "epoch": 1.191033265799621,
-      "grad_norm": 0.4596615433692932,
-      "learning_rate": 6.771894196893878e-06,
-      "loss": 0.3566,
-      "step": 18224
-    },
-    {
-      "epoch": 1.191098621005163,
-      "grad_norm": 0.44246578216552734,
-      "learning_rate": 6.771567664320241e-06,
-      "loss": 0.3383,
-      "step": 18225
-    },
-    {
-      "epoch": 1.1911639762107051,
-      "grad_norm": 0.4468989670276642,
-      "learning_rate": 6.7712411231060535e-06,
-      "loss": 0.3472,
-      "step": 18226
-    },
-    {
-      "epoch": 1.1912293314162472,
-      "grad_norm": 0.4473535418510437,
-      "learning_rate": 6.770914573252911e-06,
-      "loss": 0.3263,
-      "step": 18227
-    },
-    {
-      "epoch": 1.1912946866217895,
-      "grad_norm": 0.4716193974018097,
-      "learning_rate": 6.770588014762403e-06,
-      "loss": 0.3683,
-      "step": 18228
-    },
-    {
-      "epoch": 1.1913600418273316,
-      "grad_norm": 0.41069507598876953,
-      "learning_rate": 6.770261447636126e-06,
-      "loss": 0.3095,
-      "step": 18229
-    },
-    {
-      "epoch": 1.1914253970328736,
-      "grad_norm": 0.41994649171829224,
-      "learning_rate": 6.76993487187567e-06,
-      "loss": 0.2926,
-      "step": 18230
-    },
-    {
-      "epoch": 1.1914907522384157,
-      "grad_norm": 0.4434123635292053,
-      "learning_rate": 6.769608287482627e-06,
-      "loss": 0.3388,
-      "step": 18231
-    },
-    {
-      "epoch": 1.191556107443958,
-      "grad_norm": 0.5141905546188354,
-      "learning_rate": 6.769281694458593e-06,
-      "loss": 0.4069,
-      "step": 18232
-    },
-    {
-      "epoch": 1.1916214626495,
-      "grad_norm": 0.44772619009017944,
-      "learning_rate": 6.768955092805158e-06,
-      "loss": 0.3512,
-      "step": 18233
-    },
-    {
-      "epoch": 1.1916868178550422,
-      "grad_norm": 0.4542248249053955,
-      "learning_rate": 6.768628482523918e-06,
-      "loss": 0.3474,
-      "step": 18234
-    },
-    {
-      "epoch": 1.1917521730605842,
-      "grad_norm": 0.43119528889656067,
-      "learning_rate": 6.768301863616462e-06,
-      "loss": 0.2986,
-      "step": 18235
-    },
-    {
-      "epoch": 1.1918175282661263,
-      "grad_norm": 0.45107516646385193,
-      "learning_rate": 6.767975236084387e-06,
-      "loss": 0.3603,
-      "step": 18236
-    },
-    {
-      "epoch": 1.1918828834716686,
-      "grad_norm": 0.4860307574272156,
-      "learning_rate": 6.767648599929284e-06,
-      "loss": 0.4286,
-      "step": 18237
-    },
-    {
-      "epoch": 1.1919482386772107,
-      "grad_norm": 0.4243297576904297,
-      "learning_rate": 6.767321955152746e-06,
-      "loss": 0.2967,
-      "step": 18238
-    },
-    {
-      "epoch": 1.1920135938827527,
-      "grad_norm": 0.4643009305000305,
-      "learning_rate": 6.766995301756366e-06,
-      "loss": 0.3699,
-      "step": 18239
-    },
-    {
-      "epoch": 1.1920789490882948,
-      "grad_norm": 0.43786120414733887,
-      "learning_rate": 6.766668639741738e-06,
-      "loss": 0.3363,
-      "step": 18240
-    },
-    {
-      "epoch": 1.192144304293837,
-      "grad_norm": 0.4591931700706482,
-      "learning_rate": 6.766341969110457e-06,
-      "loss": 0.3099,
-      "step": 18241
-    },
-    {
-      "epoch": 1.1922096594993792,
-      "grad_norm": 0.41595813632011414,
-      "learning_rate": 6.766015289864112e-06,
-      "loss": 0.3128,
-      "step": 18242
-    },
-    {
-      "epoch": 1.1922750147049213,
-      "grad_norm": 0.5058223009109497,
-      "learning_rate": 6.765688602004299e-06,
-      "loss": 0.4074,
-      "step": 18243
-    },
-    {
-      "epoch": 1.1923403699104633,
-      "grad_norm": 0.4456130564212799,
-      "learning_rate": 6.76536190553261e-06,
-      "loss": 0.3295,
-      "step": 18244
-    },
-    {
-      "epoch": 1.1924057251160054,
-      "grad_norm": 0.45912837982177734,
-      "learning_rate": 6.765035200450641e-06,
-      "loss": 0.3172,
-      "step": 18245
-    },
-    {
-      "epoch": 1.1924710803215477,
-      "grad_norm": 0.4306924045085907,
-      "learning_rate": 6.764708486759984e-06,
-      "loss": 0.32,
-      "step": 18246
-    },
-    {
-      "epoch": 1.1925364355270898,
-      "grad_norm": 0.46732673048973083,
-      "learning_rate": 6.764381764462231e-06,
-      "loss": 0.3679,
-      "step": 18247
-    },
-    {
-      "epoch": 1.1926017907326318,
-      "grad_norm": 0.42428305745124817,
-      "learning_rate": 6.764055033558978e-06,
-      "loss": 0.3328,
-      "step": 18248
-    },
-    {
-      "epoch": 1.192667145938174,
-      "grad_norm": 0.4748455584049225,
-      "learning_rate": 6.763728294051817e-06,
-      "loss": 0.327,
-      "step": 18249
-    },
-    {
-      "epoch": 1.1927325011437162,
-      "grad_norm": 0.4488120973110199,
-      "learning_rate": 6.763401545942343e-06,
-      "loss": 0.357,
-      "step": 18250
-    },
-    {
-      "epoch": 1.1927978563492583,
-      "grad_norm": 0.4734101891517639,
-      "learning_rate": 6.763074789232147e-06,
-      "loss": 0.3517,
-      "step": 18251
-    },
-    {
-      "epoch": 1.1928632115548004,
-      "grad_norm": 0.4629245400428772,
-      "learning_rate": 6.762748023922826e-06,
-      "loss": 0.3853,
-      "step": 18252
-    },
-    {
-      "epoch": 1.1929285667603424,
-      "grad_norm": 0.45737552642822266,
-      "learning_rate": 6.762421250015971e-06,
-      "loss": 0.3353,
-      "step": 18253
-    },
-    {
-      "epoch": 1.1929939219658845,
-      "grad_norm": 0.42586827278137207,
-      "learning_rate": 6.762094467513179e-06,
-      "loss": 0.314,
-      "step": 18254
-    },
-    {
-      "epoch": 1.1930592771714268,
-      "grad_norm": 0.4478102922439575,
-      "learning_rate": 6.76176767641604e-06,
-      "loss": 0.3367,
-      "step": 18255
-    },
-    {
-      "epoch": 1.1931246323769689,
-      "grad_norm": 0.4543229937553406,
-      "learning_rate": 6.761440876726151e-06,
-      "loss": 0.3543,
-      "step": 18256
-    },
-    {
-      "epoch": 1.193189987582511,
-      "grad_norm": 0.42153725028038025,
-      "learning_rate": 6.761114068445104e-06,
-      "loss": 0.3132,
-      "step": 18257
-    },
-    {
-      "epoch": 1.193255342788053,
-      "grad_norm": 0.4443165063858032,
-      "learning_rate": 6.760787251574492e-06,
-      "loss": 0.331,
-      "step": 18258
