{
  "best_metric": 0.8061292767524719,
  "best_model_checkpoint": "miner_id_24/checkpoint-500",
  "epoch": 0.2516039753428104,
  "eval_steps": 100,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005032079506856208,
      "grad_norm": 0.38297849893569946,
      "learning_rate": 5e-06,
      "loss": 0.5872,
      "step": 1
    },
    {
      "epoch": 0.0005032079506856208,
      "eval_loss": 1.0933938026428223,
      "eval_runtime": 104.1253,
      "eval_samples_per_second": 32.144,
      "eval_steps_per_second": 8.038,
      "step": 1
    },
    {
      "epoch": 0.0010064159013712416,
      "grad_norm": 0.4977751076221466,
      "learning_rate": 1e-05,
      "loss": 0.7907,
      "step": 2
    },
    {
      "epoch": 0.0015096238520568624,
      "grad_norm": 0.4053085148334503,
      "learning_rate": 1.5e-05,
      "loss": 0.7258,
      "step": 3
    },
    {
      "epoch": 0.0020128318027424832,
      "grad_norm": 0.3187407851219177,
      "learning_rate": 2e-05,
      "loss": 0.6408,
      "step": 4
    },
    {
      "epoch": 0.002516039753428104,
      "grad_norm": 0.3603987395763397,
      "learning_rate": 2.5e-05,
      "loss": 0.8964,
      "step": 5
    },
    {
      "epoch": 0.003019247704113725,
      "grad_norm": 0.4167505204677582,
      "learning_rate": 3e-05,
      "loss": 0.7977,
      "step": 6
    },
    {
      "epoch": 0.0035224556547993457,
      "grad_norm": 0.33529505133628845,
      "learning_rate": 3.5e-05,
      "loss": 0.741,
      "step": 7
    },
    {
      "epoch": 0.0040256636054849665,
      "grad_norm": 0.3075070083141327,
      "learning_rate": 4e-05,
      "loss": 0.69,
      "step": 8
    },
    {
      "epoch": 0.004528871556170587,
      "grad_norm": 0.374713271856308,
      "learning_rate": 4.5e-05,
      "loss": 0.7692,
      "step": 9
    },
    {
      "epoch": 0.005032079506856208,
      "grad_norm": 0.42237210273742676,
      "learning_rate": 5e-05,
      "loss": 0.8035,
      "step": 10
    },
    {
      "epoch": 0.005535287457541829,
      "grad_norm": 0.44411763548851013,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.8701,
      "step": 11
    },
    {
      "epoch": 0.00603849540822745,
      "grad_norm": 0.41626623272895813,
      "learning_rate": 6e-05,
      "loss": 0.8624,
      "step": 12
    },
    {
      "epoch": 0.0065417033589130705,
      "grad_norm": 0.40315064787864685,
      "learning_rate": 6.500000000000001e-05,
      "loss": 0.8356,
      "step": 13
    },
    {
      "epoch": 0.007044911309598691,
      "grad_norm": 0.40788719058036804,
      "learning_rate": 7e-05,
      "loss": 0.9516,
      "step": 14
    },
    {
      "epoch": 0.007548119260284312,
      "grad_norm": 0.40332308411598206,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.9631,
      "step": 15
    },
    {
      "epoch": 0.008051327210969933,
      "grad_norm": 0.4551430344581604,
      "learning_rate": 8e-05,
      "loss": 0.9224,
      "step": 16
    },
    {
      "epoch": 0.008554535161655554,
      "grad_norm": 0.4207383096218109,
      "learning_rate": 8.5e-05,
      "loss": 0.8631,
      "step": 17
    },
    {
      "epoch": 0.009057743112341175,
      "grad_norm": 0.45070353150367737,
      "learning_rate": 9e-05,
      "loss": 0.867,
      "step": 18
    },
    {
      "epoch": 0.009560951063026795,
      "grad_norm": 0.41204211115837097,
      "learning_rate": 9.5e-05,
      "loss": 0.8123,
      "step": 19
    },
    {
      "epoch": 0.010064159013712416,
      "grad_norm": 0.4532608985900879,
      "learning_rate": 0.0001,
      "loss": 0.8152,
      "step": 20
    },
    {
      "epoch": 0.010567366964398037,
      "grad_norm": 0.42522719502449036,
      "learning_rate": 9.999892908320647e-05,
      "loss": 0.9145,
      "step": 21
    },
    {
      "epoch": 0.011070574915083658,
      "grad_norm": 0.43336769938468933,
      "learning_rate": 9.999571637870036e-05,
      "loss": 0.9232,
      "step": 22
    },
    {
      "epoch": 0.011573782865769279,
      "grad_norm": 0.43486812710762024,
      "learning_rate": 9.999036202410325e-05,
      "loss": 0.9012,
      "step": 23
    },
    {
      "epoch": 0.0120769908164549,
      "grad_norm": 0.43434756994247437,
      "learning_rate": 9.998286624877786e-05,
      "loss": 0.8315,
      "step": 24
    },
    {
      "epoch": 0.01258019876714052,
      "grad_norm": 0.4306849539279938,
      "learning_rate": 9.997322937381829e-05,
      "loss": 0.9571,
      "step": 25
    },
    {
      "epoch": 0.013083406717826141,
      "grad_norm": 0.4672311246395111,
      "learning_rate": 9.996145181203615e-05,
      "loss": 0.8713,
      "step": 26
    },
    {
      "epoch": 0.013586614668511762,
      "grad_norm": 0.4402594268321991,
      "learning_rate": 9.994753406794301e-05,
      "loss": 0.8639,
      "step": 27
    },
    {
      "epoch": 0.014089822619197383,
      "grad_norm": 0.45539745688438416,
      "learning_rate": 9.99314767377287e-05,
      "loss": 0.8925,
      "step": 28
    },
    {
      "epoch": 0.014593030569883004,
      "grad_norm": 0.48189201951026917,
      "learning_rate": 9.991328050923581e-05,
      "loss": 0.9734,
      "step": 29
    },
    {
      "epoch": 0.015096238520568624,
      "grad_norm": 0.44766247272491455,
      "learning_rate": 9.989294616193017e-05,
      "loss": 0.8371,
      "step": 30
    },
    {
      "epoch": 0.015599446471254245,
      "grad_norm": 0.5342682003974915,
      "learning_rate": 9.98704745668676e-05,
      "loss": 0.9424,
      "step": 31
    },
    {
      "epoch": 0.016102654421939866,
      "grad_norm": 0.5079587697982788,
      "learning_rate": 9.98458666866564e-05,
      "loss": 0.8545,
      "step": 32
    },
    {
      "epoch": 0.01660586237262549,
      "grad_norm": 0.5153644680976868,
      "learning_rate": 9.981912357541627e-05,
      "loss": 0.8507,
      "step": 33
    },
    {
      "epoch": 0.017109070323311108,
      "grad_norm": 0.5304479002952576,
      "learning_rate": 9.97902463787331e-05,
      "loss": 0.7365,
      "step": 34
    },
    {
      "epoch": 0.01761227827399673,
      "grad_norm": 0.5784561634063721,
      "learning_rate": 9.975923633360985e-05,
      "loss": 0.8531,
      "step": 35
    },
    {
      "epoch": 0.01811548622468235,
      "grad_norm": 0.6525458097457886,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.9531,
      "step": 36
    },
    {
      "epoch": 0.018618694175367972,
      "grad_norm": 0.59874427318573,
      "learning_rate": 9.969082310281891e-05,
      "loss": 0.8825,
      "step": 37
    },
    {
      "epoch": 0.01912190212605359,
      "grad_norm": 0.6657208800315857,
      "learning_rate": 9.965342284774632e-05,
      "loss": 1.013,
      "step": 38
    },
    {
      "epoch": 0.019625110076739213,
      "grad_norm": 0.673851490020752,
      "learning_rate": 9.961389560529836e-05,
      "loss": 1.0122,
      "step": 39
    },
    {
      "epoch": 0.020128318027424832,
      "grad_norm": 0.7437425255775452,
      "learning_rate": 9.957224306869053e-05,
      "loss": 0.9996,
      "step": 40
    },
    {
      "epoch": 0.020631525978110455,
      "grad_norm": 0.6888524293899536,
      "learning_rate": 9.952846702217886e-05,
      "loss": 0.9479,
      "step": 41
    },
    {
      "epoch": 0.021134733928796074,
      "grad_norm": 0.665365993976593,
      "learning_rate": 9.948256934098352e-05,
      "loss": 0.8076,
      "step": 42
    },
    {
      "epoch": 0.021637941879481697,
      "grad_norm": 0.9384682774543762,
      "learning_rate": 9.943455199120837e-05,
      "loss": 1.1964,
      "step": 43
    },
    {
      "epoch": 0.022141149830167316,
      "grad_norm": 1.06331205368042,
      "learning_rate": 9.938441702975689e-05,
      "loss": 1.1962,
      "step": 44
    },
    {
      "epoch": 0.022644357780852938,
      "grad_norm": 1.0956038236618042,
      "learning_rate": 9.933216660424395e-05,
      "loss": 1.0293,
      "step": 45
    },
    {
      "epoch": 0.023147565731538557,
      "grad_norm": 1.0600272417068481,
      "learning_rate": 9.927780295290389e-05,
      "loss": 1.2817,
      "step": 46
    },
    {
      "epoch": 0.02365077368222418,
      "grad_norm": 1.032386064529419,
      "learning_rate": 9.922132840449459e-05,
      "loss": 1.1316,
      "step": 47
    },
    {
      "epoch": 0.0241539816329098,
      "grad_norm": 1.1418523788452148,
      "learning_rate": 9.916274537819775e-05,
      "loss": 0.9598,
      "step": 48
    },
    {
      "epoch": 0.02465718958359542,
      "grad_norm": 1.591659426689148,
      "learning_rate": 9.91020563835152e-05,
      "loss": 1.1625,
      "step": 49
    },
    {
      "epoch": 0.02516039753428104,
      "grad_norm": 1.7732504606246948,
      "learning_rate": 9.903926402016153e-05,
      "loss": 1.0553,
      "step": 50
    },
    {
      "epoch": 0.025663605484966663,
      "grad_norm": 0.5284194946289062,
      "learning_rate": 9.897437097795257e-05,
      "loss": 0.6376,
      "step": 51
    },
    {
      "epoch": 0.026166813435652282,
      "grad_norm": 0.6791232824325562,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.5686,
      "step": 52
    },
    {
      "epoch": 0.026670021386337905,
      "grad_norm": 0.4423704445362091,
      "learning_rate": 9.883829406604363e-05,
      "loss": 0.6714,
      "step": 53
    },
    {
      "epoch": 0.027173229337023524,
      "grad_norm": 0.4226953387260437,
      "learning_rate": 9.876711602542563e-05,
      "loss": 0.7254,
      "step": 54
    },
    {
      "epoch": 0.027676437287709146,
      "grad_norm": 0.3696565628051758,
      "learning_rate": 9.869384896386668e-05,
      "loss": 0.8249,
      "step": 55
    },
    {
      "epoch": 0.028179645238394765,
      "grad_norm": 0.2843206822872162,
      "learning_rate": 9.861849601988383e-05,
      "loss": 0.6783,
      "step": 56
    },
    {
      "epoch": 0.028682853189080388,
      "grad_norm": 0.3041861057281494,
      "learning_rate": 9.854106042134641e-05,
      "loss": 0.7137,
      "step": 57
    },
    {
      "epoch": 0.029186061139766007,
      "grad_norm": 0.39011338353157043,
      "learning_rate": 9.846154548533773e-05,
      "loss": 0.8799,
      "step": 58
    },
    {
      "epoch": 0.02968926909045163,
      "grad_norm": 0.40477269887924194,
      "learning_rate": 9.837995461801299e-05,
      "loss": 0.8491,
      "step": 59
    },
    {
      "epoch": 0.03019247704113725,
      "grad_norm": 0.3500373363494873,
      "learning_rate": 9.829629131445342e-05,
      "loss": 0.7849,
      "step": 60
    },
    {
      "epoch": 0.03069568499182287,
      "grad_norm": 0.3607657849788666,
      "learning_rate": 9.821055915851647e-05,
      "loss": 0.8036,
      "step": 61
    },
    {
      "epoch": 0.03119889294250849,
      "grad_norm": 0.3407435417175293,
      "learning_rate": 9.812276182268236e-05,
      "loss": 0.8677,
      "step": 62
    },
    {
      "epoch": 0.03170210089319411,
      "grad_norm": 0.34367769956588745,
      "learning_rate": 9.803290306789676e-05,
