diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6656681644200366,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006656681644200366,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 1.9001,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013313363288400732,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 1.1836,
+      "step": 2
+    },
+    {
+      "epoch": 0.00199700449326011,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0,
+      "loss": 2.0432,
+      "step": 3
+    },
+    {
+      "epoch": 0.0026626726576801465,
+      "grad_norm": Infinity,
+      "learning_rate": 0.0,
+      "loss": 2.7364,
+      "step": 4
+    },
+    {
+      "epoch": 0.003328340822100183,
+      "grad_norm": 305.05010986328125,
+      "learning_rate": 0.0001,
+      "loss": 2.7278,
+      "step": 5
+    },
+    {
+      "epoch": 0.00399400898652022,
+      "grad_norm": 195.2991180419922,
+      "learning_rate": 0.0002,
+      "loss": 1.5935,
+      "step": 6
+    },
+    {
+      "epoch": 0.004659677150940256,
+      "grad_norm": 135.4779815673828,
+      "learning_rate": 0.00019995559502664298,
+      "loss": 1.558,
+      "step": 7
+    },
+    {
+      "epoch": 0.005325345315360293,
+      "grad_norm": 175.23486328125,
+      "learning_rate": 0.00019991119005328598,
+      "loss": 1.8278,
+      "step": 8
+    },
+    {
+      "epoch": 0.005991013479780329,
+      "grad_norm": 70.18768310546875,
+      "learning_rate": 0.00019986678507992895,
+      "loss": 1.2821,
+      "step": 9
+    },
+    {
+      "epoch": 0.006656681644200366,
+      "grad_norm": 91.9547348022461,
+      "learning_rate": 0.00019982238010657195,
+      "loss": 1.3009,
+      "step": 10
+    },
+    {
+      "epoch": 0.007322349808620403,
+      "grad_norm": 140.80667114257812,
+      "learning_rate": 0.00019977797513321492,
+      "loss": 1.4739,
+      "step": 11
+    },
+    {
+      "epoch": 0.00798801797304044,
+      "grad_norm": 117.13510131835938,
+      "learning_rate": 0.00019973357015985792,
+      "loss": 1.0922,
+      "step": 12
+    },
+    {
+      "epoch": 0.008653686137460476,
+      "grad_norm": 87.54082489013672,
+      "learning_rate": 0.0001996891651865009,
+      "loss": 1.6633,
+      "step": 13
+    },
+    {
+      "epoch": 0.009319354301880512,
+      "grad_norm": 178.7975616455078,
+      "learning_rate": 0.0001996447602131439,
+      "loss": 1.7614,
+      "step": 14
+    },
+    {
+      "epoch": 0.00998502246630055,
+      "grad_norm": 199.108154296875,
+      "learning_rate": 0.00019960035523978686,
+      "loss": 1.4873,
+      "step": 15
+    },
+    {
+      "epoch": 0.010650690630720586,
+      "grad_norm": 97.36341857910156,
+      "learning_rate": 0.00019955595026642986,
+      "loss": 1.2921,
+      "step": 16
+    },
+    {
+      "epoch": 0.011316358795140622,
+      "grad_norm": 145.5021209716797,
+      "learning_rate": 0.00019951154529307283,
+      "loss": 1.4667,
+      "step": 17
+    },
+    {
+      "epoch": 0.011982026959560658,
+      "grad_norm": 163.09628295898438,
+      "learning_rate": 0.00019946714031971583,
+      "loss": 1.4959,
+      "step": 18
+    },
+    {
+      "epoch": 0.012647695123980696,
+      "grad_norm": 181.53123474121094,
+      "learning_rate": 0.0001994227353463588,
+      "loss": 1.4931,
+      "step": 19
+    },
+    {
+      "epoch": 0.013313363288400732,
+      "grad_norm": 134.01087951660156,
+      "learning_rate": 0.00019937833037300177,
+      "loss": 1.1797,
+      "step": 20
+    },
+    {
+      "epoch": 0.013979031452820768,
+      "grad_norm": 138.4213409423828,
+      "learning_rate": 0.00019933392539964477,
+      "loss": 1.3021,
+      "step": 21
+    },
+    {
+      "epoch": 0.014644699617240806,
+      "grad_norm": 94.12645721435547,
+      "learning_rate": 0.00019928952042628774,
+      "loss": 1.2099,
+      "step": 22
+    },
+    {
+      "epoch": 0.015310367781660842,
+      "grad_norm": 68.36018371582031,
+      "learning_rate": 0.00019924511545293074,
+      "loss": 1.0593,
+      "step": 23
+    },
+    {
+      "epoch": 0.01597603594608088,
+      "grad_norm": 83.23944854736328,
+      "learning_rate": 0.0001992007104795737,
+      "loss": 1.1152,
+      "step": 24
+    },
+    {
+      "epoch": 0.016641704110500914,
+      "grad_norm": 145.29953002929688,
+      "learning_rate": 0.0001991563055062167,
+      "loss": 1.194,
+      "step": 25
+    },
+    {
+      "epoch": 0.017307372274920952,
+      "grad_norm": 71.01487731933594,
+      "learning_rate": 0.00019911190053285968,
+      "loss": 1.0201,
+      "step": 26
+    },
+    {
+      "epoch": 0.01797304043934099,
+      "grad_norm": 82.99989318847656,
+      "learning_rate": 0.00019906749555950268,
+      "loss": 1.3319,
+      "step": 27
+    },
+    {
+      "epoch": 0.018638708603761024,
+      "grad_norm": 51.65462112426758,
+      "learning_rate": 0.00019902309058614565,
+      "loss": 0.8622,
+      "step": 28
+    },
+    {
+      "epoch": 0.019304376768181062,
+      "grad_norm": 106.69747161865234,
+      "learning_rate": 0.00019897868561278865,
+      "loss": 1.0051,
+      "step": 29
+    },
+    {
+      "epoch": 0.0199700449326011,
+      "grad_norm": 85.53784942626953,
+      "learning_rate": 0.00019893428063943162,
+      "loss": 0.7687,
+      "step": 30
+    },
+    {
+      "epoch": 0.020635713097021134,
+      "grad_norm": 141.33482360839844,
+      "learning_rate": 0.00019888987566607462,
+      "loss": 0.9697,
+      "step": 31
+    },
+    {
+      "epoch": 0.021301381261441172,
+      "grad_norm": 110.15571594238281,
+      "learning_rate": 0.0001988454706927176,
+      "loss": 0.6955,
+      "step": 32
+    },
+    {
+      "epoch": 0.02196704942586121,
+      "grad_norm": 123.28990173339844,
+      "learning_rate": 0.00019880106571936056,
+      "loss": 0.9405,
+      "step": 33
+    },
+    {
+      "epoch": 0.022632717590281244,
+      "grad_norm": 104.15376281738281,
+      "learning_rate": 0.00019875666074600356,
+      "loss": 1.1353,
+      "step": 34
+    },
+    {
+      "epoch": 0.023298385754701282,
+      "grad_norm": 122.83297729492188,
+      "learning_rate": 0.00019871225577264653,
+      "loss": 0.9912,
+      "step": 35
+    },
+    {
+      "epoch": 0.023964053919121316,
+      "grad_norm": 137.1550750732422,
+      "learning_rate": 0.00019866785079928953,
+      "loss": 1.0156,
+      "step": 36
+    },
+    {
+      "epoch": 0.024629722083541354,
+      "grad_norm": 95.7591781616211,
+      "learning_rate": 0.0001986234458259325,
+      "loss": 0.9295,
+      "step": 37
+    },
+    {
+      "epoch": 0.025295390247961392,
+      "grad_norm": 163.35743713378906,
+      "learning_rate": 0.0001985790408525755,
+      "loss": 0.9697,
+      "step": 38
+    },
+    {
+      "epoch": 0.025961058412381426,
+      "grad_norm": 42.57448959350586,
+      "learning_rate": 0.00019853463587921847,
+      "loss": 0.8329,
+      "step": 39
+    },
+    {
+      "epoch": 0.026626726576801464,
+      "grad_norm": 161.05844116210938,
+      "learning_rate": 0.00019849023090586147,
+      "loss": 0.8588,
+      "step": 40
+    },
+    {
+      "epoch": 0.027292394741221502,
+      "grad_norm": 123.78165435791016,
+      "learning_rate": 0.00019844582593250444,
+      "loss": 0.8082,
+      "step": 41
+    },
+    {
+      "epoch": 0.027958062905641536,
+      "grad_norm": 117.62382507324219,
+      "learning_rate": 0.00019840142095914744,
+      "loss": 1.0692,
+      "step": 42
+    },
+    {
+      "epoch": 0.028623731070061574,
+      "grad_norm": NaN,
+      "learning_rate": 0.00019840142095914744,
+      "loss": 1.316,
+      "step": 43
+    },
+    {
+      "epoch": 0.029289399234481612,
+      "grad_norm": 63.35919952392578,
+      "learning_rate": 0.0001983570159857904,
+      "loss": 0.6028,
+      "step": 44
+    },
+    {
+      "epoch": 0.029955067398901646,
+      "grad_norm": 277.1460266113281,
+      "learning_rate": 0.00019831261101243338,
+      "loss": 1.4349,
+      "step": 45
+    },
+    {
+      "epoch": 0.030620735563321684,
+      "grad_norm": NaN,
+      "learning_rate": 0.00019831261101243338,
+      "loss": 1.816,
+      "step": 46
+    },
+    {
+      "epoch": 0.03128640372774172,
+      "grad_norm": 206.0849151611328,
+      "learning_rate": 0.00019826820603907638,
+      "loss": 1.7035,
+      "step": 47
+    },
+    {
+      "epoch": 0.03195207189216176,
+      "grad_norm": 154.10183715820312,
+      "learning_rate": 0.00019822380106571935,
+      "loss": 0.7897,
+      "step": 48
+    },
+    {
+      "epoch": 0.032617740056581794,
+      "grad_norm": 182.32650756835938,
+      "learning_rate": 0.00019817939609236235,
+      "loss": 1.2611,
+      "step": 49
+    },
+    {
+      "epoch": 0.03328340822100183,
+      "grad_norm": 100.30379486083984,
+      "learning_rate": 0.00019813499111900532,
+      "loss": 0.8746,
+      "step": 50
+    },
+    {
+      "epoch": 0.03394907638542187,
+      "grad_norm": 99.05520629882812,
+      "learning_rate": 0.00019809058614564832,
+      "loss": 0.6544,
+      "step": 51
+    },
+    {
+      "epoch": 0.034614744549841904,
+      "grad_norm": 75.6047592163086,
+      "learning_rate": 0.0001980461811722913,
+      "loss": 0.5308,
+      "step": 52
+    },
+    {
+      "epoch": 0.03528041271426194,
+      "grad_norm": 81.81549835205078,
+      "learning_rate": 0.0001980017761989343,
+      "loss": 1.0103,
+      "step": 53
+    },
+    {
+      "epoch": 0.03594608087868198,
+      "grad_norm": 305.5655517578125,
+      "learning_rate": 0.00019795737122557726,
+      "loss": 2.0811,
+      "step": 54
+    },
+    {
+      "epoch": 0.036611749043102014,
+      "grad_norm": 254.49461364746094,
+      "learning_rate": 0.00019791296625222026,
+      "loss": 1.5316,
+      "step": 55
+    },
+    {
+      "epoch": 0.03727741720752205,
+      "grad_norm": 249.2161407470703,
+      "learning_rate": 0.00019786856127886323,
+      "loss": 1.4851,
+      "step": 56
+    },
+    {
+      "epoch": 0.03794308537194209,
+      "grad_norm": 187.69805908203125,
+      "learning_rate": 0.00019782415630550623,
+      "loss": 1.0635,
+      "step": 57
+    },
+    {
+      "epoch": 0.038608753536362124,
+      "grad_norm": 135.42141723632812,
+      "learning_rate": 0.0001977797513321492,
+      "loss": 0.8954,
+      "step": 58
+    },
+    {
+      "epoch": 0.03927442170078216,
+      "grad_norm": 198.7463836669922,
+      "learning_rate": 0.0001977353463587922,
+      "loss": 1.2765,
+      "step": 59
+    },
+    {
+      "epoch": 0.0399400898652022,
+      "grad_norm": 256.67279052734375,
+      "learning_rate": 0.00019769094138543517,
+      "loss": 1.4927,
+      "step": 60
+    },
+    {
+      "epoch": 0.040605758029622234,
+      "grad_norm": 64.07623291015625,
+      "learning_rate": 0.00019764653641207817,
+      "loss": 0.7874,
+      "step": 61
+    },
+    {
+      "epoch": 0.04127142619404227,
+      "grad_norm": 147.7261962890625,
+      "learning_rate": 0.00019760213143872114,
+      "loss": 0.7924,
+      "step": 62
+    },
+    {
+      "epoch": 0.04193709435846231,
+      "grad_norm": 122.39730072021484,
+      "learning_rate": 0.00019755772646536413,
+      "loss": 0.7656,
+      "step": 63
+    },
+    {
+      "epoch": 0.042602762522882344,
+      "grad_norm": 94.13729095458984,
+      "learning_rate": 0.0001975133214920071,
+      "loss": 0.9051,
+      "step": 64
+    },
+    {
+      "epoch": 0.04326843068730238,
+      "grad_norm": 48.220184326171875,
+      "learning_rate": 0.0001974689165186501,
+      "loss": 0.6262,
+      "step": 65
+    },
+    {
+      "epoch": 0.04393409885172242,
+      "grad_norm": 32.33949661254883,
+      "learning_rate": 0.00019742451154529308,
+      "loss": 0.544,
+      "step": 66
+    },
+    {
+      "epoch": 0.044599767016142454,
+      "grad_norm": 133.78097534179688,
+      "learning_rate": 0.00019738010657193607,
+      "loss": 0.6472,
+      "step": 67
+    },
+    {
+      "epoch": 0.04526543518056249,
+      "grad_norm": 105.52882385253906,
+      "learning_rate": 0.00019733570159857907,
+      "loss": 0.7272,
+      "step": 68
+    },
+    {
+      "epoch": 0.04593110334498253,
+      "grad_norm": 40.91939926147461,
+      "learning_rate": 0.00019729129662522204,
+      "loss": 0.6232,
+      "step": 69
+    },
+    {
+      "epoch": 0.046596771509402564,
+      "grad_norm": 48.693763732910156,
+      "learning_rate": 0.00019724689165186504,
+      "loss": 0.3698,
+      "step": 70
+    },
+    {
+      "epoch": 0.0472624396738226,
+      "grad_norm": 98.95121765136719,
+      "learning_rate": 0.00019720248667850801,
+      "loss": 0.4914,
+      "step": 71
+    },
+    {
+      "epoch": 0.04792810783824263,
+      "grad_norm": 217.00401306152344,
+      "learning_rate": 0.00019715808170515098,
+      "loss": 1.3837,
+      "step": 72
+    },
+    {
+      "epoch": 0.048593776002662674,
+      "grad_norm": 97.31861877441406,
+      "learning_rate": 0.00019711367673179398,
+      "loss": 0.7037,
+      "step": 73
+    },
+    {
+      "epoch": 0.04925944416708271,
+      "grad_norm": 101.28430938720703,
+      "learning_rate": 0.00019706927175843695,
+      "loss": 1.3999,
+      "step": 74
+    },
+    {
+      "epoch": 0.04992511233150274,
+      "grad_norm": 86.35800170898438,
+      "learning_rate": 0.00019702486678507995,
+      "loss": 0.8045,
+      "step": 75
+    },
+    {
+      "epoch": 0.050590780495922784,
+      "grad_norm": 102.23155212402344,
+      "learning_rate": 0.00019698046181172292,
+      "loss": 0.7521,
+      "step": 76
+    },
+    {
+      "epoch": 0.05125644866034282,
+      "grad_norm": 125.80261993408203,
+      "learning_rate": 0.00019693605683836592,
+      "loss": 0.7933,
+      "step": 77
+    },
+    {
+      "epoch": 0.05192211682476285,
+      "grad_norm": 160.28123474121094,
+      "learning_rate": 0.0001968916518650089,
+      "loss": 0.6355,
+      "step": 78
+    },
+    {
+      "epoch": 0.052587784989182894,
+      "grad_norm": 42.919464111328125,
+      "learning_rate": 0.0001968472468916519,
+      "loss": 0.2693,
+      "step": 79
+    },
+    {
+      "epoch": 0.05325345315360293,
+      "grad_norm": 56.61810302734375,
+      "learning_rate": 0.00019680284191829486,
+      "loss": 0.3141,
+      "step": 80
+    },
+    {
+      "epoch": 0.05391912131802296,
+      "grad_norm": 299.3166198730469,
+      "learning_rate": 0.00019675843694493786,
+      "loss": 1.7119,
+      "step": 81
+    },
+    {
+      "epoch": 0.054584789482443004,
+      "grad_norm": 133.17459106445312,
+      "learning_rate": 0.00019671403197158083,
+      "loss": 1.2885,
+      "step": 82
+    },
+    {
+      "epoch": 0.05525045764686304,
+      "grad_norm": 131.6356964111328,
+      "learning_rate": 0.00019666962699822383,
+      "loss": 0.7976,
+      "step": 83
+    },
+    {
+      "epoch": 0.05591612581128307,
+      "grad_norm": 122.43846130371094,
+      "learning_rate": 0.0001966252220248668,
+      "loss": 0.7031,
+      "step": 84
+    },
+    {
+      "epoch": 0.056581793975703114,
+      "grad_norm": 51.365901947021484,
+      "learning_rate": 0.00019658081705150977,
+      "loss": 0.6684,
+      "step": 85
+    },
+    {
+      "epoch": 0.05724746214012315,
+      "grad_norm": 251.70664978027344,
+      "learning_rate": 0.00019653641207815277,
+      "loss": 1.2329,
+      "step": 86
+    },
+    {
+      "epoch": 0.05791313030454318,
+      "grad_norm": 94.94196319580078,
+      "learning_rate": 0.00019649200710479574,
+      "loss": 0.746,
+      "step": 87
+    },
+    {
+      "epoch": 0.058578798468963224,
+      "grad_norm": 106.89917755126953,
+      "learning_rate": 0.00019644760213143874,
+      "loss": 0.8735,
+      "step": 88
+    },
+    {
+      "epoch": 0.05924446663338326,
+      "grad_norm": 130.65577697753906,
+      "learning_rate": 0.00019640319715808171,
+      "loss": 1.0905,
+      "step": 89
+    },
+    {
+      "epoch": 0.05991013479780329,
+      "grad_norm": 37.6524543762207,
+      "learning_rate": 0.0001963587921847247,
+      "loss": 0.447,
+      "step": 90
+    },
+    {
+      "epoch": 0.060575802962223334,
+      "grad_norm": 55.81675720214844,
+      "learning_rate": 0.00019631438721136768,
+      "loss": 0.8305,
+      "step": 91
+    },
+    {
+      "epoch": 0.06124147112664337,
+      "grad_norm": 56.32381057739258,
+      "learning_rate": 0.00019626998223801068,
+      "loss": 0.6437,
+      "step": 92
+    },
+    {
+      "epoch": 0.0619071392910634,
+      "grad_norm": 107.72248840332031,
+      "learning_rate": 0.00019622557726465365,
+      "loss": 0.7331,
+      "step": 93
+    },
+    {
+      "epoch": 0.06257280745548344,
+      "grad_norm": 60.965084075927734,
+      "learning_rate": 0.00019618117229129665,
+      "loss": 0.1975,
+      "step": 94
+    },
+    {
+      "epoch": 0.06323847561990348,
+      "grad_norm": 236.83071899414062,
+      "learning_rate": 0.00019613676731793962,
+      "loss": 1.3465,
+      "step": 95
+    },
+    {
+      "epoch": 0.06390414378432352,
+      "grad_norm": 32.12284851074219,
+      "learning_rate": 0.0001960923623445826,
+      "loss": 0.4818,
+      "step": 96
+    },
+    {
+      "epoch": 0.06456981194874355,
+      "grad_norm": 139.45631408691406,
+      "learning_rate": 0.0001960479573712256,
+      "loss": 0.7907,
+      "step": 97
+    },
+    {
+      "epoch": 0.06523548011316359,
+      "grad_norm": 85.41361999511719,
+      "learning_rate": 0.00019600355239786856,
+      "loss": 0.5648,
+      "step": 98
+    },
+    {
+      "epoch": 0.06590114827758363,
+      "grad_norm": 79.85189056396484,
+      "learning_rate": 0.00019595914742451156,
+      "loss": 0.7735,
+      "step": 99
+    },
+    {
+      "epoch": 0.06656681644200366,
+      "grad_norm": 64.68962860107422,
+      "learning_rate": 0.00019591474245115453,
+      "loss": 0.4128,
+      "step": 100
+    },
+    {
+      "epoch": 0.0672324846064237,
+      "grad_norm": 58.707054138183594,
+      "learning_rate": 0.00019587033747779753,
+      "loss": 0.8143,
+      "step": 101
+    },
+    {
+      "epoch": 0.06789815277084374,
+      "grad_norm": 107.83698272705078,
+      "learning_rate": 0.0001958259325044405,
+      "loss": 0.8662,
+      "step": 102
+    },
+    {
+      "epoch": 0.06856382093526377,
+      "grad_norm": 221.64857482910156,
+      "learning_rate": 0.0001957815275310835,
+      "loss": 1.2667,
+      "step": 103
+    },
+    {
+      "epoch": 0.06922948909968381,
+      "grad_norm": 75.43016815185547,
+      "learning_rate": 0.00019573712255772647,
+      "loss": 0.6056,
+      "step": 104
+    },
+    {
+      "epoch": 0.06989515726410385,
+      "grad_norm": 81.48358917236328,
+      "learning_rate": 0.00019569271758436947,
+      "loss": 0.4954,
+      "step": 105
+    },
+    {
+      "epoch": 0.07056082542852388,
+      "grad_norm": 100.88373565673828,
+      "learning_rate": 0.00019564831261101244,
+      "loss": 0.8129,
+      "step": 106
+    },
+    {
+      "epoch": 0.07122649359294392,
+      "grad_norm": 56.06926727294922,
+      "learning_rate": 0.00019560390763765544,
+      "loss": 0.5694,
+      "step": 107
+    },
+    {
+      "epoch": 0.07189216175736396,
+      "grad_norm": 119.89936828613281,
+      "learning_rate": 0.0001955595026642984,
+      "loss": 0.8492,
+      "step": 108
+    },
+    {
+      "epoch": 0.07255782992178399,
+      "grad_norm": 226.14344787597656,
+      "learning_rate": 0.00019551509769094138,
+      "loss": 1.5915,
+      "step": 109
+    },
+    {
+      "epoch": 0.07322349808620403,
+      "grad_norm": 167.6849365234375,
+      "learning_rate": 0.00019547069271758438,
+      "loss": 0.8734,
+      "step": 110
+    },
+    {
+      "epoch": 0.07388916625062407,
+      "grad_norm": 191.29388427734375,
+      "learning_rate": 0.00019542628774422735,
+      "loss": 1.1915,
+      "step": 111
+    },
+    {
+      "epoch": 0.0745548344150441,
+      "grad_norm": 74.29916381835938,
+      "learning_rate": 0.00019538188277087035,
+      "loss": 0.399,
+      "step": 112
+    },
+    {
+      "epoch": 0.07522050257946414,
+      "grad_norm": 193.20899963378906,
+      "learning_rate": 0.00019533747779751332,
+      "loss": 1.0812,
+      "step": 113
+    },
+    {
+      "epoch": 0.07588617074388418,
+      "grad_norm": 104.37834930419922,
+      "learning_rate": 0.00019529307282415632,
+      "loss": 0.7821,
+      "step": 114
+    },
+    {
+      "epoch": 0.0765518389083042,
+      "grad_norm": 63.434146881103516,
+      "learning_rate": 0.0001952486678507993,
+      "loss": 0.5636,
+      "step": 115
+    },
+    {
+      "epoch": 0.07721750707272425,
+      "grad_norm": 163.28578186035156,
+      "learning_rate": 0.0001952042628774423,
+      "loss": 0.8423,
+      "step": 116
+    },
+    {
+      "epoch": 0.07788317523714429,
+      "grad_norm": 94.21780395507812,
+      "learning_rate": 0.00019515985790408526,
+      "loss": 0.6704,
+      "step": 117
+    },
+    {
+      "epoch": 0.07854884340156432,
+      "grad_norm": 97.0186767578125,
+      "learning_rate": 0.00019511545293072826,
+      "loss": 0.4906,
+      "step": 118
+    },
+    {
+      "epoch": 0.07921451156598436,
+      "grad_norm": 67.47974395751953,
+      "learning_rate": 0.00019507104795737123,
+      "loss": 0.5328,
+      "step": 119
+    },
+    {
+      "epoch": 0.0798801797304044,
+      "grad_norm": 113.68756103515625,
+      "learning_rate": 0.00019502664298401423,
+      "loss": 0.7459,
+      "step": 120
+    },
+    {
+      "epoch": 0.08054584789482443,
+      "grad_norm": 71.1500244140625,
+      "learning_rate": 0.0001949822380106572,
+      "loss": 0.6177,
+      "step": 121
+    },
+    {
+      "epoch": 0.08121151605924447,
+      "grad_norm": 131.37168884277344,
+      "learning_rate": 0.00019493783303730017,
+      "loss": 0.7955,
+      "step": 122
+    },
+    {
+      "epoch": 0.08187718422366451,