-    },
-    {
-      "epoch": 1.1933206979935953,
-      "grad_norm": 0.4544033408164978,
-      "learning_rate": 6.760460426115913e-06,
-      "loss": 0.3367,
-      "step": 18259
-    },
-    {
-      "epoch": 1.1933860531991374,
-      "grad_norm": 0.46163734793663025,
-      "learning_rate": 6.7601335920709566e-06,
-      "loss": 0.3337,
-      "step": 18260
-    },
-    {
-      "epoch": 1.1934514084046794,
-      "grad_norm": 0.46515271067619324,
-      "learning_rate": 6.759806749441222e-06,
-      "loss": 0.3807,
-      "step": 18261
-    },
-    {
-      "epoch": 1.1935167636102215,
-      "grad_norm": 0.42006489634513855,
-      "learning_rate": 6.759479898228297e-06,
-      "loss": 0.3045,
-      "step": 18262
-    },
-    {
-      "epoch": 1.1935821188157636,
-      "grad_norm": 0.4615705907344818,
-      "learning_rate": 6.759153038433781e-06,
-      "loss": 0.3286,
-      "step": 18263
-    },
-    {
-      "epoch": 1.1936474740213059,
-      "grad_norm": 0.4246770739555359,
-      "learning_rate": 6.758826170059265e-06,
-      "loss": 0.3131,
-      "step": 18264
-    },
-    {
-      "epoch": 1.193712829226848,
-      "grad_norm": 0.4041743874549866,
-      "learning_rate": 6.758499293106345e-06,
-      "loss": 0.287,
-      "step": 18265
-    },
-    {
-      "epoch": 1.19377818443239,
-      "grad_norm": 0.44723162055015564,
-      "learning_rate": 6.758172407576614e-06,
-      "loss": 0.3509,
-      "step": 18266
-    },
-    {
-      "epoch": 1.193843539637932,
-      "grad_norm": 0.4195310175418854,
-      "learning_rate": 6.757845513471668e-06,
-      "loss": 0.312,
-      "step": 18267
-    },
-    {
-      "epoch": 1.1939088948434744,
-      "grad_norm": 0.45519694685935974,
-      "learning_rate": 6.7575186107931e-06,
-      "loss": 0.3444,
-      "step": 18268
-    },
-    {
-      "epoch": 1.1939742500490165,
-      "grad_norm": 0.46090462803840637,
-      "learning_rate": 6.757191699542505e-06,
-      "loss": 0.3635,
-      "step": 18269
-    },
-    {
-      "epoch": 1.1940396052545585,
-      "grad_norm": 0.41247397661209106,
-      "learning_rate": 6.756864779721477e-06,
-      "loss": 0.2704,
-      "step": 18270
-    },
-    {
-      "epoch": 1.1941049604601006,
-      "grad_norm": 0.44314804673194885,
-      "learning_rate": 6.756537851331611e-06,
-      "loss": 0.3595,
-      "step": 18271
-    },
-    {
-      "epoch": 1.1941703156656427,
-      "grad_norm": 0.5021336078643799,
-      "learning_rate": 6.756210914374501e-06,
-      "loss": 0.4201,
-      "step": 18272
-    },
-    {
-      "epoch": 1.194235670871185,
-      "grad_norm": 0.4680226445198059,
-      "learning_rate": 6.755883968851743e-06,
-      "loss": 0.3646,
-      "step": 18273
-    },
-    {
-      "epoch": 1.194301026076727,
-      "grad_norm": 0.42127227783203125,
-      "learning_rate": 6.75555701476493e-06,
-      "loss": 0.3272,
-      "step": 18274
-    },
-    {
-      "epoch": 1.1943663812822691,
-      "grad_norm": 0.4428028166294098,
-      "learning_rate": 6.7552300521156576e-06,
-      "loss": 0.3445,
-      "step": 18275
-    },
-    {
-      "epoch": 1.1944317364878112,
-      "grad_norm": 0.4234282374382019,
-      "learning_rate": 6.754903080905519e-06,
-      "loss": 0.3154,
-      "step": 18276
-    },
-    {
-      "epoch": 1.1944970916933535,
-      "grad_norm": 0.4369146525859833,
-      "learning_rate": 6.754576101136112e-06,
-      "loss": 0.3119,
-      "step": 18277
-    },
-    {
-      "epoch": 1.1945624468988956,
-      "grad_norm": 0.4304760992527008,
-      "learning_rate": 6.754249112809028e-06,
-      "loss": 0.3306,
-      "step": 18278
-    },
-    {
-      "epoch": 1.1946278021044376,
-      "grad_norm": 0.41174158453941345,
-      "learning_rate": 6.753922115925864e-06,
-      "loss": 0.2967,
-      "step": 18279
-    },
-    {
-      "epoch": 1.1946931573099797,
-      "grad_norm": 0.44075703620910645,
-      "learning_rate": 6.753595110488214e-06,
-      "loss": 0.3228,
-      "step": 18280
-    },
-    {
-      "epoch": 1.1947585125155218,
-      "grad_norm": 0.4231216013431549,
-      "learning_rate": 6.753268096497674e-06,
-      "loss": 0.3147,
-      "step": 18281
-    },
-    {
-      "epoch": 1.1948238677210639,
-      "grad_norm": 0.4534105062484741,
-      "learning_rate": 6.752941073955837e-06,
-      "loss": 0.3398,
-      "step": 18282
-    },
-    {
-      "epoch": 1.1948892229266062,
-      "grad_norm": 0.47695019841194153,
-      "learning_rate": 6.752614042864301e-06,
-      "loss": 0.3946,
-      "step": 18283
-    },
-    {
-      "epoch": 1.1949545781321482,
-      "grad_norm": 0.44449788331985474,
-      "learning_rate": 6.752287003224656e-06,
-      "loss": 0.3492,
-      "step": 18284
-    },
-    {
-      "epoch": 1.1950199333376903,
-      "grad_norm": 0.4293326437473297,
-      "learning_rate": 6.751959955038503e-06,
-      "loss": 0.3232,
-      "step": 18285
-    },
-    {
-      "epoch": 1.1950852885432324,
-      "grad_norm": 0.4535491168498993,
-      "learning_rate": 6.751632898307432e-06,
-      "loss": 0.3344,
-      "step": 18286
-    },
-    {
-      "epoch": 1.1951506437487747,
-      "grad_norm": 0.4369806945323944,
-      "learning_rate": 6.751305833033041e-06,
-      "loss": 0.3412,
-      "step": 18287
-    },
-    {
-      "epoch": 1.1952159989543167,
-      "grad_norm": 0.45930948853492737,
-      "learning_rate": 6.750978759216928e-06,
-      "loss": 0.3445,
-      "step": 18288
-    },
-    {
-      "epoch": 1.1952813541598588,
-      "grad_norm": 0.4697064459323883,
-      "learning_rate": 6.750651676860681e-06,
-      "loss": 0.3336,
-      "step": 18289
-    },
-    {
-      "epoch": 1.1953467093654009,
-      "grad_norm": 0.44547131657600403,
-      "learning_rate": 6.7503245859659014e-06,
-      "loss": 0.3166,
-      "step": 18290
-    },
-    {
-      "epoch": 1.195412064570943,
-      "grad_norm": 0.5118493437767029,
-      "learning_rate": 6.7499974865341815e-06,
-      "loss": 0.3937,
-      "step": 18291
-    },
-    {
-      "epoch": 1.1954774197764853,
-      "grad_norm": 0.4325563311576843,
-      "learning_rate": 6.749670378567117e-06,
-      "loss": 0.3135,
-      "step": 18292
-    },
-    {
-      "epoch": 1.1955427749820273,
-      "grad_norm": 0.4876798987388611,
-      "learning_rate": 6.749343262066304e-06,
-      "loss": 0.3801,