      "loss": 0.7972,
      "step": 63
    },
    {
      "epoch": 0.03220530884387973,
      "grad_norm": 0.3393014371395111,
      "learning_rate": 9.794098674340965e-05,
      "loss": 0.7821,
      "step": 64
    },
    {
      "epoch": 0.03270851679456535,
      "grad_norm": 0.399403840303421,
      "learning_rate": 9.784701678661045e-05,
      "loss": 0.9752,
      "step": 65
    },
    {
      "epoch": 0.03321172474525098,
      "grad_norm": 0.336550772190094,
      "learning_rate": 9.775099722285935e-05,
      "loss": 0.8358,
      "step": 66
    },
    {
      "epoch": 0.033714932695936596,
      "grad_norm": 0.42282477021217346,
      "learning_rate": 9.765293216531486e-05,
      "loss": 0.8486,
      "step": 67
    },
    {
      "epoch": 0.034218140646622215,
      "grad_norm": 0.3811139464378357,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.8015,
      "step": 68
    },
    {
      "epoch": 0.034721348597307834,
      "grad_norm": 0.38650673627853394,
      "learning_rate": 9.74506824594107e-05,
      "loss": 0.8441,
      "step": 69
    },
    {
      "epoch": 0.03522455654799346,
      "grad_norm": 0.39965903759002686,
      "learning_rate": 9.73465064747553e-05,
      "loss": 0.9519,
      "step": 70
    },
    {
      "epoch": 0.03572776449867908,
      "grad_norm": 0.3865853548049927,
      "learning_rate": 9.724030232334391e-05,
      "loss": 0.8355,
      "step": 71
    },
    {
      "epoch": 0.0362309724493647,
      "grad_norm": 0.3812100887298584,
      "learning_rate": 9.713207455460894e-05,
      "loss": 0.815,
      "step": 72
    },
    {
      "epoch": 0.03673418040005032,
      "grad_norm": 0.40402311086654663,
      "learning_rate": 9.702182780466775e-05,
      "loss": 0.7945,
      "step": 73
    },
    {
      "epoch": 0.037237388350735943,
      "grad_norm": 0.43912094831466675,
      "learning_rate": 9.690956679612421e-05,
      "loss": 0.9382,
      "step": 74
    },
    {
      "epoch": 0.03774059630142156,
      "grad_norm": 0.48043292760849,
      "learning_rate": 9.67952963378663e-05,
      "loss": 0.9166,
      "step": 75
    },
    {
      "epoch": 0.03824380425210718,
      "grad_norm": 0.47702670097351074,
      "learning_rate": 9.667902132486009e-05,
      "loss": 0.8208,
      "step": 76
    },
    {
      "epoch": 0.0387470122027928,
      "grad_norm": 0.4606092870235443,
      "learning_rate": 9.656074673794018e-05,
      "loss": 0.7812,
      "step": 77
    },
    {
      "epoch": 0.03925022015347843,
      "grad_norm": 0.410383403301239,
      "learning_rate": 9.644047764359622e-05,
      "loss": 0.7814,
      "step": 78
    },
    {
      "epoch": 0.039753428104164046,
      "grad_norm": 0.43723535537719727,
      "learning_rate": 9.631821919375591e-05,
      "loss": 0.7831,
      "step": 79
    },
    {
      "epoch": 0.040256636054849665,
      "grad_norm": 0.47842466831207275,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.8328,
      "step": 80
    },
    {
      "epoch": 0.04075984400553529,
      "grad_norm": 0.5565553903579712,
      "learning_rate": 9.606775526115963e-05,
      "loss": 0.8564,
      "step": 81
    },
    {
      "epoch": 0.04126305195622091,
      "grad_norm": 0.5389965772628784,
      "learning_rate": 9.593956050744492e-05,
      "loss": 0.9588,
      "step": 82
    },
    {
      "epoch": 0.04176625990690653,
      "grad_norm": 0.5707739591598511,
      "learning_rate": 9.580939785585681e-05,
      "loss": 0.8665,
      "step": 83
    },
    {
      "epoch": 0.04226946785759215,
      "grad_norm": 0.5401300191879272,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.8702,
      "step": 84
    },
    {
      "epoch": 0.042772675808277774,
      "grad_norm": 0.5842973589897156,
      "learning_rate": 9.554319124605879e-05,
      "loss": 0.8452,
      "step": 85
    },
    {
      "epoch": 0.04327588375896339,
      "grad_norm": 0.526501476764679,
      "learning_rate": 9.540715869125407e-05,
      "loss": 0.8399,
      "step": 86
    },
    {
      "epoch": 0.04377909170964901,
      "grad_norm": 0.609856128692627,
      "learning_rate": 9.526918104489777e-05,
      "loss": 1.1022,
      "step": 87
    },
    {
      "epoch": 0.04428229966033463,
      "grad_norm": 0.6386977434158325,
      "learning_rate": 9.512926421749304e-05,
      "loss": 0.9333,
      "step": 88
    },
    {
      "epoch": 0.04478550761102026,
      "grad_norm": 0.6148085594177246,
      "learning_rate": 9.498741420261108e-05,
      "loss": 0.8722,
      "step": 89
    },
    {
      "epoch": 0.045288715561705876,
      "grad_norm": 0.610058069229126,
      "learning_rate": 9.484363707663442e-05,
      "loss": 0.9483,
      "step": 90
    },
    {
      "epoch": 0.045791923512391496,
      "grad_norm": 0.6424174904823303,
      "learning_rate": 9.469793899849661e-05,
      "loss": 0.9546,
      "step": 91
    },
    {
      "epoch": 0.046295131463077115,
      "grad_norm": 0.6571527719497681,
      "learning_rate": 9.45503262094184e-05,
      "loss": 0.84,
      "step": 92
    },
    {
      "epoch": 0.04679833941376274,
      "grad_norm": 0.9009180665016174,
      "learning_rate": 9.440080503264037e-05,
      "loss": 1.2002,
      "step": 93
    },
    {
      "epoch": 0.04730154736444836,
      "grad_norm": 0.8972412347793579,
      "learning_rate": 9.42493818731521e-05,
      "loss": 0.9899,
      "step": 94
    },
    {
      "epoch": 0.04780475531513398,
      "grad_norm": 0.8759328722953796,
      "learning_rate": 9.409606321741775e-05,
      "loss": 0.9973,
      "step": 95
    },
    {
      "epoch": 0.0483079632658196,
      "grad_norm": 0.880595326423645,
      "learning_rate": 9.394085563309827e-05,
      "loss": 0.9668,
      "step": 96
    },
    {
      "epoch": 0.048811171216505224,
      "grad_norm": 1.0916500091552734,
      "learning_rate": 9.378376576876999e-05,
      "loss": 1.0941,
      "step": 97
    },
    {
      "epoch": 0.04931437916719084,
      "grad_norm": 0.9800620675086975,
      "learning_rate": 9.362480035363986e-05,
      "loss": 1.0244,
      "step": 98
    },
    {
      "epoch": 0.04981758711787646,
      "grad_norm": 1.3642239570617676,
      "learning_rate": 9.34639661972572e-05,
      "loss": 1.1874,
      "step": 99
    },
    {
      "epoch": 0.05032079506856208,
      "grad_norm": 1.7741835117340088,
      "learning_rate": 9.330127018922194e-05,
      "loss": 1.1594,
      "step": 100
    },
    {
      "epoch": 0.05032079506856208,
      "eval_loss": 0.8690917491912842,
      "eval_runtime": 104.3719,
      "eval_samples_per_second": 32.068,
      "eval_steps_per_second": 8.019,
      "step": 100
    },
    {
      "epoch": 0.05082400301924771,
      "grad_norm": 0.33244970440864563,
      "learning_rate": 9.31367192988896e-05,
      "loss": 0.6948,
      "step": 101
    },
    {
      "epoch": 0.051327210969933326,
      "grad_norm": 0.3542371392250061,
      "learning_rate": 9.297032057507264e-05,
      "loss": 0.6499,
      "step": 102
    },
    {
      "epoch": 0.051830418920618945,
      "grad_norm": 0.31864750385284424,
      "learning_rate": 9.280208114573859e-05,
      "loss": 0.5946,
      "step": 103
    },
    {
      "epoch": 0.052333626871304564,
      "grad_norm": 0.31783977150917053,
      "learning_rate": 9.263200821770461e-05,
      "loss": 0.7341,
      "step": 104
    },
    {
      "epoch": 0.05283683482199019,
      "grad_norm": 0.27859652042388916,
      "learning_rate": 9.246010907632895e-05,
      "loss": 0.6446,
      "step": 105
    },
    {
      "epoch": 0.05334004277267581,
      "grad_norm": 0.29917001724243164,
      "learning_rate": 9.228639108519868e-05,
      "loss": 0.735,
      "step": 106
    },
    {
      "epoch": 0.05384325072336143,
      "grad_norm": 0.2935086190700531,
      "learning_rate": 9.211086168581433e-05,
      "loss": 0.6817,
      "step": 107
    },
    {
      "epoch": 0.05434645867404705,
      "grad_norm": 0.35257166624069214,
      "learning_rate": 9.193352839727121e-05,
      "loss": 0.7623,
      "step": 108
    },
    {
      "epoch": 0.054849666624732674,
      "grad_norm": 0.37438997626304626,
      "learning_rate": 9.175439881593716e-05,
      "loss": 0.8425,
      "step": 109
    },
    {
      "epoch": 0.05535287457541829,
      "grad_norm": 0.3695034980773926,
      "learning_rate": 9.157348061512727e-05,
      "loss": 0.7458,
      "step": 110
    },
    {
      "epoch": 0.05585608252610391,
      "grad_norm": 0.3363094925880432,
      "learning_rate": 9.139078154477512e-05,
      "loss": 0.8181,
      "step": 111
    },
    {
      "epoch": 0.05635929047678953,
      "grad_norm": 0.3058965504169464,
      "learning_rate": 9.120630943110077e-05,
      "loss": 0.6645,
      "step": 112
    },
    {
      "epoch": 0.05686249842747516,
      "grad_norm": 0.31257951259613037,
      "learning_rate": 9.102007217627568e-05,
      "loss": 0.7269,
      "step": 113
    },
    {
      "epoch": 0.057365706378160776,
      "grad_norm": 0.3484824299812317,
      "learning_rate": 9.083207775808396e-05,
      "loss": 0.8735,
      "step": 114
    },
    {
      "epoch": 0.057868914328846395,
      "grad_norm": 0.3367786109447479,
      "learning_rate": 9.064233422958077e-05,
      "loss": 0.766,
      "step": 115
    },
    {
      "epoch": 0.058372122279532014,
      "grad_norm": 0.371697336435318,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.7952,
      "step": 116
    },
    {
      "epoch": 0.05887533023021764,
      "grad_norm": 0.3939005136489868,
      "learning_rate": 9.025763242814291e-05,
      "loss": 0.895,
      "step": 117
    },
    {
      "epoch": 0.05937853818090326,
      "grad_norm": 0.40599775314331055,
      "learning_rate": 9.006269063455304e-05,
      "loss": 0.9353,
      "step": 118
    },
    {
      "epoch": 0.05988174613158888,
      "grad_norm": 0.4013155698776245,
      "learning_rate": 8.986603268863536e-05,
      "loss": 0.9384,
      "step": 119
    },
    {
      "epoch": 0.0603849540822745,
      "grad_norm": 0.41658180952072144,
      "learning_rate": 8.966766701456177e-05,
      "loss": 0.9412,
      "step": 120
    },
    {
      "epoch": 0.06088816203296012,
      "grad_norm": 0.38339975476264954,
      "learning_rate": 8.94676021096575e-05,
      "loss": 0.8672,
      "step": 121
    },
    {
      "epoch": 0.06139136998364574,
      "grad_norm": 0.4122770428657532,
      "learning_rate": 8.926584654403724e-05,
      "loss": 0.9392,
      "step": 122
    },
    {
      "epoch": 0.06189457793433136,
      "grad_norm": 0.4064811170101166,
      "learning_rate": 8.906240896023794e-05,
      "loss": 0.8886,
      "step": 123
    },
    {
      "epoch": 0.06239778588501698,
      "grad_norm": 0.40755781531333923,
      "learning_rate": 8.885729807284856e-05,
      "loss": 0.9177,
      "step": 124
    },
    {
      "epoch": 0.0629009938357026,
      "grad_norm": 0.4159265160560608,
      "learning_rate": 8.865052266813685e-05,
      "loss": 0.8445,
      "step": 125
    },
    {