+      "grad_norm": 72.7484359741211,
+      "learning_rate": 0.00019489342806394317,
+      "loss": 0.7026,
+      "step": 123
+    },
+    {
+      "epoch": 0.08254285238808454,
+      "grad_norm": 34.95873260498047,
+      "learning_rate": 0.00019484902309058614,
+      "loss": 0.3607,
+      "step": 124
+    },
+    {
+      "epoch": 0.08320852055250458,
+      "grad_norm": 23.322694778442383,
+      "learning_rate": 0.00019480461811722914,
+      "loss": 0.4757,
+      "step": 125
+    },
+    {
+      "epoch": 0.08387418871692462,
+      "grad_norm": 102.11334228515625,
+      "learning_rate": 0.0001947602131438721,
+      "loss": 0.9689,
+      "step": 126
+    },
+    {
+      "epoch": 0.08453985688134465,
+      "grad_norm": 70.20091247558594,
+      "learning_rate": 0.0001947158081705151,
+      "loss": 0.5523,
+      "step": 127
+    },
+    {
+      "epoch": 0.08520552504576469,
+      "grad_norm": 108.84575653076172,
+      "learning_rate": 0.00019467140319715808,
+      "loss": 0.7221,
+      "step": 128
+    },
+    {
+      "epoch": 0.08587119321018473,
+      "grad_norm": 74.28465270996094,
+      "learning_rate": 0.00019462699822380108,
+      "loss": 0.3408,
+      "step": 129
+    },
+    {
+      "epoch": 0.08653686137460476,
+      "grad_norm": 45.54611587524414,
+      "learning_rate": 0.00019458259325044405,
+      "loss": 0.3064,
+      "step": 130
+    },
+    {
+      "epoch": 0.0872025295390248,
+      "grad_norm": 107.5720443725586,
+      "learning_rate": 0.00019453818827708705,
+      "loss": 0.4285,
+      "step": 131
+    },
+    {
+      "epoch": 0.08786819770344484,
+      "grad_norm": 113.14404296875,
+      "learning_rate": 0.00019449378330373002,
+      "loss": 0.7893,
+      "step": 132
+    },
+    {
+      "epoch": 0.08853386586786487,
+      "grad_norm": 41.36758041381836,
+      "learning_rate": 0.00019444937833037302,
+      "loss": 0.2135,
+      "step": 133
+    },
+    {
+      "epoch": 0.08919953403228491,
+      "grad_norm": 111.0009765625,
+      "learning_rate": 0.000194404973357016,
+      "loss": 0.4909,
+      "step": 134
+    },
+    {
+      "epoch": 0.08986520219670495,
+      "grad_norm": 119.50690460205078,
+      "learning_rate": 0.00019436056838365896,
+      "loss": 0.8744,
+      "step": 135
+    },
+    {
+      "epoch": 0.09053087036112498,
+      "grad_norm": 83.16583251953125,
+      "learning_rate": 0.00019431616341030196,
+      "loss": 0.5229,
+      "step": 136
+    },
+    {
+      "epoch": 0.09119653852554502,
+      "grad_norm": 96.94984436035156,
+      "learning_rate": 0.00019427175843694493,
+      "loss": 0.6746,
+      "step": 137
+    },
+    {
+      "epoch": 0.09186220668996506,
+      "grad_norm": 65.90393829345703,
+      "learning_rate": 0.00019422735346358793,
+      "loss": 0.3112,
+      "step": 138
+    },
+    {
+      "epoch": 0.09252787485438509,
+      "grad_norm": 92.0575180053711,
+      "learning_rate": 0.0001941829484902309,
+      "loss": 0.5264,
+      "step": 139
+    },
+    {
+      "epoch": 0.09319354301880513,
+      "grad_norm": 63.846744537353516,
+      "learning_rate": 0.0001941385435168739,
+      "loss": 0.8869,
+      "step": 140
+    },
+    {
+      "epoch": 0.09385921118322517,
+      "grad_norm": 129.04086303710938,
+      "learning_rate": 0.00019409413854351687,
+      "loss": 0.7077,
+      "step": 141
+    },
+    {
+      "epoch": 0.0945248793476452,
+      "grad_norm": 38.49999237060547,
+      "learning_rate": 0.00019404973357015987,
+      "loss": 0.1854,
+      "step": 142
+    },
+    {
+      "epoch": 0.09519054751206524,
+      "grad_norm": 142.83619689941406,
+      "learning_rate": 0.00019400532859680284,
+      "loss": 1.3549,
+      "step": 143
+    },
+    {
+      "epoch": 0.09585621567648527,
+      "grad_norm": 58.420223236083984,
+      "learning_rate": 0.00019396092362344584,
+      "loss": 0.3978,
+      "step": 144
+    },
+    {
+      "epoch": 0.0965218838409053,
+      "grad_norm": 82.57951354980469,
+      "learning_rate": 0.0001939165186500888,
+      "loss": 0.9771,
+      "step": 145
+    },
+    {
+      "epoch": 0.09718755200532535,
+      "grad_norm": 52.69244384765625,
+      "learning_rate": 0.00019387211367673178,
+      "loss": 0.389,
+      "step": 146
+    },
+    {
+      "epoch": 0.09785322016974538,
+      "grad_norm": 99.57772827148438,
+      "learning_rate": 0.00019382770870337478,
+      "loss": 0.938,
+      "step": 147
+    },
+    {
+      "epoch": 0.09851888833416542,
+      "grad_norm": 66.55644226074219,
+      "learning_rate": 0.00019378330373001775,
+      "loss": 0.4984,
+      "step": 148
+    },
+    {
+      "epoch": 0.09918455649858546,
+      "grad_norm": 64.8500747680664,
+      "learning_rate": 0.00019373889875666075,
+      "loss": 0.4189,
+      "step": 149
+    },
+    {
+      "epoch": 0.09985022466300549,
+      "grad_norm": 113.6292495727539,
+      "learning_rate": 0.00019369449378330372,
+      "loss": 0.486,
+      "step": 150
+    },
+    {
+      "epoch": 0.10051589282742553,
+      "grad_norm": 83.64993286132812,
+      "learning_rate": 0.00019365008880994672,
+      "loss": 0.3288,
+      "step": 151
+    },
+    {
+      "epoch": 0.10118156099184557,
+      "grad_norm": 70.0993423461914,
+      "learning_rate": 0.0001936056838365897,
+      "loss": 0.7637,
+      "step": 152
+    },
+    {
+      "epoch": 0.1018472291562656,
+      "grad_norm": 68.05122375488281,
+      "learning_rate": 0.0001935612788632327,
+      "loss": 0.3255,
+      "step": 153
+    },
+    {
+      "epoch": 0.10251289732068564,
+      "grad_norm": 56.31193923950195,
+      "learning_rate": 0.00019351687388987566,
+      "loss": 0.1978,
+      "step": 154
+    },
+    {
+      "epoch": 0.10317856548510568,
+      "grad_norm": 131.5350799560547,
+      "learning_rate": 0.00019347246891651866,
+      "loss": 0.7034,
+      "step": 155
+    },
+    {
+      "epoch": 0.1038442336495257,
+      "grad_norm": 143.9272003173828,
+      "learning_rate": 0.00019342806394316163,
+      "loss": 0.6122,
+      "step": 156
+    },
+    {
+      "epoch": 0.10450990181394575,
+      "grad_norm": 56.10287094116211,
+      "learning_rate": 0.00019338365896980463,
+      "loss": 0.5982,
+      "step": 157
+    },
+    {
+      "epoch": 0.10517556997836579,
+      "grad_norm": 113.03327941894531,
+      "learning_rate": 0.0001933392539964476,
+      "loss": 0.5578,
+      "step": 158
+    },
+    {
+      "epoch": 0.10584123814278582,
+      "grad_norm": 156.54730224609375,
+      "learning_rate": 0.0001932948490230906,
+      "loss": 0.9944,
+      "step": 159
+    },
+    {
+      "epoch": 0.10650690630720586,
+      "grad_norm": 117.35420989990234,
+      "learning_rate": 0.00019325044404973357,
+      "loss": 0.9612,
+      "step": 160
+    },
+    {
+      "epoch": 0.1071725744716259,
+      "grad_norm": 137.27517700195312,
+      "learning_rate": 0.00019320603907637657,
+      "loss": 0.8892,
+      "step": 161
+    },
+    {
+      "epoch": 0.10783824263604593,
+      "grad_norm": 121.9662094116211,
+      "learning_rate": 0.00019316163410301954,
+      "loss": 0.8541,
+      "step": 162
+    },
+    {
+      "epoch": 0.10850391080046597,
+      "grad_norm": 50.96897506713867,
+      "learning_rate": 0.00019311722912966254,
+      "loss": 0.4626,
+      "step": 163
+    },
+    {
+      "epoch": 0.10916957896488601,
+      "grad_norm": 86.26678466796875,
+      "learning_rate": 0.0001930728241563055,
+      "loss": 0.4941,
+      "step": 164
+    },
+    {
+      "epoch": 0.10983524712930604,
+      "grad_norm": 92.44398498535156,
+      "learning_rate": 0.0001930284191829485,
+      "loss": 0.6016,
+      "step": 165
+    },
+    {
+      "epoch": 0.11050091529372608,
+      "grad_norm": 76.84557342529297,
+      "learning_rate": 0.00019298401420959148,
+      "loss": 0.465,
+      "step": 166
+    },
+    {
+      "epoch": 0.11116658345814612,
+      "grad_norm": 120.3193588256836,
+      "learning_rate": 0.00019293960923623448,
+      "loss": 0.7597,
+      "step": 167
+    },
+    {
+      "epoch": 0.11183225162256615,
+      "grad_norm": 98.53234100341797,
+      "learning_rate": 0.00019289520426287745,
+      "loss": 0.4161,
+      "step": 168
+    },
+    {
+      "epoch": 0.11249791978698619,
+      "grad_norm": 136.8874053955078,
+      "learning_rate": 0.00019285079928952045,
+      "loss": 1.0431,
+      "step": 169
+    },
+    {
+      "epoch": 0.11316358795140623,
+      "grad_norm": 145.96075439453125,
+      "learning_rate": 0.00019280639431616342,
+      "loss": 0.8223,
+      "step": 170
+    },
+    {
+      "epoch": 0.11382925611582626,
+      "grad_norm": 95.09919738769531,
+      "learning_rate": 0.00019276198934280642,
+      "loss": 0.4007,
+      "step": 171
+    },
+    {
+      "epoch": 0.1144949242802463,
+      "grad_norm": 95.6208267211914,
+      "learning_rate": 0.0001927175843694494,
+      "loss": 0.4496,
+      "step": 172
+    },
+    {
+      "epoch": 0.11516059244466634,
+      "grad_norm": 131.4598846435547,
+      "learning_rate": 0.00019267317939609239,
+      "loss": 0.7203,
+      "step": 173
+    },
+    {
+      "epoch": 0.11582626060908637,
+      "grad_norm": 183.2200469970703,
+      "learning_rate": 0.00019262877442273536,
+      "loss": 0.8879,
+      "step": 174
+    },
+    {
+      "epoch": 0.1164919287735064,
+      "grad_norm": 65.30191802978516,
+      "learning_rate": 0.00019258436944937836,
+      "loss": 0.2161,
+      "step": 175
+    },
+    {
+      "epoch": 0.11715759693792645,
+      "grad_norm": 100.97784423828125,
+      "learning_rate": 0.00019253996447602133,
+      "loss": 0.8095,
+      "step": 176
+    },
+    {
+      "epoch": 0.11782326510234647,
+      "grad_norm": 57.705963134765625,
+      "learning_rate": 0.00019249555950266432,
+      "loss": 0.3389,
+      "step": 177
+    },
+    {
+      "epoch": 0.11848893326676652,
+      "grad_norm": 119.22857666015625,
+      "learning_rate": 0.0001924511545293073,
+      "loss": 0.5287,
+      "step": 178
+    },
+    {
+      "epoch": 0.11915460143118656,
+      "grad_norm": 97.59465026855469,
+      "learning_rate": 0.0001924067495559503,
+      "loss": 0.4567,
+      "step": 179
+    },
+    {
+      "epoch": 0.11982026959560658,
+      "grad_norm": 47.80794906616211,
+      "learning_rate": 0.00019236234458259327,
+      "loss": 0.3795,
+      "step": 180
+    },
+    {
+      "epoch": 0.12048593776002663,
+      "grad_norm": 93.19955444335938,
+      "learning_rate": 0.00019231793960923626,
+      "loss": 1.0376,
+      "step": 181
+    },
+    {
+      "epoch": 0.12115160592444667,
+      "grad_norm": 57.24619674682617,
+      "learning_rate": 0.00019227353463587924,
+      "loss": 0.3167,
+      "step": 182
+    },
+    {
+      "epoch": 0.1218172740888667,
+      "grad_norm": 71.7004165649414,
+      "learning_rate": 0.00019222912966252223,
+      "loss": 0.3515,
+      "step": 183
+    },
+    {
+      "epoch": 0.12248294225328674,
+      "grad_norm": 42.93682098388672,
+      "learning_rate": 0.0001921847246891652,
+      "loss": 0.3272,
+      "step": 184
+    },
+    {
+      "epoch": 0.12314861041770678,
+      "grad_norm": 88.22802734375,
+      "learning_rate": 0.00019214031971580818,
+      "loss": 0.5751,
+      "step": 185
+    },
+    {
+      "epoch": 0.1238142785821268,
+      "grad_norm": 88.80204772949219,
+      "learning_rate": 0.00019209591474245117,
+      "loss": 0.5306,
+      "step": 186
+    },
+    {
+      "epoch": 0.12447994674654685,
+      "grad_norm": 92.46751403808594,
+      "learning_rate": 0.00019205150976909415,
+      "loss": 0.4262,
+      "step": 187
+    },
+    {
+      "epoch": 0.12514561491096687,
+      "grad_norm": 55.72347640991211,
+      "learning_rate": 0.00019200710479573714,
+      "loss": 0.3939,
+      "step": 188
+    },
+    {
+      "epoch": 0.12581128307538691,
+      "grad_norm": 72.36625671386719,
+      "learning_rate": 0.00019196269982238012,
+      "loss": 0.3739,
+      "step": 189
+    },
+    {
+      "epoch": 0.12647695123980696,
+      "grad_norm": 96.56300354003906,
+      "learning_rate": 0.00019191829484902311,
+      "loss": 0.4574,
+      "step": 190
+    },
+    {
+      "epoch": 0.127142619404227,
+      "grad_norm": 122.78588104248047,
+      "learning_rate": 0.00019187388987566609,
+      "loss": 0.7646,
+      "step": 191
+    },
+    {
+      "epoch": 0.12780828756864704,
+      "grad_norm": 74.9114990234375,
+      "learning_rate": 0.00019182948490230908,
+      "loss": 0.2156,
+      "step": 192
+    },
+    {
+      "epoch": 0.12847395573306705,
+      "grad_norm": 61.17184829711914,
+      "learning_rate": 0.00019178507992895206,
+      "loss": 0.7067,
+      "step": 193
+    },
+    {
+      "epoch": 0.1291396238974871,
+      "grad_norm": 120.3477783203125,
+      "learning_rate": 0.00019174067495559505,
+      "loss": 0.7562,
+      "step": 194
+    },
+    {
+      "epoch": 0.12980529206190713,
+      "grad_norm": 204.0528564453125,
+      "learning_rate": 0.00019169626998223802,
+      "loss": 1.2893,
+      "step": 195
+    },
+    {
+      "epoch": 0.13047096022632718,
+      "grad_norm": 91.55975341796875,
+      "learning_rate": 0.000191651865008881,
+      "loss": 1.025,
+      "step": 196
+    },
+    {
+      "epoch": 0.13113662839074722,
+      "grad_norm": 111.15852355957031,
+      "learning_rate": 0.000191607460035524,
+      "loss": 0.4063,
+      "step": 197
+    },
+    {
+      "epoch": 0.13180229655516726,
+      "grad_norm": 90.34394073486328,
+      "learning_rate": 0.00019156305506216697,
+      "loss": 0.6954,
+      "step": 198
+    },
+    {
+      "epoch": 0.13246796471958727,
+      "grad_norm": 95.49404907226562,
+      "learning_rate": 0.00019151865008880996,
+      "loss": 0.6267,
+      "step": 199
+    },
+    {
+      "epoch": 0.1331336328840073,
+      "grad_norm": 41.02389144897461,
+      "learning_rate": 0.00019147424511545294,
+      "loss": 0.2385,
+      "step": 200
+    },
+    {
+      "epoch": 0.13379930104842735,
+      "grad_norm": 91.21604919433594,
+      "learning_rate": 0.00019142984014209593,
+      "loss": 0.4531,
+      "step": 201
+    },
+    {
+      "epoch": 0.1344649692128474,
+      "grad_norm": 49.900184631347656,
+      "learning_rate": 0.0001913854351687389,
+      "loss": 0.1657,
+      "step": 202
+    },
+    {
+      "epoch": 0.13513063737726744,
+      "grad_norm": 144.78623962402344,
+      "learning_rate": 0.0001913410301953819,
+      "loss": 0.7326,
+      "step": 203
+    },
+    {
+      "epoch": 0.13579630554168748,
+      "grad_norm": 143.92132568359375,
+      "learning_rate": 0.00019129662522202487,
+      "loss": 1.0342,
+      "step": 204
+    },
+    {
+      "epoch": 0.1364619737061075,
+      "grad_norm": 107.93486022949219,
+      "learning_rate": 0.00019125222024866787,
+      "loss": 0.3643,
+      "step": 205
+    },
+    {
+      "epoch": 0.13712764187052753,
+      "grad_norm": 69.4767074584961,
+      "learning_rate": 0.00019120781527531084,
+      "loss": 0.4302,
+      "step": 206
+    },
+    {
+      "epoch": 0.13779331003494757,
+      "grad_norm": 102.22624206542969,
+      "learning_rate": 0.00019116341030195384,
+      "loss": 0.7252,
+      "step": 207
+    },
+    {
+      "epoch": 0.13845897819936762,
+      "grad_norm": 39.15114974975586,
+      "learning_rate": 0.00019111900532859681,
+      "loss": 0.376,
+      "step": 208
+    },
+    {
+      "epoch": 0.13912464636378766,
+      "grad_norm": 65.12492370605469,
+      "learning_rate": 0.00019107460035523979,
+      "loss": 0.6005,
+      "step": 209
+    },
+    {
+      "epoch": 0.1397903145282077,
+      "grad_norm": 95.04318237304688,
+      "learning_rate": 0.00019103019538188278,
+      "loss": 0.8371,
+      "step": 210
+    },
+    {
+      "epoch": 0.1404559826926277,
+      "grad_norm": 119.12702941894531,
+      "learning_rate": 0.00019098579040852576,
+      "loss": 0.5701,
+      "step": 211
+    },
+    {
+      "epoch": 0.14112165085704775,
+      "grad_norm": 56.38259506225586,
+      "learning_rate": 0.00019094138543516875,
+      "loss": 0.6285,
+      "step": 212
+    },
+    {
+      "epoch": 0.1417873190214678,
+      "grad_norm": 29.10639762878418,
+      "learning_rate": 0.00019089698046181172,
+      "loss": 0.4639,
+      "step": 213
+    },
+    {
+      "epoch": 0.14245298718588784,
+      "grad_norm": 60.445865631103516,
+      "learning_rate": 0.00019085257548845472,
+      "loss": 0.6099,
+      "step": 214
+    },
+    {
+      "epoch": 0.14311865535030788,
+      "grad_norm": 83.33357238769531,
+      "learning_rate": 0.0001908081705150977,
+      "loss": 0.6992,
+      "step": 215
+    },
+    {
+      "epoch": 0.14378432351472792,
+      "grad_norm": 50.539451599121094,
+      "learning_rate": 0.0001907637655417407,
+      "loss": 0.673,
+      "step": 216
+    },
+    {
+      "epoch": 0.14444999167914793,
+      "grad_norm": 108.4279556274414,
+      "learning_rate": 0.00019071936056838366,
+      "loss": 0.7867,
+      "step": 217
+    },
+    {
+      "epoch": 0.14511565984356797,
+      "grad_norm": 71.25292205810547,
+      "learning_rate": 0.00019067495559502666,
+      "loss": 0.2414,
+      "step": 218
+    },
+    {
+      "epoch": 0.14578132800798801,
+      "grad_norm": 96.37142944335938,
+      "learning_rate": 0.00019063055062166963,
+      "loss": 0.4556,
+      "step": 219
+    },
+    {
+      "epoch": 0.14644699617240806,
+      "grad_norm": 105.84346008300781,
+      "learning_rate": 0.00019058614564831263,
+      "loss": 0.5288,
+      "step": 220
+    },
+    {
+      "epoch": 0.1471126643368281,
+      "grad_norm": 53.66656494140625,
+      "learning_rate": 0.0001905417406749556,
+      "loss": 0.2665,
+      "step": 221
+    },
+    {
+      "epoch": 0.14777833250124814,
+      "grad_norm": 73.43135070800781,
+      "learning_rate": 0.00019049733570159857,
+      "loss": 0.465,
+      "step": 222
+    },
+    {
+      "epoch": 0.14844400066566815,
+      "grad_norm": 108.94127655029297,
+      "learning_rate": 0.00019045293072824157,
+      "loss": 0.6516,
+      "step": 223
+    },
+    {
+      "epoch": 0.1491096688300882,
+      "grad_norm": 23.10163688659668,
+      "learning_rate": 0.00019040852575488454,
+      "loss": 0.1401,
+      "step": 224
+    },
+    {
+      "epoch": 0.14977533699450823,
+      "grad_norm": 182.62322998046875,
+      "learning_rate": 0.00019036412078152754,
+      "loss": 0.7953,
+      "step": 225
+    },
+    {
+      "epoch": 0.15044100515892828,
+      "grad_norm": 104.31300354003906,
+      "learning_rate": 0.00019031971580817051,
+      "loss": 0.5117,
+      "step": 226
+    },
+    {
+      "epoch": 0.15110667332334832,
+      "grad_norm": 122.88026428222656,
+      "learning_rate": 0.0001902753108348135,
+      "loss": 0.8029,
+      "step": 227
+    },
+    {
+      "epoch": 0.15177234148776836,
+      "grad_norm": 72.17605590820312,
+      "learning_rate": 0.00019023090586145648,
+      "loss": 0.3403,
+      "step": 228
+    },
+    {
+      "epoch": 0.15243800965218837,
+      "grad_norm": 55.70381164550781,
+      "learning_rate": 0.00019018650088809948,
+      "loss": 0.2811,
+      "step": 229
+    },
+    {
+      "epoch": 0.1531036778166084,
+      "grad_norm": 80.8816146850586,
+      "learning_rate": 0.00019014209591474245,
+      "loss": 1.0904,
+      "step": 230
+    },
+    {
+      "epoch": 0.15376934598102845,
+      "grad_norm": 77.0992431640625,
+      "learning_rate": 0.00019009769094138545,
+      "loss": 0.8617,
+      "step": 231
+    },
+    {
+      "epoch": 0.1544350141454485,
+      "grad_norm": 47.77476119995117,
+      "learning_rate": 0.00019005328596802842,
+      "loss": 0.3185,
+      "step": 232
+    },
+    {
+      "epoch": 0.15510068230986854,
+      "grad_norm": 77.98711395263672,
+      "learning_rate": 0.00019000888099467142,
+      "loss": 0.2655,
+      "step": 233
+    },
+    {
+      "epoch": 0.15576635047428858,
+      "grad_norm": 63.54255294799805,
+      "learning_rate": 0.0001899644760213144,
+      "loss": 0.3955,
+      "step": 234
+    },
+    {
+      "epoch": 0.1564320186387086,
+      "grad_norm": 190.26271057128906,
+      "learning_rate": 0.00018992007104795736,
+      "loss": 1.2886,
+      "step": 235
+    },
+    {
+      "epoch": 0.15709768680312863,
+      "grad_norm": 117.44766998291016,
+      "learning_rate": 0.00018987566607460036,
+      "loss": 0.7142,
+      "step": 236
+    },
+    {
+      "epoch": 0.15776335496754867,
+      "grad_norm": 133.50717163085938,
+      "learning_rate": 0.00018983126110124333,
+      "loss": 0.8898,
+      "step": 237
+    },
+    {
+      "epoch": 0.15842902313196872,
+      "grad_norm": 62.771507263183594,
+      "learning_rate": 0.00018978685612788633,
+      "loss": 0.3062,
+      "step": 238
+    },
+    {
+      "epoch": 0.15909469129638876,
+      "grad_norm": 74.36737060546875,
+      "learning_rate": 0.0001897424511545293,
+      "loss": 0.5336,
+      "step": 239
+    },
+    {
+      "epoch": 0.1597603594608088,
+      "grad_norm": 137.45458984375,
+      "learning_rate": 0.0001896980461811723,
+      "loss": 0.7025,
+      "step": 240
+    },
+    {
+      "epoch": 0.1604260276252288,
+      "grad_norm": 54.84712600708008,
+      "learning_rate": 0.00018965364120781527,
+      "loss": 0.338,
+      "step": 241
+    },
+    {
+      "epoch": 0.16109169578964885,
+      "grad_norm": 127.16575622558594,
+      "learning_rate": 0.00018960923623445827,
+      "loss": 1.0382,
+      "step": 242
+    },
+    {
+      "epoch": 0.1617573639540689,
+      "grad_norm": 93.72176361083984,
+      "learning_rate": 0.00018956483126110124,
+      "loss": 1.0084,
+      "step": 243
+    },
+    {
+      "epoch": 0.16242303211848894,
+      "grad_norm": 56.54580307006836,
+      "learning_rate": 0.00018952042628774424,