-      "step": 18293
-    },
-    {
-      "epoch": 1.1956081301875694,
-      "grad_norm": 0.41790640354156494,
-      "learning_rate": 6.74901613703334e-06,
-      "loss": 0.3113,
-      "step": 18294
-    },
-    {
-      "epoch": 1.1956734853931115,
-      "grad_norm": 0.4941082298755646,
-      "learning_rate": 6.748689003469817e-06,
-      "loss": 0.3767,
-      "step": 18295
-    },
-    {
-      "epoch": 1.1957388405986538,
-      "grad_norm": 0.4460056722164154,
-      "learning_rate": 6.748361861377331e-06,
-      "loss": 0.3448,
-      "step": 18296
-    },
-    {
-      "epoch": 1.1958041958041958,
-      "grad_norm": 0.41867968440055847,
-      "learning_rate": 6.748034710757481e-06,
-      "loss": 0.3002,
-      "step": 18297
-    },
-    {
-      "epoch": 1.195869551009738,
-      "grad_norm": 0.46213915944099426,
-      "learning_rate": 6.747707551611857e-06,
-      "loss": 0.3563,
-      "step": 18298
-    },
-    {
-      "epoch": 1.19593490621528,
-      "grad_norm": 0.43928977847099304,
-      "learning_rate": 6.74738038394206e-06,
-      "loss": 0.336,
-      "step": 18299
-    },
-    {
-      "epoch": 1.196000261420822,
-      "grad_norm": 0.4726327657699585,
-      "learning_rate": 6.747053207749683e-06,
-      "loss": 0.3712,
-      "step": 18300
-    },
-    {
-      "epoch": 1.1960656166263643,
-      "grad_norm": 0.488456666469574,
-      "learning_rate": 6.746726023036323e-06,
-      "loss": 0.3862,
-      "step": 18301
-    },
-    {
-      "epoch": 1.1961309718319064,
-      "grad_norm": 0.435140997171402,
-      "learning_rate": 6.746398829803574e-06,
-      "loss": 0.3398,
-      "step": 18302
-    },
-    {
-      "epoch": 1.1961963270374485,
-      "grad_norm": 0.4667286276817322,
-      "learning_rate": 6.746071628053033e-06,
-      "loss": 0.3788,
-      "step": 18303
-    },
-    {
-      "epoch": 1.1962616822429906,
-      "grad_norm": 0.41344180703163147,
-      "learning_rate": 6.745744417786297e-06,
-      "loss": 0.3298,
-      "step": 18304
-    },
-    {
-      "epoch": 1.1963270374485329,
-      "grad_norm": 0.4400014281272888,
-      "learning_rate": 6.745417199004959e-06,
-      "loss": 0.3286,
-      "step": 18305
-    },
-    {
-      "epoch": 1.196392392654075,
-      "grad_norm": 0.46907752752304077,
-      "learning_rate": 6.745089971710618e-06,
-      "loss": 0.383,
-      "step": 18306
-    },
-    {
-      "epoch": 1.196457747859617,
-      "grad_norm": 0.4877195656299591,
-      "learning_rate": 6.744762735904867e-06,
-      "loss": 0.3518,
-      "step": 18307
-    },
-    {
-      "epoch": 1.196523103065159,
-      "grad_norm": 0.4745230972766876,
-      "learning_rate": 6.744435491589305e-06,
-      "loss": 0.3855,
-      "step": 18308
-    },
-    {
-      "epoch": 1.1965884582707011,
-      "grad_norm": 0.42876559495925903,
-      "learning_rate": 6.7441082387655255e-06,
-      "loss": 0.3208,
-      "step": 18309
-    },
-    {
-      "epoch": 1.1966538134762434,
-      "grad_norm": 0.43249624967575073,
-      "learning_rate": 6.743780977435128e-06,
-      "loss": 0.3201,
-      "step": 18310
-    },
-    {
-      "epoch": 1.1967191686817855,
-      "grad_norm": 0.48484688997268677,
-      "learning_rate": 6.743453707599704e-06,
-      "loss": 0.3531,
-      "step": 18311
-    },
-    {
-      "epoch": 1.1967845238873276,
-      "grad_norm": 0.5050974488258362,
-      "learning_rate": 6.743126429260855e-06,
-      "loss": 0.4019,
-      "step": 18312
-    },
-    {
-      "epoch": 1.1968498790928697,
-      "grad_norm": 0.45557287335395813,
-      "learning_rate": 6.742799142420172e-06,
-      "loss": 0.3717,
-      "step": 18313
-    },
-    {
-      "epoch": 1.196915234298412,
-      "grad_norm": 0.4626365900039673,
-      "learning_rate": 6.742471847079255e-06,
-      "loss": 0.3663,
-      "step": 18314
-    },
-    {
-      "epoch": 1.196980589503954,
-      "grad_norm": 0.4385165274143219,
-      "learning_rate": 6.742144543239701e-06,
-      "loss": 0.3638,
-      "step": 18315
-    },
-    {
-      "epoch": 1.197045944709496,
-      "grad_norm": 0.456708699464798,
-      "learning_rate": 6.741817230903102e-06,
-      "loss": 0.3552,
-      "step": 18316
-    },
-    {
-      "epoch": 1.1971112999150382,
-      "grad_norm": 0.4330587685108185,
-      "learning_rate": 6.741489910071057e-06,
-      "loss": 0.339,
-      "step": 18317
-    },
-    {
-      "epoch": 1.1971766551205802,
-      "grad_norm": 0.44074746966362,
-      "learning_rate": 6.741162580745163e-06,
-      "loss": 0.3683,
-      "step": 18318
-    },
-    {
-      "epoch": 1.1972420103261225,
-      "grad_norm": 0.4403582811355591,
-      "learning_rate": 6.740835242927016e-06,
-      "loss": 0.3602,
-      "step": 18319
-    },
-    {
-      "epoch": 1.1973073655316646,
-      "grad_norm": 0.42535579204559326,
-      "learning_rate": 6.740507896618211e-06,
-      "loss": 0.3044,
-      "step": 18320
-    },
-    {
-      "epoch": 1.1973727207372067,
-      "grad_norm": 0.47823721170425415,
-      "learning_rate": 6.74018054182035e-06,
-      "loss": 0.379,
-      "step": 18321
-    },
-    {
-      "epoch": 1.1974380759427488,
-      "grad_norm": 0.46522536873817444,
-      "learning_rate": 6.739853178535022e-06,
-      "loss": 0.3186,
-      "step": 18322
-    },
-    {
-      "epoch": 1.197503431148291,
-      "grad_norm": 0.4860043525695801,
-      "learning_rate": 6.739525806763828e-06,
-      "loss": 0.3255,
-      "step": 18323
-    },
-    {
-      "epoch": 1.1975687863538331,
-      "grad_norm": 0.4254266321659088,
-      "learning_rate": 6.739198426508364e-06,
-      "loss": 0.3066,
-      "step": 18324
-    },
-    {
-      "epoch": 1.1976341415593752,
-      "grad_norm": 0.4584939777851105,
-      "learning_rate": 6.738871037770228e-06,
-      "loss": 0.3675,
-      "step": 18325
-    },
-    {
-      "epoch": 1.1976994967649173,
-      "grad_norm": 0.4622330665588379,
-      "learning_rate": 6.738543640551015e-06,
-      "loss": 0.3752,
-      "step": 18326
-    },
-    {
-      "epoch": 1.1977648519704593,
-      "grad_norm": 0.47860780358314514,
-      "learning_rate": 6.738216234852321e-06,
-      "loss": 0.386,
-      "step": 18327
-    },
-    {
-      "epoch": 1.1978302071760016,
-      "grad_norm": 0.456599622964859,
-      "learning_rate": 6.737888820675747e-06,