      "epoch": 0.06340420178638823,
      "grad_norm": 0.403283953666687,
      "learning_rate": 8.844209160367299e-05,
      "loss": 0.8543,
      "step": 126
    },
    {
      "epoch": 0.06390740973707384,
      "grad_norm": 0.47934430837631226,
      "learning_rate": 8.823201380795001e-05,
      "loss": 0.8296,
      "step": 127
    },
    {
      "epoch": 0.06441061768775946,
      "grad_norm": 0.47067782282829285,
      "learning_rate": 8.802029828000156e-05,
      "loss": 0.936,
      "step": 128
    },
    {
      "epoch": 0.06491382563844508,
      "grad_norm": 0.501133382320404,
      "learning_rate": 8.780695408901613e-05,
      "loss": 0.8566,
      "step": 129
    },
    {
      "epoch": 0.0654170335891307,
      "grad_norm": 0.4946805536746979,
      "learning_rate": 8.759199037394887e-05,
      "loss": 0.9518,
      "step": 130
    },
    {
      "epoch": 0.06592024153981633,
      "grad_norm": 0.47323423624038696,
      "learning_rate": 8.737541634312985e-05,
      "loss": 0.8213,
      "step": 131
    },
    {
      "epoch": 0.06642344949050195,
      "grad_norm": 0.4894852340221405,
      "learning_rate": 8.715724127386972e-05,
      "loss": 0.8871,
      "step": 132
    },
    {
      "epoch": 0.06692665744118757,
      "grad_norm": 0.5523199439048767,
      "learning_rate": 8.693747451206232e-05,
      "loss": 0.9287,
      "step": 133
    },
    {
      "epoch": 0.06742986539187319,
      "grad_norm": 0.5813436508178711,
      "learning_rate": 8.671612547178428e-05,
      "loss": 0.8782,
      "step": 134
    },
    {
      "epoch": 0.06793307334255881,
      "grad_norm": 0.5135183334350586,
      "learning_rate": 8.649320363489179e-05,
      "loss": 0.7929,
      "step": 135
    },
    {
      "epoch": 0.06843628129324443,
      "grad_norm": 0.6729757189750671,
      "learning_rate": 8.626871855061438e-05,
      "loss": 1.014,
      "step": 136
    },
    {
      "epoch": 0.06893948924393005,
      "grad_norm": 0.5609135031700134,
      "learning_rate": 8.604267983514594e-05,
      "loss": 0.8092,
      "step": 137
    },
    {
      "epoch": 0.06944269719461567,
      "grad_norm": 0.5602811574935913,
      "learning_rate": 8.581509717123273e-05,
      "loss": 0.8407,
      "step": 138
    },
    {
      "epoch": 0.0699459051453013,
      "grad_norm": 0.5880002379417419,
      "learning_rate": 8.558598030775857e-05,
      "loss": 0.8237,
      "step": 139
    },
    {
      "epoch": 0.07044911309598692,
      "grad_norm": 0.6810115575790405,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.9806,
      "step": 140
    },
    {
      "epoch": 0.07095232104667254,
      "grad_norm": 0.7699592113494873,
      "learning_rate": 8.51231833058426e-05,
      "loss": 0.9699,
      "step": 141
    },
    {
      "epoch": 0.07145552899735816,
      "grad_norm": 0.7383102178573608,
      "learning_rate": 8.488952299208401e-05,
      "loss": 0.9188,
      "step": 142
    },
    {
      "epoch": 0.07195873694804378,
      "grad_norm": 0.784283459186554,
      "learning_rate": 8.46543681272818e-05,
      "loss": 0.9962,
      "step": 143
    },
    {
      "epoch": 0.0724619448987294,
      "grad_norm": 0.8139053583145142,
      "learning_rate": 8.44177287846877e-05,
      "loss": 0.9982,
      "step": 144
    },
    {
      "epoch": 0.07296515284941502,
      "grad_norm": 0.9453422427177429,
      "learning_rate": 8.417961510114356e-05,
      "loss": 1.1149,
      "step": 145
    },
    {
      "epoch": 0.07346836080010063,
      "grad_norm": 0.8534306883811951,
      "learning_rate": 8.39400372766471e-05,
      "loss": 1.1168,
      "step": 146
    },
    {
      "epoch": 0.07397156875078627,
      "grad_norm": 1.0746036767959595,
      "learning_rate": 8.36990055739149e-05,
      "loss": 1.1099,
      "step": 147
    },
    {
      "epoch": 0.07447477670147189,
      "grad_norm": 1.0740514993667603,
      "learning_rate": 8.345653031794292e-05,
      "loss": 1.0318,
      "step": 148
    },
    {
      "epoch": 0.0749779846521575,
      "grad_norm": 1.3433160781860352,
      "learning_rate": 8.321262189556409e-05,
      "loss": 1.2778,
      "step": 149
    },
    {
      "epoch": 0.07548119260284313,
      "grad_norm": 1.709749698638916,
      "learning_rate": 8.296729075500344e-05,
      "loss": 0.9436,
      "step": 150
    },
    {
      "epoch": 0.07598440055352874,
      "grad_norm": 0.2686924636363983,
      "learning_rate": 8.272054740543052e-05,
      "loss": 0.5435,
      "step": 151
    },
    {
      "epoch": 0.07648760850421436,
      "grad_norm": 0.3222774565219879,
      "learning_rate": 8.247240241650918e-05,
      "loss": 0.6375,
      "step": 152
    },
    {
      "epoch": 0.07699081645489998,
      "grad_norm": 0.30507364869117737,
      "learning_rate": 8.222286641794488e-05,
      "loss": 0.7499,
      "step": 153
    },
    {
      "epoch": 0.0774940244055856,
      "grad_norm": 0.269182950258255,
      "learning_rate": 8.197195009902924e-05,
      "loss": 0.5893,
      "step": 154
    },
    {
      "epoch": 0.07799723235627123,
      "grad_norm": 0.2954126298427582,
      "learning_rate": 8.171966420818228e-05,
      "loss": 0.6649,
      "step": 155
    },
    {
      "epoch": 0.07850044030695685,
      "grad_norm": 0.3091728091239929,
      "learning_rate": 8.146601955249188e-05,
      "loss": 0.6618,
      "step": 156
    },
    {
      "epoch": 0.07900364825764247,
      "grad_norm": 0.2694949805736542,
      "learning_rate": 8.121102699725089e-05,
      "loss": 0.6362,
      "step": 157
    },
    {
      "epoch": 0.07950685620832809,
      "grad_norm": 0.3082178235054016,
      "learning_rate": 8.095469746549172e-05,
      "loss": 0.6229,
      "step": 158
    },
    {
      "epoch": 0.08001006415901371,
      "grad_norm": 0.3117753565311432,
      "learning_rate": 8.069704193751832e-05,
      "loss": 0.7307,
      "step": 159
    },
    {
      "epoch": 0.08051327210969933,
      "grad_norm": 0.32263603806495667,
      "learning_rate": 8.043807145043604e-05,
      "loss": 0.7396,
      "step": 160
    },
    {
      "epoch": 0.08101648006038495,
      "grad_norm": 0.406573086977005,
      "learning_rate": 8.017779709767858e-05,
      "loss": 0.924,
      "step": 161
    },
    {
      "epoch": 0.08151968801107058,
      "grad_norm": 0.3405095040798187,
      "learning_rate": 7.991623002853296e-05,
      "loss": 0.7669,
      "step": 162
    },
    {
      "epoch": 0.0820228959617562,
      "grad_norm": 0.3500613272190094,
      "learning_rate": 7.965338144766186e-05,
      "loss": 0.7785,
      "step": 163
    },
    {
      "epoch": 0.08252610391244182,
      "grad_norm": 0.3334035575389862,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.7282,
      "step": 164
    },
    {
      "epoch": 0.08302931186312744,
      "grad_norm": 0.3296736180782318,
      "learning_rate": 7.912388484339012e-05,
      "loss": 0.714,
      "step": 165
    },
    {
      "epoch": 0.08353251981381306,
      "grad_norm": 0.3632432222366333,
      "learning_rate": 7.88572595018617e-05,
      "loss": 0.7388,
      "step": 166
    },
    {
      "epoch": 0.08403572776449868,
      "grad_norm": 0.38017383217811584,
      "learning_rate": 7.858939801138061e-05,
      "loss": 0.8051,
      "step": 167
    },
    {
      "epoch": 0.0845389357151843,
      "grad_norm": 0.41629263758659363,
      "learning_rate": 7.832031184624164e-05,
      "loss": 0.9006,
      "step": 168
    },
    {
      "epoch": 0.08504214366586992,
      "grad_norm": 0.40468400716781616,
      "learning_rate": 7.80500125332005e-05,
      "loss": 0.9159,
      "step": 169
    },
    {
      "epoch": 0.08554535161655555,
      "grad_norm": 0.4251197576522827,
      "learning_rate": 7.777851165098012e-05,
      "loss": 0.9539,
      "step": 170
    },
    {
      "epoch": 0.08604855956724117,
      "grad_norm": 0.3863302171230316,
      "learning_rate": 7.750582082977467e-05,
      "loss": 0.7704,
      "step": 171
    },
    {
      "epoch": 0.08655176751792679,
      "grad_norm": 0.38953644037246704,
      "learning_rate": 7.723195175075136e-05,
      "loss": 0.8375,
      "step": 172
    },
    {
      "epoch": 0.0870549754686124,
      "grad_norm": 0.39719176292419434,
      "learning_rate": 7.695691614555003e-05,
      "loss": 0.8864,
      "step": 173
    },
    {
      "epoch": 0.08755818341929802,
      "grad_norm": 0.4050215482711792,
      "learning_rate": 7.668072579578058e-05,
      "loss": 0.8163,
      "step": 174
    },
    {
      "epoch": 0.08806139136998364,
      "grad_norm": 0.42585161328315735,
      "learning_rate": 7.64033925325184e-05,
      "loss": 0.8266,
      "step": 175
    },
    {
      "epoch": 0.08856459932066926,
      "grad_norm": 0.4436732232570648,
      "learning_rate": 7.612492823579745e-05,
      "loss": 0.9099,
      "step": 176
    },
    {
      "epoch": 0.08906780727135488,
      "grad_norm": 0.41037729382514954,
      "learning_rate": 7.584534483410137e-05,
      "loss": 0.9768,
      "step": 177
    },
    {
      "epoch": 0.08957101522204051,
      "grad_norm": 0.413972407579422,
      "learning_rate": 7.55646543038526e-05,
      "loss": 0.761,
      "step": 178
    },
    {
      "epoch": 0.09007422317272613,
      "grad_norm": 0.445989727973938,
      "learning_rate": 7.528286866889924e-05,
      "loss": 0.8737,
      "step": 179
    },
    {
      "epoch": 0.09057743112341175,
      "grad_norm": 0.4886978566646576,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.8799,
      "step": 180
    },
    {
      "epoch": 0.09108063907409737,
      "grad_norm": 0.515007495880127,
      "learning_rate": 7.471606041430723e-05,
      "loss": 0.9363,
      "step": 181
    },
    {
      "epoch": 0.09158384702478299,
      "grad_norm": 0.5086027979850769,
      "learning_rate": 7.443106207484776e-05,
      "loss": 0.8112,
      "step": 182
    },
    {
      "epoch": 0.09208705497546861,
      "grad_norm": 0.5021002292633057,
      "learning_rate": 7.414501719000187e-05,
      "loss": 0.8402,
      "step": 183
    },
    {
      "epoch": 0.09259026292615423,
      "grad_norm": 0.5450283288955688,
      "learning_rate": 7.385793801298042e-05,
      "loss": 0.7513,
      "step": 184
    },
    {
      "epoch": 0.09309347087683985,
      "grad_norm": 0.5646865963935852,
      "learning_rate": 7.35698368412999e-05,
      "loss": 0.847,
      "step": 185
    },
    {
      "epoch": 0.09359667882752548,
      "grad_norm": 0.5946149230003357,
      "learning_rate": 7.328072601625557e-05,
      "loss": 0.8666,
      "step": 186
    },
    {
      "epoch": 0.0940998867782111,
      "grad_norm": 0.5803913474082947,
      "learning_rate": 7.2990617922393e-05,
      "loss": 0.9721,
      "step": 187
    },
    {
      "epoch": 0.09460309472889672,
      "grad_norm": 0.6186138391494751,
      "learning_rate": 7.269952498697734e-05,
      "loss": 0.8006,
      "step": 188
    },
    {
      "epoch": 0.09510630267958234,
      "grad_norm": 0.6833581328392029,