+      "loss": 0.5353,
+      "step": 244
+    },
+    {
+      "epoch": 0.16308870028290898,
+      "grad_norm": 76.96385955810547,
+      "learning_rate": 0.0001894760213143872,
+      "loss": 0.397,
+      "step": 245
+    },
+    {
+      "epoch": 0.16375436844732902,
+      "grad_norm": 91.15630340576172,
+      "learning_rate": 0.00018943161634103018,
+      "loss": 0.4533,
+      "step": 246
+    },
+    {
+      "epoch": 0.16442003661174903,
+      "grad_norm": 103.4432373046875,
+      "learning_rate": 0.00018938721136767318,
+      "loss": 0.6514,
+      "step": 247
+    },
+    {
+      "epoch": 0.16508570477616907,
+      "grad_norm": 60.81359100341797,
+      "learning_rate": 0.00018934280639431615,
+      "loss": 0.4884,
+      "step": 248
+    },
+    {
+      "epoch": 0.16575137294058911,
+      "grad_norm": 46.649139404296875,
+      "learning_rate": 0.00018929840142095915,
+      "loss": 0.271,
+      "step": 249
+    },
+    {
+      "epoch": 0.16641704110500916,
+      "grad_norm": 58.7072868347168,
+      "learning_rate": 0.00018925399644760212,
+      "loss": 0.27,
+      "step": 250
+    },
+    {
+      "epoch": 0.1670827092694292,
+      "grad_norm": 79.08338928222656,
+      "learning_rate": 0.00018920959147424512,
+      "loss": 0.4671,
+      "step": 251
+    },
+    {
+      "epoch": 0.16774837743384924,
+      "grad_norm": 98.57723236083984,
+      "learning_rate": 0.0001891651865008881,
+      "loss": 0.3235,
+      "step": 252
+    },
+    {
+      "epoch": 0.16841404559826925,
+      "grad_norm": 131.836181640625,
+      "learning_rate": 0.0001891207815275311,
+      "loss": 0.4688,
+      "step": 253
+    },
+    {
+      "epoch": 0.1690797137626893,
+      "grad_norm": 67.51493835449219,
+      "learning_rate": 0.00018907637655417406,
+      "loss": 0.6911,
+      "step": 254
+    },
+    {
+      "epoch": 0.16974538192710933,
+      "grad_norm": 51.76738357543945,
+      "learning_rate": 0.00018903197158081706,
+      "loss": 0.1837,
+      "step": 255
+    },
+    {
+      "epoch": 0.17041105009152938,
+      "grad_norm": 52.23995590209961,
+      "learning_rate": 0.00018898756660746003,
+      "loss": 0.71,
+      "step": 256
+    },
+    {
+      "epoch": 0.17107671825594942,
+      "grad_norm": 49.984336853027344,
+      "learning_rate": 0.00018894316163410303,
+      "loss": 0.3221,
+      "step": 257
+    },
+    {
+      "epoch": 0.17174238642036946,
+      "grad_norm": 48.08108901977539,
+      "learning_rate": 0.000188898756660746,
+      "loss": 0.3334,
+      "step": 258
+    },
+    {
+      "epoch": 0.17240805458478947,
+      "grad_norm": 24.143003463745117,
+      "learning_rate": 0.00018885435168738897,
+      "loss": 0.2096,
+      "step": 259
+    },
+    {
+      "epoch": 0.1730737227492095,
+      "grad_norm": 116.61510467529297,
+      "learning_rate": 0.00018880994671403197,
+      "loss": 0.8852,
+      "step": 260
+    },
+    {
+      "epoch": 0.17373939091362955,
+      "grad_norm": 41.168052673339844,
+      "learning_rate": 0.00018876554174067494,
+      "loss": 0.2541,
+      "step": 261
+    },
+    {
+      "epoch": 0.1744050590780496,
+      "grad_norm": 19.256343841552734,
+      "learning_rate": 0.00018872113676731794,
+      "loss": 0.2278,
+      "step": 262
+    },
+    {
+      "epoch": 0.17507072724246964,
+      "grad_norm": 51.543418884277344,
+      "learning_rate": 0.0001886767317939609,
+      "loss": 0.2494,
+      "step": 263
+    },
+    {
+      "epoch": 0.17573639540688968,
+      "grad_norm": 44.67826461791992,
+      "learning_rate": 0.0001886323268206039,
+      "loss": 0.3112,
+      "step": 264
+    },
+    {
+      "epoch": 0.1764020635713097,
+      "grad_norm": 41.30339431762695,
+      "learning_rate": 0.00018858792184724688,
+      "loss": 0.5114,
+      "step": 265
+    },
+    {
+      "epoch": 0.17706773173572973,
+      "grad_norm": 12.575431823730469,
+      "learning_rate": 0.00018854351687388988,
+      "loss": 0.0375,
+      "step": 266
+    },
+    {
+      "epoch": 0.17773339990014977,
+      "grad_norm": 71.91178894042969,
+      "learning_rate": 0.00018849911190053285,
+      "loss": 0.5989,
+      "step": 267
+    },
+    {
+      "epoch": 0.17839906806456982,
+      "grad_norm": 83.31620788574219,
+      "learning_rate": 0.00018845470692717585,
+      "loss": 0.51,
+      "step": 268
+    },
+    {
+      "epoch": 0.17906473622898986,
+      "grad_norm": 83.13764190673828,
+      "learning_rate": 0.00018841030195381885,
+      "loss": 1.1146,
+      "step": 269
+    },
+    {
+      "epoch": 0.1797304043934099,
+      "grad_norm": 151.9849853515625,
+      "learning_rate": 0.00018836589698046182,
+      "loss": 1.1219,
+      "step": 270
+    },
+    {
+      "epoch": 0.1803960725578299,
+      "grad_norm": 121.70146179199219,
+      "learning_rate": 0.00018832149200710482,
+      "loss": 0.9397,
+      "step": 271
+    },
+    {
+      "epoch": 0.18106174072224995,
+      "grad_norm": 51.62641906738281,
+      "learning_rate": 0.0001882770870337478,
+      "loss": 0.2968,
+      "step": 272
+    },
+    {
+      "epoch": 0.18172740888667,
+      "grad_norm": 66.82881164550781,
+      "learning_rate": 0.0001882326820603908,
+      "loss": 0.5365,
+      "step": 273
+    },
+    {
+      "epoch": 0.18239307705109004,
+      "grad_norm": 78.71537017822266,
+      "learning_rate": 0.00018818827708703376,
+      "loss": 0.8149,
+      "step": 274
+    },
+    {
+      "epoch": 0.18305874521551008,
+      "grad_norm": 137.4581756591797,
+      "learning_rate": 0.00018814387211367676,
+      "loss": 0.6777,
+      "step": 275
+    },
+    {
+      "epoch": 0.18372441337993012,
+      "grad_norm": 64.81407165527344,
+      "learning_rate": 0.00018809946714031973,
+      "loss": 0.7677,
+      "step": 276
+    },
+    {
+      "epoch": 0.18439008154435013,
+      "grad_norm": 96.37416076660156,
+      "learning_rate": 0.00018805506216696273,
+      "loss": 1.332,
+      "step": 277
+    },
+    {
+      "epoch": 0.18505574970877017,
+      "grad_norm": 47.11737823486328,
+      "learning_rate": 0.0001880106571936057,
+      "loss": 0.29,
+      "step": 278
+    },
+    {
+      "epoch": 0.18572141787319021,
+      "grad_norm": 111.5100326538086,
+      "learning_rate": 0.0001879662522202487,
+      "loss": 0.5466,
+      "step": 279
+    },
+    {
+      "epoch": 0.18638708603761026,
+      "grad_norm": 123.91747283935547,
+      "learning_rate": 0.00018792184724689167,
+      "loss": 0.6733,
+      "step": 280
+    },
+    {
+      "epoch": 0.1870527542020303,
+      "grad_norm": 107.9887924194336,
+      "learning_rate": 0.00018787744227353467,
+      "loss": 0.4885,
+      "step": 281
+    },
+    {
+      "epoch": 0.18771842236645034,
+      "grad_norm": 76.66865539550781,
+      "learning_rate": 0.00018783303730017764,
+      "loss": 0.6393,
+      "step": 282
+    },
+    {
+      "epoch": 0.18838409053087035,
+      "grad_norm": 67.36553192138672,
+      "learning_rate": 0.00018778863232682064,
+      "loss": 0.4087,
+      "step": 283
+    },
+    {
+      "epoch": 0.1890497586952904,
+      "grad_norm": 50.0120849609375,
+      "learning_rate": 0.0001877442273534636,
+      "loss": 0.3911,
+      "step": 284
+    },
+    {
+      "epoch": 0.18971542685971043,
+      "grad_norm": 109.88864135742188,
+      "learning_rate": 0.00018769982238010658,
+      "loss": 0.6727,
+      "step": 285
+    },
+    {
+      "epoch": 0.19038109502413048,
+      "grad_norm": 114.64707946777344,
+      "learning_rate": 0.00018765541740674958,
+      "loss": 0.8044,
+      "step": 286
+    },
+    {
+      "epoch": 0.19104676318855052,
+      "grad_norm": 130.69219970703125,
+      "learning_rate": 0.00018761101243339255,
+      "loss": 0.6119,
+      "step": 287
+    },
+    {
+      "epoch": 0.19171243135297053,
+      "grad_norm": 50.73808670043945,
+      "learning_rate": 0.00018756660746003555,
+      "loss": 0.4887,
+      "step": 288
+    },
+    {
+      "epoch": 0.19237809951739057,
+      "grad_norm": 143.2826385498047,
+      "learning_rate": 0.00018752220248667852,
+      "loss": 0.8228,
+      "step": 289
+    },
+    {
+      "epoch": 0.1930437676818106,
+      "grad_norm": 132.3501739501953,
+      "learning_rate": 0.00018747779751332152,
+      "loss": 0.808,
+      "step": 290
+    },
+    {
+      "epoch": 0.19370943584623065,
+      "grad_norm": 74.29531860351562,
+      "learning_rate": 0.0001874333925399645,
+      "loss": 0.7815,
+      "step": 291
+    },
+    {
+      "epoch": 0.1943751040106507,
+      "grad_norm": 81.61593627929688,
+      "learning_rate": 0.00018738898756660749,
+      "loss": 0.5732,
+      "step": 292
+    },
+    {
+      "epoch": 0.19504077217507074,
+      "grad_norm": 86.1830825805664,
+      "learning_rate": 0.00018734458259325046,
+      "loss": 0.5226,
+      "step": 293
+    },
+    {
+      "epoch": 0.19570644033949075,
+      "grad_norm": 23.495864868164062,
+      "learning_rate": 0.00018730017761989346,
+      "loss": 0.1476,
+      "step": 294
+    },
+    {
+      "epoch": 0.1963721085039108,
+      "grad_norm": 45.22040939331055,
+      "learning_rate": 0.00018725577264653643,
+      "loss": 0.4335,
+      "step": 295
+    },
+    {
+      "epoch": 0.19703777666833083,
+      "grad_norm": 102.33810424804688,
+      "learning_rate": 0.0001872113676731794,
+      "loss": 0.6778,
+      "step": 296
+    },
+    {
+      "epoch": 0.19770344483275087,
+      "grad_norm": 90.49777221679688,
+      "learning_rate": 0.0001871669626998224,
+      "loss": 0.7652,
+      "step": 297
+    },
+    {
+      "epoch": 0.19836911299717092,
+      "grad_norm": 14.38054084777832,
+      "learning_rate": 0.00018712255772646537,
+      "loss": 0.0598,
+      "step": 298
+    },
+    {
+      "epoch": 0.19903478116159096,
+      "grad_norm": 24.415199279785156,
+      "learning_rate": 0.00018707815275310837,
+      "loss": 0.3666,
+      "step": 299
+    },
+    {
+      "epoch": 0.19970044932601097,
+      "grad_norm": 37.869869232177734,
+      "learning_rate": 0.00018703374777975134,
+      "loss": 0.1511,
+      "step": 300
+    },
+    {
+      "epoch": 0.200366117490431,
+      "grad_norm": 115.4464111328125,
+      "learning_rate": 0.00018698934280639434,
+      "loss": 0.7441,
+      "step": 301
+    },
+    {
+      "epoch": 0.20103178565485105,
+      "grad_norm": 124.49742889404297,
+      "learning_rate": 0.0001869449378330373,
+      "loss": 0.6898,
+      "step": 302
+    },
+    {
+      "epoch": 0.2016974538192711,
+      "grad_norm": 45.529197692871094,
+      "learning_rate": 0.0001869005328596803,
+      "loss": 0.2891,
+      "step": 303
+    },
+    {
+      "epoch": 0.20236312198369114,
+      "grad_norm": 107.60458374023438,
+      "learning_rate": 0.00018685612788632328,
+      "loss": 0.6037,
+      "step": 304
+    },
+    {
+      "epoch": 0.20302879014811118,
+      "grad_norm": 72.92047119140625,
+      "learning_rate": 0.00018681172291296628,
+      "loss": 0.8364,
+      "step": 305
+    },
+    {
+      "epoch": 0.2036944583125312,
+      "grad_norm": 80.35264587402344,
+      "learning_rate": 0.00018676731793960925,
+      "loss": 0.6218,
+      "step": 306
+    },
+    {
+      "epoch": 0.20436012647695123,
+      "grad_norm": 181.0091552734375,
+      "learning_rate": 0.00018672291296625225,
+      "loss": 0.8943,
+      "step": 307
+    },
+    {
+      "epoch": 0.20502579464137127,
+      "grad_norm": 54.508758544921875,
+      "learning_rate": 0.00018667850799289522,
+      "loss": 0.3581,
+      "step": 308
+    },
+    {
+      "epoch": 0.20569146280579131,
+      "grad_norm": 22.76518440246582,
+      "learning_rate": 0.0001866341030195382,
+      "loss": 0.0707,
+      "step": 309
+    },
+    {
+      "epoch": 0.20635713097021136,
+      "grad_norm": 144.45138549804688,
+      "learning_rate": 0.00018658969804618119,
+      "loss": 1.199,
+      "step": 310
+    },
+    {
+      "epoch": 0.2070227991346314,
+      "grad_norm": 51.15176010131836,
+      "learning_rate": 0.00018654529307282416,
+      "loss": 0.2966,
+      "step": 311
+    },
+    {
+      "epoch": 0.2076884672990514,
+      "grad_norm": 84.0188980102539,
+      "learning_rate": 0.00018650088809946716,
+      "loss": 0.3478,
+      "step": 312
+    },
+    {
+      "epoch": 0.20835413546347145,
+      "grad_norm": 74.65904235839844,
+      "learning_rate": 0.00018645648312611013,
+      "loss": 0.5906,
+      "step": 313
+    },
+    {
+      "epoch": 0.2090198036278915,
+      "grad_norm": 76.40025329589844,
+      "learning_rate": 0.00018641207815275313,
+      "loss": 0.5313,
+      "step": 314
+    },
+    {
+      "epoch": 0.20968547179231153,
+      "grad_norm": 67.46442413330078,
+      "learning_rate": 0.0001863676731793961,
+      "loss": 0.5431,
+      "step": 315
+    },
+    {
+      "epoch": 0.21035113995673158,
+      "grad_norm": 153.51947021484375,
+      "learning_rate": 0.0001863232682060391,
+      "loss": 1.4613,
+      "step": 316
+    },
+    {
+      "epoch": 0.21101680812115162,
+      "grad_norm": 141.46417236328125,
+      "learning_rate": 0.00018627886323268207,
+      "loss": 0.6173,
+      "step": 317
+    },
+    {
+      "epoch": 0.21168247628557163,
+      "grad_norm": 117.94355773925781,
+      "learning_rate": 0.00018623445825932506,
+      "loss": 0.703,
+      "step": 318
+    },
+    {
+      "epoch": 0.21234814444999167,
+      "grad_norm": 72.00027465820312,
+      "learning_rate": 0.00018619005328596804,
+      "loss": 0.3786,
+      "step": 319
+    },
+    {
+      "epoch": 0.2130138126144117,
+      "grad_norm": 54.18503189086914,
+      "learning_rate": 0.00018614564831261103,
+      "loss": 0.3633,
+      "step": 320
+    },
+    {
+      "epoch": 0.21367948077883175,
+      "grad_norm": 121.49830627441406,
+      "learning_rate": 0.000186101243339254,
+      "loss": 0.7459,
+      "step": 321
+    },
+    {
+      "epoch": 0.2143451489432518,
+      "grad_norm": 176.50872802734375,
+      "learning_rate": 0.00018605683836589698,
+      "loss": 0.8166,
+      "step": 322
+    },
+    {
+      "epoch": 0.21501081710767184,
+      "grad_norm": 151.5113067626953,
+      "learning_rate": 0.00018601243339253998,
+      "loss": 0.7919,
+      "step": 323
+    },
+    {
+      "epoch": 0.21567648527209185,
+      "grad_norm": 84.63894653320312,
+      "learning_rate": 0.00018596802841918295,
+      "loss": 0.7106,
+      "step": 324
+    },
+    {
+      "epoch": 0.2163421534365119,
+      "grad_norm": 102.90968322753906,
+      "learning_rate": 0.00018592362344582595,
+      "loss": 0.689,
+      "step": 325
+    },
+    {
+      "epoch": 0.21700782160093193,
+      "grad_norm": 61.89885330200195,
+      "learning_rate": 0.00018587921847246892,
+      "loss": 0.3393,
+      "step": 326
+    },
+    {
+      "epoch": 0.21767348976535197,
+      "grad_norm": 106.49290466308594,
+      "learning_rate": 0.00018583481349911191,
+      "loss": 0.3928,
+      "step": 327
+    },
+    {
+      "epoch": 0.21833915792977202,
+      "grad_norm": 43.59565353393555,
+      "learning_rate": 0.00018579040852575489,
+      "loss": 0.3696,
+      "step": 328
+    },
+    {
+      "epoch": 0.21900482609419206,
+      "grad_norm": 67.05834197998047,
+      "learning_rate": 0.00018574600355239788,
+      "loss": 0.8861,
+      "step": 329
+    },
+    {
+      "epoch": 0.21967049425861207,
+      "grad_norm": 108.7596435546875,
+      "learning_rate": 0.00018570159857904086,
+      "loss": 1.1147,
+      "step": 330
+    },
+    {
+      "epoch": 0.2203361624230321,
+      "grad_norm": 45.871212005615234,
+      "learning_rate": 0.00018565719360568385,
+      "loss": 0.2634,
+      "step": 331
+    },
+    {
+      "epoch": 0.22100183058745215,
+      "grad_norm": 97.05673217773438,
+      "learning_rate": 0.00018561278863232683,
+      "loss": 0.7176,
+      "step": 332
+    },
+    {
+      "epoch": 0.2216674987518722,
+      "grad_norm": 37.64607238769531,
+      "learning_rate": 0.00018556838365896982,
+      "loss": 0.405,
+      "step": 333
+    },
+    {
+      "epoch": 0.22233316691629224,
+      "grad_norm": 82.15957641601562,
+      "learning_rate": 0.0001855239786856128,
+      "loss": 0.3589,
+      "step": 334
+    },
+    {
+      "epoch": 0.22299883508071228,
+      "grad_norm": 48.32238006591797,
+      "learning_rate": 0.00018547957371225577,
+      "loss": 0.2946,
+      "step": 335
+    },
+    {
+      "epoch": 0.2236645032451323,
+      "grad_norm": 33.3898811340332,
+      "learning_rate": 0.00018543516873889876,
+      "loss": 0.388,
+      "step": 336
+    },
+    {
+      "epoch": 0.22433017140955233,
+      "grad_norm": 42.87260818481445,
+      "learning_rate": 0.00018539076376554174,
+      "loss": 0.2302,
+      "step": 337
+    },
+    {
+      "epoch": 0.22499583957397237,
+      "grad_norm": 58.45392608642578,
+      "learning_rate": 0.00018534635879218473,
+      "loss": 0.6616,
+      "step": 338
+    },
+    {
+      "epoch": 0.22566150773839241,
+      "grad_norm": 28.114885330200195,
+      "learning_rate": 0.0001853019538188277,
+      "loss": 0.1042,
+      "step": 339
+    },
+    {
+      "epoch": 0.22632717590281246,
+      "grad_norm": 51.35744857788086,
+      "learning_rate": 0.0001852575488454707,
+      "loss": 0.4782,
+      "step": 340
+    },
+    {
+      "epoch": 0.2269928440672325,
+      "grad_norm": 82.15337371826172,
+      "learning_rate": 0.00018521314387211368,
+      "loss": 0.3746,
+      "step": 341
+    },
+    {
+      "epoch": 0.2276585122316525,
+      "grad_norm": 109.32012176513672,
+      "learning_rate": 0.00018516873889875667,
+      "loss": 0.6589,
+      "step": 342
+    },
+    {
+      "epoch": 0.22832418039607255,
+      "grad_norm": 155.4207000732422,
+      "learning_rate": 0.00018512433392539965,
+      "loss": 1.1968,
+      "step": 343
+    },
+    {
+      "epoch": 0.2289898485604926,
+      "grad_norm": 33.03417205810547,
+      "learning_rate": 0.00018507992895204264,
+      "loss": 0.3845,
+      "step": 344
+    },
+    {
+      "epoch": 0.22965551672491263,
+      "grad_norm": 37.060585021972656,
+      "learning_rate": 0.00018503552397868561,
+      "loss": 0.1341,
+      "step": 345
+    },
+    {
+      "epoch": 0.23032118488933268,
+      "grad_norm": 106.52214050292969,
+      "learning_rate": 0.00018499111900532859,
+      "loss": 0.7823,
+      "step": 346
+    },
+    {
+      "epoch": 0.23098685305375272,
+      "grad_norm": 48.00297164916992,
+      "learning_rate": 0.00018494671403197158,
+      "loss": 0.6375,
+      "step": 347
+    },
+    {
+      "epoch": 0.23165252121817273,
+      "grad_norm": 51.31806564331055,
+      "learning_rate": 0.00018490230905861456,
+      "loss": 0.4176,
+      "step": 348
+    },
+    {
+      "epoch": 0.23231818938259277,
+      "grad_norm": 57.65751647949219,
+      "learning_rate": 0.00018485790408525755,
+      "loss": 0.2368,
+      "step": 349
+    },
+    {
+      "epoch": 0.2329838575470128,
+      "grad_norm": 46.041908264160156,
+      "learning_rate": 0.00018481349911190053,
+      "loss": 0.4219,
+      "step": 350
+    },
+    {
+      "epoch": 0.23364952571143285,
+      "grad_norm": 97.23726654052734,
+      "learning_rate": 0.00018476909413854352,
+      "loss": 0.8618,
+      "step": 351
+    },
+    {
+      "epoch": 0.2343151938758529,
+      "grad_norm": 39.60933303833008,
+      "learning_rate": 0.0001847246891651865,
+      "loss": 0.2446,
+      "step": 352
+    },
+    {
+      "epoch": 0.23498086204027294,
+      "grad_norm": 83.39789581298828,
+      "learning_rate": 0.0001846802841918295,
+      "loss": 0.5283,
+      "step": 353
+    },
+    {
+      "epoch": 0.23564653020469295,
+      "grad_norm": 23.28626823425293,
+      "learning_rate": 0.00018463587921847246,
+      "loss": 0.2269,
+      "step": 354
+    },
+    {
+      "epoch": 0.236312198369113,
+      "grad_norm": 75.43800354003906,
+      "learning_rate": 0.00018459147424511546,
+      "loss": 0.5959,
+      "step": 355
+    },
+    {
+      "epoch": 0.23697786653353303,
+      "grad_norm": 33.72787094116211,
+      "learning_rate": 0.00018454706927175843,
+      "loss": 0.2795,
+      "step": 356
+    },
+    {
+      "epoch": 0.23764353469795307,
+      "grad_norm": 56.4429817199707,
+      "learning_rate": 0.00018450266429840143,
+      "loss": 0.6191,
+      "step": 357
+    },
+    {
+      "epoch": 0.23830920286237312,
+      "grad_norm": 97.20005798339844,
+      "learning_rate": 0.0001844582593250444,
+      "loss": 0.6252,
+      "step": 358
+    },
+    {
+      "epoch": 0.23897487102679316,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001844582593250444,
+      "loss": 1.1109,
+      "step": 359
+    },
+    {
+      "epoch": 0.23964053919121317,
+      "grad_norm": 103.0977783203125,
+      "learning_rate": 0.00018441385435168738,
+      "loss": 0.5326,
+      "step": 360
+    },
+    {
+      "epoch": 0.2403062073556332,
+      "grad_norm": 33.12345504760742,
+      "learning_rate": 0.00018436944937833037,
+      "loss": 0.4465,
+      "step": 361
+    },
+    {
+      "epoch": 0.24097187552005325,
+      "grad_norm": 29.757606506347656,
+      "learning_rate": 0.00018432504440497335,
+      "loss": 0.3853,
+      "step": 362
+    },
+    {
+      "epoch": 0.2416375436844733,
+      "grad_norm": 52.76167678833008,
+      "learning_rate": 0.00018428063943161634,
+      "loss": 0.3521,
+      "step": 363
+    },
+    {
+      "epoch": 0.24230321184889334,
+      "grad_norm": 80.51199340820312,
+      "learning_rate": 0.00018423623445825931,
+      "loss": 0.4551,
+      "step": 364
+    },
+    {
+      "epoch": 0.24296888001331338,
+      "grad_norm": 47.3682861328125,
+      "learning_rate": 0.0001841918294849023,
+      "loss": 0.2393,
+      "step": 365
+    },
+    {
+      "epoch": 0.2436345481777334,
+      "grad_norm": 80.18375396728516,