-      "loss": 0.3106,
-      "step": 18328
-    },
-    {
-      "epoch": 1.1978955623815437,
-      "grad_norm": 0.45784202218055725,
-      "learning_rate": 6.737561398022884e-06,
-      "loss": 0.3593,
-      "step": 18329
-    },
-    {
-      "epoch": 1.1979609175870858,
-      "grad_norm": 0.5206227898597717,
-      "learning_rate": 6.7372339668953335e-06,
-      "loss": 0.3107,
-      "step": 18330
-    },
-    {
-      "epoch": 1.1980262727926279,
-      "grad_norm": 0.41274091601371765,
-      "learning_rate": 6.736906527294691e-06,
-      "loss": 0.2937,
-      "step": 18331
-    },
-    {
-      "epoch": 1.1980916279981701,
-      "grad_norm": 0.485586941242218,
-      "learning_rate": 6.736579079222554e-06,
-      "loss": 0.3808,
-      "step": 18332
-    },
-    {
-      "epoch": 1.1981569832037122,
-      "grad_norm": 0.40877220034599304,
-      "learning_rate": 6.73625162268052e-06,
-      "loss": 0.2818,
-      "step": 18333
-    },
-    {
-      "epoch": 1.1982223384092543,
-      "grad_norm": 0.4090054929256439,
-      "learning_rate": 6.735924157670184e-06,
-      "loss": 0.2949,
-      "step": 18334
-    },
-    {
-      "epoch": 1.1982876936147964,
-      "grad_norm": 0.45550888776779175,
-      "learning_rate": 6.7355966841931445e-06,
-      "loss": 0.3454,
-      "step": 18335
-    },
-    {
-      "epoch": 1.1983530488203384,
-      "grad_norm": 0.408974826335907,
-      "learning_rate": 6.735269202250998e-06,
-      "loss": 0.3092,
-      "step": 18336
-    },
-    {
-      "epoch": 1.1984184040258807,
-      "grad_norm": 0.46963611245155334,
-      "learning_rate": 6.734941711845344e-06,
-      "loss": 0.3839,
-      "step": 18337
-    },
-    {
-      "epoch": 1.1984837592314228,
-      "grad_norm": 0.4983595907688141,
-      "learning_rate": 6.734614212977777e-06,
-      "loss": 0.3335,
-      "step": 18338
-    },
-    {
-      "epoch": 1.1985491144369649,
-      "grad_norm": 0.5197293162345886,
-      "learning_rate": 6.7342867056498975e-06,
-      "loss": 0.4474,
-      "step": 18339
-    },
-    {
-      "epoch": 1.198614469642507,
-      "grad_norm": 0.41194507479667664,
-      "learning_rate": 6.7339591898633e-06,
-      "loss": 0.3003,
-      "step": 18340
-    },
-    {
-      "epoch": 1.1986798248480492,
-      "grad_norm": 0.5035253763198853,
-      "learning_rate": 6.733631665619582e-06,
-      "loss": 0.4243,
-      "step": 18341
-    },
-    {
-      "epoch": 1.1987451800535913,
-      "grad_norm": 0.4093713164329529,
-      "learning_rate": 6.733304132920342e-06,
-      "loss": 0.3176,
-      "step": 18342
-    },
-    {
-      "epoch": 1.1988105352591334,
-      "grad_norm": 0.4392082393169403,
-      "learning_rate": 6.732976591767177e-06,
-      "loss": 0.3054,
-      "step": 18343
-    },
-    {
-      "epoch": 1.1988758904646755,
-      "grad_norm": 0.45759931206703186,
-      "learning_rate": 6.732649042161686e-06,
-      "loss": 0.357,
-      "step": 18344
-    },
-    {
-      "epoch": 1.1989412456702175,
-      "grad_norm": 0.4255608916282654,
-      "learning_rate": 6.732321484105465e-06,
-      "loss": 0.3113,
-      "step": 18345
-    },
-    {
-      "epoch": 1.1990066008757598,
-      "grad_norm": 0.47021111845970154,
-      "learning_rate": 6.731993917600113e-06,
-      "loss": 0.3626,
-      "step": 18346
-    },
-    {
-      "epoch": 1.199071956081302,
-      "grad_norm": 0.48122110962867737,
-      "learning_rate": 6.731666342647225e-06,
-      "loss": 0.3777,
-      "step": 18347
-    },
-    {
-      "epoch": 1.199137311286844,
-      "grad_norm": 0.45146089792251587,
-      "learning_rate": 6.7313387592484e-06,
-      "loss": 0.3636,
-      "step": 18348
-    },
-    {
-      "epoch": 1.199202666492386,
-      "grad_norm": 0.4471517503261566,
-      "learning_rate": 6.731011167405237e-06,
-      "loss": 0.3177,
-      "step": 18349
-    },
-    {
-      "epoch": 1.1992680216979283,
-      "grad_norm": 0.41821399331092834,
-      "learning_rate": 6.7306835671193325e-06,
-      "loss": 0.2989,
-      "step": 18350
-    },
-    {
-      "epoch": 1.1993333769034704,
-      "grad_norm": 0.451188325881958,
-      "learning_rate": 6.730355958392285e-06,
-      "loss": 0.3262,
-      "step": 18351
-    },
-    {
-      "epoch": 1.1993987321090125,
-      "grad_norm": 0.4433478116989136,
-      "learning_rate": 6.730028341225692e-06,
-      "loss": 0.3204,
-      "step": 18352
-    },
-    {
-      "epoch": 1.1994640873145546,
-      "grad_norm": 0.48006168007850647,
-      "learning_rate": 6.7297007156211516e-06,
-      "loss": 0.3424,
-      "step": 18353
-    },
-    {
-      "epoch": 1.1995294425200966,
-      "grad_norm": 0.4221143424510956,
-      "learning_rate": 6.72937308158026e-06,
-      "loss": 0.297,
-      "step": 18354
-    },
-    {
-      "epoch": 1.199594797725639,
-      "grad_norm": 0.42380595207214355,
-      "learning_rate": 6.729045439104619e-06,
-      "loss": 0.3239,
-      "step": 18355
-    },
-    {
-      "epoch": 1.199660152931181,
-      "grad_norm": 0.47803959250450134,
-      "learning_rate": 6.728717788195823e-06,
-      "loss": 0.3746,
-      "step": 18356
-    },
-    {
-      "epoch": 1.199725508136723,
-      "grad_norm": 0.40294092893600464,
-      "learning_rate": 6.728390128855472e-06,
-      "loss": 0.313,
-      "step": 18357
-    },
-    {
-      "epoch": 1.1997908633422651,
-      "grad_norm": 0.4656482934951782,
-      "learning_rate": 6.728062461085163e-06,
-      "loss": 0.3661,
-      "step": 18358
-    },
-    {
-      "epoch": 1.1998562185478074,
-      "grad_norm": 0.49004656076431274,
-      "learning_rate": 6.727734784886496e-06,
-      "loss": 0.3781,
-      "step": 18359
-    },
-    {
-      "epoch": 1.1999215737533495,
-      "grad_norm": 0.46724215149879456,
-      "learning_rate": 6.7274071002610675e-06,
-      "loss": 0.4008,
-      "step": 18360
-    },
-    {
-      "epoch": 1.1999869289588916,
-      "grad_norm": 0.4653272032737732,
-      "learning_rate": 6.727079407210475e-06,
-      "loss": 0.3873,
-      "step": 18361
-    },
-    {
-      "epoch": 1.2000522841644337,
-      "grad_norm": 0.42921003699302673,
-      "learning_rate": 6.72675170573632e-06,
-      "loss": 0.2975,
-      "step": 18362
-    },
-    {
-      "epoch": 1.2001176393699757,
-      "grad_norm": 0.4449147880077362,
-      "learning_rate": 6.726423995840197e-06,