      "learning_rate": 7.240745967946113e-05,
      "loss": 0.9072,
      "step": 189
    },
    {
      "epoch": 0.09560951063026796,
      "grad_norm": 0.804831326007843,
      "learning_rate": 7.211443451095007e-05,
      "loss": 0.8989,
      "step": 190
    },
    {
      "epoch": 0.09611271858095358,
      "grad_norm": 0.7454904317855835,
      "learning_rate": 7.18204620336671e-05,
      "loss": 0.9415,
      "step": 191
    },
    {
      "epoch": 0.0966159265316392,
      "grad_norm": 0.7386621236801147,
      "learning_rate": 7.152555484041476e-05,
      "loss": 1.1014,
      "step": 192
    },
    {
      "epoch": 0.09711913448232481,
      "grad_norm": 0.8027065396308899,
      "learning_rate": 7.122972556403567e-05,
      "loss": 0.9651,
      "step": 193
    },
    {
      "epoch": 0.09762234243301045,
      "grad_norm": 0.843412458896637,
      "learning_rate": 7.09329868768714e-05,
      "loss": 0.9255,
      "step": 194
    },
    {
      "epoch": 0.09812555038369607,
      "grad_norm": 0.8859095573425293,
      "learning_rate": 7.063535149021973e-05,
      "loss": 1.0216,
      "step": 195
    },
    {
      "epoch": 0.09862875833438169,
      "grad_norm": 0.9696882367134094,
      "learning_rate": 7.033683215379002e-05,
      "loss": 1.0431,
      "step": 196
    },
    {
      "epoch": 0.0991319662850673,
      "grad_norm": 0.9128344655036926,
      "learning_rate": 7.003744165515705e-05,
      "loss": 0.9979,
      "step": 197
    },
    {
      "epoch": 0.09963517423575292,
      "grad_norm": 1.0365312099456787,
      "learning_rate": 6.973719281921335e-05,
      "loss": 1.0488,
      "step": 198
    },
    {
      "epoch": 0.10013838218643854,
      "grad_norm": 1.1797194480895996,
      "learning_rate": 6.943609850761979e-05,
      "loss": 1.0059,
      "step": 199
    },
    {
      "epoch": 0.10064159013712416,
      "grad_norm": 2.792616367340088,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.9833,
      "step": 200
    },
    {
      "epoch": 0.10064159013712416,
      "eval_loss": 0.8332151174545288,
      "eval_runtime": 104.6119,
      "eval_samples_per_second": 31.994,
      "eval_steps_per_second": 8.001,
      "step": 200
    },
    {
      "epoch": 0.10114479808780978,
      "grad_norm": 0.22519482672214508,
      "learning_rate": 6.883142508466054e-05,
      "loss": 0.4304,
      "step": 201
    },
    {
      "epoch": 0.10164800603849541,
      "grad_norm": 0.29026907682418823,
      "learning_rate": 6.852787187549182e-05,
      "loss": 0.5447,
      "step": 202
    },
    {
      "epoch": 0.10215121398918103,
      "grad_norm": 0.24368657171726227,
      "learning_rate": 6.82235249939575e-05,
      "loss": 0.5302,
      "step": 203
    },
    {
      "epoch": 0.10265442193986665,
      "grad_norm": 0.34774377942085266,
      "learning_rate": 6.7918397477265e-05,
      "loss": 0.6745,
      "step": 204
    },
    {
      "epoch": 0.10315762989055227,
      "grad_norm": 0.3001360595226288,
      "learning_rate": 6.761250239606169e-05,
      "loss": 0.7076,
      "step": 205
    },
    {
      "epoch": 0.10366083784123789,
      "grad_norm": 0.299600750207901,
      "learning_rate": 6.730585285387465e-05,
      "loss": 0.7122,
      "step": 206
    },
    {
      "epoch": 0.10416404579192351,
      "grad_norm": 0.291136771440506,
      "learning_rate": 6.699846198654971e-05,
      "loss": 0.6374,
      "step": 207
    },
    {
      "epoch": 0.10466725374260913,
      "grad_norm": 0.2981280982494354,
      "learning_rate": 6.669034296168855e-05,
      "loss": 0.7013,
      "step": 208
    },
    {
      "epoch": 0.10517046169329475,
      "grad_norm": 0.3012051582336426,
      "learning_rate": 6.638150897808468e-05,
      "loss": 0.7481,
      "step": 209
    },
    {
      "epoch": 0.10567366964398038,
      "grad_norm": 0.31151890754699707,
      "learning_rate": 6.607197326515808e-05,
      "loss": 0.6905,
      "step": 210
    },
    {
      "epoch": 0.106176877594666,
      "grad_norm": 0.3481593132019043,
      "learning_rate": 6.57617490823885e-05,
      "loss": 0.8155,
      "step": 211
    },
    {
      "epoch": 0.10668008554535162,
      "grad_norm": 0.3384002149105072,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.6831,
      "step": 212
    },
    {
      "epoch": 0.10718329349603724,
      "grad_norm": 0.3902130722999573,
      "learning_rate": 6.513928849212873e-05,
      "loss": 0.8484,
      "step": 213
    },
    {
      "epoch": 0.10768650144672286,
      "grad_norm": 0.32079294323921204,
      "learning_rate": 6.482707874877854e-05,
      "loss": 0.7408,
      "step": 214
    },
    {
      "epoch": 0.10818970939740848,
      "grad_norm": 0.3088681101799011,
      "learning_rate": 6.451423386272312e-05,
      "loss": 0.7263,
      "step": 215
    },
    {
      "epoch": 0.1086929173480941,
      "grad_norm": 0.33241790533065796,
      "learning_rate": 6.420076723519614e-05,
      "loss": 0.7808,
      "step": 216
    },
    {
      "epoch": 0.10919612529877971,
      "grad_norm": 0.36814364790916443,
      "learning_rate": 6.388669229406462e-05,
      "loss": 0.8254,
      "step": 217
    },
    {
      "epoch": 0.10969933324946535,
      "grad_norm": 0.367464542388916,
      "learning_rate": 6.357202249325371e-05,
      "loss": 0.9331,
      "step": 218
    },
    {
      "epoch": 0.11020254120015097,
      "grad_norm": 0.3247632086277008,
      "learning_rate": 6.32567713121704e-05,
      "loss": 0.8049,
      "step": 219
    },
    {
      "epoch": 0.11070574915083659,
      "grad_norm": 0.3686283230781555,
      "learning_rate": 6.294095225512603e-05,
      "loss": 0.8438,
      "step": 220
    },
    {
      "epoch": 0.1112089571015222,
      "grad_norm": 0.35872358083724976,
      "learning_rate": 6.26245788507579e-05,
      "loss": 0.8997,
      "step": 221
    },
    {
      "epoch": 0.11171216505220782,
      "grad_norm": 0.4434812068939209,
      "learning_rate": 6.230766465144967e-05,
      "loss": 0.8421,
      "step": 222
    },
    {
      "epoch": 0.11221537300289344,
      "grad_norm": 0.3434706926345825,
      "learning_rate": 6.199022323275083e-05,
      "loss": 0.7275,
      "step": 223
    },
    {
      "epoch": 0.11271858095357906,
      "grad_norm": 0.3942730724811554,
      "learning_rate": 6.167226819279528e-05,
      "loss": 0.7783,
      "step": 224
    },
    {
      "epoch": 0.11322178890426468,
      "grad_norm": 0.3891355097293854,
      "learning_rate": 6.135381315171867e-05,
      "loss": 0.7945,
      "step": 225
    },
    {
      "epoch": 0.11372499685495031,
      "grad_norm": 0.41171687841415405,
      "learning_rate": 6.103487175107507e-05,
      "loss": 0.9494,
      "step": 226
    },
    {
      "epoch": 0.11422820480563593,
      "grad_norm": 0.41461703181266785,
      "learning_rate": 6.071545765325254e-05,
      "loss": 0.7896,
      "step": 227
    },
    {
      "epoch": 0.11473141275632155,
      "grad_norm": 0.4718533754348755,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 0.8729,
      "step": 228
    },
    {
      "epoch": 0.11523462070700717,
      "grad_norm": 0.49657323956489563,
      "learning_rate": 6.007526611628086e-05,
      "loss": 0.8849,
      "step": 229
    },
    {
      "epoch": 0.11573782865769279,
      "grad_norm": 0.4768148958683014,
      "learning_rate": 5.9754516100806423e-05,
      "loss": 0.9346,
      "step": 230
    },
    {
      "epoch": 0.11624103660837841,
      "grad_norm": 0.4757716953754425,
      "learning_rate": 5.9433348234327765e-05,
      "loss": 0.8979,
      "step": 231
    },
    {
      "epoch": 0.11674424455906403,
      "grad_norm": 0.5215921998023987,
      "learning_rate": 5.911177627460739e-05,
      "loss": 0.7759,
      "step": 232
    },
    {
      "epoch": 0.11724745250974965,
      "grad_norm": 0.5007186532020569,
      "learning_rate": 5.8789813996717736e-05,
      "loss": 0.7752,
      "step": 233
    },
    {
      "epoch": 0.11775066046043528,
      "grad_norm": 0.5211498737335205,
      "learning_rate": 5.8467475192451226e-05,
      "loss": 0.7625,
      "step": 234
    },
    {
      "epoch": 0.1182538684111209,
      "grad_norm": 0.5580277442932129,
      "learning_rate": 5.814477366972945e-05,
      "loss": 0.8294,
      "step": 235
    },
    {
      "epoch": 0.11875707636180652,
      "grad_norm": 0.6667461395263672,
      "learning_rate": 5.782172325201155e-05,
      "loss": 0.9982,
      "step": 236
    },
    {
      "epoch": 0.11926028431249214,
      "grad_norm": 0.6224911212921143,
      "learning_rate": 5.749833777770225e-05,
      "loss": 0.9011,
      "step": 237
    },
    {
      "epoch": 0.11976349226317776,
      "grad_norm": 0.6415894627571106,
      "learning_rate": 5.717463109955896e-05,
      "loss": 0.8895,
      "step": 238
    },
    {
      "epoch": 0.12026670021386338,
      "grad_norm": 0.6843252778053284,
      "learning_rate": 5.685061708409841e-05,
      "loss": 0.9872,
      "step": 239
    },
    {
      "epoch": 0.120769908164549,
      "grad_norm": 0.6954467296600342,
      "learning_rate": 5.6526309611002594e-05,
      "loss": 0.8853,
      "step": 240
    },
    {
      "epoch": 0.12127311611523463,
      "grad_norm": 0.7757971286773682,
      "learning_rate": 5.6201722572524275e-05,
      "loss": 0.911,
      "step": 241
    },
    {
      "epoch": 0.12177632406592025,
      "grad_norm": 0.7121040225028992,
      "learning_rate": 5.587686987289189e-05,
      "loss": 0.8526,
      "step": 242
    },
    {
      "epoch": 0.12227953201660587,
      "grad_norm": 0.7902366518974304,
      "learning_rate": 5.5551765427713884e-05,
      "loss": 0.9972,
      "step": 243
    },
    {
      "epoch": 0.12278273996729148,
      "grad_norm": 0.8224020600318909,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.993,
      "step": 244
    },
    {
      "epoch": 0.1232859479179771,
      "grad_norm": 0.9266722798347473,
      "learning_rate": 5.490085701647805e-05,
      "loss": 1.0755,
      "step": 245
    },
    {
      "epoch": 0.12378915586866272,
      "grad_norm": 1.0139940977096558,
      "learning_rate": 5.457508093317013e-05,
      "loss": 1.1419,
      "step": 246
    },
    {
      "epoch": 0.12429236381934834,
      "grad_norm": 1.0336477756500244,
      "learning_rate": 5.4249108868622086e-05,
      "loss": 1.0416,
      "step": 247
    },
    {
      "epoch": 0.12479557177003396,
      "grad_norm": 1.122233510017395,
      "learning_rate": 5.392295478639225e-05,
      "loss": 1.0255,
      "step": 248
    },
    {
      "epoch": 0.1252987797207196,
      "grad_norm": 1.1999965906143188,
      "learning_rate": 5.359663265783598e-05,
      "loss": 1.035,
      "step": 249
    },
    {
      "epoch": 0.1258019876714052,
      "grad_norm": 1.8675717115402222,
      "learning_rate": 5.327015646150716e-05,
      "loss": 1.2304,
      "step": 250
    },
    {
      "epoch": 0.12630519562209083,
      "grad_norm": 0.25763022899627686,
      "learning_rate": 5.294354018255945e-05,