+      "learning_rate": 0.00018414742451154528,
+      "loss": 0.9122,
+      "step": 366
+    },
+    {
+      "epoch": 0.24430021634215343,
+      "grad_norm": 65.6572265625,
+      "learning_rate": 0.00018410301953818828,
+      "loss": 0.3345,
+      "step": 367
+    },
+    {
+      "epoch": 0.24496588450657347,
+      "grad_norm": 108.21941375732422,
+      "learning_rate": 0.00018405861456483125,
+      "loss": 0.5405,
+      "step": 368
+    },
+    {
+      "epoch": 0.2456315526709935,
+      "grad_norm": 57.63541030883789,
+      "learning_rate": 0.00018401420959147425,
+      "loss": 0.2502,
+      "step": 369
+    },
+    {
+      "epoch": 0.24629722083541356,
+      "grad_norm": 35.54652786254883,
+      "learning_rate": 0.00018396980461811722,
+      "loss": 0.5846,
+      "step": 370
+    },
+    {
+      "epoch": 0.2469628889998336,
+      "grad_norm": 55.58504867553711,
+      "learning_rate": 0.00018392539964476022,
+      "loss": 0.2154,
+      "step": 371
+    },
+    {
+      "epoch": 0.2476285571642536,
+      "grad_norm": 55.62077331542969,
+      "learning_rate": 0.0001838809946714032,
+      "loss": 0.4984,
+      "step": 372
+    },
+    {
+      "epoch": 0.24829422532867365,
+      "grad_norm": 87.00872039794922,
+      "learning_rate": 0.0001838365896980462,
+      "loss": 0.4538,
+      "step": 373
+    },
+    {
+      "epoch": 0.2489598934930937,
+      "grad_norm": 46.597068786621094,
+      "learning_rate": 0.00018379218472468916,
+      "loss": 0.648,
+      "step": 374
+    },
+    {
+      "epoch": 0.24962556165751373,
+      "grad_norm": 99.84630584716797,
+      "learning_rate": 0.00018374777975133216,
+      "loss": 0.7931,
+      "step": 375
+    },
+    {
+      "epoch": 0.25029122982193375,
+      "grad_norm": 54.02254104614258,
+      "learning_rate": 0.00018370337477797513,
+      "loss": 0.2499,
+      "step": 376
+    },
+    {
+      "epoch": 0.2509568979863538,
+      "grad_norm": 29.319000244140625,
+      "learning_rate": 0.00018365896980461813,
+      "loss": 0.4007,
+      "step": 377
+    },
+    {
+      "epoch": 0.25162256615077383,
+      "grad_norm": 113.27352905273438,
+      "learning_rate": 0.0001836145648312611,
+      "loss": 0.7379,
+      "step": 378
+    },
+    {
+      "epoch": 0.25228823431519387,
+      "grad_norm": 112.12903594970703,
+      "learning_rate": 0.0001835701598579041,
+      "loss": 0.6326,
+      "step": 379
+    },
+    {
+      "epoch": 0.2529539024796139,
+      "grad_norm": 123.96865844726562,
+      "learning_rate": 0.00018352575488454707,
+      "loss": 0.5521,
+      "step": 380
+    },
+    {
+      "epoch": 0.25361957064403395,
+      "grad_norm": 42.07058334350586,
+      "learning_rate": 0.00018348134991119007,
+      "loss": 0.4259,
+      "step": 381
+    },
+    {
+      "epoch": 0.254285238808454,
+      "grad_norm": 55.002960205078125,
+      "learning_rate": 0.00018343694493783304,
+      "loss": 0.4797,
+      "step": 382
+    },
+    {
+      "epoch": 0.25495090697287404,
+      "grad_norm": 33.028038024902344,
+      "learning_rate": 0.00018339253996447604,
+      "loss": 0.2893,
+      "step": 383
+    },
+    {
+      "epoch": 0.2556165751372941,
+      "grad_norm": 54.003570556640625,
+      "learning_rate": 0.000183348134991119,
+      "loss": 0.4897,
+      "step": 384
+    },
+    {
+      "epoch": 0.2562822433017141,
+      "grad_norm": 80.12110137939453,
+      "learning_rate": 0.000183303730017762,
+      "loss": 0.2902,
+      "step": 385
+    },
+    {
+      "epoch": 0.2569479114661341,
+      "grad_norm": 46.586753845214844,
+      "learning_rate": 0.00018325932504440498,
+      "loss": 0.3721,
+      "step": 386
+    },
+    {
+      "epoch": 0.25761357963055415,
+      "grad_norm": 45.596553802490234,
+      "learning_rate": 0.00018321492007104798,
+      "loss": 0.2719,
+      "step": 387
+    },
+    {
+      "epoch": 0.2582792477949742,
+      "grad_norm": 90.06937408447266,
+      "learning_rate": 0.00018317051509769095,
+      "loss": 0.6424,
+      "step": 388
+    },
+    {
+      "epoch": 0.25894491595939423,
+      "grad_norm": 44.34123611450195,
+      "learning_rate": 0.00018312611012433395,
+      "loss": 0.2926,
+      "step": 389
+    },
+    {
+      "epoch": 0.25961058412381427,
+      "grad_norm": 123.05774688720703,
+      "learning_rate": 0.00018308170515097692,
+      "loss": 0.8703,
+      "step": 390
+    },
+    {
+      "epoch": 0.2602762522882343,
+      "grad_norm": 132.67784118652344,
+      "learning_rate": 0.00018303730017761992,
+      "loss": 0.492,
+      "step": 391
+    },
+    {
+      "epoch": 0.26094192045265435,
+      "grad_norm": 65.75527954101562,
+      "learning_rate": 0.0001829928952042629,
+      "loss": 0.2371,
+      "step": 392
+    },
+    {
+      "epoch": 0.2616075886170744,
+      "grad_norm": 70.54750061035156,
+      "learning_rate": 0.0001829484902309059,
+      "loss": 0.5192,
+      "step": 393
+    },
+    {
+      "epoch": 0.26227325678149443,
+      "grad_norm": 85.00897216796875,
+      "learning_rate": 0.00018290408525754886,
+      "loss": 0.7114,
+      "step": 394
+    },
+    {
+      "epoch": 0.2629389249459145,
+      "grad_norm": 96.78849792480469,
+      "learning_rate": 0.00018285968028419186,
+      "loss": 0.4515,
+      "step": 395
+    },
+    {
+      "epoch": 0.2636045931103345,
+      "grad_norm": 21.13523292541504,
+      "learning_rate": 0.00018281527531083483,
+      "loss": 0.3526,
+      "step": 396
+    },
+    {
+      "epoch": 0.26427026127475456,
+      "grad_norm": 29.27351188659668,
+      "learning_rate": 0.0001827708703374778,
+      "loss": 0.5151,
+      "step": 397
+    },
+    {
+      "epoch": 0.26493592943917454,
+      "grad_norm": 35.63692855834961,
+      "learning_rate": 0.0001827264653641208,
+      "loss": 0.2374,
+      "step": 398
+    },
+    {
+      "epoch": 0.2656015976035946,
+      "grad_norm": 66.08139038085938,
+      "learning_rate": 0.00018268206039076377,
+      "loss": 0.4133,
+      "step": 399
+    },
+    {
+      "epoch": 0.2662672657680146,
+      "grad_norm": 34.3465461730957,
+      "learning_rate": 0.00018263765541740677,
+      "loss": 0.3082,
+      "step": 400
+    },
+    {
+      "epoch": 0.26693293393243467,
+      "grad_norm": 100.62875366210938,
+      "learning_rate": 0.00018259325044404974,
+      "loss": 0.6884,
+      "step": 401
+    },
+    {
+      "epoch": 0.2675986020968547,
+      "grad_norm": 139.31932067871094,
+      "learning_rate": 0.00018254884547069274,
+      "loss": 1.0101,
+      "step": 402
+    },
+    {
+      "epoch": 0.26826427026127475,
+      "grad_norm": 118.92940521240234,
+      "learning_rate": 0.0001825044404973357,
+      "loss": 0.9706,
+      "step": 403
+    },
+    {
+      "epoch": 0.2689299384256948,
+      "grad_norm": 151.63421630859375,
+      "learning_rate": 0.0001824600355239787,
+      "loss": 0.961,
+      "step": 404
+    },
+    {
+      "epoch": 0.26959560659011483,
+      "grad_norm": 27.14082145690918,
+      "learning_rate": 0.00018241563055062168,
+      "loss": 0.227,
+      "step": 405
+    },
+    {
+      "epoch": 0.2702612747545349,
+      "grad_norm": 80.59782409667969,
+      "learning_rate": 0.00018237122557726468,
+      "loss": 0.6828,
+      "step": 406
+    },
+    {
+      "epoch": 0.2709269429189549,
+      "grad_norm": 49.46958541870117,
+      "learning_rate": 0.00018232682060390765,
+      "loss": 0.3184,
+      "step": 407
+    },
+    {
+      "epoch": 0.27159261108337496,
+      "grad_norm": 73.98738098144531,
+      "learning_rate": 0.00018228241563055065,
+      "loss": 0.6858,
+      "step": 408
+    },
+    {
+      "epoch": 0.272258279247795,
+      "grad_norm": 86.27637481689453,
+      "learning_rate": 0.00018223801065719362,
+      "loss": 0.5358,
+      "step": 409
+    },
+    {
+      "epoch": 0.272923947412215,
+      "grad_norm": 145.42340087890625,
+      "learning_rate": 0.0001821936056838366,
+      "loss": 1.0992,
+      "step": 410
+    },
+    {
+      "epoch": 0.273589615576635,
+      "grad_norm": 41.17599105834961,
+      "learning_rate": 0.0001821492007104796,
+      "loss": 0.369,
+      "step": 411
+    },
+    {
+      "epoch": 0.27425528374105507,
+      "grad_norm": 55.49694061279297,
+      "learning_rate": 0.00018210479573712256,
+      "loss": 0.2842,
+      "step": 412
+    },
+    {
+      "epoch": 0.2749209519054751,
+      "grad_norm": 47.01858139038086,
+      "learning_rate": 0.00018206039076376556,
+      "loss": 0.3368,
+      "step": 413
+    },
+    {
+      "epoch": 0.27558662006989515,
+      "grad_norm": 40.550880432128906,
+      "learning_rate": 0.00018201598579040853,
+      "loss": 0.2099,
+      "step": 414
+    },
+    {
+      "epoch": 0.2762522882343152,
+      "grad_norm": 38.81869125366211,
+      "learning_rate": 0.00018197158081705153,
+      "loss": 0.28,
+      "step": 415
+    },
+    {
+      "epoch": 0.27691795639873523,
+      "grad_norm": 65.78675842285156,
+      "learning_rate": 0.0001819271758436945,
+      "loss": 0.521,
+      "step": 416
+    },
+    {
+      "epoch": 0.2775836245631553,
+      "grad_norm": 73.13219451904297,
+      "learning_rate": 0.0001818827708703375,
+      "loss": 0.7937,
+      "step": 417
+    },
+    {
+      "epoch": 0.2782492927275753,
+      "grad_norm": 65.54695129394531,
+      "learning_rate": 0.00018183836589698047,
+      "loss": 0.4079,
+      "step": 418
+    },
+    {
+      "epoch": 0.27891496089199536,
+      "grad_norm": 34.30622863769531,
+      "learning_rate": 0.00018179396092362347,
+      "loss": 0.0964,
+      "step": 419
+    },
+    {
+      "epoch": 0.2795806290564154,
+      "grad_norm": 67.12702941894531,
+      "learning_rate": 0.00018174955595026644,
+      "loss": 1.4374,
+      "step": 420
+    },
+    {
+      "epoch": 0.28024629722083544,
+      "grad_norm": 33.8336181640625,
+      "learning_rate": 0.00018170515097690944,
+      "loss": 0.1509,
+      "step": 421
+    },
+    {
+      "epoch": 0.2809119653852554,
+      "grad_norm": 66.7548599243164,
+      "learning_rate": 0.0001816607460035524,
+      "loss": 0.2623,
+      "step": 422
+    },
+    {
+      "epoch": 0.28157763354967547,
+      "grad_norm": 57.463253021240234,
+      "learning_rate": 0.00018161634103019538,
+      "loss": 0.3713,
+      "step": 423
+    },
+    {
+      "epoch": 0.2822433017140955,
+      "grad_norm": 32.50680923461914,
+      "learning_rate": 0.00018157193605683838,
+      "loss": 0.1945,
+      "step": 424
+    },
+    {
+      "epoch": 0.28290896987851555,
+      "grad_norm": 72.0036849975586,
+      "learning_rate": 0.00018152753108348135,
+      "loss": 0.3324,
+      "step": 425
+    },
+    {
+      "epoch": 0.2835746380429356,
+      "grad_norm": 38.52781295776367,
+      "learning_rate": 0.00018148312611012435,
+      "loss": 0.185,
+      "step": 426
+    },
+    {
+      "epoch": 0.28424030620735563,
+      "grad_norm": 74.13616943359375,
+      "learning_rate": 0.00018143872113676732,
+      "loss": 0.2684,
+      "step": 427
+    },
+    {
+      "epoch": 0.28490597437177567,
+      "grad_norm": 76.92218017578125,
+      "learning_rate": 0.00018139431616341032,
+      "loss": 0.1456,
+      "step": 428
+    },
+    {
+      "epoch": 0.2855716425361957,
+      "grad_norm": 131.1004638671875,
+      "learning_rate": 0.0001813499111900533,
+      "loss": 0.6252,
+      "step": 429
+    },
+    {
+      "epoch": 0.28623731070061575,
+      "grad_norm": 112.32926940917969,
+      "learning_rate": 0.0001813055062166963,
+      "loss": 0.8337,
+      "step": 430
+    },
+    {
+      "epoch": 0.2869029788650358,
+      "grad_norm": 125.7912368774414,
+      "learning_rate": 0.00018126110124333926,
+      "loss": 0.3741,
+      "step": 431
+    },
+    {
+      "epoch": 0.28756864702945584,
+      "grad_norm": 164.3785858154297,
+      "learning_rate": 0.00018121669626998226,
+      "loss": 0.9616,
+      "step": 432
+    },
+    {
+      "epoch": 0.2882343151938759,
+      "grad_norm": 191.42367553710938,
+      "learning_rate": 0.00018117229129662523,
+      "loss": 0.8257,
+      "step": 433
+    },
+    {
+      "epoch": 0.28889998335829586,
+      "grad_norm": 79.08208465576172,
+      "learning_rate": 0.00018112788632326823,
+      "loss": 0.5202,
+      "step": 434
+    },
+    {
+      "epoch": 0.2895656515227159,
+      "grad_norm": 109.46979522705078,
+      "learning_rate": 0.0001810834813499112,
+      "loss": 0.941,
+      "step": 435
+    },
+    {
+      "epoch": 0.29023131968713595,
+      "grad_norm": 74.53866577148438,
+      "learning_rate": 0.00018103907637655417,
+      "loss": 0.7279,
+      "step": 436
+    },
+    {
+      "epoch": 0.290896987851556,
+      "grad_norm": 137.80848693847656,
+      "learning_rate": 0.00018099467140319717,
+      "loss": 1.4923,
+      "step": 437
+    },
+    {
+      "epoch": 0.29156265601597603,
+      "grad_norm": 126.38627624511719,
+      "learning_rate": 0.00018095026642984014,
+      "loss": 0.721,
+      "step": 438
+    },
+    {
+      "epoch": 0.29222832418039607,
+      "grad_norm": 98.57434844970703,
+      "learning_rate": 0.00018090586145648314,
+      "loss": 0.5846,
+      "step": 439
+    },
+    {
+      "epoch": 0.2928939923448161,
+      "grad_norm": 33.60892868041992,
+      "learning_rate": 0.0001808614564831261,
+      "loss": 0.4007,
+      "step": 440
+    },
+    {
+      "epoch": 0.29355966050923615,
+      "grad_norm": 84.4782943725586,
+      "learning_rate": 0.0001808170515097691,
+      "loss": 0.6742,
+      "step": 441
+    },
+    {
+      "epoch": 0.2942253286736562,
+      "grad_norm": 78.77970886230469,
+      "learning_rate": 0.00018077264653641208,
+      "loss": 0.5823,
+      "step": 442
+    },
+    {
+      "epoch": 0.29489099683807624,
+      "grad_norm": 157.4933319091797,
+      "learning_rate": 0.00018072824156305508,
+      "loss": 0.7795,
+      "step": 443
+    },
+    {
+      "epoch": 0.2955566650024963,
+      "grad_norm": 18.716060638427734,
+      "learning_rate": 0.00018068383658969805,
+      "loss": 0.3715,
+      "step": 444
+    },
+    {
+      "epoch": 0.2962223331669163,
+      "grad_norm": 85.15080261230469,
+      "learning_rate": 0.00018063943161634105,
+      "loss": 0.606,
+      "step": 445
+    },
+    {
+      "epoch": 0.2968880013313363,
+      "grad_norm": 83.07491302490234,
+      "learning_rate": 0.00018059502664298402,
+      "loss": 0.4614,
+      "step": 446
+    },
+    {
+      "epoch": 0.29755366949575635,
+      "grad_norm": 52.511085510253906,
+      "learning_rate": 0.000180550621669627,
+      "loss": 0.4219,
+      "step": 447
+    },
+    {
+      "epoch": 0.2982193376601764,
+      "grad_norm": 126.2627944946289,
+      "learning_rate": 0.00018050621669627,
+      "loss": 0.6889,
+      "step": 448
+    },
+    {
+      "epoch": 0.29888500582459643,
+      "grad_norm": 73.53312683105469,
+      "learning_rate": 0.00018046181172291296,
+      "loss": 0.577,
+      "step": 449
+    },
+    {
+      "epoch": 0.29955067398901647,
+      "grad_norm": 103.77356719970703,
+      "learning_rate": 0.00018041740674955596,
+      "loss": 0.7924,
+      "step": 450
+    },
+    {
+      "epoch": 0.3002163421534365,
+      "grad_norm": 89.16803741455078,
+      "learning_rate": 0.00018037300177619893,
+      "loss": 0.5117,
+      "step": 451
+    },
+    {
+      "epoch": 0.30088201031785655,
+      "grad_norm": 116.20187377929688,
+      "learning_rate": 0.00018032859680284193,
+      "loss": 0.6893,
+      "step": 452
+    },
+    {
+      "epoch": 0.3015476784822766,
+      "grad_norm": 130.5649871826172,
+      "learning_rate": 0.0001802841918294849,
+      "loss": 0.6782,
+      "step": 453
+    },
+    {
+      "epoch": 0.30221334664669663,
+      "grad_norm": 71.50609588623047,
+      "learning_rate": 0.0001802397868561279,
+      "loss": 0.4608,
+      "step": 454
+    },
+    {
+      "epoch": 0.3028790148111167,
+      "grad_norm": 101.877197265625,
+      "learning_rate": 0.00018019538188277087,
+      "loss": 0.5512,
+      "step": 455
+    },
+    {
+      "epoch": 0.3035446829755367,
+      "grad_norm": 135.80545043945312,
+      "learning_rate": 0.00018015097690941387,
+      "loss": 0.6684,
+      "step": 456
+    },
+    {
+      "epoch": 0.30421035113995676,
+      "grad_norm": 83.38865661621094,
+      "learning_rate": 0.00018010657193605684,
+      "loss": 0.4582,
+      "step": 457
+    },
+    {
+      "epoch": 0.30487601930437674,
+      "grad_norm": 93.57701110839844,
+      "learning_rate": 0.00018006216696269984,
+      "loss": 0.7277,
+      "step": 458
+    },
+    {
+      "epoch": 0.3055416874687968,
+      "grad_norm": 72.99111938476562,
+      "learning_rate": 0.0001800177619893428,
+      "loss": 0.5737,
+      "step": 459
+    },
+    {
+      "epoch": 0.3062073556332168,
+      "grad_norm": 64.9205551147461,
+      "learning_rate": 0.00017997335701598578,
+      "loss": 0.3831,
+      "step": 460
+    },
+    {
+      "epoch": 0.30687302379763687,
+      "grad_norm": 36.749935150146484,
+      "learning_rate": 0.00017992895204262878,
+      "loss": 0.4547,
+      "step": 461
+    },
+    {
+      "epoch": 0.3075386919620569,
+      "grad_norm": 77.0653076171875,
+      "learning_rate": 0.00017988454706927175,
+      "loss": 0.8629,
+      "step": 462
+    },
+    {
+      "epoch": 0.30820436012647695,
+      "grad_norm": 48.22278594970703,
+      "learning_rate": 0.00017984014209591475,
+      "loss": 0.3634,
+      "step": 463
+    },
+    {
+      "epoch": 0.308870028290897,
+      "grad_norm": 39.0225715637207,
+      "learning_rate": 0.00017979573712255772,
+      "loss": 0.317,
+      "step": 464
+    },
+    {
+      "epoch": 0.30953569645531703,
+      "grad_norm": 68.747314453125,
+      "learning_rate": 0.00017975133214920072,
+      "loss": 0.3624,
+      "step": 465
+    },
+    {
+      "epoch": 0.3102013646197371,
+      "grad_norm": 71.55371856689453,
+      "learning_rate": 0.0001797069271758437,
+      "loss": 0.3992,
+      "step": 466
+    },
+    {
+      "epoch": 0.3108670327841571,
+      "grad_norm": 36.87495422363281,
+      "learning_rate": 0.00017966252220248669,
+      "loss": 0.2098,
+      "step": 467
+    },
+    {
+      "epoch": 0.31153270094857716,
+      "grad_norm": 10.664894104003906,
+      "learning_rate": 0.00017961811722912966,
+      "loss": 0.0542,
+      "step": 468
+    },
+    {
+      "epoch": 0.3121983691129972,
+      "grad_norm": 34.32514953613281,
+      "learning_rate": 0.00017957371225577265,
+      "loss": 0.3147,
+      "step": 469
+    },
+    {
+      "epoch": 0.3128640372774172,
+      "grad_norm": 24.026748657226562,
+      "learning_rate": 0.00017952930728241563,
+      "loss": 0.0797,
+      "step": 470
+    },
+    {
+      "epoch": 0.3135297054418372,
+      "grad_norm": 23.764530181884766,
+      "learning_rate": 0.00017948490230905862,
+      "loss": 0.0965,
+      "step": 471
+    },
+    {
+      "epoch": 0.31419537360625727,
+      "grad_norm": 5.237178325653076,
+      "learning_rate": 0.0001794404973357016,
+      "loss": 0.0106,
+      "step": 472
+    },
+    {
+      "epoch": 0.3148610417706773,
+      "grad_norm": 20.264196395874023,
+      "learning_rate": 0.0001793960923623446,
+      "loss": 0.0546,
+      "step": 473
+    },
+    {
+      "epoch": 0.31552670993509735,
+      "grad_norm": 79.56272888183594,
+      "learning_rate": 0.00017935168738898757,
+      "loss": 0.5564,
+      "step": 474
+    },
+    {
+      "epoch": 0.3161923780995174,
+      "grad_norm": 37.59955596923828,
+      "learning_rate": 0.00017930728241563056,
+      "loss": 0.3443,
+      "step": 475
+    },
+    {
+      "epoch": 0.31685804626393743,
+      "grad_norm": 49.22467803955078,
+      "learning_rate": 0.00017926287744227354,
+      "loss": 0.3749,
+      "step": 476
+    },
+    {
+      "epoch": 0.3175237144283575,
+      "grad_norm": 175.10850524902344,
+      "learning_rate": 0.00017921847246891653,
+      "loss": 1.7856,
+      "step": 477
+    },
+    {
+      "epoch": 0.3181893825927775,
+      "grad_norm": 47.71009826660156,
+      "learning_rate": 0.0001791740674955595,
+      "loss": 0.7082,
+      "step": 478
+    },
+    {
+      "epoch": 0.31885505075719756,
+      "grad_norm": 82.27182006835938,
+      "learning_rate": 0.0001791296625222025,
+      "loss": 0.9351,
+      "step": 479
+    },
+    {
+      "epoch": 0.3195207189216176,
+      "grad_norm": 79.33952331542969,
+      "learning_rate": 0.00017908525754884547,
+      "loss": 0.3707,
+      "step": 480
+    },
+    {
+      "epoch": 0.3201863870860376,
+      "grad_norm": 88.69058227539062,
+      "learning_rate": 0.00017904085257548847,
+      "loss": 0.85,
+      "step": 481
+    },
+    {
+      "epoch": 0.3208520552504576,
+      "grad_norm": 110.95427703857422,
+      "learning_rate": 0.00017899644760213144,
+      "loss": 0.3507,
+      "step": 482
+    },
+    {
+      "epoch": 0.32151772341487767,
+      "grad_norm": 101.72648620605469,
+      "learning_rate": 0.00017895204262877444,
+      "loss": 0.8123,
+      "step": 483
+    },
+    {
+      "epoch": 0.3221833915792977,
+      "grad_norm": 172.9845733642578,
+      "learning_rate": 0.00017890763765541741,
+      "loss": 1.1193,
+      "step": 484
+    },
+    {
+      "epoch": 0.32284905974371775,
+      "grad_norm": 40.887081146240234,
+      "learning_rate": 0.0001788632326820604,
+      "loss": 0.1791,
+      "step": 485
+    },
+    {
+      "epoch": 0.3235147279081378,
+      "grad_norm": 89.45999145507812,
+      "learning_rate": 0.00017881882770870338,
+      "loss": 0.6104,
+      "step": 486
+    },
+    {
+      "epoch": 0.32418039607255783,
+      "grad_norm": 101.5306167602539,
+      "learning_rate": 0.00017877442273534638,
+      "loss": 0.5197,
+      "step": 487
+    },
+    {
+      "epoch": 0.32484606423697787,