-      "loss": 0.3383,
-      "step": 18363
-    },
-    {
-      "epoch": 1.200182994575518,
-      "grad_norm": 0.4797692894935608,
-      "learning_rate": 6.726096277523706e-06,
-      "loss": 0.3716,
-      "step": 18364
-    },
-    {
-      "epoch": 1.20024834978106,
-      "grad_norm": 0.4608686566352844,
-      "learning_rate": 6.725768550788446e-06,
-      "loss": 0.3385,
-      "step": 18365
-    },
-    {
-      "epoch": 1.2003137049866022,
-      "grad_norm": 0.44967734813690186,
-      "learning_rate": 6.725440815636015e-06,
-      "loss": 0.3531,
-      "step": 18366
-    },
-    {
-      "epoch": 1.2003790601921442,
-      "grad_norm": 0.46174466609954834,
-      "learning_rate": 6.725113072068011e-06,
-      "loss": 0.3359,
-      "step": 18367
-    },
-    {
-      "epoch": 1.2004444153976865,
-      "grad_norm": 0.4304310083389282,
-      "learning_rate": 6.724785320086034e-06,
-      "loss": 0.3457,
-      "step": 18368
-    },
-    {
-      "epoch": 1.2005097706032286,
-      "grad_norm": 0.4110058546066284,
-      "learning_rate": 6.724457559691679e-06,
-      "loss": 0.3237,
-      "step": 18369
-    },
-    {
-      "epoch": 1.2005751258087707,
-      "grad_norm": 0.4807645380496979,
-      "learning_rate": 6.72412979088655e-06,
-      "loss": 0.364,
-      "step": 18370
-    },
-    {
-      "epoch": 1.2006404810143128,
-      "grad_norm": 0.45398521423339844,
-      "learning_rate": 6.723802013672243e-06,
-      "loss": 0.3271,
-      "step": 18371
-    },
-    {
-      "epoch": 1.2007058362198548,
-      "grad_norm": 0.4776749312877655,
-      "learning_rate": 6.723474228050353e-06,
-      "loss": 0.3146,
-      "step": 18372
-    },
-    {
-      "epoch": 1.2007711914253971,
-      "grad_norm": 0.507209837436676,
-      "learning_rate": 6.723146434022485e-06,
-      "loss": 0.4261,
-      "step": 18373
-    },
-    {
-      "epoch": 1.2008365466309392,
-      "grad_norm": 0.4721026122570038,
-      "learning_rate": 6.7228186315902335e-06,
-      "loss": 0.3839,
-      "step": 18374
-    },
-    {
-      "epoch": 1.2009019018364813,
-      "grad_norm": 0.4585725665092468,
-      "learning_rate": 6.722490820755199e-06,
-      "loss": 0.3017,
-      "step": 18375
-    },
-    {
-      "epoch": 1.2009672570420233,
-      "grad_norm": 0.4542023539543152,
-      "learning_rate": 6.72216300151898e-06,
-      "loss": 0.3204,
-      "step": 18376
-    },
-    {
-      "epoch": 1.2010326122475656,
-      "grad_norm": 0.45127978920936584,
-      "learning_rate": 6.721835173883175e-06,
-      "loss": 0.3538,
-      "step": 18377
-    },
-    {
-      "epoch": 1.2010979674531077,
-      "grad_norm": 0.411732941865921,
-      "learning_rate": 6.721507337849383e-06,
-      "loss": 0.2875,
-      "step": 18378
-    },
-    {
-      "epoch": 1.2011633226586498,
-      "grad_norm": 0.49450406432151794,
-      "learning_rate": 6.721179493419205e-06,
-      "loss": 0.3799,
-      "step": 18379
-    },
-    {
-      "epoch": 1.2012286778641919,
-      "grad_norm": 0.45164230465888977,
-      "learning_rate": 6.720851640594238e-06,
-      "loss": 0.3679,
-      "step": 18380
-    },
-    {
-      "epoch": 1.201294033069734,
-      "grad_norm": 0.4650866687297821,
-      "learning_rate": 6.72052377937608e-06,
-      "loss": 0.3671,
-      "step": 18381
-    },
-    {
-      "epoch": 1.2013593882752762,
-      "grad_norm": 0.44707417488098145,
-      "learning_rate": 6.720195909766333e-06,
-      "loss": 0.3516,
-      "step": 18382
-    },
-    {
-      "epoch": 1.2014247434808183,
-      "grad_norm": 0.4069920480251312,
-      "learning_rate": 6.719868031766593e-06,
-      "loss": 0.2836,
-      "step": 18383
-    },
-    {
-      "epoch": 1.2014900986863604,
-      "grad_norm": 0.5384685397148132,
-      "learning_rate": 6.719540145378463e-06,
-      "loss": 0.3649,
-      "step": 18384
-    },
-    {
-      "epoch": 1.2015554538919024,
-      "grad_norm": 0.44358280301094055,
-      "learning_rate": 6.719212250603537e-06,
-      "loss": 0.3455,
-      "step": 18385
-    },
-    {
-      "epoch": 1.2016208090974447,
-      "grad_norm": 0.4112396836280823,
-      "learning_rate": 6.718884347443422e-06,
-      "loss": 0.289,
-      "step": 18386
-    },
-    {
-      "epoch": 1.2016861643029868,
-      "grad_norm": 0.46106982231140137,
-      "learning_rate": 6.718556435899708e-06,
-      "loss": 0.3395,
-      "step": 18387
-    },
-    {
-      "epoch": 1.2017515195085289,
-      "grad_norm": 0.45882683992385864,
-      "learning_rate": 6.718228515974001e-06,
-      "loss": 0.3475,
-      "step": 18388
-    },
-    {
-      "epoch": 1.201816874714071,
-      "grad_norm": 0.4271091818809509,
-      "learning_rate": 6.717900587667898e-06,
-      "loss": 0.2913,
-      "step": 18389
-    },
-    {
-      "epoch": 1.201882229919613,
-      "grad_norm": 0.43459373712539673,
-      "learning_rate": 6.717572650982998e-06,
-      "loss": 0.3317,
-      "step": 18390
-    },
-    {
-      "epoch": 1.2019475851251553,
-      "grad_norm": 0.4594324231147766,
-      "learning_rate": 6.717244705920902e-06,
-      "loss": 0.3937,
-      "step": 18391
-    },
-    {
-      "epoch": 1.2020129403306974,
-      "grad_norm": 0.4645597040653229,
-      "learning_rate": 6.716916752483208e-06,
-      "loss": 0.3541,
-      "step": 18392
-    },
-    {
-      "epoch": 1.2020782955362395,
-      "grad_norm": 0.47792962193489075,
-      "learning_rate": 6.716588790671516e-06,
-      "loss": 0.3837,
-      "step": 18393
-    },
-    {
-      "epoch": 1.2021436507417815,
-      "grad_norm": 0.4746323227882385,
-      "learning_rate": 6.716260820487427e-06,
-      "loss": 0.3692,
-      "step": 18394
-    },
-    {
-      "epoch": 1.2022090059473238,
-      "grad_norm": 0.44234099984169006,
-      "learning_rate": 6.715932841932539e-06,
-      "loss": 0.3298,
-      "step": 18395
-    },
-    {
-      "epoch": 1.202274361152866,
-      "grad_norm": 0.4199508726596832,
-      "learning_rate": 6.715604855008451e-06,
-      "loss": 0.2807,
-      "step": 18396
-    },
-    {
-      "epoch": 1.202339716358408,
-      "grad_norm": 0.45626598596572876,
-      "learning_rate": 6.715276859716765e-06,
-      "loss": 0.3825,
-      "step": 18397
-    },
-    {
-      "epoch": 1.20240507156395,
-      "grad_norm": 0.44778361916542053,