      "loss": 0.5091,
      "step": 251
    },
    {
      "epoch": 0.12680840357277645,
      "grad_norm": 0.3074914216995239,
      "learning_rate": 5.26167978121472e-05,
      "loss": 0.5562,
      "step": 252
    },
    {
      "epoch": 0.12731161152346207,
      "grad_norm": 0.2566604018211365,
      "learning_rate": 5.228994334682604e-05,
      "loss": 0.6086,
      "step": 253
    },
    {
      "epoch": 0.1278148194741477,
      "grad_norm": 0.30195143818855286,
      "learning_rate": 5.196299078795344e-05,
      "loss": 0.6226,
      "step": 254
    },
    {
      "epoch": 0.1283180274248333,
      "grad_norm": 0.22067229449748993,
      "learning_rate": 5.1635954141088813e-05,
      "loss": 0.5459,
      "step": 255
    },
    {
      "epoch": 0.12882123537551893,
      "grad_norm": 0.25044503808021545,
      "learning_rate": 5.1308847415393666e-05,
      "loss": 0.6079,
      "step": 256
    },
    {
      "epoch": 0.12932444332620455,
      "grad_norm": 0.3060043454170227,
      "learning_rate": 5.0981684623031415e-05,
      "loss": 0.6744,
      "step": 257
    },
    {
      "epoch": 0.12982765127689017,
      "grad_norm": 0.312330961227417,
      "learning_rate": 5.0654479778567223e-05,
      "loss": 0.7216,
      "step": 258
    },
    {
      "epoch": 0.13033085922757578,
      "grad_norm": 0.29661911725997925,
      "learning_rate": 5.0327246898367597e-05,
      "loss": 0.6644,
      "step": 259
    },
    {
      "epoch": 0.1308340671782614,
      "grad_norm": 0.32180383801460266,
      "learning_rate": 5e-05,
      "loss": 0.8465,
      "step": 260
    },
    {
      "epoch": 0.13133727512894705,
      "grad_norm": 0.34142372012138367,
      "learning_rate": 4.9672753101632415e-05,
      "loss": 0.8359,
      "step": 261
    },
    {
      "epoch": 0.13184048307963267,
      "grad_norm": 0.31370797753334045,
      "learning_rate": 4.934552022143279e-05,
      "loss": 0.7117,
      "step": 262
    },
    {
      "epoch": 0.1323436910303183,
      "grad_norm": 0.31379300355911255,
      "learning_rate": 4.901831537696859e-05,
      "loss": 0.7076,
      "step": 263
    },
    {
      "epoch": 0.1328468989810039,
      "grad_norm": 0.2920902669429779,
      "learning_rate": 4.869115258460635e-05,
      "loss": 0.6233,
      "step": 264
    },
    {
      "epoch": 0.13335010693168953,
      "grad_norm": 0.3320544362068176,
      "learning_rate": 4.83640458589112e-05,
      "loss": 0.8579,
      "step": 265
    },
    {
      "epoch": 0.13385331488237515,
      "grad_norm": 0.3355720341205597,
      "learning_rate": 4.8037009212046586e-05,
      "loss": 0.7268,
      "step": 266
    },
    {
      "epoch": 0.13435652283306077,
      "grad_norm": 0.3696479797363281,
      "learning_rate": 4.7710056653173976e-05,
      "loss": 0.8072,
      "step": 267
    },
    {
      "epoch": 0.13485973078374638,
      "grad_norm": 0.36613941192626953,
      "learning_rate": 4.738320218785281e-05,
      "loss": 0.7789,
      "step": 268
    },
    {
      "epoch": 0.135362938734432,
      "grad_norm": 0.4317779541015625,
      "learning_rate": 4.7056459817440544e-05,
      "loss": 0.8655,
      "step": 269
    },
    {
      "epoch": 0.13586614668511762,
      "grad_norm": 0.3596924841403961,
      "learning_rate": 4.6729843538492847e-05,
      "loss": 0.7526,
      "step": 270
    },
    {
      "epoch": 0.13636935463580324,
      "grad_norm": 0.39818912744522095,
      "learning_rate": 4.640336734216403e-05,
      "loss": 0.8869,
      "step": 271
    },
    {
      "epoch": 0.13687256258648886,
      "grad_norm": 0.39517346024513245,
      "learning_rate": 4.607704521360776e-05,
      "loss": 0.7689,
      "step": 272
    },
    {
      "epoch": 0.13737577053717448,
      "grad_norm": 0.37257814407348633,
      "learning_rate": 4.575089113137792e-05,
      "loss": 0.8039,
      "step": 273
    },
    {
      "epoch": 0.1378789784878601,
      "grad_norm": 0.39023667573928833,
      "learning_rate": 4.542491906682989e-05,
      "loss": 0.8586,
      "step": 274
    },
    {
      "epoch": 0.13838218643854572,
      "grad_norm": 0.41638848185539246,
      "learning_rate": 4.509914298352197e-05,
      "loss": 0.8507,
      "step": 275
    },
    {
      "epoch": 0.13888539438923134,
      "grad_norm": 0.3964393436908722,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.7734,
      "step": 276
    },
    {
      "epoch": 0.13938860233991698,
      "grad_norm": 0.4602956473827362,
      "learning_rate": 4.444823457228612e-05,
      "loss": 0.9757,
      "step": 277
    },
    {
      "epoch": 0.1398918102906026,
      "grad_norm": 0.420833557844162,
      "learning_rate": 4.412313012710813e-05,
      "loss": 0.7962,
      "step": 278
    },
    {
      "epoch": 0.14039501824128822,
      "grad_norm": 0.4108290374279022,
      "learning_rate": 4.379827742747575e-05,
      "loss": 0.7666,
      "step": 279
    },
    {
      "epoch": 0.14089822619197384,
      "grad_norm": 0.44668057560920715,
      "learning_rate": 4.347369038899744e-05,
      "loss": 0.9031,
      "step": 280
    },
    {
      "epoch": 0.14140143414265946,
      "grad_norm": 0.4494163990020752,
      "learning_rate": 4.3149382915901606e-05,
      "loss": 0.7926,
      "step": 281
    },
    {
      "epoch": 0.14190464209334508,
      "grad_norm": 0.46361368894577026,
      "learning_rate": 4.282536890044104e-05,
      "loss": 0.8512,
      "step": 282
    },
    {
      "epoch": 0.1424078500440307,
      "grad_norm": 0.4348432421684265,
      "learning_rate": 4.250166222229774e-05,
      "loss": 0.7417,
      "step": 283
    },
    {
      "epoch": 0.14291105799471632,
      "grad_norm": 0.5306094288825989,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 0.791,
      "step": 284
    },
    {
      "epoch": 0.14341426594540194,
      "grad_norm": 0.5647591948509216,
      "learning_rate": 4.185522633027057e-05,
      "loss": 0.8814,
      "step": 285
    },
    {
      "epoch": 0.14391747389608756,
      "grad_norm": 0.6206047534942627,
      "learning_rate": 4.153252480754877e-05,
      "loss": 0.9626,
      "step": 286
    },
    {
      "epoch": 0.14442068184677317,
      "grad_norm": 0.5752646923065186,
      "learning_rate": 4.1210186003282275e-05,
      "loss": 0.9403,
      "step": 287
    },
    {
      "epoch": 0.1449238897974588,
      "grad_norm": 0.6283671259880066,
      "learning_rate": 4.088822372539263e-05,
      "loss": 0.8473,
      "step": 288
    },
    {
      "epoch": 0.1454270977481444,
      "grad_norm": 0.598839282989502,
      "learning_rate": 4.0566651765672246e-05,
      "loss": 0.7755,
      "step": 289
    },
    {
      "epoch": 0.14593030569883003,
      "grad_norm": 0.6881972551345825,
      "learning_rate": 4.0245483899193595e-05,
      "loss": 0.9936,
      "step": 290
    },
    {
      "epoch": 0.14643351364951565,
      "grad_norm": 0.7141215801239014,
      "learning_rate": 3.992473388371915e-05,
      "loss": 0.9047,
      "step": 291
    },
    {
      "epoch": 0.14693672160020127,
      "grad_norm": 0.6933887600898743,
      "learning_rate": 3.960441545911204e-05,
      "loss": 0.8822,
      "step": 292
    },
    {
      "epoch": 0.14743992955088692,
      "grad_norm": 0.8213725090026855,
      "learning_rate": 3.928454234674747e-05,
      "loss": 0.9608,
      "step": 293
    },
    {
      "epoch": 0.14794313750157254,
      "grad_norm": 0.7432293891906738,
      "learning_rate": 3.896512824892495e-05,
      "loss": 0.9556,
      "step": 294
    },
    {
      "epoch": 0.14844634545225815,
      "grad_norm": 0.8309504389762878,
      "learning_rate": 3.864618684828134e-05,
      "loss": 1.0111,
      "step": 295
    },
    {
      "epoch": 0.14894955340294377,
      "grad_norm": 0.9679251313209534,
      "learning_rate": 3.832773180720475e-05,
      "loss": 1.0554,
      "step": 296
    },
    {
      "epoch": 0.1494527613536294,
      "grad_norm": 0.952207088470459,
      "learning_rate": 3.800977676724919e-05,
      "loss": 1.1046,
      "step": 297
    },
    {
      "epoch": 0.149955969304315,
      "grad_norm": 1.2538658380508423,
      "learning_rate": 3.769233534855035e-05,
      "loss": 1.0074,
      "step": 298
    },
    {
      "epoch": 0.15045917725500063,
      "grad_norm": 1.117653250694275,
      "learning_rate": 3.73754211492421e-05,
      "loss": 1.1258,
      "step": 299
    },
    {
      "epoch": 0.15096238520568625,
      "grad_norm": 2.037066698074341,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.9562,
      "step": 300
    },
    {
      "epoch": 0.15096238520568625,
      "eval_loss": 0.8160807490348816,
      "eval_runtime": 104.4474,
      "eval_samples_per_second": 32.045,
      "eval_steps_per_second": 8.014,
      "step": 300
    },
    {
      "epoch": 0.15146559315637187,
      "grad_norm": 0.26211196184158325,
      "learning_rate": 3.6743228687829595e-05,
      "loss": 0.5602,
      "step": 301
    },
    {
      "epoch": 0.1519688011070575,
      "grad_norm": 0.26436537504196167,
      "learning_rate": 3.642797750674629e-05,
      "loss": 0.4846,
      "step": 302
    },
    {
      "epoch": 0.1524720090577431,
      "grad_norm": 0.2748914361000061,
      "learning_rate": 3.6113307705935396e-05,
      "loss": 0.6021,
      "step": 303
    },
    {
      "epoch": 0.15297521700842873,
      "grad_norm": 0.22355616092681885,
      "learning_rate": 3.579923276480387e-05,
      "loss": 0.4879,
      "step": 304
    },
    {
      "epoch": 0.15347842495911435,
      "grad_norm": 0.27402985095977783,
      "learning_rate": 3.5485766137276894e-05,
      "loss": 0.7196,
      "step": 305
    },
    {
      "epoch": 0.15398163290979996,
      "grad_norm": 0.23475830256938934,
      "learning_rate": 3.5172921251221455e-05,
      "loss": 0.5369,
      "step": 306
    },
    {
      "epoch": 0.15448484086048558,
      "grad_norm": 0.28415268659591675,
      "learning_rate": 3.486071150787128e-05,
      "loss": 0.6017,
      "step": 307
    },
    {
      "epoch": 0.1549880488111712,
      "grad_norm": 0.27458059787750244,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.6019,
      "step": 308
    },
    {
      "epoch": 0.15549125676185685,
      "grad_norm": 0.32678642868995667,
      "learning_rate": 3.423825091761153e-05,
      "loss": 0.8207,
      "step": 309
    },
    {
      "epoch": 0.15599446471254247,
      "grad_norm": 0.3003508746623993,
      "learning_rate": 3.392802673484193e-05,
      "loss": 0.7746,
      "step": 310
    },
    {
      "epoch": 0.1564976726632281,
      "grad_norm": 0.321545273065567,
      "learning_rate": 3.361849102191533e-05,
      "loss": 0.8657,
      "step": 311
    },
    {
      "epoch": 0.1570008806139137,
      "grad_norm": 0.3106718361377716,
      "learning_rate": 3.330965703831146e-05,
      "loss": 0.6584,
      "step": 312
    },
    {
      "epoch": 0.15750408856459933,
      "grad_norm": 0.32135796546936035,
      "learning_rate": 3.300153801345028e-05,
      "loss": 0.7068,