+      "grad_norm": 25.401044845581055,
+      "learning_rate": 0.00017873001776198935,
+      "loss": 0.2328,
+      "step": 488
+    },
+    {
+      "epoch": 0.3255117324013979,
+      "grad_norm": 122.61588287353516,
+      "learning_rate": 0.00017868561278863235,
+      "loss": 0.9357,
+      "step": 489
+    },
+    {
+      "epoch": 0.32617740056581795,
+      "grad_norm": 51.50986099243164,
+      "learning_rate": 0.00017864120781527532,
+      "loss": 0.506,
+      "step": 490
+    },
+    {
+      "epoch": 0.326843068730238,
+      "grad_norm": 64.19971466064453,
+      "learning_rate": 0.00017859680284191832,
+      "loss": 0.2713,
+      "step": 491
+    },
+    {
+      "epoch": 0.32750873689465804,
+      "grad_norm": 19.61275863647461,
+      "learning_rate": 0.0001785523978685613,
+      "loss": 0.1013,
+      "step": 492
+    },
+    {
+      "epoch": 0.328174405059078,
+      "grad_norm": 66.36785125732422,
+      "learning_rate": 0.0001785079928952043,
+      "loss": 0.3726,
+      "step": 493
+    },
+    {
+      "epoch": 0.32884007322349806,
+      "grad_norm": 87.43013763427734,
+      "learning_rate": 0.00017846358792184726,
+      "loss": 0.4851,
+      "step": 494
+    },
+    {
+      "epoch": 0.3295057413879181,
+      "grad_norm": 63.96004867553711,
+      "learning_rate": 0.00017841918294849026,
+      "loss": 0.2111,
+      "step": 495
+    },
+    {
+      "epoch": 0.33017140955233815,
+      "grad_norm": 81.01036834716797,
+      "learning_rate": 0.00017837477797513323,
+      "loss": 0.6494,
+      "step": 496
+    },
+    {
+      "epoch": 0.3308370777167582,
+      "grad_norm": 63.75148010253906,
+      "learning_rate": 0.0001783303730017762,
+      "loss": 0.3041,
+      "step": 497
+    },
+    {
+      "epoch": 0.33150274588117823,
+      "grad_norm": 91.487060546875,
+      "learning_rate": 0.0001782859680284192,
+      "loss": 0.689,
+      "step": 498
+    },
+    {
+      "epoch": 0.33216841404559827,
+      "grad_norm": 58.02032470703125,
+      "learning_rate": 0.00017824156305506217,
+      "loss": 0.5346,
+      "step": 499
+    },
+    {
+      "epoch": 0.3328340822100183,
+      "grad_norm": 73.02722930908203,
+      "learning_rate": 0.00017819715808170517,
+      "loss": 0.541,
+      "step": 500
+    },
+    {
+      "epoch": 0.33349975037443835,
+      "grad_norm": 118.74954986572266,
+      "learning_rate": 0.00017815275310834814,
+      "loss": 0.5354,
+      "step": 501
+    },
+    {
+      "epoch": 0.3341654185388584,
+      "grad_norm": 40.1008415222168,
+      "learning_rate": 0.00017810834813499114,
+      "loss": 0.2462,
+      "step": 502
+    },
+    {
+      "epoch": 0.33483108670327844,
+      "grad_norm": 63.85152053833008,
+      "learning_rate": 0.0001780639431616341,
+      "loss": 0.5912,
+      "step": 503
+    },
+    {
+      "epoch": 0.3354967548676985,
+      "grad_norm": 48.006710052490234,
+      "learning_rate": 0.0001780195381882771,
+      "loss": 0.5213,
+      "step": 504
+    },
+    {
+      "epoch": 0.33616242303211846,
+      "grad_norm": 72.3712387084961,
+      "learning_rate": 0.00017797513321492008,
+      "loss": 0.3234,
+      "step": 505
+    },
+    {
+      "epoch": 0.3368280911965385,
+      "grad_norm": 31.964515686035156,
+      "learning_rate": 0.00017793072824156308,
+      "loss": 0.1817,
+      "step": 506
+    },
+    {
+      "epoch": 0.33749375936095855,
+      "grad_norm": 113.52684783935547,
+      "learning_rate": 0.00017788632326820605,
+      "loss": 0.6327,
+      "step": 507
+    },
+    {
+      "epoch": 0.3381594275253786,
+      "grad_norm": 50.74462127685547,
+      "learning_rate": 0.00017784191829484905,
+      "loss": 0.3449,
+      "step": 508
+    },
+    {
+      "epoch": 0.3388250956897986,
+      "grad_norm": 46.287391662597656,
+      "learning_rate": 0.00017779751332149202,
+      "loss": 0.2474,
+      "step": 509
+    },
+    {
+      "epoch": 0.33949076385421867,
+      "grad_norm": 62.21271514892578,
+      "learning_rate": 0.000177753108348135,
+      "loss": 0.2743,
+      "step": 510
+    },
+    {
+      "epoch": 0.3401564320186387,
+      "grad_norm": 52.463260650634766,
+      "learning_rate": 0.000177708703374778,
+      "loss": 0.3754,
+      "step": 511
+    },
+    {
+      "epoch": 0.34082210018305875,
+      "grad_norm": 70.85062408447266,
+      "learning_rate": 0.00017766429840142096,
+      "loss": 0.3937,
+      "step": 512
+    },
+    {
+      "epoch": 0.3414877683474788,
+      "grad_norm": 86.9253158569336,
+      "learning_rate": 0.00017761989342806396,
+      "loss": 0.3501,
+      "step": 513
+    },
+    {
+      "epoch": 0.34215343651189883,
+      "grad_norm": 51.851112365722656,
+      "learning_rate": 0.00017757548845470693,
+      "loss": 0.2259,
+      "step": 514
+    },
+    {
+      "epoch": 0.3428191046763189,
+      "grad_norm": 68.38616943359375,
+      "learning_rate": 0.00017753108348134993,
+      "loss": 0.4034,
+      "step": 515
+    },
+    {
+      "epoch": 0.3434847728407389,
+      "grad_norm": 189.30650329589844,
+      "learning_rate": 0.0001774866785079929,
+      "loss": 0.8131,
+      "step": 516
+    },
+    {
+      "epoch": 0.3441504410051589,
+      "grad_norm": 30.575939178466797,
+      "learning_rate": 0.0001774422735346359,
+      "loss": 0.1302,
+      "step": 517
+    },
+    {
+      "epoch": 0.34481610916957894,
+      "grad_norm": 14.13813304901123,
+      "learning_rate": 0.00017739786856127887,
+      "loss": 0.1055,
+      "step": 518
+    },
+    {
+      "epoch": 0.345481777333999,
+      "grad_norm": 124.42664337158203,
+      "learning_rate": 0.00017735346358792187,
+      "loss": 0.4918,
+      "step": 519
+    },
+    {
+      "epoch": 0.346147445498419,
+      "grad_norm": 62.3731575012207,
+      "learning_rate": 0.00017730905861456484,
+      "loss": 0.2553,
+      "step": 520
+    },
+    {
+      "epoch": 0.34681311366283907,
+      "grad_norm": 91.36651611328125,
+      "learning_rate": 0.00017726465364120784,
+      "loss": 0.3425,
+      "step": 521
+    },
+    {
+      "epoch": 0.3474787818272591,
+      "grad_norm": 102.5283432006836,
+      "learning_rate": 0.0001772202486678508,
+      "loss": 0.4142,
+      "step": 522
+    },
+    {
+      "epoch": 0.34814444999167915,
+      "grad_norm": 48.83149337768555,
+      "learning_rate": 0.00017717584369449378,
+      "loss": 0.327,
+      "step": 523
+    },
+    {
+      "epoch": 0.3488101181560992,
+      "grad_norm": 68.21099090576172,
+      "learning_rate": 0.00017713143872113678,
+      "loss": 0.5076,
+      "step": 524
+    },
+    {
+      "epoch": 0.34947578632051923,
+      "grad_norm": 50.855072021484375,
+      "learning_rate": 0.00017708703374777975,
+      "loss": 0.7123,
+      "step": 525
+    },
+    {
+      "epoch": 0.3501414544849393,
+      "grad_norm": 95.03412628173828,
+      "learning_rate": 0.00017704262877442275,
+      "loss": 0.4952,
+      "step": 526
+    },
+    {
+      "epoch": 0.3508071226493593,
+      "grad_norm": 45.51231384277344,
+      "learning_rate": 0.00017699822380106572,
+      "loss": 0.2897,
+      "step": 527
+    },
+    {
+      "epoch": 0.35147279081377936,
+      "grad_norm": 83.111083984375,
+      "learning_rate": 0.00017695381882770872,
+      "loss": 0.3769,
+      "step": 528
+    },
+    {
+      "epoch": 0.35213845897819934,
+      "grad_norm": 222.33758544921875,
+      "learning_rate": 0.0001769094138543517,
+      "loss": 1.4976,
+      "step": 529
+    },
+    {
+      "epoch": 0.3528041271426194,
+      "grad_norm": 117.77812957763672,
+      "learning_rate": 0.0001768650088809947,
+      "loss": 1.0537,
+      "step": 530
+    },
+    {
+      "epoch": 0.3534697953070394,
+      "grad_norm": 143.2985382080078,
+      "learning_rate": 0.00017682060390763766,
+      "loss": 1.206,
+      "step": 531
+    },
+    {
+      "epoch": 0.35413546347145947,
+      "grad_norm": 78.69023132324219,
+      "learning_rate": 0.00017677619893428066,
+      "loss": 0.4841,
+      "step": 532
+    },
+    {
+      "epoch": 0.3548011316358795,
+      "grad_norm": 30.563365936279297,
+      "learning_rate": 0.00017673179396092363,
+      "loss": 0.213,
+      "step": 533
+    },
+    {
+      "epoch": 0.35546679980029955,
+      "grad_norm": 128.74452209472656,
+      "learning_rate": 0.00017668738898756663,
+      "loss": 0.9367,
+      "step": 534
+    },
+    {
+      "epoch": 0.3561324679647196,
+      "grad_norm": 68.94506072998047,
+      "learning_rate": 0.0001766429840142096,
+      "loss": 0.671,
+      "step": 535
+    },
+    {
+      "epoch": 0.35679813612913963,
+      "grad_norm": 87.38471984863281,
+      "learning_rate": 0.00017659857904085257,
+      "loss": 0.4509,
+      "step": 536
+    },
+    {
+      "epoch": 0.3574638042935597,
+      "grad_norm": 127.65184783935547,
+      "learning_rate": 0.00017655417406749557,
+      "loss": 0.6365,
+      "step": 537
+    },
+    {
+      "epoch": 0.3581294724579797,
+      "grad_norm": 23.673858642578125,
+      "learning_rate": 0.00017650976909413854,
+      "loss": 0.1011,
+      "step": 538
+    },
+    {
+      "epoch": 0.35879514062239976,
+      "grad_norm": 41.28850173950195,
+      "learning_rate": 0.00017646536412078154,
+      "loss": 1.0469,
+      "step": 539
+    },
+    {
+      "epoch": 0.3594608087868198,
+      "grad_norm": 66.7071304321289,
+      "learning_rate": 0.0001764209591474245,
+      "loss": 0.382,
+      "step": 540
+    },
+    {
+      "epoch": 0.3601264769512398,
+      "grad_norm": 41.335899353027344,
+      "learning_rate": 0.0001763765541740675,
+      "loss": 0.3328,
+      "step": 541
+    },
+    {
+      "epoch": 0.3607921451156598,
+      "grad_norm": 65.9492416381836,
+      "learning_rate": 0.00017633214920071048,
+      "loss": 0.2534,
+      "step": 542
+    },
+    {
+      "epoch": 0.36145781328007986,
+      "grad_norm": 51.26821517944336,
+      "learning_rate": 0.00017628774422735348,
+      "loss": 0.2136,
+      "step": 543
+    },
+    {
+      "epoch": 0.3621234814444999,
+      "grad_norm": 80.04798126220703,
+      "learning_rate": 0.00017624333925399645,
+      "loss": 0.3201,
+      "step": 544
+    },
+    {
+      "epoch": 0.36278914960891995,
+      "grad_norm": 29.61306381225586,
+      "learning_rate": 0.00017619893428063945,
+      "loss": 0.361,
+      "step": 545
+    },
+    {
+      "epoch": 0.36345481777334,
+      "grad_norm": 57.36445617675781,
+      "learning_rate": 0.00017615452930728242,
+      "loss": 0.3203,
+      "step": 546
+    },
+    {
+      "epoch": 0.36412048593776003,
+      "grad_norm": 22.750394821166992,
+      "learning_rate": 0.0001761101243339254,
+      "loss": 0.0615,
+      "step": 547
+    },
+    {
+      "epoch": 0.36478615410218007,
+      "grad_norm": 92.89879608154297,
+      "learning_rate": 0.0001760657193605684,
+      "loss": 0.7158,
+      "step": 548
+    },
+    {
+      "epoch": 0.3654518222666001,
+      "grad_norm": 75.37178039550781,
+      "learning_rate": 0.00017602131438721136,
+      "loss": 0.2596,
+      "step": 549
+    },
+    {
+      "epoch": 0.36611749043102015,
+      "grad_norm": 77.78299713134766,
+      "learning_rate": 0.00017597690941385436,
+      "loss": 0.5387,
+      "step": 550
+    },
+    {
+      "epoch": 0.3667831585954402,
+      "grad_norm": 106.28474426269531,
+      "learning_rate": 0.00017593250444049733,
+      "loss": 0.5178,
+      "step": 551
+    },
+    {
+      "epoch": 0.36744882675986024,
+      "grad_norm": 67.27828979492188,
+      "learning_rate": 0.00017588809946714033,
+      "loss": 0.4135,
+      "step": 552
+    },
+    {
+      "epoch": 0.3681144949242802,
+      "grad_norm": 128.6137237548828,
+      "learning_rate": 0.0001758436944937833,
+      "loss": 0.591,
+      "step": 553
+    },
+    {
+      "epoch": 0.36878016308870026,
+      "grad_norm": 69.33638763427734,
+      "learning_rate": 0.0001757992895204263,
+      "loss": 0.653,
+      "step": 554
+    },
+    {
+      "epoch": 0.3694458312531203,
+      "grad_norm": 50.19843292236328,
+      "learning_rate": 0.00017575488454706927,
+      "loss": 0.2516,
+      "step": 555
+    },
+    {
+      "epoch": 0.37011149941754035,
+      "grad_norm": 16.58024787902832,
+      "learning_rate": 0.00017571047957371227,
+      "loss": 0.26,
+      "step": 556
+    },
+    {
+      "epoch": 0.3707771675819604,
+      "grad_norm": 25.605676651000977,
+      "learning_rate": 0.00017566607460035524,
+      "loss": 0.249,
+      "step": 557
+    },
+    {
+      "epoch": 0.37144283574638043,
+      "grad_norm": 56.728515625,
+      "learning_rate": 0.00017562166962699824,
+      "loss": 0.3381,
+      "step": 558
+    },
+    {
+      "epoch": 0.37210850391080047,
+      "grad_norm": 27.297210693359375,
+      "learning_rate": 0.0001755772646536412,
+      "loss": 0.111,
+      "step": 559
+    },
+    {
+      "epoch": 0.3727741720752205,
+      "grad_norm": 85.54072570800781,
+      "learning_rate": 0.00017553285968028418,
+      "loss": 0.4148,
+      "step": 560
+    },
+    {
+      "epoch": 0.37343984023964055,
+      "grad_norm": 80.2583236694336,
+      "learning_rate": 0.00017548845470692718,
+      "loss": 0.4845,
+      "step": 561
+    },
+    {
+      "epoch": 0.3741055084040606,
+      "grad_norm": 33.99844741821289,
+      "learning_rate": 0.00017544404973357015,
+      "loss": 0.2521,
+      "step": 562
+    },
+    {
+      "epoch": 0.37477117656848064,
+      "grad_norm": 55.42804718017578,
+      "learning_rate": 0.00017539964476021315,
+      "loss": 0.1915,
+      "step": 563
+    },
+    {
+      "epoch": 0.3754368447329007,
+      "grad_norm": 34.7382926940918,
+      "learning_rate": 0.00017535523978685612,
+      "loss": 0.1789,
+      "step": 564
+    },
+    {
+      "epoch": 0.37610251289732066,
+      "grad_norm": 67.93021392822266,
+      "learning_rate": 0.00017531083481349912,
+      "loss": 0.4769,
+      "step": 565
+    },
+    {
+      "epoch": 0.3767681810617407,
+      "grad_norm": 83.29434967041016,
+      "learning_rate": 0.0001752664298401421,
+      "loss": 0.9316,
+      "step": 566
+    },
+    {
+      "epoch": 0.37743384922616074,
+      "grad_norm": 29.34658432006836,
+      "learning_rate": 0.0001752220248667851,
+      "loss": 0.1181,
+      "step": 567
+    },
+    {
+      "epoch": 0.3780995173905808,
+      "grad_norm": 73.09173583984375,
+      "learning_rate": 0.00017517761989342806,
+      "loss": 0.2341,
+      "step": 568
+    },
+    {
+      "epoch": 0.3787651855550008,
+      "grad_norm": 100.28826141357422,
+      "learning_rate": 0.00017513321492007106,
+      "loss": 0.7953,
+      "step": 569
+    },
+    {
+      "epoch": 0.37943085371942087,
+      "grad_norm": 34.65471267700195,
+      "learning_rate": 0.00017508880994671403,
+      "loss": 0.1516,
+      "step": 570
+    },
+    {
+      "epoch": 0.3800965218838409,
+      "grad_norm": 40.57853317260742,
+      "learning_rate": 0.00017504440497335703,
+      "loss": 0.4384,
+      "step": 571
+    },
+    {
+      "epoch": 0.38076219004826095,
+      "grad_norm": 42.92374038696289,
+      "learning_rate": 0.000175,
+      "loss": 0.7244,
+      "step": 572
+    },
+    {
+      "epoch": 0.381427858212681,
+      "grad_norm": 43.757568359375,
+      "learning_rate": 0.00017495559502664297,
+      "loss": 0.2562,
+      "step": 573
+    },
+    {
+      "epoch": 0.38209352637710103,
+      "grad_norm": 53.40153503417969,
+      "learning_rate": 0.00017491119005328597,
+      "loss": 0.272,
+      "step": 574
+    },
+    {
+      "epoch": 0.3827591945415211,
+      "grad_norm": 58.91749572753906,
+      "learning_rate": 0.00017486678507992894,
+      "loss": 0.2955,
+      "step": 575
+    },
+    {
+      "epoch": 0.38342486270594106,
+      "grad_norm": 45.21612548828125,
+      "learning_rate": 0.00017482238010657194,
+      "loss": 0.1861,
+      "step": 576
+    },
+    {
+      "epoch": 0.3840905308703611,
+      "grad_norm": 179.73159790039062,
+      "learning_rate": 0.0001747779751332149,
+      "loss": 1.1993,
+      "step": 577
+    },
+    {
+      "epoch": 0.38475619903478114,
+      "grad_norm": 129.5082550048828,
+      "learning_rate": 0.0001747335701598579,
+      "loss": 0.8181,
+      "step": 578
+    },
+    {
+      "epoch": 0.3854218671992012,
+      "grad_norm": 39.42813491821289,
+      "learning_rate": 0.00017468916518650088,
+      "loss": 0.5695,
+      "step": 579
+    },
+    {
+      "epoch": 0.3860875353636212,
+      "grad_norm": 57.18281173706055,
+      "learning_rate": 0.00017464476021314388,
+      "loss": 0.2304,
+      "step": 580
+    },
+    {
+      "epoch": 0.38675320352804127,
+      "grad_norm": 43.29937744140625,
+      "learning_rate": 0.00017460035523978685,
+      "loss": 0.235,
+      "step": 581
+    },
+    {
+      "epoch": 0.3874188716924613,
+      "grad_norm": 60.49991989135742,
+      "learning_rate": 0.00017455595026642985,
+      "loss": 0.1998,
+      "step": 582
+    },
+    {
+      "epoch": 0.38808453985688135,
+      "grad_norm": 82.13057708740234,
+      "learning_rate": 0.00017451154529307282,
+      "loss": 0.4337,
+      "step": 583
+    },
+    {
+      "epoch": 0.3887502080213014,
+      "grad_norm": 29.797197341918945,
+      "learning_rate": 0.00017446714031971582,
+      "loss": 0.1105,
+      "step": 584
+    },
+    {
+      "epoch": 0.38941587618572143,
+      "grad_norm": 55.97276306152344,
+      "learning_rate": 0.0001744227353463588,
+      "loss": 0.3703,
+      "step": 585
+    },
+    {
+      "epoch": 0.3900815443501415,
+      "grad_norm": 59.474639892578125,
+      "learning_rate": 0.00017437833037300179,
+      "loss": 0.3482,
+      "step": 586
+    },
+    {
+      "epoch": 0.3907472125145615,
+      "grad_norm": 49.60660934448242,
+      "learning_rate": 0.00017433392539964478,
+      "loss": 0.5983,
+      "step": 587
+    },
+    {
+      "epoch": 0.3914128806789815,
+      "grad_norm": 108.58460998535156,
+      "learning_rate": 0.00017428952042628776,
+      "loss": 0.9933,
+      "step": 588
+    },
+    {
+      "epoch": 0.39207854884340154,
+      "grad_norm": 49.046234130859375,
+      "learning_rate": 0.00017424511545293075,
+      "loss": 0.1581,
+      "step": 589
+    },
+    {
+      "epoch": 0.3927442170078216,
+      "grad_norm": 62.635292053222656,
+      "learning_rate": 0.00017420071047957373,
+      "loss": 0.6443,
+      "step": 590
+    },
+    {
+      "epoch": 0.3934098851722416,
+      "grad_norm": 65.41645812988281,
+      "learning_rate": 0.00017415630550621672,
+      "loss": 0.5636,
+      "step": 591
+    },
+    {
+      "epoch": 0.39407555333666167,
+      "grad_norm": 84.65705871582031,
+      "learning_rate": 0.0001741119005328597,
+      "loss": 0.5554,
+      "step": 592
+    },
+    {
+      "epoch": 0.3947412215010817,
+      "grad_norm": 58.44446563720703,
+      "learning_rate": 0.0001740674955595027,
+      "loss": 0.2876,
+      "step": 593
+    },
+    {
+      "epoch": 0.39540688966550175,
+      "grad_norm": 91.17931365966797,
+      "learning_rate": 0.00017402309058614566,
+      "loss": 0.4579,
+      "step": 594
+    },
+    {
+      "epoch": 0.3960725578299218,
+      "grad_norm": 49.55839920043945,
+      "learning_rate": 0.00017397868561278866,
+      "loss": 0.5159,
+      "step": 595
+    },
+    {
+      "epoch": 0.39673822599434183,
+      "grad_norm": 38.55497360229492,
+      "learning_rate": 0.00017393428063943163,
+      "loss": 0.2211,
+      "step": 596
+    },
+    {
+      "epoch": 0.3974038941587619,
+      "grad_norm": 72.52485656738281,
+      "learning_rate": 0.00017388987566607463,
+      "loss": 0.8013,
+      "step": 597
+    },
+    {
+      "epoch": 0.3980695623231819,
+      "grad_norm": 42.55357360839844,
+      "learning_rate": 0.0001738454706927176,
+      "loss": 0.1123,
+      "step": 598
+    },
+    {
+      "epoch": 0.39873523048760195,
+      "grad_norm": 75.76644897460938,
+      "learning_rate": 0.00017380106571936058,
+      "loss": 0.6063,
+      "step": 599
+    },
+    {
+      "epoch": 0.39940089865202194,
+      "grad_norm": 49.05486297607422,
+      "learning_rate": 0.00017375666074600357,
+      "loss": 0.3929,
+      "step": 600
+    },
+    {
+      "epoch": 0.400066566816442,
+      "grad_norm": 54.81421661376953,
+      "learning_rate": 0.00017371225577264654,
+      "loss": 0.3706,
+      "step": 601
+    },
+    {
+      "epoch": 0.400732234980862,
+      "grad_norm": 45.12424087524414,
+      "learning_rate": 0.00017366785079928954,
+      "loss": 0.1521,
+      "step": 602
+    },
+    {
+      "epoch": 0.40139790314528206,
+      "grad_norm": 159.4431915283203,
+      "learning_rate": 0.00017362344582593251,
+      "loss": 0.9427,
+      "step": 603
+    },
+    {
+      "epoch": 0.4020635713097021,
+      "grad_norm": 37.86516189575195,
+      "learning_rate": 0.0001735790408525755,
+      "loss": 0.3125,
+      "step": 604
+    },
+    {
+      "epoch": 0.40272923947412215,
+      "grad_norm": 122.04438781738281,
+      "learning_rate": 0.00017353463587921848,
+      "loss": 0.6369,
+      "step": 605
+    },
+    {
+      "epoch": 0.4033949076385422,
+      "grad_norm": 29.7852725982666,
+      "learning_rate": 0.00017349023090586148,
+      "loss": 0.0806,
+      "step": 606
+    },
+    {
+      "epoch": 0.40406057580296223,
+      "grad_norm": 87.71656799316406,
+      "learning_rate": 0.00017344582593250445,
+      "loss": 0.624,
+      "step": 607
+    },
+    {
+      "epoch": 0.40472624396738227,
+      "grad_norm": 91.41803741455078,
+      "learning_rate": 0.00017340142095914745,
+      "loss": 0.9653,
+      "step": 608
+    },
+    {
+      "epoch": 0.4053919121318023,
+      "grad_norm": 116.02742004394531,
+      "learning_rate": 0.00017335701598579042,
+      "loss": 0.6518,
+      "step": 609
+    },
+    {
+      "epoch": 0.40605758029622235,