-      "learning_rate": 6.714948856059079e-06,
-      "loss": 0.36,
-      "step": 18398
-    },
-    {
-      "epoch": 1.2024704267694921,
-      "grad_norm": 0.4117039740085602,
-      "learning_rate": 6.714620844036993e-06,
-      "loss": 0.3113,
-      "step": 18399
-    },
-    {
-      "epoch": 1.2025357819750342,
-      "grad_norm": 0.46973392367362976,
-      "learning_rate": 6.714292823652109e-06,
-      "loss": 0.3927,
-      "step": 18400
-    },
-    {
-      "epoch": 1.2026011371805765,
-      "grad_norm": 0.4537544846534729,
-      "learning_rate": 6.7139647949060224e-06,
-      "loss": 0.3453,
-      "step": 18401
-    },
-    {
-      "epoch": 1.2026664923861186,
-      "grad_norm": 0.4220884144306183,
-      "learning_rate": 6.713636757800337e-06,
-      "loss": 0.3525,
-      "step": 18402
-    },
-    {
-      "epoch": 1.2027318475916606,
-      "grad_norm": 0.4099014699459076,
-      "learning_rate": 6.71330871233665e-06,
-      "loss": 0.2811,
-      "step": 18403
-    },
-    {
-      "epoch": 1.2027972027972027,
-      "grad_norm": 0.44728389382362366,
-      "learning_rate": 6.7129806585165654e-06,
-      "loss": 0.3036,
-      "step": 18404
-    },
-    {
-      "epoch": 1.202862558002745,
-      "grad_norm": 0.4400940537452698,
-      "learning_rate": 6.712652596341679e-06,
-      "loss": 0.3257,
-      "step": 18405
-    },
-    {
-      "epoch": 1.202927913208287,
-      "grad_norm": 0.4527154564857483,
-      "learning_rate": 6.712324525813594e-06,
-      "loss": 0.313,
-      "step": 18406
-    },
-    {
-      "epoch": 1.2029932684138291,
-      "grad_norm": 0.47343170642852783,
-      "learning_rate": 6.711996446933908e-06,
-      "loss": 0.3562,
-      "step": 18407
-    },
-    {
-      "epoch": 1.2030586236193712,
-      "grad_norm": 0.47466331720352173,
-      "learning_rate": 6.711668359704223e-06,
-      "loss": 0.3581,
-      "step": 18408
-    },
-    {
-      "epoch": 1.2031239788249133,
-      "grad_norm": 0.44942548871040344,
-      "learning_rate": 6.7113402641261376e-06,
-      "loss": 0.3391,
-      "step": 18409
-    },
-    {
-      "epoch": 1.2031893340304556,
-      "grad_norm": 0.48919522762298584,
-      "learning_rate": 6.711012160201253e-06,
-      "loss": 0.3793,
-      "step": 18410
-    },
-    {
-      "epoch": 1.2032546892359977,
-      "grad_norm": 0.504368007183075,
-      "learning_rate": 6.71068404793117e-06,
-      "loss": 0.3952,
-      "step": 18411
-    },
-    {
-      "epoch": 1.2033200444415397,
-      "grad_norm": 0.4561106562614441,
-      "learning_rate": 6.710355927317487e-06,
-      "loss": 0.3273,
-      "step": 18412
-    },
-    {
-      "epoch": 1.2033853996470818,
-      "grad_norm": 0.4240868389606476,
-      "learning_rate": 6.710027798361807e-06,
-      "loss": 0.312,
-      "step": 18413
-    },
-    {
-      "epoch": 1.203450754852624,
-      "grad_norm": 0.4330175817012787,
-      "learning_rate": 6.709699661065727e-06,
-      "loss": 0.3233,
-      "step": 18414
-    },
-    {
-      "epoch": 1.2035161100581662,
-      "grad_norm": 0.4198254942893982,
-      "learning_rate": 6.7093715154308516e-06,
-      "loss": 0.314,
-      "step": 18415
-    },
-    {
-      "epoch": 1.2035814652637082,
-      "grad_norm": 0.6476282477378845,
-      "learning_rate": 6.709043361458778e-06,
-      "loss": 0.3672,
-      "step": 18416
-    },
-    {
-      "epoch": 1.2036468204692503,
-      "grad_norm": 0.46822741627693176,
-      "learning_rate": 6.708715199151108e-06,
-      "loss": 0.3261,
-      "step": 18417
-    },
-    {
-      "epoch": 1.2037121756747924,
-      "grad_norm": 0.4359154999256134,
-      "learning_rate": 6.708387028509442e-06,
-      "loss": 0.367,
-      "step": 18418
-    },
-    {
-      "epoch": 1.2037775308803347,
-      "grad_norm": 0.46764418482780457,
-      "learning_rate": 6.708058849535382e-06,
-      "loss": 0.3664,
-      "step": 18419
-    },
-    {
-      "epoch": 1.2038428860858768,
-      "grad_norm": 0.4774992763996124,
-      "learning_rate": 6.707730662230525e-06,
-      "loss": 0.3881,
-      "step": 18420
-    },
-    {
-      "epoch": 1.2039082412914188,
-      "grad_norm": 0.4189451336860657,
-      "learning_rate": 6.707402466596475e-06,
-      "loss": 0.3069,
-      "step": 18421
-    },
-    {
-      "epoch": 1.203973596496961,
-      "grad_norm": 0.44668343663215637,
-      "learning_rate": 6.7070742626348314e-06,
-      "loss": 0.3215,
-      "step": 18422
-    },
-    {
-      "epoch": 1.2040389517025032,
-      "grad_norm": 0.4437035322189331,
-      "learning_rate": 6.706746050347195e-06,
-      "loss": 0.343,
-      "step": 18423
-    },
-    {
-      "epoch": 1.2041043069080453,
-      "grad_norm": 0.40654462575912476,
-      "learning_rate": 6.706417829735168e-06,
-      "loss": 0.3015,
-      "step": 18424
-    },
-    {
-      "epoch": 1.2041696621135873,
-      "grad_norm": 0.42640653252601624,
-      "learning_rate": 6.706089600800349e-06,
-      "loss": 0.3275,
-      "step": 18425
-    },
-    {
-      "epoch": 1.2042350173191294,
-      "grad_norm": 0.4433779716491699,
-      "learning_rate": 6.705761363544341e-06,
-      "loss": 0.325,
-      "step": 18426
-    },
-    {
-      "epoch": 1.2043003725246715,
-      "grad_norm": 0.4791772663593292,
-      "learning_rate": 6.705433117968744e-06,
-      "loss": 0.3717,
-      "step": 18427
-    },
-    {
-      "epoch": 1.2043657277302138,
-      "grad_norm": 0.4121028482913971,
-      "learning_rate": 6.705104864075158e-06,
-      "loss": 0.3148,
-      "step": 18428
-    },
-    {
-      "epoch": 1.2044310829357558,
-      "grad_norm": 0.45768189430236816,
-      "learning_rate": 6.7047766018651864e-06,
-      "loss": 0.3427,
-      "step": 18429
-    },
-    {
-      "epoch": 1.204496438141298,
-      "grad_norm": 0.4377157390117645,
-      "learning_rate": 6.704448331340427e-06,
-      "loss": 0.3247,
-      "step": 18430
-    },
-    {
-      "epoch": 1.20456179334684,
-      "grad_norm": 0.41707608103752136,
-      "learning_rate": 6.704120052502483e-06,
-      "loss": 0.33,
-      "step": 18431
-    },
-    {
-      "epoch": 1.2046271485523823,
-      "grad_norm": 0.442849725484848,
-      "learning_rate": 6.703791765352954e-06,
-      "loss": 0.3614,
-      "step": 18432
-    },
-    {
-      "epoch": 1.2046925037579244,
-      "grad_norm": 0.467950701713562,