      "step": 313
    },
    {
      "epoch": 0.15800729651528495,
      "grad_norm": 0.3328196108341217,
      "learning_rate": 3.2694147146125345e-05,
      "loss": 0.809,
      "step": 314
    },
    {
      "epoch": 0.15851050446597056,
      "grad_norm": 0.33247265219688416,
      "learning_rate": 3.2387497603938326e-05,
      "loss": 0.7724,
      "step": 315
    },
    {
      "epoch": 0.15901371241665618,
      "grad_norm": 0.3445248305797577,
      "learning_rate": 3.2081602522734986e-05,
      "loss": 0.729,
      "step": 316
    },
    {
      "epoch": 0.1595169203673418,
      "grad_norm": 0.3597284257411957,
      "learning_rate": 3.177647500604252e-05,
      "loss": 0.8055,
      "step": 317
    },
    {
      "epoch": 0.16002012831802742,
      "grad_norm": 0.36206623911857605,
      "learning_rate": 3.147212812450819e-05,
      "loss": 0.8012,
      "step": 318
    },
    {
      "epoch": 0.16052333626871304,
      "grad_norm": 0.34603044390678406,
      "learning_rate": 3.116857491533947e-05,
      "loss": 0.8477,
      "step": 319
    },
    {
      "epoch": 0.16102654421939866,
      "grad_norm": 0.3707107901573181,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.7539,
      "step": 320
    },
    {
      "epoch": 0.16152975217008428,
      "grad_norm": 0.3770596385002136,
      "learning_rate": 3.056390149238022e-05,
      "loss": 0.8248,
      "step": 321
    },
    {
      "epoch": 0.1620329601207699,
      "grad_norm": 0.3955375552177429,
      "learning_rate": 3.0262807180786647e-05,
      "loss": 0.8374,
      "step": 322
    },
    {
      "epoch": 0.16253616807145552,
      "grad_norm": 0.3951384127140045,
      "learning_rate": 2.996255834484296e-05,
      "loss": 0.8669,
      "step": 323
    },
    {
      "epoch": 0.16303937602214116,
      "grad_norm": 0.39075663685798645,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 0.8174,
      "step": 324
    },
    {
      "epoch": 0.16354258397282678,
      "grad_norm": 0.41376301646232605,
      "learning_rate": 2.936464850978027e-05,
      "loss": 0.87,
      "step": 325
    },
    {
      "epoch": 0.1640457919235124,
      "grad_norm": 0.37442007660865784,
      "learning_rate": 2.9067013123128613e-05,
      "loss": 0.7344,
      "step": 326
    },
    {
      "epoch": 0.16454899987419802,
      "grad_norm": 0.4042944610118866,
      "learning_rate": 2.8770274435964355e-05,
      "loss": 0.7885,
      "step": 327
    },
    {
      "epoch": 0.16505220782488364,
      "grad_norm": 0.4240545332431793,
      "learning_rate": 2.8474445159585235e-05,
      "loss": 0.8146,
      "step": 328
    },
    {
      "epoch": 0.16555541577556926,
      "grad_norm": 0.4908733665943146,
      "learning_rate": 2.8179537966332887e-05,
      "loss": 0.9611,
      "step": 329
    },
    {
      "epoch": 0.16605862372625488,
      "grad_norm": 0.46376368403434753,
      "learning_rate": 2.7885565489049946e-05,
      "loss": 0.7945,
      "step": 330
    },
    {
      "epoch": 0.1665618316769405,
      "grad_norm": 0.5365279912948608,
      "learning_rate": 2.759254032053888e-05,
      "loss": 0.9014,
      "step": 331
    },
    {
      "epoch": 0.16706503962762612,
      "grad_norm": 0.5090306997299194,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 0.8049,
      "step": 332
    },
    {
      "epoch": 0.16756824757831174,
      "grad_norm": 0.536597728729248,
      "learning_rate": 2.700938207760701e-05,
      "loss": 0.7892,
      "step": 333
    },
    {
      "epoch": 0.16807145552899735,
      "grad_norm": 0.5863428115844727,
      "learning_rate": 2.671927398374443e-05,
      "loss": 0.7805,
      "step": 334
    },
    {
      "epoch": 0.16857466347968297,
      "grad_norm": 0.49151116609573364,
      "learning_rate": 2.6430163158700115e-05,
      "loss": 0.7139,
      "step": 335
    },
    {
      "epoch": 0.1690778714303686,
      "grad_norm": 0.5732700824737549,
      "learning_rate": 2.6142061987019577e-05,
      "loss": 0.7674,
      "step": 336
    },
    {
      "epoch": 0.1695810793810542,
      "grad_norm": 0.7063980102539062,
      "learning_rate": 2.5854982809998153e-05,
      "loss": 0.8623,
      "step": 337
    },
    {
      "epoch": 0.17008428733173983,
      "grad_norm": 0.5727605223655701,
      "learning_rate": 2.556893792515227e-05,
      "loss": 0.7973,
      "step": 338
    },
    {
      "epoch": 0.17058749528242545,
      "grad_norm": 0.6519417762756348,
      "learning_rate": 2.5283939585692783e-05,
      "loss": 0.9101,
      "step": 339
    },
    {
      "epoch": 0.1710907032331111,
      "grad_norm": 0.6710662245750427,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.9189,
      "step": 340
    },
    {
      "epoch": 0.17159391118379672,
      "grad_norm": 0.7912317514419556,
      "learning_rate": 2.471713133110078e-05,
      "loss": 0.9465,
      "step": 341
    },
    {
      "epoch": 0.17209711913448233,
      "grad_norm": 0.9441129565238953,
      "learning_rate": 2.4435345696147403e-05,
      "loss": 1.2199,
      "step": 342
    },
    {
      "epoch": 0.17260032708516795,
      "grad_norm": 0.80022794008255,
      "learning_rate": 2.4154655165898627e-05,
      "loss": 0.9379,
      "step": 343
    },
    {
      "epoch": 0.17310353503585357,
      "grad_norm": 0.7774207592010498,
      "learning_rate": 2.3875071764202563e-05,
      "loss": 0.8809,
      "step": 344
    },
    {
      "epoch": 0.1736067429865392,
      "grad_norm": 0.8378373980522156,
      "learning_rate": 2.3596607467481603e-05,
      "loss": 0.9557,
      "step": 345
    },
    {
      "epoch": 0.1741099509372248,
      "grad_norm": 0.8967252373695374,
      "learning_rate": 2.3319274204219428e-05,
      "loss": 0.8607,
      "step": 346
    },
    {
      "epoch": 0.17461315888791043,
      "grad_norm": 0.870495080947876,
      "learning_rate": 2.3043083854449988e-05,
      "loss": 0.8724,
      "step": 347
    },
    {
      "epoch": 0.17511636683859605,
      "grad_norm": 1.1094945669174194,
      "learning_rate": 2.2768048249248648e-05,
      "loss": 0.8732,
      "step": 348
    },
    {
      "epoch": 0.17561957478928167,
      "grad_norm": 1.1429485082626343,
      "learning_rate": 2.2494179170225333e-05,
      "loss": 0.8457,
      "step": 349
    },
    {
      "epoch": 0.1761227827399673,
      "grad_norm": 1.5905135869979858,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.9063,
      "step": 350
    },
    {
      "epoch": 0.1766259906906529,
      "grad_norm": 0.23821896314620972,
      "learning_rate": 2.194998746679952e-05,
      "loss": 0.4251,
      "step": 351
    },
    {
      "epoch": 0.17712919864133853,
      "grad_norm": 0.23636475205421448,
      "learning_rate": 2.167968815375837e-05,
      "loss": 0.3922,
      "step": 352
    },
    {
      "epoch": 0.17763240659202414,
      "grad_norm": 0.25628796219825745,
      "learning_rate": 2.1410601988619394e-05,
      "loss": 0.5204,
      "step": 353
    },
    {
      "epoch": 0.17813561454270976,
      "grad_norm": 0.2599431574344635,
      "learning_rate": 2.1142740498138324e-05,
      "loss": 0.683,
      "step": 354
    },
    {
      "epoch": 0.17863882249339538,
      "grad_norm": 0.26432743668556213,
      "learning_rate": 2.08761151566099e-05,
      "loss": 0.6276,
      "step": 355
    },
    {
      "epoch": 0.17914203044408103,
      "grad_norm": 0.23032639920711517,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.5113,
      "step": 356
    },
    {
      "epoch": 0.17964523839476665,
      "grad_norm": 0.27185848355293274,
      "learning_rate": 2.034661855233815e-05,
      "loss": 0.6249,
      "step": 357
    },
    {
      "epoch": 0.18014844634545227,
      "grad_norm": 0.3025369942188263,
      "learning_rate": 2.008376997146705e-05,
      "loss": 0.753,
      "step": 358
    },
    {
      "epoch": 0.1806516542961379,
      "grad_norm": 0.31796345114707947,
      "learning_rate": 1.982220290232143e-05,
      "loss": 0.6984,
      "step": 359
    },
    {
      "epoch": 0.1811548622468235,
      "grad_norm": 0.31694695353507996,
      "learning_rate": 1.9561928549563968e-05,
      "loss": 0.7719,
      "step": 360
    },
    {
      "epoch": 0.18165807019750912,
      "grad_norm": 0.30704811215400696,
      "learning_rate": 1.9302958062481673e-05,
      "loss": 0.6763,
      "step": 361
    },
    {
      "epoch": 0.18216127814819474,
      "grad_norm": 0.32100924849510193,
      "learning_rate": 1.9045302534508297e-05,
      "loss": 0.6178,
      "step": 362
    },
    {
      "epoch": 0.18266448609888036,
      "grad_norm": 0.3147682845592499,
      "learning_rate": 1.8788973002749112e-05,
      "loss": 0.7238,
      "step": 363
    },
    {
      "epoch": 0.18316769404956598,
      "grad_norm": 0.34390130639076233,
      "learning_rate": 1.8533980447508137e-05,
      "loss": 0.793,
      "step": 364
    },
    {
      "epoch": 0.1836709020002516,
      "grad_norm": 0.36561378836631775,
      "learning_rate": 1.8280335791817733e-05,
      "loss": 0.7969,
      "step": 365
    },
    {
      "epoch": 0.18417410995093722,
      "grad_norm": 0.3845183551311493,
      "learning_rate": 1.8028049900970767e-05,
      "loss": 0.8577,
      "step": 366
    },
    {
      "epoch": 0.18467731790162284,
      "grad_norm": 0.35048288106918335,
      "learning_rate": 1.777713358205514e-05,
      "loss": 0.7802,
      "step": 367
    },
    {
      "epoch": 0.18518052585230846,
      "grad_norm": 0.40498775243759155,
      "learning_rate": 1.7527597583490822e-05,
      "loss": 0.9165,
      "step": 368
    },
    {
      "epoch": 0.18568373380299408,
      "grad_norm": 0.38008689880371094,
      "learning_rate": 1.7279452594569483e-05,
      "loss": 0.8561,
      "step": 369
    },
    {
      "epoch": 0.1861869417536797,
      "grad_norm": 0.39767763018608093,
      "learning_rate": 1.703270924499656e-05,
      "loss": 0.7432,
      "step": 370
    },
    {
      "epoch": 0.18669014970436532,
      "grad_norm": 0.40831029415130615,
      "learning_rate": 1.678737810443593e-05,
      "loss": 0.8613,
      "step": 371
    },
    {
      "epoch": 0.18719335765505096,
      "grad_norm": 0.37755683064460754,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.9187,
      "step": 372
    },
    {
      "epoch": 0.18769656560573658,
      "grad_norm": 0.41120558977127075,
      "learning_rate": 1.6300994426085103e-05,
      "loss": 0.851,
      "step": 373
    },
    {
      "epoch": 0.1881997735564222,
      "grad_norm": 0.3905021846294403,
      "learning_rate": 1.605996272335291e-05,
      "loss": 0.7259,
      "step": 374
    },
    {
      "epoch": 0.18870298150710782,
      "grad_norm": 0.8140775561332703,
      "learning_rate": 1.5820384898856434e-05,
      "loss": 0.9068,
      "step": 375
    },
    {
      "epoch": 0.18920618945779344,
      "grad_norm": 0.40314650535583496,
      "learning_rate": 1.5582271215312294e-05,
      "loss": 0.7781,
      "step": 376
    },
    {
      "epoch": 0.18970939740847906,