+      "grad_norm": 39.57069778442383,
+      "learning_rate": 0.0001733126110124334,
+      "loss": 0.1943,
+      "step": 610
+    },
+    {
+      "epoch": 0.4067232484606424,
+      "grad_norm": 65.77499389648438,
+      "learning_rate": 0.0001732682060390764,
+      "loss": 0.379,
+      "step": 611
+    },
+    {
+      "epoch": 0.4073889166250624,
+      "grad_norm": 110.7475814819336,
+      "learning_rate": 0.00017322380106571936,
+      "loss": 0.618,
+      "step": 612
+    },
+    {
+      "epoch": 0.4080545847894824,
+      "grad_norm": 43.67226791381836,
+      "learning_rate": 0.00017317939609236236,
+      "loss": 0.2807,
+      "step": 613
+    },
+    {
+      "epoch": 0.40872025295390246,
+      "grad_norm": 64.67245483398438,
+      "learning_rate": 0.00017313499111900533,
+      "loss": 0.4116,
+      "step": 614
+    },
+    {
+      "epoch": 0.4093859211183225,
+      "grad_norm": 37.3428840637207,
+      "learning_rate": 0.00017309058614564833,
+      "loss": 0.7732,
+      "step": 615
+    },
+    {
+      "epoch": 0.41005158928274255,
+      "grad_norm": 106.12779998779297,
+      "learning_rate": 0.0001730461811722913,
+      "loss": 0.7123,
+      "step": 616
+    },
+    {
+      "epoch": 0.4107172574471626,
+      "grad_norm": 89.30298614501953,
+      "learning_rate": 0.0001730017761989343,
+      "loss": 0.4929,
+      "step": 617
+    },
+    {
+      "epoch": 0.41138292561158263,
+      "grad_norm": 119.3560791015625,
+      "learning_rate": 0.00017295737122557727,
+      "loss": 1.2403,
+      "step": 618
+    },
+    {
+      "epoch": 0.41204859377600267,
+      "grad_norm": 40.7421760559082,
+      "learning_rate": 0.00017291296625222027,
+      "loss": 0.2883,
+      "step": 619
+    },
+    {
+      "epoch": 0.4127142619404227,
+      "grad_norm": 108.93859100341797,
+      "learning_rate": 0.00017286856127886324,
+      "loss": 0.829,
+      "step": 620
+    },
+    {
+      "epoch": 0.41337993010484275,
+      "grad_norm": 67.42476654052734,
+      "learning_rate": 0.00017282415630550624,
+      "loss": 0.5903,
+      "step": 621
+    },
+    {
+      "epoch": 0.4140455982692628,
+      "grad_norm": 46.429893493652344,
+      "learning_rate": 0.0001727797513321492,
+      "loss": 0.3084,
+      "step": 622
+    },
+    {
+      "epoch": 0.41471126643368283,
+      "grad_norm": 82.67963409423828,
+      "learning_rate": 0.00017273534635879218,
+      "loss": 0.4691,
+      "step": 623
+    },
+    {
+      "epoch": 0.4153769345981028,
+      "grad_norm": 93.61371612548828,
+      "learning_rate": 0.00017269094138543518,
+      "loss": 0.4056,
+      "step": 624
+    },
+    {
+      "epoch": 0.41604260276252286,
+      "grad_norm": 133.24710083007812,
+      "learning_rate": 0.00017264653641207815,
+      "loss": 0.5845,
+      "step": 625
+    },
+    {
+      "epoch": 0.4167082709269429,
+      "grad_norm": 65.41972351074219,
+      "learning_rate": 0.00017260213143872115,
+      "loss": 0.3826,
+      "step": 626
+    },
+    {
+      "epoch": 0.41737393909136294,
+      "grad_norm": 103.12113189697266,
+      "learning_rate": 0.00017255772646536412,
+      "loss": 0.3715,
+      "step": 627
+    },
+    {
+      "epoch": 0.418039607255783,
+      "grad_norm": 57.001365661621094,
+      "learning_rate": 0.00017251332149200712,
+      "loss": 0.5263,
+      "step": 628
+    },
+    {
+      "epoch": 0.418705275420203,
+      "grad_norm": 67.60276794433594,
+      "learning_rate": 0.0001724689165186501,
+      "loss": 0.9148,
+      "step": 629
+    },
+    {
+      "epoch": 0.41937094358462307,
+      "grad_norm": 42.20960235595703,
+      "learning_rate": 0.0001724245115452931,
+      "loss": 0.2265,
+      "step": 630
+    },
+    {
+      "epoch": 0.4200366117490431,
+      "grad_norm": 61.50966262817383,
+      "learning_rate": 0.00017238010657193606,
+      "loss": 0.2724,
+      "step": 631
+    },
+    {
+      "epoch": 0.42070227991346315,
+      "grad_norm": 154.45419311523438,
+      "learning_rate": 0.00017233570159857906,
+      "loss": 0.7941,
+      "step": 632
+    },
+    {
+      "epoch": 0.4213679480778832,
+      "grad_norm": 50.674625396728516,
+      "learning_rate": 0.00017229129662522203,
+      "loss": 0.5776,
+      "step": 633
+    },
+    {
+      "epoch": 0.42203361624230323,
+      "grad_norm": 93.19532012939453,
+      "learning_rate": 0.00017224689165186503,
+      "loss": 0.442,
+      "step": 634
+    },
+    {
+      "epoch": 0.4226992844067233,
+      "grad_norm": 47.145652770996094,
+      "learning_rate": 0.000172202486678508,
+      "loss": 0.3244,
+      "step": 635
+    },
+    {
+      "epoch": 0.42336495257114326,
+      "grad_norm": 54.49449157714844,
+      "learning_rate": 0.00017215808170515097,
+      "loss": 0.3336,
+      "step": 636
+    },
+    {
+      "epoch": 0.4240306207355633,
+      "grad_norm": 57.44892120361328,
+      "learning_rate": 0.00017211367673179397,
+      "loss": 0.3192,
+      "step": 637
+    },
+    {
+      "epoch": 0.42469628889998334,
+      "grad_norm": 32.6402587890625,
+      "learning_rate": 0.00017206927175843694,
+      "loss": 0.3978,
+      "step": 638
+    },
+    {
+      "epoch": 0.4253619570644034,
+      "grad_norm": 62.53096389770508,
+      "learning_rate": 0.00017202486678507994,
+      "loss": 0.2868,
+      "step": 639
+    },
+    {
+      "epoch": 0.4260276252288234,
+      "grad_norm": 29.77239227294922,
+      "learning_rate": 0.0001719804618117229,
+      "loss": 0.1912,
+      "step": 640
+    },
+    {
+      "epoch": 0.42669329339324347,
+      "grad_norm": 23.30158042907715,
+      "learning_rate": 0.0001719360568383659,
+      "loss": 0.0966,
+      "step": 641
+    },
+    {
+      "epoch": 0.4273589615576635,
+      "grad_norm": 74.98001861572266,
+      "learning_rate": 0.00017189165186500888,
+      "loss": 0.342,
+      "step": 642
+    },
+    {
+      "epoch": 0.42802462972208355,
+      "grad_norm": 57.433895111083984,
+      "learning_rate": 0.00017184724689165188,
+      "loss": 0.1998,
+      "step": 643
+    },
+    {
+      "epoch": 0.4286902978865036,
+      "grad_norm": 26.67135238647461,
+      "learning_rate": 0.00017180284191829485,
+      "loss": 0.2448,
+      "step": 644
+    },
+    {
+      "epoch": 0.42935596605092363,
+      "grad_norm": 103.6978988647461,
+      "learning_rate": 0.00017175843694493785,
+      "loss": 0.5228,
+      "step": 645
+    },
+    {
+      "epoch": 0.4300216342153437,
+      "grad_norm": 95.85309600830078,
+      "learning_rate": 0.00017171403197158082,
+      "loss": 0.4366,
+      "step": 646
+    },
+    {
+      "epoch": 0.4306873023797637,
+      "grad_norm": 41.5027961730957,
+      "learning_rate": 0.00017166962699822382,
+      "loss": 0.2968,
+      "step": 647
+    },
+    {
+      "epoch": 0.4313529705441837,
+      "grad_norm": 20.21183967590332,
+      "learning_rate": 0.0001716252220248668,
+      "loss": 0.055,
+      "step": 648
+    },
+    {
+      "epoch": 0.43201863870860374,
+      "grad_norm": 74.72583770751953,
+      "learning_rate": 0.00017158081705150976,
+      "loss": 0.9381,
+      "step": 649
+    },
+    {
+      "epoch": 0.4326843068730238,
+      "grad_norm": 77.46644592285156,
+      "learning_rate": 0.00017153641207815276,
+      "loss": 0.3423,
+      "step": 650
+    },
+    {
+      "epoch": 0.4333499750374438,
+      "grad_norm": 37.057373046875,
+      "learning_rate": 0.00017149200710479573,
+      "loss": 0.6322,
+      "step": 651
+    },
+    {
+      "epoch": 0.43401564320186387,
+      "grad_norm": 70.27009582519531,
+      "learning_rate": 0.00017144760213143873,
+      "loss": 0.3602,
+      "step": 652
+    },
+    {
+      "epoch": 0.4346813113662839,
+      "grad_norm": 49.69029998779297,
+      "learning_rate": 0.0001714031971580817,
+      "loss": 0.2872,
+      "step": 653
+    },
+    {
+      "epoch": 0.43534697953070395,
+      "grad_norm": 37.20336151123047,
+      "learning_rate": 0.0001713587921847247,
+      "loss": 0.2243,
+      "step": 654
+    },
+    {
+      "epoch": 0.436012647695124,
+      "grad_norm": 46.541385650634766,
+      "learning_rate": 0.00017131438721136767,
+      "loss": 0.5596,
+      "step": 655
+    },
+    {
+      "epoch": 0.43667831585954403,
+      "grad_norm": 96.3812026977539,
+      "learning_rate": 0.00017126998223801067,
+      "loss": 0.7193,
+      "step": 656
+    },
+    {
+      "epoch": 0.43734398402396407,
+      "grad_norm": 39.73197555541992,
+      "learning_rate": 0.00017122557726465364,
+      "loss": 0.3243,
+      "step": 657
+    },
+    {
+      "epoch": 0.4380096521883841,
+      "grad_norm": 64.15336608886719,
+      "learning_rate": 0.00017118117229129664,
+      "loss": 0.4194,
+      "step": 658
+    },
+    {
+      "epoch": 0.43867532035280415,
+      "grad_norm": 38.59345245361328,
+      "learning_rate": 0.0001711367673179396,
+      "loss": 0.5887,
+      "step": 659
+    },
+    {
+      "epoch": 0.43934098851722414,
+      "grad_norm": 20.948026657104492,
+      "learning_rate": 0.00017109236234458258,
+      "loss": 0.2583,
+      "step": 660
+    },
+    {
+      "epoch": 0.4400066566816442,
+      "grad_norm": 50.797611236572266,
+      "learning_rate": 0.00017104795737122558,
+      "loss": 0.4461,
+      "step": 661
+    },
+    {
+      "epoch": 0.4406723248460642,
+      "grad_norm": 90.01778411865234,
+      "learning_rate": 0.00017100355239786855,
+      "loss": 0.264,
+      "step": 662
+    },
+    {
+      "epoch": 0.44133799301048426,
+      "grad_norm": 66.13937377929688,
+      "learning_rate": 0.00017095914742451155,
+      "loss": 0.9227,
+      "step": 663
+    },
+    {
+      "epoch": 0.4420036611749043,
+      "grad_norm": 17.219995498657227,
+      "learning_rate": 0.00017091474245115452,
+      "loss": 0.1018,
+      "step": 664
+    },
+    {
+      "epoch": 0.44266932933932435,
+      "grad_norm": 51.22856903076172,
+      "learning_rate": 0.00017087033747779752,
+      "loss": 0.2328,
+      "step": 665
+    },
+    {
+      "epoch": 0.4433349975037444,
+      "grad_norm": 109.7330551147461,
+      "learning_rate": 0.0001708259325044405,
+      "loss": 0.6561,
+      "step": 666
+    },
+    {
+      "epoch": 0.44400066566816443,
+      "grad_norm": 46.88761901855469,
+      "learning_rate": 0.0001707815275310835,
+      "loss": 0.1331,
+      "step": 667
+    },
+    {
+      "epoch": 0.44466633383258447,
+      "grad_norm": 61.69406509399414,
+      "learning_rate": 0.00017073712255772646,
+      "loss": 0.4561,
+      "step": 668
+    },
+    {
+      "epoch": 0.4453320019970045,
+      "grad_norm": 75.9193115234375,
+      "learning_rate": 0.00017069271758436946,
+      "loss": 0.3184,
+      "step": 669
+    },
+    {
+      "epoch": 0.44599767016142455,
+      "grad_norm": 44.89881896972656,
+      "learning_rate": 0.00017064831261101243,
+      "loss": 0.3697,
+      "step": 670
+    },
+    {
+      "epoch": 0.44666333832584454,
+      "grad_norm": 78.53019714355469,
+      "learning_rate": 0.00017060390763765543,
+      "loss": 0.7524,
+      "step": 671
+    },
+    {
+      "epoch": 0.4473290064902646,
+      "grad_norm": 61.14173889160156,
+      "learning_rate": 0.0001705595026642984,
+      "loss": 0.465,
+      "step": 672
+    },
+    {
+      "epoch": 0.4479946746546846,
+      "grad_norm": 30.314422607421875,
+      "learning_rate": 0.00017051509769094137,
+      "loss": 0.3398,
+      "step": 673
+    },
+    {
+      "epoch": 0.44866034281910466,
+      "grad_norm": 68.44725799560547,
+      "learning_rate": 0.00017047069271758437,
+      "loss": 0.3817,
+      "step": 674
+    },
+    {
+      "epoch": 0.4493260109835247,
+      "grad_norm": 66.30414581298828,
+      "learning_rate": 0.00017042628774422734,
+      "loss": 0.3968,
+      "step": 675
+    },
+    {
+      "epoch": 0.44999167914794475,
+      "grad_norm": 192.29794311523438,
+      "learning_rate": 0.00017038188277087034,
+      "loss": 1.2593,
+      "step": 676
+    },
+    {
+      "epoch": 0.4506573473123648,
+      "grad_norm": 90.28983306884766,
+      "learning_rate": 0.0001703374777975133,
+      "loss": 0.3282,
+      "step": 677
+    },
+    {
+      "epoch": 0.45132301547678483,
+      "grad_norm": 24.365171432495117,
+      "learning_rate": 0.0001702930728241563,
+      "loss": 0.0878,
+      "step": 678
+    },
+    {
+      "epoch": 0.45198868364120487,
+      "grad_norm": 57.6705322265625,
+      "learning_rate": 0.00017024866785079928,
+      "loss": 0.2243,
+      "step": 679
+    },
+    {
+      "epoch": 0.4526543518056249,
+      "grad_norm": 83.08112335205078,
+      "learning_rate": 0.00017020426287744228,
+      "loss": 0.4434,
+      "step": 680
+    },
+    {
+      "epoch": 0.45332001997004495,
+      "grad_norm": 66.54988098144531,
+      "learning_rate": 0.00017015985790408525,
+      "loss": 0.6353,
+      "step": 681
+    },
+    {
+      "epoch": 0.453985688134465,
+      "grad_norm": 62.16245651245117,
+      "learning_rate": 0.00017011545293072825,
+      "loss": 0.5275,
+      "step": 682
+    },
+    {
+      "epoch": 0.454651356298885,
+      "grad_norm": 119.9114761352539,
+      "learning_rate": 0.00017007104795737122,
+      "loss": 0.7318,
+      "step": 683
+    },
+    {
+      "epoch": 0.455317024463305,
+      "grad_norm": 22.712766647338867,
+      "learning_rate": 0.00017002664298401422,
+      "loss": 0.1221,
+      "step": 684
+    },
+    {
+      "epoch": 0.45598269262772506,
+      "grad_norm": 30.54381561279297,
+      "learning_rate": 0.0001699822380106572,
+      "loss": 0.1475,
+      "step": 685
+    },
+    {
+      "epoch": 0.4566483607921451,
+      "grad_norm": 50.76087951660156,
+      "learning_rate": 0.0001699378330373002,
+      "loss": 0.2274,
+      "step": 686
+    },
+    {
+      "epoch": 0.45731402895656514,
+      "grad_norm": 48.10842514038086,
+      "learning_rate": 0.00016989342806394316,
+      "loss": 0.2283,
+      "step": 687
+    },
+    {
+      "epoch": 0.4579796971209852,
+      "grad_norm": 123.0356216430664,
+      "learning_rate": 0.00016984902309058616,
+      "loss": 0.3123,
+      "step": 688
+    },
+    {
+      "epoch": 0.4586453652854052,
+      "grad_norm": 84.65762329101562,
+      "learning_rate": 0.00016980461811722913,
+      "loss": 0.2491,
+      "step": 689
+    },
+    {
+      "epoch": 0.45931103344982527,
+      "grad_norm": 30.20061683654785,
+      "learning_rate": 0.00016976021314387213,
+      "loss": 0.0821,
+      "step": 690
+    },
+    {
+      "epoch": 0.4599767016142453,
+      "grad_norm": 14.239197731018066,
+      "learning_rate": 0.0001697158081705151,
+      "loss": 0.0399,
+      "step": 691
+    },
+    {
+      "epoch": 0.46064236977866535,
+      "grad_norm": 17.447071075439453,
+      "learning_rate": 0.0001696714031971581,
+      "loss": 0.1809,
+      "step": 692
+    },
+    {
+      "epoch": 0.4613080379430854,
+      "grad_norm": 120.60710144042969,
+      "learning_rate": 0.00016962699822380107,
+      "loss": 0.6193,
+      "step": 693
+    },
+    {
+      "epoch": 0.46197370610750543,
+      "grad_norm": 89.69656372070312,
+      "learning_rate": 0.00016958259325044407,
+      "loss": 0.7936,
+      "step": 694
+    },
+    {
+      "epoch": 0.4626393742719254,
+      "grad_norm": 60.408546447753906,
+      "learning_rate": 0.00016953818827708704,
+      "loss": 0.2803,
+      "step": 695
+    },
+    {
+      "epoch": 0.46330504243634546,
+      "grad_norm": 72.99823760986328,
+      "learning_rate": 0.00016949378330373004,
+      "loss": 0.436,
+      "step": 696
+    },
+    {
+      "epoch": 0.4639707106007655,
+      "grad_norm": 117.71556091308594,
+      "learning_rate": 0.000169449378330373,
+      "loss": 1.5063,
+      "step": 697
+    },
+    {
+      "epoch": 0.46463637876518554,
+      "grad_norm": 45.7595100402832,
+      "learning_rate": 0.000169404973357016,
+      "loss": 0.361,
+      "step": 698
+    },
+    {
+      "epoch": 0.4653020469296056,
+      "grad_norm": 73.0155029296875,
+      "learning_rate": 0.00016936056838365898,
+      "loss": 0.81,
+      "step": 699
+    },
+    {
+      "epoch": 0.4659677150940256,
+      "grad_norm": 99.09642791748047,
+      "learning_rate": 0.00016931616341030198,
+      "loss": 0.925,
+      "step": 700
+    },
+    {
+      "epoch": 0.46663338325844567,
+      "grad_norm": 77.26228332519531,
+      "learning_rate": 0.00016927175843694495,
+      "loss": 0.5596,
+      "step": 701
+    },
+    {
+      "epoch": 0.4672990514228657,
+      "grad_norm": 34.49489212036133,
+      "learning_rate": 0.00016922735346358795,
+      "loss": 0.2152,
+      "step": 702
+    },
+    {
+      "epoch": 0.46796471958728575,
+      "grad_norm": 65.58250427246094,
+      "learning_rate": 0.00016918294849023092,
+      "loss": 0.3207,
+      "step": 703
+    },
+    {
+      "epoch": 0.4686303877517058,
+      "grad_norm": 69.55067443847656,
+      "learning_rate": 0.00016913854351687392,
+      "loss": 0.4024,
+      "step": 704
+    },
+    {
+      "epoch": 0.46929605591612583,
+      "grad_norm": 36.643272399902344,
+      "learning_rate": 0.0001690941385435169,
+      "loss": 0.3132,
+      "step": 705
+    },
+    {
+      "epoch": 0.4699617240805459,
+      "grad_norm": 69.05078887939453,
+      "learning_rate": 0.00016904973357015989,
+      "loss": 0.4488,
+      "step": 706
+    },
+    {
+      "epoch": 0.47062739224496586,
+      "grad_norm": 30.151975631713867,
+      "learning_rate": 0.00016900532859680286,
+      "loss": 0.3472,
+      "step": 707
+    },
+    {
+      "epoch": 0.4712930604093859,
+      "grad_norm": 79.70284271240234,
+      "learning_rate": 0.00016896092362344585,
+      "loss": 0.6793,
+      "step": 708
+    },
+    {
+      "epoch": 0.47195872857380594,
+      "grad_norm": 54.25303649902344,
+      "learning_rate": 0.00016891651865008883,
+      "loss": 0.2562,
+      "step": 709
+    },
+    {
+      "epoch": 0.472624396738226,
+      "grad_norm": 62.36642837524414,
+      "learning_rate": 0.0001688721136767318,
+      "loss": 0.3446,
+      "step": 710
+    },
+    {
+      "epoch": 0.473290064902646,
+      "grad_norm": 43.8344612121582,
+      "learning_rate": 0.0001688277087033748,
+      "loss": 0.3185,
+      "step": 711
+    },
+    {
+      "epoch": 0.47395573306706607,
+      "grad_norm": 27.90556526184082,
+      "learning_rate": 0.00016878330373001777,
+      "loss": 0.2685,
+      "step": 712
+    },
+    {
+      "epoch": 0.4746214012314861,
+      "grad_norm": 27.03131866455078,
+      "learning_rate": 0.00016873889875666077,
+      "loss": 0.0934,
+      "step": 713
+    },
+    {
+      "epoch": 0.47528706939590615,
+      "grad_norm": 123.10067749023438,
+      "learning_rate": 0.00016869449378330374,
+      "loss": 0.7965,
+      "step": 714
+    },
+    {
+      "epoch": 0.4759527375603262,
+      "grad_norm": 79.507568359375,
+      "learning_rate": 0.00016865008880994673,
+      "loss": 0.6379,
+      "step": 715
+    },
+    {
+      "epoch": 0.47661840572474623,
+      "grad_norm": 62.209964752197266,
+      "learning_rate": 0.0001686056838365897,
+      "loss": 0.187,
+      "step": 716
+    },
+    {
+      "epoch": 0.47728407388916627,
+      "grad_norm": 92.40242767333984,
+      "learning_rate": 0.0001685612788632327,
+      "loss": 0.4563,
+      "step": 717
+    },
+    {
+      "epoch": 0.4779497420535863,
+      "grad_norm": 54.89033126831055,
+      "learning_rate": 0.00016851687388987568,
+      "loss": 0.3967,
+      "step": 718
+    },
+    {
+      "epoch": 0.4786154102180063,
+      "grad_norm": 18.528785705566406,
+      "learning_rate": 0.00016847246891651867,
+      "loss": 0.0683,
+      "step": 719
+    },
+    {
+      "epoch": 0.47928107838242634,
+      "grad_norm": 87.40145874023438,
+      "learning_rate": 0.00016842806394316165,
+      "loss": 0.7017,
+      "step": 720
+    },
+    {
+      "epoch": 0.4799467465468464,
+      "grad_norm": 61.1949462890625,
+      "learning_rate": 0.00016838365896980464,
+      "loss": 0.7981,
+      "step": 721
+    },
+    {
+      "epoch": 0.4806124147112664,
+      "grad_norm": 90.75920104980469,
+      "learning_rate": 0.00016833925399644762,
+      "loss": 0.8007,
+      "step": 722
+    },
+    {
+      "epoch": 0.48127808287568646,
+      "grad_norm": 31.92529296875,
+      "learning_rate": 0.0001682948490230906,
+      "loss": 0.1569,
+      "step": 723
+    },
+    {
+      "epoch": 0.4819437510401065,
+      "grad_norm": 81.36243438720703,
+      "learning_rate": 0.00016825044404973358,
+      "loss": 0.2877,
+      "step": 724
+    },
+    {
+      "epoch": 0.48260941920452655,
+      "grad_norm": 45.11757278442383,
+      "learning_rate": 0.00016820603907637656,
+      "loss": 0.4982,
+      "step": 725
+    },
+    {
+      "epoch": 0.4832750873689466,
+      "grad_norm": 47.99403762817383,
+      "learning_rate": 0.00016816163410301955,
+      "loss": 0.1855,
+      "step": 726
+    },
+    {
+      "epoch": 0.48394075553336663,
+      "grad_norm": 103.36270141601562,
+      "learning_rate": 0.00016811722912966253,
+      "loss": 0.3909,
+      "step": 727
+    },
+    {
+      "epoch": 0.48460642369778667,
+      "grad_norm": 11.007644653320312,
+      "learning_rate": 0.00016807282415630552,
+      "loss": 0.0956,
+      "step": 728
+    },
+    {
+      "epoch": 0.4852720918622067,
+      "grad_norm": 51.31671905517578,
+      "learning_rate": 0.0001680284191829485,
+      "loss": 0.2221,
+      "step": 729
+    },
+    {
+      "epoch": 0.48593776002662675,
+      "grad_norm": 36.494293212890625,
+      "learning_rate": 0.0001679840142095915,
+      "loss": 0.1623,
+      "step": 730
+    },
+    {
+      "epoch": 0.48660342819104674,
+      "grad_norm": 100.02330017089844,
+      "learning_rate": 0.00016793960923623447,
+      "loss": 0.4674,
+      "step": 731
+    },
+    {
+      "epoch": 0.4872690963554668,