-      "learning_rate": 6.703463469893443e-06,
-      "loss": 0.3797,
-      "step": 18433
-    },
-    {
-      "epoch": 1.2047578589634664,
-      "grad_norm": 0.4545117914676666,
-      "learning_rate": 6.703135166125552e-06,
-      "loss": 0.3355,
-      "step": 18434
-    },
-    {
-      "epoch": 1.2048232141690085,
-      "grad_norm": 0.4704337418079376,
-      "learning_rate": 6.70280685405088e-06,
-      "loss": 0.3615,
-      "step": 18435
-    },
-    {
-      "epoch": 1.2048885693745506,
-      "grad_norm": 0.4516729414463043,
-      "learning_rate": 6.702478533671028e-06,
-      "loss": 0.3353,
-      "step": 18436
-    },
-    {
-      "epoch": 1.2049539245800929,
-      "grad_norm": 0.468779593706131,
-      "learning_rate": 6.702150204987598e-06,
-      "loss": 0.3744,
-      "step": 18437
-    },
-    {
-      "epoch": 1.205019279785635,
-      "grad_norm": 0.4292548596858978,
-      "learning_rate": 6.701821868002194e-06,
-      "loss": 0.329,
-      "step": 18438
-    },
-    {
-      "epoch": 1.205084634991177,
-      "grad_norm": 0.46684563159942627,
-      "learning_rate": 6.701493522716414e-06,
-      "loss": 0.368,
-      "step": 18439
-    },
-    {
-      "epoch": 1.205149990196719,
-      "grad_norm": 0.44545918703079224,
-      "learning_rate": 6.7011651691318615e-06,
-      "loss": 0.3256,
-      "step": 18440
-    },
-    {
-      "epoch": 1.2052153454022614,
-      "grad_norm": 0.48918187618255615,
-      "learning_rate": 6.700836807250135e-06,
-      "loss": 0.3876,
-      "step": 18441
-    },
-    {
-      "epoch": 1.2052807006078035,
-      "grad_norm": 0.48126718401908875,
-      "learning_rate": 6.70050843707284e-06,
-      "loss": 0.3348,
-      "step": 18442
-    },
-    {
-      "epoch": 1.2053460558133455,
-      "grad_norm": 0.5070162415504456,
-      "learning_rate": 6.700180058601576e-06,
-      "loss": 0.393,
-      "step": 18443
-    },
-    {
-      "epoch": 1.2054114110188876,
-      "grad_norm": 0.44988730549812317,
-      "learning_rate": 6.699851671837945e-06,
-      "loss": 0.3106,
-      "step": 18444
-    },
-    {
-      "epoch": 1.2054767662244297,
-      "grad_norm": 0.44040974974632263,
-      "learning_rate": 6.699523276783548e-06,
-      "loss": 0.3293,
-      "step": 18445
-    },
-    {
-      "epoch": 1.205542121429972,
-      "grad_norm": 0.48614758253097534,
-      "learning_rate": 6.699194873439987e-06,
-      "loss": 0.4042,
-      "step": 18446
-    },
-    {
-      "epoch": 1.205607476635514,
-      "grad_norm": 0.43726658821105957,
-      "learning_rate": 6.698866461808865e-06,
-      "loss": 0.3358,
-      "step": 18447
-    },
-    {
-      "epoch": 1.2056728318410561,
-      "grad_norm": 0.495114803314209,
-      "learning_rate": 6.698538041891781e-06,
-      "loss": 0.3923,
-      "step": 18448
-    },
-    {
-      "epoch": 1.2057381870465982,
-      "grad_norm": 0.42943546175956726,
-      "learning_rate": 6.698209613690341e-06,
-      "loss": 0.3211,
-      "step": 18449
-    },
-    {
-      "epoch": 1.2058035422521405,
-      "grad_norm": 0.4599703550338745,
-      "learning_rate": 6.697881177206143e-06,
-      "loss": 0.3527,
-      "step": 18450
-    },
-    {
-      "epoch": 1.2058688974576826,
-      "grad_norm": 0.41305315494537354,
-      "learning_rate": 6.697552732440791e-06,
-      "loss": 0.3041,
-      "step": 18451
-    },
-    {
-      "epoch": 1.2059342526632246,
-      "grad_norm": 0.4313637316226959,
-      "learning_rate": 6.697224279395884e-06,
-      "loss": 0.3675,
-      "step": 18452
-    },
-    {
-      "epoch": 1.2059996078687667,
-      "grad_norm": 0.48010942339897156,
-      "learning_rate": 6.696895818073028e-06,
-      "loss": 0.3661,
-      "step": 18453
-    },
-    {
-      "epoch": 1.2060649630743088,
-      "grad_norm": 0.42056751251220703,
-      "learning_rate": 6.696567348473823e-06,
-      "loss": 0.3159,
-      "step": 18454
-    },
-    {
-      "epoch": 1.206130318279851,
-      "grad_norm": 0.4515971541404724,
-      "learning_rate": 6.696238870599871e-06,
-      "loss": 0.3728,
-      "step": 18455
-    },
-    {
-      "epoch": 1.2061956734853931,
-      "grad_norm": 0.42346087098121643,
-      "learning_rate": 6.695910384452775e-06,
-      "loss": 0.3413,
-      "step": 18456
-    },
-    {
-      "epoch": 1.2062610286909352,
-      "grad_norm": 0.45464929938316345,
-      "learning_rate": 6.695581890034136e-06,
-      "loss": 0.3215,
-      "step": 18457
-    },
-    {
-      "epoch": 1.2063263838964773,
-      "grad_norm": 0.4392387866973877,
-      "learning_rate": 6.695253387345557e-06,
-      "loss": 0.3225,
-      "step": 18458
-    },
-    {
-      "epoch": 1.2063917391020196,
-      "grad_norm": 0.4536644220352173,
-      "learning_rate": 6.69492487638864e-06,
-      "loss": 0.3767,
-      "step": 18459
-    },
-    {
-      "epoch": 1.2064570943075617,
-      "grad_norm": 0.43927276134490967,
-      "learning_rate": 6.694596357164986e-06,
-      "loss": 0.3384,
-      "step": 18460
-    },
-    {
-      "epoch": 1.2065224495131037,
-      "grad_norm": 0.5052178502082825,
-      "learning_rate": 6.6942678296762e-06,
-      "loss": 0.4275,
-      "step": 18461
-    },
-    {
-      "epoch": 1.2065878047186458,
-      "grad_norm": 0.4483092725276947,
-      "learning_rate": 6.693939293923883e-06,
-      "loss": 0.3586,
-      "step": 18462
-    },
-    {
-      "epoch": 1.2066531599241879,
-      "grad_norm": 0.4657086431980133,
-      "learning_rate": 6.693610749909636e-06,
-      "loss": 0.3051,
-      "step": 18463
-    },
-    {
-      "epoch": 1.2067185151297302,
-      "grad_norm": 0.4663565754890442,
-      "learning_rate": 6.693282197635063e-06,
-      "loss": 0.3384,
-      "step": 18464
-    },
-    {
-      "epoch": 1.2067838703352722,
-      "grad_norm": 0.4573187232017517,
-      "learning_rate": 6.692953637101766e-06,
-      "loss": 0.3564,
-      "step": 18465
-    },
-    {
-      "epoch": 1.2068492255408143,
-      "grad_norm": 0.42286911606788635,
-      "learning_rate": 6.692625068311349e-06,
-      "loss": 0.3141,
-      "step": 18466
-    },
-    {
-      "epoch": 1.2069145807463564,
-      "grad_norm": 0.4303493797779083,
-      "learning_rate": 6.692296491265412e-06,
-      "loss": 0.3104,
-      "step": 18467
-    },
-    {
-      "epoch": 1.2069799359518987,