      "grad_norm": 0.4199778735637665,
      "learning_rate": 1.5345631872718214e-05,
      "loss": 0.8189,
      "step": 377
    },
    {
      "epoch": 0.19021260535916468,
      "grad_norm": 0.44229626655578613,
      "learning_rate": 1.5110477007916001e-05,
      "loss": 0.8987,
      "step": 378
    },
    {
      "epoch": 0.1907158133098503,
      "grad_norm": 0.4618576467037201,
      "learning_rate": 1.4876816694157419e-05,
      "loss": 0.7793,
      "step": 379
    },
    {
      "epoch": 0.19121902126053592,
      "grad_norm": 0.47276002168655396,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.787,
      "step": 380
    },
    {
      "epoch": 0.19172222921122153,
      "grad_norm": 0.50196772813797,
      "learning_rate": 1.4414019692241437e-05,
      "loss": 0.8903,
      "step": 381
    },
    {
      "epoch": 0.19222543716190715,
      "grad_norm": 0.4731937348842621,
      "learning_rate": 1.4184902828767287e-05,
      "loss": 0.7957,
      "step": 382
    },
    {
      "epoch": 0.19272864511259277,
      "grad_norm": 0.5056723952293396,
      "learning_rate": 1.3957320164854059e-05,
      "loss": 0.8921,
      "step": 383
    },
    {
      "epoch": 0.1932318530632784,
      "grad_norm": 0.5013747215270996,
      "learning_rate": 1.373128144938563e-05,
      "loss": 0.7901,
      "step": 384
    },
    {
      "epoch": 0.193735061013964,
      "grad_norm": 0.521843671798706,
      "learning_rate": 1.3506796365108232e-05,
      "loss": 0.7771,
      "step": 385
    },
    {
      "epoch": 0.19423826896464963,
      "grad_norm": 0.5642167329788208,
      "learning_rate": 1.3283874528215733e-05,
      "loss": 0.7303,
      "step": 386
    },
    {
      "epoch": 0.19474147691533525,
      "grad_norm": 0.5664670467376709,
      "learning_rate": 1.3062525487937699e-05,
      "loss": 0.7348,
      "step": 387
    },
    {
      "epoch": 0.1952446848660209,
      "grad_norm": 0.6890117526054382,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 0.9545,
      "step": 388
    },
    {
      "epoch": 0.19574789281670651,
      "grad_norm": 0.6803514957427979,
      "learning_rate": 1.2624583656870154e-05,
      "loss": 0.9202,
      "step": 389
    },
    {
      "epoch": 0.19625110076739213,
      "grad_norm": 0.6082613468170166,
      "learning_rate": 1.2408009626051137e-05,
      "loss": 0.8661,
      "step": 390
    },
    {
      "epoch": 0.19675430871807775,
      "grad_norm": 0.70882648229599,
      "learning_rate": 1.2193045910983863e-05,
      "loss": 1.0095,
      "step": 391
    },
    {
      "epoch": 0.19725751666876337,
      "grad_norm": 0.6985040307044983,
      "learning_rate": 1.1979701719998453e-05,
      "loss": 0.8489,
      "step": 392
    },
    {
      "epoch": 0.197760724619449,
      "grad_norm": 0.6823087930679321,
      "learning_rate": 1.1767986192049984e-05,
      "loss": 0.8537,
      "step": 393
    },
    {
      "epoch": 0.1982639325701346,
      "grad_norm": 0.8139995336532593,
      "learning_rate": 1.1557908396327028e-05,
      "loss": 0.9329,
      "step": 394
    },
    {
      "epoch": 0.19876714052082023,
      "grad_norm": 0.8146170377731323,
      "learning_rate": 1.134947733186315e-05,
      "loss": 1.0423,
      "step": 395
    },
    {
      "epoch": 0.19927034847150585,
      "grad_norm": 0.8821521401405334,
      "learning_rate": 1.1142701927151456e-05,
      "loss": 0.9563,
      "step": 396
    },
    {
      "epoch": 0.19977355642219147,
      "grad_norm": 1.0392780303955078,
      "learning_rate": 1.0937591039762085e-05,
      "loss": 1.01,
      "step": 397
    },
    {
      "epoch": 0.2002767643728771,
      "grad_norm": 0.8813369870185852,
      "learning_rate": 1.0734153455962765e-05,
      "loss": 1.0518,
      "step": 398
    },
    {
      "epoch": 0.2007799723235627,
      "grad_norm": 1.0717155933380127,
      "learning_rate": 1.0532397890342505e-05,
      "loss": 0.9553,
      "step": 399
    },
    {
      "epoch": 0.20128318027424832,
      "grad_norm": 1.5049222707748413,
      "learning_rate": 1.0332332985438248e-05,
      "loss": 0.9136,
      "step": 400
    },
    {
      "epoch": 0.20128318027424832,
      "eval_loss": 0.8091070055961609,
      "eval_runtime": 104.4834,
      "eval_samples_per_second": 32.034,
      "eval_steps_per_second": 8.011,
      "step": 400
    },
    {
      "epoch": 0.20178638822493394,
      "grad_norm": 0.2686651647090912,
      "learning_rate": 1.013396731136465e-05,
      "loss": 0.5779,
      "step": 401
    },
    {
      "epoch": 0.20228959617561956,
      "grad_norm": 0.3213135600090027,
      "learning_rate": 9.937309365446973e-06,
      "loss": 0.5277,
      "step": 402
    },
    {
      "epoch": 0.2027928041263052,
      "grad_norm": 0.22786889970302582,
      "learning_rate": 9.742367571857091e-06,
      "loss": 0.595,
      "step": 403
    },
    {
      "epoch": 0.20329601207699083,
      "grad_norm": 0.23553386330604553,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.5259,
      "step": 404
    },
    {
      "epoch": 0.20379922002767645,
      "grad_norm": 0.21647228300571442,
      "learning_rate": 9.357665770419244e-06,
      "loss": 0.4914,
      "step": 405
    },
    {
      "epoch": 0.20430242797836207,
      "grad_norm": 0.24146980047225952,
      "learning_rate": 9.167922241916055e-06,
      "loss": 0.5754,
      "step": 406
    },
    {
      "epoch": 0.20480563592904769,
      "grad_norm": 0.25054699182510376,
      "learning_rate": 8.97992782372432e-06,
      "loss": 0.6017,
      "step": 407
    },
    {
      "epoch": 0.2053088438797333,
      "grad_norm": 0.3012691140174866,
      "learning_rate": 8.793690568899216e-06,
      "loss": 0.7198,
      "step": 408
    },
    {
      "epoch": 0.20581205183041892,
      "grad_norm": 0.3064156174659729,
      "learning_rate": 8.609218455224893e-06,
      "loss": 0.7218,
      "step": 409
    },
    {
      "epoch": 0.20631525978110454,
      "grad_norm": 0.3092438578605652,
      "learning_rate": 8.426519384872733e-06,
      "loss": 0.7895,
      "step": 410
    },
    {
      "epoch": 0.20681846773179016,
      "grad_norm": 0.3434111475944519,
      "learning_rate": 8.245601184062852e-06,
      "loss": 0.7085,
      "step": 411
    },
    {
      "epoch": 0.20732167568247578,
      "grad_norm": 0.37145957350730896,
      "learning_rate": 8.066471602728803e-06,
      "loss": 0.7766,
      "step": 412
    },
    {
      "epoch": 0.2078248836331614,
      "grad_norm": 0.32149067521095276,
      "learning_rate": 7.889138314185678e-06,
      "loss": 0.6546,
      "step": 413
    },
    {
      "epoch": 0.20832809158384702,
      "grad_norm": 0.33787834644317627,
      "learning_rate": 7.71360891480134e-06,
      "loss": 0.7035,
      "step": 414
    },
    {
      "epoch": 0.20883129953453264,
      "grad_norm": 0.35734379291534424,
      "learning_rate": 7.539890923671062e-06,
      "loss": 0.7837,
      "step": 415
    },
    {
      "epoch": 0.20933450748521826,
      "grad_norm": 0.40482255816459656,
      "learning_rate": 7.367991782295391e-06,
      "loss": 0.7856,
      "step": 416
    },
    {
      "epoch": 0.20983771543590388,
      "grad_norm": 0.35036709904670715,
      "learning_rate": 7.197918854261432e-06,
      "loss": 0.7238,
      "step": 417
    },
    {
      "epoch": 0.2103409233865895,
      "grad_norm": 0.33897969126701355,
      "learning_rate": 7.029679424927365e-06,
      "loss": 0.7213,
      "step": 418
    },
    {
      "epoch": 0.21084413133727514,
      "grad_norm": 0.4156653881072998,
      "learning_rate": 6.863280701110408e-06,
      "loss": 0.8968,
      "step": 419
    },
    {
      "epoch": 0.21134733928796076,
      "grad_norm": 0.4062703251838684,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.8871,
      "step": 420
    },
    {
      "epoch": 0.21185054723864638,
      "grad_norm": 0.3825936019420624,
      "learning_rate": 6.536033802742813e-06,
      "loss": 0.8309,
      "step": 421
    },
    {
      "epoch": 0.212353755189332,
      "grad_norm": 0.3812710642814636,
      "learning_rate": 6.375199646360142e-06,
      "loss": 0.7463,
      "step": 422
    },
    {
      "epoch": 0.21285696314001762,
      "grad_norm": 0.39211493730545044,
      "learning_rate": 6.216234231230012e-06,
      "loss": 0.8161,
      "step": 423
    },
    {
      "epoch": 0.21336017109070324,
      "grad_norm": 0.38219356536865234,
      "learning_rate": 6.059144366901736e-06,
      "loss": 0.7659,
      "step": 424
    },
    {
      "epoch": 0.21386337904138886,
      "grad_norm": 0.4277362823486328,
      "learning_rate": 5.903936782582253e-06,
      "loss": 0.7985,
      "step": 425
    },
    {
      "epoch": 0.21436658699207448,
      "grad_norm": 0.43157392740249634,
      "learning_rate": 5.750618126847912e-06,
      "loss": 0.9045,
      "step": 426
    },
    {
      "epoch": 0.2148697949427601,
      "grad_norm": 0.3815480172634125,
      "learning_rate": 5.599194967359639e-06,
      "loss": 0.6918,
      "step": 427
    },
    {
      "epoch": 0.21537300289344571,
      "grad_norm": 0.43258634209632874,
      "learning_rate": 5.449673790581611e-06,
      "loss": 0.8096,
      "step": 428
    },
    {
      "epoch": 0.21587621084413133,
      "grad_norm": 0.4353589415550232,
      "learning_rate": 5.302061001503394e-06,
      "loss": 0.8579,
      "step": 429
    },
    {
      "epoch": 0.21637941879481695,
      "grad_norm": 0.4448375999927521,
      "learning_rate": 5.156362923365588e-06,
      "loss": 0.7577,
      "step": 430
    },
    {
      "epoch": 0.21688262674550257,
      "grad_norm": 0.5607813000679016,
      "learning_rate": 5.012585797388936e-06,
      "loss": 0.9384,
      "step": 431
    },
    {
      "epoch": 0.2173858346961882,
      "grad_norm": 0.514840304851532,
      "learning_rate": 4.87073578250698e-06,
      "loss": 0.9132,
      "step": 432
    },
    {
      "epoch": 0.2178890426468738,
      "grad_norm": 0.4784408211708069,
      "learning_rate": 4.730818955102234e-06,
      "loss": 0.7861,
      "step": 433
    },
    {
      "epoch": 0.21839225059755943,
      "grad_norm": 0.5339136123657227,
      "learning_rate": 4.592841308745932e-06,
      "loss": 0.8646,
      "step": 434
    },
    {
      "epoch": 0.21889545854824508,
      "grad_norm": 0.5781549215316772,
      "learning_rate": 4.456808753941205e-06,
      "loss": 0.8948,
      "step": 435
    },
    {
      "epoch": 0.2193986664989307,
      "grad_norm": 0.5680283308029175,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.8141,
      "step": 436
    },
    {
      "epoch": 0.2199018744496163,
      "grad_norm": 0.5979980826377869,
      "learning_rate": 4.190602144143207e-06,
      "loss": 0.8601,
      "step": 437
    },
    {
      "epoch": 0.22040508240030193,
      "grad_norm": 0.60759037733078,
      "learning_rate": 4.06043949255509e-06,
      "loss": 0.7886,
      "step": 438
    },
    {
      "epoch": 0.22090829035098755,