+      "grad_norm": 41.67621612548828,
+      "learning_rate": 0.00016789520426287746,
+      "loss": 0.1248,
+      "step": 732
+    },
+    {
+      "epoch": 0.4879347645198868,
+      "grad_norm": 114.75006103515625,
+      "learning_rate": 0.00016785079928952043,
+      "loss": 0.4371,
+      "step": 733
+    },
+    {
+      "epoch": 0.48860043268430686,
+      "grad_norm": 41.169837951660156,
+      "learning_rate": 0.00016780639431616343,
+      "loss": 0.1483,
+      "step": 734
+    },
+    {
+      "epoch": 0.4892661008487269,
+      "grad_norm": 77.92648315429688,
+      "learning_rate": 0.0001677619893428064,
+      "loss": 0.4915,
+      "step": 735
+    },
+    {
+      "epoch": 0.48993176901314694,
+      "grad_norm": 97.74514770507812,
+      "learning_rate": 0.00016771758436944938,
+      "loss": 0.3256,
+      "step": 736
+    },
+    {
+      "epoch": 0.490597437177567,
+      "grad_norm": 63.13382339477539,
+      "learning_rate": 0.00016767317939609237,
+      "loss": 0.4301,
+      "step": 737
+    },
+    {
+      "epoch": 0.491263105341987,
+      "grad_norm": 104.03643798828125,
+      "learning_rate": 0.00016762877442273535,
+      "loss": 0.6005,
+      "step": 738
+    },
+    {
+      "epoch": 0.49192877350640707,
+      "grad_norm": 102.05120086669922,
+      "learning_rate": 0.00016758436944937834,
+      "loss": 1.173,
+      "step": 739
+    },
+    {
+      "epoch": 0.4925944416708271,
+      "grad_norm": 75.59339141845703,
+      "learning_rate": 0.00016753996447602132,
+      "loss": 0.7547,
+      "step": 740
+    },
+    {
+      "epoch": 0.49326010983524715,
+      "grad_norm": 59.76469802856445,
+      "learning_rate": 0.00016749555950266431,
+      "loss": 0.302,
+      "step": 741
+    },
+    {
+      "epoch": 0.4939257779996672,
+      "grad_norm": 69.68283081054688,
+      "learning_rate": 0.00016745115452930728,
+      "loss": 0.4141,
+      "step": 742
+    },
+    {
+      "epoch": 0.4945914461640872,
+      "grad_norm": 8.145694732666016,
+      "learning_rate": 0.00016740674955595028,
+      "loss": 0.0285,
+      "step": 743
+    },
+    {
+      "epoch": 0.4952571143285072,
+      "grad_norm": 32.968955993652344,
+      "learning_rate": 0.00016736234458259325,
+      "loss": 0.1136,
+      "step": 744
+    },
+    {
+      "epoch": 0.49592278249292726,
+      "grad_norm": 73.82479095458984,
+      "learning_rate": 0.00016731793960923625,
+      "loss": 0.261,
+      "step": 745
+    },
+    {
+      "epoch": 0.4965884506573473,
+      "grad_norm": 108.1319808959961,
+      "learning_rate": 0.00016727353463587922,
+      "loss": 0.6509,
+      "step": 746
+    },
+    {
+      "epoch": 0.49725411882176734,
+      "grad_norm": 129.9436798095703,
+      "learning_rate": 0.00016722912966252222,
+      "loss": 1.5656,
+      "step": 747
+    },
+    {
+      "epoch": 0.4979197869861874,
+      "grad_norm": 42.64841842651367,
+      "learning_rate": 0.0001671847246891652,
+      "loss": 0.0873,
+      "step": 748
+    },
+    {
+      "epoch": 0.4985854551506074,
+      "grad_norm": 74.73552703857422,
+      "learning_rate": 0.00016714031971580817,
+      "loss": 0.4388,
+      "step": 749
+    },
+    {
+      "epoch": 0.49925112331502747,
+      "grad_norm": 97.19401550292969,
+      "learning_rate": 0.00016709591474245116,
+      "loss": 0.6456,
+      "step": 750
+    },
+    {
+      "epoch": 0.4999167914794475,
+      "grad_norm": 31.533082962036133,
+      "learning_rate": 0.00016705150976909413,
+      "loss": 0.123,
+      "step": 751
+    },
+    {
+      "epoch": 0.5005824596438675,
+      "grad_norm": 62.29495620727539,
+      "learning_rate": 0.00016700710479573713,
+      "loss": 0.4031,
+      "step": 752
+    },
+    {
+      "epoch": 0.5012481278082875,
+      "grad_norm": 49.27903366088867,
+      "learning_rate": 0.0001669626998223801,
+      "loss": 0.3012,
+      "step": 753
+    },
+    {
+      "epoch": 0.5019137959727076,
+      "grad_norm": 24.742130279541016,
+      "learning_rate": 0.0001669182948490231,
+      "loss": 0.0864,
+      "step": 754
+    },
+    {
+      "epoch": 0.5025794641371276,
+      "grad_norm": 71.86795806884766,
+      "learning_rate": 0.00016687388987566607,
+      "loss": 1.0555,
+      "step": 755
+    },
+    {
+      "epoch": 0.5032451323015477,
+      "grad_norm": 26.747037887573242,
+      "learning_rate": 0.00016682948490230907,
+      "loss": 0.1599,
+      "step": 756
+    },
+    {
+      "epoch": 0.5039108004659677,
+      "grad_norm": 71.43767547607422,
+      "learning_rate": 0.00016678507992895204,
+      "loss": 0.255,
+      "step": 757
+    },
+    {
+      "epoch": 0.5045764686303877,
+      "grad_norm": 45.117950439453125,
+      "learning_rate": 0.00016674067495559504,
+      "loss": 0.4431,
+      "step": 758
+    },
+    {
+      "epoch": 0.5052421367948078,
+      "grad_norm": 44.90398025512695,
+      "learning_rate": 0.00016669626998223801,
+      "loss": 0.291,
+      "step": 759
+    },
+    {
+      "epoch": 0.5059078049592278,
+      "grad_norm": 24.768451690673828,
+      "learning_rate": 0.00016665186500888098,
+      "loss": 0.109,
+      "step": 760
+    },
+    {
+      "epoch": 0.5065734731236479,
+      "grad_norm": 46.24464797973633,
+      "learning_rate": 0.00016660746003552398,
+      "loss": 0.3432,
+      "step": 761
+    },
+    {
+      "epoch": 0.5072391412880679,
+      "grad_norm": 19.104442596435547,
+      "learning_rate": 0.00016656305506216695,
+      "loss": 0.091,
+      "step": 762
+    },
+    {
+      "epoch": 0.507904809452488,
+      "grad_norm": 95.51792907714844,
+      "learning_rate": 0.00016651865008880995,
+      "loss": 0.5077,
+      "step": 763
+    },
+    {
+      "epoch": 0.508570477616908,
+      "grad_norm": 33.28221130371094,
+      "learning_rate": 0.00016647424511545292,
+      "loss": 0.1207,
+      "step": 764
+    },
+    {
+      "epoch": 0.509236145781328,
+      "grad_norm": 48.90406036376953,
+      "learning_rate": 0.00016642984014209592,
+      "loss": 0.4793,
+      "step": 765
+    },
+    {
+      "epoch": 0.5099018139457481,
+      "grad_norm": 29.814743041992188,
+      "learning_rate": 0.0001663854351687389,
+      "loss": 0.0951,
+      "step": 766
+    },
+    {
+      "epoch": 0.5105674821101681,
+      "grad_norm": 96.55838012695312,
+      "learning_rate": 0.0001663410301953819,
+      "loss": 1.0379,
+      "step": 767
+    },
+    {
+      "epoch": 0.5112331502745882,
+      "grad_norm": 100.2326431274414,
+      "learning_rate": 0.00016629662522202486,
+      "loss": 0.5686,
+      "step": 768
+    },
+    {
+      "epoch": 0.5118988184390082,
+      "grad_norm": 81.53526306152344,
+      "learning_rate": 0.00016625222024866786,
+      "loss": 0.4334,
+      "step": 769
+    },
+    {
+      "epoch": 0.5125644866034282,
+      "grad_norm": 50.92794418334961,
+      "learning_rate": 0.00016620781527531083,
+      "loss": 0.4187,
+      "step": 770
+    },
+    {
+      "epoch": 0.5132301547678483,
+      "grad_norm": 37.605533599853516,
+      "learning_rate": 0.00016616341030195383,
+      "loss": 0.1725,
+      "step": 771
+    },
+    {
+      "epoch": 0.5138958229322682,
+      "grad_norm": 48.62330627441406,
+      "learning_rate": 0.0001661190053285968,
+      "loss": 0.226,
+      "step": 772
+    },
+    {
+      "epoch": 0.5145614910966883,
+      "grad_norm": 131.95993041992188,
+      "learning_rate": 0.00016607460035523977,
+      "loss": 0.9214,
+      "step": 773
+    },
+    {
+      "epoch": 0.5152271592611083,
+      "grad_norm": 59.95537567138672,
+      "learning_rate": 0.00016603019538188277,
+      "loss": 0.4827,
+      "step": 774
+    },
+    {
+      "epoch": 0.5158928274255283,
+      "grad_norm": 126.53961181640625,
+      "learning_rate": 0.00016598579040852574,
+      "loss": 0.6684,
+      "step": 775
+    },
+    {
+      "epoch": 0.5165584955899484,
+      "grad_norm": 88.98424530029297,
+      "learning_rate": 0.00016594138543516874,
+      "loss": 0.5835,
+      "step": 776
+    },
+    {
+      "epoch": 0.5172241637543684,
+      "grad_norm": 63.73120880126953,
+      "learning_rate": 0.0001658969804618117,
+      "loss": 0.3389,
+      "step": 777
+    },
+    {
+      "epoch": 0.5178898319187885,
+      "grad_norm": 73.84934997558594,
+      "learning_rate": 0.0001658525754884547,
+      "loss": 0.6063,
+      "step": 778
+    },
+    {
+      "epoch": 0.5185555000832085,
+      "grad_norm": 47.694114685058594,
+      "learning_rate": 0.00016580817051509768,
+      "loss": 0.6936,
+      "step": 779
+    },
+    {
+      "epoch": 0.5192211682476285,
+      "grad_norm": 50.0419921875,
+      "learning_rate": 0.00016576376554174068,
+      "loss": 0.3097,
+      "step": 780
+    },
+    {
+      "epoch": 0.5198868364120486,
+      "grad_norm": 36.88435745239258,
+      "learning_rate": 0.00016571936056838365,
+      "loss": 0.2563,
+      "step": 781
+    },
+    {
+      "epoch": 0.5205525045764686,
+      "grad_norm": 82.27161407470703,
+      "learning_rate": 0.00016567495559502665,
+      "loss": 0.4158,
+      "step": 782
+    },
+    {
+      "epoch": 0.5212181727408887,
+      "grad_norm": 46.550270080566406,
+      "learning_rate": 0.00016563055062166962,
+      "loss": 0.2749,
+      "step": 783
+    },
+    {
+      "epoch": 0.5218838409053087,
+      "grad_norm": 69.61832427978516,
+      "learning_rate": 0.00016558614564831262,
+      "loss": 0.3109,
+      "step": 784
+    },
+    {
+      "epoch": 0.5225495090697287,
+      "grad_norm": 60.874752044677734,
+      "learning_rate": 0.0001655417406749556,
+      "loss": 0.7005,
+      "step": 785
+    },
+    {
+      "epoch": 0.5232151772341488,
+      "grad_norm": 33.250511169433594,
+      "learning_rate": 0.00016549733570159856,
+      "loss": 0.6432,
+      "step": 786
+    },
+    {
+      "epoch": 0.5238808453985688,
+      "grad_norm": 116.75189971923828,
+      "learning_rate": 0.00016545293072824156,
+      "loss": 0.5193,
+      "step": 787
+    },
+    {
+      "epoch": 0.5245465135629889,
+      "grad_norm": 88.6585922241211,
+      "learning_rate": 0.00016540852575488456,
+      "loss": 0.4463,
+      "step": 788
+    },
+    {
+      "epoch": 0.5252121817274089,
+      "grad_norm": 29.097097396850586,
+      "learning_rate": 0.00016536412078152753,
+      "loss": 0.1297,
+      "step": 789
+    },
+    {
+      "epoch": 0.525877849891829,
+      "grad_norm": 28.999595642089844,
+      "learning_rate": 0.00016531971580817053,
+      "loss": 0.1885,
+      "step": 790
+    },
+    {
+      "epoch": 0.526543518056249,
+      "grad_norm": 66.06973266601562,
+      "learning_rate": 0.0001652753108348135,
+      "loss": 0.7004,
+      "step": 791
+    },
+    {
+      "epoch": 0.527209186220669,
+      "grad_norm": 31.66838836669922,
+      "learning_rate": 0.0001652309058614565,
+      "loss": 0.1691,
+      "step": 792
+    },
+    {
+      "epoch": 0.5278748543850891,
+      "grad_norm": 23.52103042602539,
+      "learning_rate": 0.00016518650088809947,
+      "loss": 0.3266,
+      "step": 793
+    },
+    {
+      "epoch": 0.5285405225495091,
+      "grad_norm": 49.926326751708984,
+      "learning_rate": 0.00016514209591474247,
+      "loss": 0.4661,
+      "step": 794
+    },
+    {
+      "epoch": 0.5292061907139292,
+      "grad_norm": 38.95738220214844,
+      "learning_rate": 0.00016509769094138544,
+      "loss": 0.5901,
+      "step": 795
+    },
+    {
+      "epoch": 0.5298718588783491,
+      "grad_norm": 79.33973693847656,
+      "learning_rate": 0.00016505328596802844,
+      "loss": 0.4434,
+      "step": 796
+    },
+    {
+      "epoch": 0.5305375270427691,
+      "grad_norm": 59.31830596923828,
+      "learning_rate": 0.0001650088809946714,
+      "loss": 0.4059,
+      "step": 797
+    },
+    {
+      "epoch": 0.5312031952071892,
+      "grad_norm": 70.5051498413086,
+      "learning_rate": 0.0001649644760213144,
+      "loss": 0.4051,
+      "step": 798
+    },
+    {
+      "epoch": 0.5318688633716092,
+      "grad_norm": 38.054168701171875,
+      "learning_rate": 0.00016492007104795738,
+      "loss": 0.3508,
+      "step": 799
+    },
+    {
+      "epoch": 0.5325345315360293,
+      "grad_norm": 42.29998016357422,
+      "learning_rate": 0.00016487566607460038,
+      "loss": 0.1866,
+      "step": 800
+    },
+    {
+      "epoch": 0.5332001997004493,
+      "grad_norm": 36.5222282409668,
+      "learning_rate": 0.00016483126110124335,
+      "loss": 0.436,
+      "step": 801
+    },
+    {
+      "epoch": 0.5338658678648693,
+      "grad_norm": 23.83983612060547,
+      "learning_rate": 0.00016478685612788635,
+      "loss": 0.0762,
+      "step": 802
+    },
+    {
+      "epoch": 0.5345315360292894,
+      "grad_norm": 36.12353515625,
+      "learning_rate": 0.00016474245115452932,
+      "loss": 0.1777,
+      "step": 803
+    },
+    {
+      "epoch": 0.5351972041937094,
+      "grad_norm": 47.710758209228516,
+      "learning_rate": 0.00016469804618117232,
+      "loss": 0.264,
+      "step": 804
+    },
+    {
+      "epoch": 0.5358628723581295,
+      "grad_norm": 59.07710266113281,
+      "learning_rate": 0.0001646536412078153,
+      "loss": 0.48,
+      "step": 805
+    },
+    {
+      "epoch": 0.5365285405225495,
+      "grad_norm": 31.360429763793945,
+      "learning_rate": 0.0001646092362344583,
+      "loss": 0.1614,
+      "step": 806
+    },
+    {
+      "epoch": 0.5371942086869695,
+      "grad_norm": 82.94551086425781,
+      "learning_rate": 0.00016456483126110126,
+      "loss": 0.6375,
+      "step": 807
+    },
+    {
+      "epoch": 0.5378598768513896,
+      "grad_norm": 97.9576644897461,
+      "learning_rate": 0.00016452042628774426,
+      "loss": 0.6941,
+      "step": 808
+    },
+    {
+      "epoch": 0.5385255450158096,
+      "grad_norm": 55.905860900878906,
+      "learning_rate": 0.00016447602131438723,
+      "loss": 0.4596,
+      "step": 809
+    },
+    {
+      "epoch": 0.5391912131802297,
+      "grad_norm": 157.75843811035156,
+      "learning_rate": 0.0001644316163410302,
+      "loss": 0.6038,
+      "step": 810
+    },
+    {
+      "epoch": 0.5398568813446497,
+      "grad_norm": 115.88114929199219,
+      "learning_rate": 0.0001643872113676732,
+      "loss": 0.7865,
+      "step": 811
+    },
+    {
+      "epoch": 0.5405225495090697,
+      "grad_norm": 50.640113830566406,
+      "learning_rate": 0.00016434280639431617,
+      "loss": 0.5734,
+      "step": 812
+    },
+    {
+      "epoch": 0.5411882176734898,
+      "grad_norm": 51.00415802001953,
+      "learning_rate": 0.00016429840142095917,
+      "loss": 0.1352,
+      "step": 813
+    },
+    {
+      "epoch": 0.5418538858379098,
+      "grad_norm": 73.97732543945312,
+      "learning_rate": 0.00016425399644760214,
+      "loss": 0.3152,
+      "step": 814
+    },
+    {
+      "epoch": 0.5425195540023299,
+      "grad_norm": 28.134428024291992,
+      "learning_rate": 0.00016420959147424514,
+      "loss": 0.2167,
+      "step": 815
+    },
+    {
+      "epoch": 0.5431852221667499,
+      "grad_norm": 45.735347747802734,
+      "learning_rate": 0.0001641651865008881,
+      "loss": 0.239,
+      "step": 816
+    },
+    {
+      "epoch": 0.54385089033117,
+      "grad_norm": 113.84656524658203,
+      "learning_rate": 0.0001641207815275311,
+      "loss": 0.8154,
+      "step": 817
+    },
+    {
+      "epoch": 0.54451655849559,
+      "grad_norm": 33.40947723388672,
+      "learning_rate": 0.00016407637655417408,
+      "loss": 0.264,
+      "step": 818
+    },
+    {
+      "epoch": 0.54518222666001,
+      "grad_norm": 63.0031623840332,
+      "learning_rate": 0.00016403197158081708,
+      "loss": 0.4459,
+      "step": 819
+    },
+    {
+      "epoch": 0.54584789482443,
+      "grad_norm": 49.237754821777344,
+      "learning_rate": 0.00016398756660746005,
+      "loss": 0.2641,
+      "step": 820
+    },
+    {
+      "epoch": 0.54651356298885,
+      "grad_norm": 36.26042175292969,
+      "learning_rate": 0.00016394316163410305,
+      "loss": 0.2145,
+      "step": 821
+    },
+    {
+      "epoch": 0.54717923115327,
+      "grad_norm": 66.88060760498047,
+      "learning_rate": 0.00016389875666074602,
+      "loss": 0.5929,
+      "step": 822
+    },
+    {
+      "epoch": 0.5478448993176901,
+      "grad_norm": 80.92005157470703,
+      "learning_rate": 0.000163854351687389,
+      "loss": 0.4821,
+      "step": 823
+    },
+    {
+      "epoch": 0.5485105674821101,
+      "grad_norm": 46.16059494018555,
+      "learning_rate": 0.000163809946714032,
+      "loss": 0.1213,
+      "step": 824
+    },
+    {
+      "epoch": 0.5491762356465302,
+      "grad_norm": 86.4364013671875,
+      "learning_rate": 0.00016376554174067496,
+      "loss": 0.5759,
+      "step": 825
+    },
+    {
+      "epoch": 0.5498419038109502,
+      "grad_norm": 70.4107437133789,
+      "learning_rate": 0.00016372113676731796,
+      "loss": 0.4036,
+      "step": 826
+    },
+    {
+      "epoch": 0.5505075719753703,
+      "grad_norm": 129.24647521972656,
+      "learning_rate": 0.00016367673179396093,
+      "loss": 0.6871,
+      "step": 827
+    },
+    {
+      "epoch": 0.5511732401397903,
+      "grad_norm": 34.301849365234375,
+      "learning_rate": 0.00016363232682060393,
+      "loss": 0.2726,
+      "step": 828
+    },
+    {
+      "epoch": 0.5518389083042103,
+      "grad_norm": 57.288421630859375,
+      "learning_rate": 0.0001635879218472469,
+      "loss": 0.1683,
+      "step": 829
+    },
+    {
+      "epoch": 0.5525045764686304,
+      "grad_norm": 71.24075317382812,
+      "learning_rate": 0.0001635435168738899,
+      "loss": 0.4904,
+      "step": 830
+    },
+    {
+      "epoch": 0.5531702446330504,
+      "grad_norm": 37.57721710205078,
+      "learning_rate": 0.00016349911190053287,
+      "loss": 0.1642,
+      "step": 831
+    },
+    {
+      "epoch": 0.5538359127974705,
+      "grad_norm": 34.01001739501953,
+      "learning_rate": 0.00016345470692717587,
+      "loss": 0.3038,
+      "step": 832
+    },
+    {
+      "epoch": 0.5545015809618905,
+      "grad_norm": 41.36484909057617,
+      "learning_rate": 0.00016341030195381884,
+      "loss": 0.5156,
+      "step": 833
+    },
+    {
+      "epoch": 0.5551672491263105,
+      "grad_norm": 24.977052688598633,
+      "learning_rate": 0.00016336589698046184,
+      "loss": 0.107,
+      "step": 834
+    },
+    {
+      "epoch": 0.5558329172907306,
+      "grad_norm": 28.269882202148438,
+      "learning_rate": 0.0001633214920071048,
+      "loss": 0.1388,
+      "step": 835
+    },
+    {
+      "epoch": 0.5564985854551506,
+      "grad_norm": 91.3974380493164,
+      "learning_rate": 0.00016327708703374778,
+      "loss": 0.2595,
+      "step": 836
+    },
+    {
+      "epoch": 0.5571642536195707,
+      "grad_norm": 60.29289245605469,
+      "learning_rate": 0.00016323268206039078,
+      "loss": 0.4556,
+      "step": 837
+    },
+    {
+      "epoch": 0.5578299217839907,
+      "grad_norm": 73.14957427978516,
+      "learning_rate": 0.00016318827708703375,
+      "loss": 0.5624,
+      "step": 838
+    },
+    {
+      "epoch": 0.5584955899484108,
+      "grad_norm": 56.330039978027344,
+      "learning_rate": 0.00016314387211367675,
+      "loss": 0.3286,
+      "step": 839
+    },
+    {
+      "epoch": 0.5591612581128308,
+      "grad_norm": 41.747867584228516,
+      "learning_rate": 0.00016309946714031972,
+      "loss": 0.2241,
+      "step": 840
+    },
+    {
+      "epoch": 0.5598269262772508,
+      "grad_norm": 111.82936096191406,
+      "learning_rate": 0.00016305506216696272,
+      "loss": 0.6143,
+      "step": 841
+    },
+    {
+      "epoch": 0.5604925944416709,
+      "grad_norm": 63.755306243896484,
+      "learning_rate": 0.0001630106571936057,
+      "loss": 0.4782,
+      "step": 842
+    },
+    {
+      "epoch": 0.5611582626060909,
+      "grad_norm": 58.69625473022461,
+      "learning_rate": 0.00016296625222024869,
+      "loss": 0.1724,
+      "step": 843
+    },
+    {
+      "epoch": 0.5618239307705108,
+      "grad_norm": 31.554664611816406,
+      "learning_rate": 0.00016292184724689166,
+      "loss": 0.2879,
+      "step": 844
+    },
+    {
+      "epoch": 0.5624895989349309,
+      "grad_norm": 49.19903564453125,
+      "learning_rate": 0.00016287744227353466,
+      "loss": 0.3343,
+      "step": 845
+    },
+    {
+      "epoch": 0.5631552670993509,
+      "grad_norm": 48.23353576660156,
+      "learning_rate": 0.00016283303730017763,
+      "loss": 0.3251,
+      "step": 846
+    },
+    {
+      "epoch": 0.563820935263771,
+      "grad_norm": 73.30461883544922,
+      "learning_rate": 0.00016278863232682062,
+      "loss": 0.5211,
+      "step": 847
+    },
+    {
+      "epoch": 0.564486603428191,
+      "grad_norm": 74.00879669189453,
+      "learning_rate": 0.0001627442273534636,
+      "loss": 0.2641,
+      "step": 848
+    },
+    {
+      "epoch": 0.565152271592611,
+      "grad_norm": 18.06185531616211,
+      "learning_rate": 0.00016269982238010657,
+      "loss": 0.0765,
+      "step": 849
+    },
+    {
+      "epoch": 0.5658179397570311,
+      "grad_norm": 38.10951614379883,
+      "learning_rate": 0.00016265541740674957,
+      "loss": 0.203,
+      "step": 850
+    },
+    {
+      "epoch": 0.5664836079214511,
+      "grad_norm": 53.78894805908203,
+      "learning_rate": 0.00016261101243339254,
+      "loss": 0.2008,
+      "step": 851
+    },
+    {
+      "epoch": 0.5671492760858712,
+      "grad_norm": 65.65164947509766,
+      "learning_rate": 0.00016256660746003554,
+      "loss": 0.4519,
+      "step": 852
+    },
+    {
+      "epoch": 0.5678149442502912,
+      "grad_norm": 68.0468521118164,
+      "learning_rate": 0.0001625222024866785,
+      "loss": 0.282,
+      "step": 853
+    },
+    {
+      "epoch": 0.5684806124147113,