-      "grad_norm": 0.450505793094635,
-      "learning_rate": 6.691967905965559e-06,
-      "loss": 0.3506,
-      "step": 18468
-    },
-    {
-      "epoch": 1.2070452911574407,
-      "grad_norm": 0.45550888776779175,
-      "learning_rate": 6.691639312413392e-06,
-      "loss": 0.3427,
-      "step": 18469
-    },
-    {
-      "epoch": 1.2071106463629828,
-      "grad_norm": 0.4892937242984772,
-      "learning_rate": 6.691310710610515e-06,
-      "loss": 0.377,
-      "step": 18470
-    },
-    {
-      "epoch": 1.207176001568525,
-      "grad_norm": 0.45304617285728455,
-      "learning_rate": 6.69098210055853e-06,
-      "loss": 0.3365,
-      "step": 18471
-    },
-    {
-      "epoch": 1.207241356774067,
-      "grad_norm": 0.4439897835254669,
-      "learning_rate": 6.690653482259038e-06,
-      "loss": 0.3241,
-      "step": 18472
-    },
-    {
-      "epoch": 1.2073067119796093,
-      "grad_norm": 0.44223734736442566,
-      "learning_rate": 6.690324855713643e-06,
-      "loss": 0.3118,
-      "step": 18473
-    },
-    {
-      "epoch": 1.2073720671851513,
-      "grad_norm": 0.4482691287994385,
-      "learning_rate": 6.689996220923949e-06,
-      "loss": 0.3248,
-      "step": 18474
-    },
-    {
-      "epoch": 1.2074374223906934,
-      "grad_norm": 0.4404681921005249,
-      "learning_rate": 6.689667577891557e-06,
-      "loss": 0.3376,
-      "step": 18475
-    },
-    {
-      "epoch": 1.2075027775962355,
-      "grad_norm": 0.4479188919067383,
-      "learning_rate": 6.689338926618073e-06,
-      "loss": 0.3424,
-      "step": 18476
-    },
-    {
-      "epoch": 1.2075681328017778,
-      "grad_norm": 0.424572229385376,
-      "learning_rate": 6.689010267105096e-06,
-      "loss": 0.3115,
-      "step": 18477
-    },
-    {
-      "epoch": 1.2076334880073198,
-      "grad_norm": 0.4301793575286865,
-      "learning_rate": 6.688681599354232e-06,
-      "loss": 0.3187,
-      "step": 18478
-    },
-    {
-      "epoch": 1.207698843212862,
-      "grad_norm": 0.5272138118743896,
-      "learning_rate": 6.688352923367081e-06,
-      "loss": 0.4015,
-      "step": 18479
-    },
-    {
-      "epoch": 1.207764198418404,
-      "grad_norm": 0.4448002576828003,
-      "learning_rate": 6.68802423914525e-06,
-      "loss": 0.3472,
-      "step": 18480
-    },
-    {
-      "epoch": 1.207829553623946,
-      "grad_norm": 0.43331530690193176,
-      "learning_rate": 6.687695546690338e-06,
-      "loss": 0.3182,
-      "step": 18481
-    },
-    {
-      "epoch": 1.2078949088294884,
-      "grad_norm": 0.49392253160476685,
-      "learning_rate": 6.687366846003952e-06,
-      "loss": 0.4234,
-      "step": 18482
-    },
-    {
-      "epoch": 1.2079602640350304,
-      "grad_norm": 0.4425753951072693,
-      "learning_rate": 6.687038137087693e-06,
-      "loss": 0.3392,
-      "step": 18483
-    },
-    {
-      "epoch": 1.2080256192405725,
-      "grad_norm": 0.44402262568473816,
-      "learning_rate": 6.686709419943163e-06,
-      "loss": 0.3223,
-      "step": 18484
-    },
-    {
-      "epoch": 1.2080909744461146,
-      "grad_norm": 0.4544754922389984,
-      "learning_rate": 6.686380694571968e-06,
-      "loss": 0.3324,
-      "step": 18485
-    },
-    {
-      "epoch": 1.2081563296516569,
-      "grad_norm": 0.4638741910457611,
-      "learning_rate": 6.68605196097571e-06,
-      "loss": 0.3662,
-      "step": 18486
-    },
-    {
-      "epoch": 1.208221684857199,
-      "grad_norm": 0.45993903279304504,
-      "learning_rate": 6.685723219155993e-06,
-      "loss": 0.3411,
-      "step": 18487
-    },
-    {
-      "epoch": 1.208287040062741,
-      "grad_norm": 0.4373694360256195,
-      "learning_rate": 6.685394469114419e-06,
-      "loss": 0.3029,
-      "step": 18488
-    },
-    {
-      "epoch": 1.208352395268283,
-      "grad_norm": 0.4439953565597534,
-      "learning_rate": 6.685065710852593e-06,
-      "loss": 0.3079,
-      "step": 18489
-    },
-    {
-      "epoch": 1.2084177504738252,
-      "grad_norm": 0.44726791977882385,
-      "learning_rate": 6.684736944372117e-06,
-      "loss": 0.3184,
-      "step": 18490
-    },
-    {
-      "epoch": 1.2084831056793675,
-      "grad_norm": 0.4733264446258545,
-      "learning_rate": 6.684408169674597e-06,
-      "loss": 0.3566,
-      "step": 18491
-    },
-    {
-      "epoch": 1.2085484608849095,
-      "grad_norm": 0.4719187915325165,
-      "learning_rate": 6.684079386761633e-06,
-      "loss": 0.3937,
-      "step": 18492
-    },
-    {
-      "epoch": 1.2086138160904516,
-      "grad_norm": 0.4557696580886841,
-      "learning_rate": 6.68375059563483e-06,
-      "loss": 0.3288,
-      "step": 18493
-    },
-    {
-      "epoch": 1.2086791712959937,
-      "grad_norm": 0.44552236795425415,
-      "learning_rate": 6.683421796295795e-06,
-      "loss": 0.3279,
-      "step": 18494
-    },
-    {
-      "epoch": 1.208744526501536,
-      "grad_norm": 0.45396193861961365,
-      "learning_rate": 6.683092988746125e-06,
-      "loss": 0.3198,
-      "step": 18495
-    },
-    {
-      "epoch": 1.208809881707078,
-      "grad_norm": 0.42722782492637634,
-      "learning_rate": 6.68276417298743e-06,
-      "loss": 0.3131,
-      "step": 18496
-    },
-    {
-      "epoch": 1.20887523691262,
-      "grad_norm": 0.48307183384895325,
-      "learning_rate": 6.6824353490213105e-06,
-      "loss": 0.4044,
-      "step": 18497
-    },
-    {
-      "epoch": 1.2089405921181622,
-      "grad_norm": 0.42966213822364807,
-      "learning_rate": 6.68210651684937e-06,
-      "loss": 0.3289,
-      "step": 18498
-    },
-    {
-      "epoch": 1.2090059473237043,
-      "grad_norm": 0.43984195590019226,
-      "learning_rate": 6.681777676473214e-06,
-      "loss": 0.345,
-      "step": 18499
-    },
-    {
-      "epoch": 1.2090713025292465,
-      "grad_norm": 0.49030372500419617,
-      "learning_rate": 6.681448827894446e-06,
-      "loss": 0.3603,
-      "step": 18500
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 45903,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
-  "save_steps": 500,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 4.037016123604992e+19,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}
+version https://git-lfs.github.com/spec/v1
+oid sha256:a350d620a7b8b47bd5794fc71ef41903558c4cfdd587518d5f3e89a4823125cf
+size 7698445