      "grad_norm": 0.5867761373519897,
      "learning_rate": 3.932244738840379e-06,
      "loss": 0.785,
      "step": 439
    },
    {
      "epoch": 0.22141149830167317,
      "grad_norm": 0.6567158699035645,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.7419,
      "step": 440
    },
    {
      "epoch": 0.2219147062523588,
      "grad_norm": 0.6796500086784363,
      "learning_rate": 3.681780806244095e-06,
      "loss": 0.8693,
      "step": 441
    },
    {
      "epoch": 0.2224179142030444,
      "grad_norm": 0.9281523823738098,
      "learning_rate": 3.5595223564037884e-06,
      "loss": 1.019,
      "step": 442
    },
    {
      "epoch": 0.22292112215373003,
      "grad_norm": 0.8401932120323181,
      "learning_rate": 3.4392532620598216e-06,
      "loss": 1.0439,
      "step": 443
    },
    {
      "epoch": 0.22342433010441565,
      "grad_norm": 0.8805630207061768,
      "learning_rate": 3.3209786751399187e-06,
      "loss": 1.0998,
      "step": 444
    },
    {
      "epoch": 0.22392753805510127,
      "grad_norm": 0.9491791725158691,
      "learning_rate": 3.2047036621337236e-06,
      "loss": 1.0169,
      "step": 445
    },
    {
      "epoch": 0.22443074600578689,
      "grad_norm": 0.8668059706687927,
      "learning_rate": 3.0904332038757977e-06,
      "loss": 0.8529,
      "step": 446
    },
    {
      "epoch": 0.2249339539564725,
      "grad_norm": 0.9326903223991394,
      "learning_rate": 2.978172195332263e-06,
      "loss": 1.1405,
      "step": 447
    },
    {
      "epoch": 0.22543716190715812,
      "grad_norm": 1.0572550296783447,
      "learning_rate": 2.8679254453910785e-06,
      "loss": 0.9234,
      "step": 448
    },
    {
      "epoch": 0.22594036985784374,
      "grad_norm": 1.223471760749817,
      "learning_rate": 2.759697676656098e-06,
      "loss": 0.8797,
      "step": 449
    },
    {
      "epoch": 0.22644357780852936,
      "grad_norm": 1.5660356283187866,
      "learning_rate": 2.653493525244721e-06,
      "loss": 0.8981,
      "step": 450
    },
    {
      "epoch": 0.226946785759215,
      "grad_norm": 0.23308148980140686,
      "learning_rate": 2.549317540589308e-06,
      "loss": 0.4266,
      "step": 451
    },
    {
      "epoch": 0.22744999370990063,
      "grad_norm": 0.25011494755744934,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.4767,
      "step": 452
    },
    {
      "epoch": 0.22795320166058625,
      "grad_norm": 0.28921380639076233,
      "learning_rate": 2.3470678346851518e-06,
      "loss": 0.6835,
      "step": 453
    },
    {
      "epoch": 0.22845640961127187,
      "grad_norm": 0.23749534785747528,
      "learning_rate": 2.2490027771406687e-06,
      "loss": 0.6596,
      "step": 454
    },
    {
      "epoch": 0.22895961756195748,
      "grad_norm": 0.2435358464717865,
      "learning_rate": 2.152983213389559e-06,
      "loss": 0.585,
      "step": 455
    },
    {
      "epoch": 0.2294628255126431,
      "grad_norm": 0.26918700337409973,
      "learning_rate": 2.0590132565903476e-06,
      "loss": 0.5973,
      "step": 456
    },
    {
      "epoch": 0.22996603346332872,
      "grad_norm": 0.2611250877380371,
      "learning_rate": 1.9670969321032407e-06,
      "loss": 0.59,
      "step": 457
    },
    {
      "epoch": 0.23046924141401434,
      "grad_norm": 0.26343515515327454,
      "learning_rate": 1.8772381773176417e-06,
      "loss": 0.5985,
      "step": 458
    },
    {
      "epoch": 0.23097244936469996,
      "grad_norm": 0.27780017256736755,
      "learning_rate": 1.7894408414835362e-06,
      "loss": 0.6332,
      "step": 459
    },
    {
      "epoch": 0.23147565731538558,
      "grad_norm": 0.3137127161026001,
      "learning_rate": 1.70370868554659e-06,
      "loss": 0.6877,
      "step": 460
    },
    {
      "epoch": 0.2319788652660712,
      "grad_norm": 0.2799319624900818,
      "learning_rate": 1.620045381987012e-06,
      "loss": 0.6106,
      "step": 461
    },
    {
      "epoch": 0.23248207321675682,
      "grad_norm": 0.33448776602745056,
      "learning_rate": 1.5384545146622852e-06,
      "loss": 0.8003,
      "step": 462
    },
    {
      "epoch": 0.23298528116744244,
      "grad_norm": 0.3402840793132782,
      "learning_rate": 1.4589395786535953e-06,
      "loss": 0.7416,
      "step": 463
    },
    {
      "epoch": 0.23348848911812806,
      "grad_norm": 0.3084582984447479,
      "learning_rate": 1.3815039801161721e-06,
      "loss": 0.7153,
      "step": 464
    },
    {
      "epoch": 0.23399169706881368,
      "grad_norm": 0.348308265209198,
      "learning_rate": 1.3061510361333185e-06,
      "loss": 0.8734,
      "step": 465
    },
    {
      "epoch": 0.2344949050194993,
      "grad_norm": 0.35534948110580444,
      "learning_rate": 1.232883974574367e-06,
      "loss": 0.8548,
      "step": 466
    },
    {
      "epoch": 0.23499811297018494,
      "grad_norm": 0.35696539282798767,
      "learning_rate": 1.1617059339563807e-06,
      "loss": 0.8498,
      "step": 467
    },
    {
      "epoch": 0.23550132092087056,
      "grad_norm": 0.4020044207572937,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.818,
      "step": 468
    },
    {
      "epoch": 0.23600452887155618,
      "grad_norm": 0.38172829151153564,
      "learning_rate": 1.0256290220474307e-06,
      "loss": 0.8157,
      "step": 469
    },
    {
      "epoch": 0.2365077368222418,
      "grad_norm": 0.3517870604991913,
      "learning_rate": 9.607359798384785e-07,
      "loss": 0.7573,
      "step": 470
    },
    {
      "epoch": 0.23701094477292742,
      "grad_norm": 0.3962482213973999,
      "learning_rate": 8.979436164848088e-07,
      "loss": 0.7767,
      "step": 471
    },
    {
      "epoch": 0.23751415272361304,
      "grad_norm": 0.42771947383880615,
      "learning_rate": 8.372546218022747e-07,
      "loss": 0.9145,
      "step": 472
    },
    {
      "epoch": 0.23801736067429866,
      "grad_norm": 0.3869607746601105,
      "learning_rate": 7.786715955054203e-07,
      "loss": 0.8365,
      "step": 473
    },
    {
      "epoch": 0.23852056862498427,
      "grad_norm": 0.3934659957885742,
      "learning_rate": 7.221970470961125e-07,
      "loss": 0.8322,
      "step": 474
    },
    {
      "epoch": 0.2390237765756699,
      "grad_norm": 0.3955768048763275,
      "learning_rate": 6.678333957560512e-07,
      "loss": 0.8821,
      "step": 475
    },
    {
      "epoch": 0.2395269845263555,
      "grad_norm": 0.4069364368915558,
      "learning_rate": 6.15582970243117e-07,
      "loss": 0.8928,
      "step": 476
    },
    {
      "epoch": 0.24003019247704113,
      "grad_norm": 0.47845911979675293,
      "learning_rate": 5.654480087916303e-07,
      "loss": 0.8978,
      "step": 477
    },
    {
      "epoch": 0.24053340042772675,
      "grad_norm": 0.46929192543029785,
      "learning_rate": 5.174306590164879e-07,
      "loss": 0.887,
      "step": 478
    },
    {
      "epoch": 0.24103660837841237,
      "grad_norm": 0.42585262656211853,
      "learning_rate": 4.715329778211375e-07,
      "loss": 0.7955,
      "step": 479
    },
    {
      "epoch": 0.241539816329098,
      "grad_norm": 0.47433245182037354,
      "learning_rate": 4.277569313094809e-07,
      "loss": 0.8784,
      "step": 480
    },
    {
      "epoch": 0.2420430242797836,
      "grad_norm": 0.4917643368244171,
      "learning_rate": 3.8610439470164737e-07,
      "loss": 0.8199,
      "step": 481
    },
    {
      "epoch": 0.24254623223046926,
      "grad_norm": 0.5189266204833984,
      "learning_rate": 3.465771522536854e-07,
      "loss": 0.8448,
      "step": 482
    },
    {
      "epoch": 0.24304944018115487,
      "grad_norm": 0.5650250315666199,
      "learning_rate": 3.09176897181096e-07,
      "loss": 0.8662,
      "step": 483
    },
    {
      "epoch": 0.2435526481318405,
      "grad_norm": 0.5354907512664795,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 0.8206,
      "step": 484
    },
    {
      "epoch": 0.2440558560825261,
      "grad_norm": 0.526403546333313,
      "learning_rate": 2.407636663901591e-07,
      "loss": 0.7682,
      "step": 485
    },
    {
      "epoch": 0.24455906403321173,
      "grad_norm": 0.5894873738288879,
      "learning_rate": 2.0975362126691712e-07,
      "loss": 0.8226,
      "step": 486
    },
    {
      "epoch": 0.24506227198389735,
      "grad_norm": 0.6889277696609497,
      "learning_rate": 1.8087642458373134e-07,
      "loss": 0.8529,
      "step": 487
    },
    {
      "epoch": 0.24556547993458297,
      "grad_norm": 0.6041770577430725,
      "learning_rate": 1.5413331334360182e-07,
      "loss": 0.839,
      "step": 488
    },
    {
      "epoch": 0.2460686878852686,
      "grad_norm": 0.6730844974517822,
      "learning_rate": 1.2952543313240472e-07,
      "loss": 0.9139,
      "step": 489
    },
    {
      "epoch": 0.2465718958359542,
      "grad_norm": 0.5944534540176392,
      "learning_rate": 1.0705383806982606e-07,
      "loss": 0.7731,
      "step": 490
    },
    {
      "epoch": 0.24707510378663983,
      "grad_norm": 0.6421075463294983,
      "learning_rate": 8.671949076420882e-08,
      "loss": 0.8907,
      "step": 491
    },
    {
      "epoch": 0.24757831173732545,
      "grad_norm": 0.704002857208252,
      "learning_rate": 6.852326227130834e-08,
      "loss": 0.8232,
      "step": 492
    },
    {
      "epoch": 0.24808151968801107,
      "grad_norm": 0.8100038766860962,
      "learning_rate": 5.246593205699424e-08,
      "loss": 0.8716,
      "step": 493
    },
    {
      "epoch": 0.24858472763869668,
      "grad_norm": 0.7674500942230225,
      "learning_rate": 3.8548187963854956e-08,
      "loss": 0.8986,
      "step": 494
    },
    {
      "epoch": 0.2490879355893823,
      "grad_norm": 0.7667043209075928,
      "learning_rate": 2.6770626181715773e-08,
      "loss": 1.008,
      "step": 495
    },
    {
      "epoch": 0.24959114354006792,
      "grad_norm": 0.891828179359436,
      "learning_rate": 1.7133751222137007e-08,
      "loss": 0.9345,
      "step": 496
    },
    {
      "epoch": 0.25009435149075354,
      "grad_norm": 0.8518345952033997,
      "learning_rate": 9.637975896759077e-09,
      "loss": 0.968,
      "step": 497
    },
    {
      "epoch": 0.2505975594414392,
      "grad_norm": 1.0205941200256348,
      "learning_rate": 4.2836212996499865e-09,
      "loss": 0.8592,
      "step": 498
    },
    {
      "epoch": 0.2511007673921248,
      "grad_norm": 1.1440129280090332,
      "learning_rate": 1.0709167935385455e-09,
      "loss": 0.9679,
      "step": 499
    },
    {
      "epoch": 0.2516039753428104,
      "grad_norm": 1.352176547050476,
      "learning_rate": 0.0,
      "loss": 0.8527,
      "step": 500
    },
    {
      "epoch": 0.2516039753428104,
      "eval_loss": 0.8061292767524719,
      "eval_runtime": 104.5326,
      "eval_samples_per_second": 32.019,
      "eval_steps_per_second": 8.007,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4280794742718464e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}