+      "grad_norm": 63.54074478149414,
+      "learning_rate": 0.0001624777975133215,
+      "loss": 0.5825,
+      "step": 854
+    },
+    {
+      "epoch": 0.5691462805791313,
+      "grad_norm": 92.11983489990234,
+      "learning_rate": 0.00016243339253996448,
+      "loss": 0.685,
+      "step": 855
+    },
+    {
+      "epoch": 0.5698119487435513,
+      "grad_norm": 16.614620208740234,
+      "learning_rate": 0.00016238898756660747,
+      "loss": 0.2696,
+      "step": 856
+    },
+    {
+      "epoch": 0.5704776169079714,
+      "grad_norm": 26.756486892700195,
+      "learning_rate": 0.00016234458259325045,
+      "loss": 0.2283,
+      "step": 857
+    },
+    {
+      "epoch": 0.5711432850723914,
+      "grad_norm": 48.378456115722656,
+      "learning_rate": 0.00016230017761989344,
+      "loss": 1.2203,
+      "step": 858
+    },
+    {
+      "epoch": 0.5718089532368115,
+      "grad_norm": 70.19944763183594,
+      "learning_rate": 0.00016225577264653642,
+      "loss": 0.6875,
+      "step": 859
+    },
+    {
+      "epoch": 0.5724746214012315,
+      "grad_norm": 41.7786750793457,
+      "learning_rate": 0.0001622113676731794,
+      "loss": 0.2601,
+      "step": 860
+    },
+    {
+      "epoch": 0.5731402895656516,
+      "grad_norm": 32.285614013671875,
+      "learning_rate": 0.00016216696269982239,
+      "loss": 0.2627,
+      "step": 861
+    },
+    {
+      "epoch": 0.5738059577300716,
+      "grad_norm": 43.81740951538086,
+      "learning_rate": 0.00016212255772646536,
+      "loss": 0.3652,
+      "step": 862
+    },
+    {
+      "epoch": 0.5744716258944916,
+      "grad_norm": 61.80424118041992,
+      "learning_rate": 0.00016207815275310836,
+      "loss": 0.2883,
+      "step": 863
+    },
+    {
+      "epoch": 0.5751372940589117,
+      "grad_norm": 59.186954498291016,
+      "learning_rate": 0.00016203374777975133,
+      "loss": 0.3879,
+      "step": 864
+    },
+    {
+      "epoch": 0.5758029622233317,
+      "grad_norm": 36.24767303466797,
+      "learning_rate": 0.00016198934280639432,
+      "loss": 0.338,
+      "step": 865
+    },
+    {
+      "epoch": 0.5764686303877518,
+      "grad_norm": 53.11578369140625,
+      "learning_rate": 0.0001619449378330373,
+      "loss": 0.5221,
+      "step": 866
+    },
+    {
+      "epoch": 0.5771342985521717,
+      "grad_norm": 85.61437225341797,
+      "learning_rate": 0.0001619005328596803,
+      "loss": 0.5565,
+      "step": 867
+    },
+    {
+      "epoch": 0.5777999667165917,
+      "grad_norm": 53.73318862915039,
+      "learning_rate": 0.00016185612788632327,
+      "loss": 0.3819,
+      "step": 868
+    },
+    {
+      "epoch": 0.5784656348810118,
+      "grad_norm": 107.41029357910156,
+      "learning_rate": 0.00016181172291296626,
+      "loss": 0.6618,
+      "step": 869
+    },
+    {
+      "epoch": 0.5791313030454318,
+      "grad_norm": 61.5377197265625,
+      "learning_rate": 0.00016176731793960924,
+      "loss": 0.3592,
+      "step": 870
+    },
+    {
+      "epoch": 0.5797969712098519,
+      "grad_norm": 39.31349182128906,
+      "learning_rate": 0.00016172291296625223,
+      "loss": 0.1912,
+      "step": 871
+    },
+    {
+      "epoch": 0.5804626393742719,
+      "grad_norm": 67.16478729248047,
+      "learning_rate": 0.0001616785079928952,
+      "loss": 0.5373,
+      "step": 872
+    },
+    {
+      "epoch": 0.5811283075386919,
+      "grad_norm": 65.04055786132812,
+      "learning_rate": 0.00016163410301953818,
+      "loss": 0.5011,
+      "step": 873
+    },
+    {
+      "epoch": 0.581793975703112,
+      "grad_norm": 30.299055099487305,
+      "learning_rate": 0.00016158969804618117,
+      "loss": 0.4038,
+      "step": 874
+    },
+    {
+      "epoch": 0.582459643867532,
+      "grad_norm": 37.89064025878906,
+      "learning_rate": 0.00016154529307282415,
+      "loss": 0.3733,
+      "step": 875
+    },
+    {
+      "epoch": 0.5831253120319521,
+      "grad_norm": 35.79871368408203,
+      "learning_rate": 0.00016150088809946714,
+      "loss": 0.4639,
+      "step": 876
+    },
+    {
+      "epoch": 0.5837909801963721,
+      "grad_norm": 41.761165618896484,
+      "learning_rate": 0.00016145648312611012,
+      "loss": 0.4087,
+      "step": 877
+    },
+    {
+      "epoch": 0.5844566483607921,
+      "grad_norm": 62.74789047241211,
+      "learning_rate": 0.00016141207815275311,
+      "loss": 0.4322,
+      "step": 878
+    },
+    {
+      "epoch": 0.5851223165252122,
+      "grad_norm": 66.2603759765625,
+      "learning_rate": 0.00016136767317939609,
+      "loss": 0.2986,
+      "step": 879
+    },
+    {
+      "epoch": 0.5857879846896322,
+      "grad_norm": 38.21004867553711,
+      "learning_rate": 0.00016132326820603908,
+      "loss": 0.2088,
+      "step": 880
+    },
+    {
+      "epoch": 0.5864536528540523,
+      "grad_norm": 24.51487922668457,
+      "learning_rate": 0.00016127886323268206,
+      "loss": 0.097,
+      "step": 881
+    },
+    {
+      "epoch": 0.5871193210184723,
+      "grad_norm": 27.898771286010742,
+      "learning_rate": 0.00016123445825932505,
+      "loss": 0.1775,
+      "step": 882
+    },
+    {
+      "epoch": 0.5877849891828923,
+      "grad_norm": 46.40385818481445,
+      "learning_rate": 0.00016119005328596802,
+      "loss": 0.2212,
+      "step": 883
+    },
+    {
+      "epoch": 0.5884506573473124,
+      "grad_norm": 56.09572982788086,
+      "learning_rate": 0.00016114564831261102,
+      "loss": 0.4678,
+      "step": 884
+    },
+    {
+      "epoch": 0.5891163255117324,
+      "grad_norm": 129.60107421875,
+      "learning_rate": 0.000161101243339254,
+      "loss": 0.809,
+      "step": 885
+    },
+    {
+      "epoch": 0.5897819936761525,
+      "grad_norm": 93.75535583496094,
+      "learning_rate": 0.00016105683836589697,
+      "loss": 1.2058,
+      "step": 886
+    },
+    {
+      "epoch": 0.5904476618405725,
+      "grad_norm": 109.20541381835938,
+      "learning_rate": 0.00016101243339253996,
+      "loss": 0.6219,
+      "step": 887
+    },
+    {
+      "epoch": 0.5911133300049926,
+      "grad_norm": 58.076332092285156,
+      "learning_rate": 0.00016096802841918294,
+      "loss": 0.4415,
+      "step": 888
+    },
+    {
+      "epoch": 0.5917789981694126,
+      "grad_norm": 91.718017578125,
+      "learning_rate": 0.00016092362344582593,
+      "loss": 0.4294,
+      "step": 889
+    },
+    {
+      "epoch": 0.5924446663338326,
+      "grad_norm": 110.23320770263672,
+      "learning_rate": 0.0001608792184724689,
+      "loss": 0.7701,
+      "step": 890
+    },
+    {
+      "epoch": 0.5931103344982526,
+      "grad_norm": 84.98243713378906,
+      "learning_rate": 0.0001608348134991119,
+      "loss": 0.5707,
+      "step": 891
+    },
+    {
+      "epoch": 0.5937760026626726,
+      "grad_norm": 82.67597961425781,
+      "learning_rate": 0.00016079040852575487,
+      "loss": 0.7084,
+      "step": 892
+    },
+    {
+      "epoch": 0.5944416708270926,
+      "grad_norm": 90.17684936523438,
+      "learning_rate": 0.00016074600355239787,
+      "loss": 0.4872,
+      "step": 893
+    },
+    {
+      "epoch": 0.5951073389915127,
+      "grad_norm": 112.2318344116211,
+      "learning_rate": 0.00016070159857904084,
+      "loss": 0.8127,
+      "step": 894
+    },
+    {
+      "epoch": 0.5957730071559327,
+      "grad_norm": 16.132047653198242,
+      "learning_rate": 0.00016065719360568384,
+      "loss": 0.0944,
+      "step": 895
+    },
+    {
+      "epoch": 0.5964386753203528,
+      "grad_norm": 69.02125549316406,
+      "learning_rate": 0.00016061278863232681,
+      "loss": 0.2236,
+      "step": 896
+    },
+    {
+      "epoch": 0.5971043434847728,
+      "grad_norm": 29.546409606933594,
+      "learning_rate": 0.0001605683836589698,
+      "loss": 0.1212,
+      "step": 897
+    },
+    {
+      "epoch": 0.5977700116491929,
+      "grad_norm": 54.70003890991211,
+      "learning_rate": 0.00016052397868561278,
+      "loss": 0.3831,
+      "step": 898
+    },
+    {
+      "epoch": 0.5984356798136129,
+      "grad_norm": 65.44963073730469,
+      "learning_rate": 0.00016047957371225578,
+      "loss": 0.4181,
+      "step": 899
+    },
+    {
+      "epoch": 0.5991013479780329,
+      "grad_norm": 109.88179779052734,
+      "learning_rate": 0.00016043516873889875,
+      "loss": 0.6356,
+      "step": 900
+    },
+    {
+      "epoch": 0.599767016142453,
+      "grad_norm": 64.11637115478516,
+      "learning_rate": 0.00016039076376554175,
+      "loss": 0.708,
+      "step": 901
+    },
+    {
+      "epoch": 0.600432684306873,
+      "grad_norm": 23.20315170288086,
+      "learning_rate": 0.00016034635879218472,
+      "loss": 0.1969,
+      "step": 902
+    },
+    {
+      "epoch": 0.6010983524712931,
+      "grad_norm": 45.66897201538086,
+      "learning_rate": 0.00016030195381882772,
+      "loss": 0.5047,
+      "step": 903
+    },
+    {
+      "epoch": 0.6017640206357131,
+      "grad_norm": 27.65083122253418,
+      "learning_rate": 0.0001602575488454707,
+      "loss": 0.1068,
+      "step": 904
+    },
+    {
+      "epoch": 0.6024296888001331,
+      "grad_norm": 49.640907287597656,
+      "learning_rate": 0.0001602131438721137,
+      "loss": 0.533,
+      "step": 905
+    },
+    {
+      "epoch": 0.6030953569645532,
+      "grad_norm": 79.02021026611328,
+      "learning_rate": 0.0001601687388987567,
+      "loss": 0.9014,
+      "step": 906
+    },
+    {
+      "epoch": 0.6037610251289732,
+      "grad_norm": 49.97978591918945,
+      "learning_rate": 0.00016012433392539966,
+      "loss": 0.5546,
+      "step": 907
+    },
+    {
+      "epoch": 0.6044266932933933,
+      "grad_norm": 70.31575775146484,
+      "learning_rate": 0.00016007992895204266,
+      "loss": 0.3117,
+      "step": 908
+    },
+    {
+      "epoch": 0.6050923614578133,
+      "grad_norm": 53.54672622680664,
+      "learning_rate": 0.00016003552397868563,
+      "loss": 0.8231,
+      "step": 909
+    },
+    {
+      "epoch": 0.6057580296222334,
+      "grad_norm": 45.44542694091797,
+      "learning_rate": 0.0001599911190053286,
+      "loss": 0.222,
+      "step": 910
+    },
+    {
+      "epoch": 0.6064236977866534,
+      "grad_norm": 27.014482498168945,
+      "learning_rate": 0.0001599467140319716,
+      "loss": 0.2923,
+      "step": 911
+    },
+    {
+      "epoch": 0.6070893659510734,
+      "grad_norm": 33.17491149902344,
+      "learning_rate": 0.00015990230905861457,
+      "loss": 0.3998,
+      "step": 912
+    },
+    {
+      "epoch": 0.6077550341154935,
+      "grad_norm": 30.22976303100586,
+      "learning_rate": 0.00015985790408525757,
+      "loss": 0.22,
+      "step": 913
+    },
+    {
+      "epoch": 0.6084207022799135,
+      "grad_norm": 92.3844223022461,
+      "learning_rate": 0.00015981349911190054,
+      "loss": 0.5543,
+      "step": 914
+    },
+    {
+      "epoch": 0.6090863704443334,
+      "grad_norm": 55.11675262451172,
+      "learning_rate": 0.00015976909413854354,
+      "loss": 0.4034,
+      "step": 915
+    },
+    {
+      "epoch": 0.6097520386087535,
+      "grad_norm": 51.71201705932617,
+      "learning_rate": 0.0001597246891651865,
+      "loss": 0.4482,
+      "step": 916
+    },
+    {
+      "epoch": 0.6104177067731735,
+      "grad_norm": 41.75859069824219,
+      "learning_rate": 0.0001596802841918295,
+      "loss": 0.2557,
+      "step": 917
+    },
+    {
+      "epoch": 0.6110833749375936,
+      "grad_norm": 47.229122161865234,
+      "learning_rate": 0.00015963587921847248,
+      "loss": 0.1506,
+      "step": 918
+    },
+    {
+      "epoch": 0.6117490431020136,
+      "grad_norm": 29.504947662353516,
+      "learning_rate": 0.00015959147424511548,
+      "loss": 0.1791,
+      "step": 919
+    },
+    {
+      "epoch": 0.6124147112664337,
+      "grad_norm": 48.473594665527344,
+      "learning_rate": 0.00015954706927175845,
+      "loss": 0.1924,
+      "step": 920
+    },
+    {
+      "epoch": 0.6130803794308537,
+      "grad_norm": 24.240800857543945,
+      "learning_rate": 0.00015950266429840145,
+      "loss": 0.1287,
+      "step": 921
+    },
+    {
+      "epoch": 0.6137460475952737,
+      "grad_norm": 55.36631393432617,
+      "learning_rate": 0.00015945825932504442,
+      "loss": 0.3015,
+      "step": 922
+    },
+    {
+      "epoch": 0.6144117157596938,
+      "grad_norm": 24.28390884399414,
+      "learning_rate": 0.0001594138543516874,
+      "loss": 0.1358,
+      "step": 923
+    },
+    {
+      "epoch": 0.6150773839241138,
+      "grad_norm": 15.040701866149902,
+      "learning_rate": 0.0001593694493783304,
+      "loss": 0.078,
+      "step": 924
+    },
+    {
+      "epoch": 0.6157430520885339,
+      "grad_norm": 80.5035171508789,
+      "learning_rate": 0.00015932504440497336,
+      "loss": 0.3659,
+      "step": 925
+    },
+    {
+      "epoch": 0.6164087202529539,
+      "grad_norm": 14.009264945983887,
+      "learning_rate": 0.00015928063943161636,
+      "loss": 0.059,
+      "step": 926
+    },
+    {
+      "epoch": 0.6170743884173739,
+      "grad_norm": 41.53221130371094,
+      "learning_rate": 0.00015923623445825933,
+      "loss": 0.2141,
+      "step": 927
+    },
+    {
+      "epoch": 0.617740056581794,
+      "grad_norm": 56.1119499206543,
+      "learning_rate": 0.00015919182948490233,
+      "loss": 0.352,
+      "step": 928
+    },
+    {
+      "epoch": 0.618405724746214,
+      "grad_norm": 150.60353088378906,
+      "learning_rate": 0.0001591474245115453,
+      "loss": 0.8479,
+      "step": 929
+    },
+    {
+      "epoch": 0.6190713929106341,
+      "grad_norm": 37.60380554199219,
+      "learning_rate": 0.0001591030195381883,
+      "loss": 0.2058,
+      "step": 930
+    },
+    {
+      "epoch": 0.6197370610750541,
+      "grad_norm": 86.55335998535156,
+      "learning_rate": 0.00015905861456483127,
+      "loss": 1.5609,
+      "step": 931
+    },
+    {
+      "epoch": 0.6204027292394741,
+      "grad_norm": 33.71519088745117,
+      "learning_rate": 0.00015901420959147427,
+      "loss": 0.2137,
+      "step": 932
+    },
+    {
+      "epoch": 0.6210683974038942,
+      "grad_norm": 120.12718200683594,
+      "learning_rate": 0.00015896980461811724,
+      "loss": 0.9256,
+      "step": 933
+    },
+    {
+      "epoch": 0.6217340655683142,
+      "grad_norm": 77.55552673339844,
+      "learning_rate": 0.00015892539964476024,
+      "loss": 0.3218,
+      "step": 934
+    },
+    {
+      "epoch": 0.6223997337327343,
+      "grad_norm": 72.75386047363281,
+      "learning_rate": 0.0001588809946714032,
+      "loss": 0.6255,
+      "step": 935
+    },
+    {
+      "epoch": 0.6230654018971543,
+      "grad_norm": 28.577308654785156,
+      "learning_rate": 0.00015883658969804618,
+      "loss": 0.5123,
+      "step": 936
+    },
+    {
+      "epoch": 0.6237310700615744,
+      "grad_norm": 64.8055419921875,
+      "learning_rate": 0.00015879218472468918,
+      "loss": 0.2593,
+      "step": 937
+    },
+    {
+      "epoch": 0.6243967382259944,
+      "grad_norm": 32.127960205078125,
+      "learning_rate": 0.00015874777975133215,
+      "loss": 0.0822,
+      "step": 938
+    },
+    {
+      "epoch": 0.6250624063904143,
+      "grad_norm": 41.353248596191406,
+      "learning_rate": 0.00015870337477797515,
+      "loss": 0.1516,
+      "step": 939
+    },
+    {
+      "epoch": 0.6257280745548344,
+      "grad_norm": 22.283430099487305,
+      "learning_rate": 0.00015865896980461812,
+      "loss": 0.2439,
+      "step": 940
+    },
+    {
+      "epoch": 0.6263937427192544,
+      "grad_norm": 87.97146606445312,
+      "learning_rate": 0.00015861456483126112,
+      "loss": 1.0962,
+      "step": 941
+    },
+    {
+      "epoch": 0.6270594108836745,
+      "grad_norm": 173.7371063232422,
+      "learning_rate": 0.0001585701598579041,
+      "loss": 1.3254,
+      "step": 942
+    },
+    {
+      "epoch": 0.6277250790480945,
+      "grad_norm": 176.40904235839844,
+      "learning_rate": 0.0001585257548845471,
+      "loss": 0.9746,
+      "step": 943
+    },
+    {
+      "epoch": 0.6283907472125145,
+      "grad_norm": 18.358781814575195,
+      "learning_rate": 0.00015848134991119006,
+      "loss": 0.0599,
+      "step": 944
+    },
+    {
+      "epoch": 0.6290564153769346,
+      "grad_norm": 37.37943649291992,
+      "learning_rate": 0.00015843694493783306,
+      "loss": 0.2559,
+      "step": 945
+    },
+    {
+      "epoch": 0.6297220835413546,
+      "grad_norm": 23.679391860961914,
+      "learning_rate": 0.00015839253996447603,
+      "loss": 0.1181,
+      "step": 946
+    },
+    {
+      "epoch": 0.6303877517057747,
+      "grad_norm": 114.88152313232422,
+      "learning_rate": 0.00015834813499111903,
+      "loss": 0.7996,
+      "step": 947
+    },
+    {
+      "epoch": 0.6310534198701947,
+      "grad_norm": 138.5634002685547,
+      "learning_rate": 0.000158303730017762,
+      "loss": 0.9386,
+      "step": 948
+    },
+    {
+      "epoch": 0.6317190880346147,
+      "grad_norm": 65.20646667480469,
+      "learning_rate": 0.00015825932504440497,
+      "loss": 0.626,
+      "step": 949
+    },
+    {
+      "epoch": 0.6323847561990348,
+      "grad_norm": 42.87594985961914,
+      "learning_rate": 0.00015821492007104797,
+      "loss": 0.4692,
+      "step": 950
+    },
+    {
+      "epoch": 0.6330504243634548,
+      "grad_norm": 18.94576644897461,
+      "learning_rate": 0.00015817051509769094,
+      "loss": 0.1272,
+      "step": 951
+    },
+    {
+      "epoch": 0.6337160925278749,
+      "grad_norm": 30.526195526123047,
+      "learning_rate": 0.00015812611012433394,
+      "loss": 0.3469,
+      "step": 952
+    },
+    {
+      "epoch": 0.6343817606922949,
+      "grad_norm": 11.924509048461914,
+      "learning_rate": 0.0001580817051509769,
+      "loss": 0.2021,
+      "step": 953
+    },
+    {
+      "epoch": 0.635047428856715,
+      "grad_norm": 64.6435775756836,
+      "learning_rate": 0.0001580373001776199,
+      "loss": 0.2674,
+      "step": 954
+    },
+    {
+      "epoch": 0.635713097021135,
+      "grad_norm": 45.49868392944336,
+      "learning_rate": 0.00015799289520426288,
+      "loss": 0.2064,
+      "step": 955
+    },
+    {
+      "epoch": 0.636378765185555,
+      "grad_norm": 65.23988342285156,
+      "learning_rate": 0.00015794849023090588,
+      "loss": 0.7653,
+      "step": 956
+    },
+    {
+      "epoch": 0.6370444333499751,
+      "grad_norm": 27.883058547973633,
+      "learning_rate": 0.00015790408525754885,
+      "loss": 0.1265,
+      "step": 957
+    },
+    {
+      "epoch": 0.6377101015143951,
+      "grad_norm": 35.20575714111328,
+      "learning_rate": 0.00015785968028419185,
+      "loss": 0.1687,
+      "step": 958
+    },
+    {
+      "epoch": 0.6383757696788152,
+      "grad_norm": 51.54057693481445,
+      "learning_rate": 0.00015781527531083482,
+      "loss": 0.9334,
+      "step": 959
+    },
+    {
+      "epoch": 0.6390414378432352,
+      "grad_norm": 46.25690841674805,
+      "learning_rate": 0.0001577708703374778,
+      "loss": 0.3797,
+      "step": 960
+    },
+    {
+      "epoch": 0.6397071060076552,
+      "grad_norm": 67.028076171875,
+      "learning_rate": 0.0001577264653641208,
+      "loss": 0.3632,
+      "step": 961
+    },
+    {
+      "epoch": 0.6403727741720752,
+      "grad_norm": 73.9373779296875,
+      "learning_rate": 0.00015768206039076376,
+      "loss": 0.631,
+      "step": 962
+    },
+    {
+      "epoch": 0.6410384423364952,
+      "grad_norm": 83.83721160888672,
+      "learning_rate": 0.00015763765541740676,
+      "loss": 0.4768,
+      "step": 963
+    },
+    {
+      "epoch": 0.6417041105009152,
+      "grad_norm": 29.81329345703125,
+      "learning_rate": 0.00015759325044404973,
+      "loss": 0.1028,
+      "step": 964
+    },
+    {
+      "epoch": 0.6423697786653353,
+      "grad_norm": 82.8477554321289,
+      "learning_rate": 0.00015754884547069273,
+      "loss": 1.1949,
+      "step": 965
+    },
+    {
+      "epoch": 0.6430354468297553,
+      "grad_norm": 55.87648010253906,
+      "learning_rate": 0.0001575044404973357,
+      "loss": 1.1176,
+      "step": 966
+    },
+    {
+      "epoch": 0.6437011149941754,
+      "grad_norm": 55.962013244628906,
+      "learning_rate": 0.0001574600355239787,
+      "loss": 0.3424,
+      "step": 967
+    },
+    {
+      "epoch": 0.6443667831585954,
+      "grad_norm": 34.9271125793457,
+      "learning_rate": 0.00015741563055062167,
+      "loss": 0.3968,
+      "step": 968
+    },
+    {
+      "epoch": 0.6450324513230155,
+      "grad_norm": 63.50864028930664,
+      "learning_rate": 0.00015737122557726467,
+      "loss": 0.7757,
+      "step": 969
+    },
+    {
+      "epoch": 0.6456981194874355,
+      "grad_norm": 13.376656532287598,
+      "learning_rate": 0.00015732682060390764,
+      "loss": 0.0667,
+      "step": 970
+    },
+    {
+      "epoch": 0.6463637876518555,
+      "grad_norm": 50.788692474365234,
+      "learning_rate": 0.00015728241563055064,
+      "loss": 0.4594,
+      "step": 971
+    },
+    {
+      "epoch": 0.6470294558162756,
+      "grad_norm": 83.55867004394531,
+      "learning_rate": 0.0001572380106571936,
+      "loss": 0.59,
+      "step": 972
+    },
+    {
+      "epoch": 0.6476951239806956,
+      "grad_norm": 44.18373107910156,
+      "learning_rate": 0.00015719360568383658,
+      "loss": 0.2554,
+      "step": 973
+    },
+    {
+      "epoch": 0.6483607921451157,
+      "grad_norm": 49.03962326049805,
+      "learning_rate": 0.00015714920071047958,
+      "loss": 0.2313,
+      "step": 974
+    },
+    {
+      "epoch": 0.6490264603095357,
+      "grad_norm": 130.43885803222656,
+      "learning_rate": 0.00015710479573712255,
+      "loss": 0.5978,
+      "step": 975
+    },
+    {
+      "epoch": 0.6496921284739557,
+      "grad_norm": 38.39082336425781,
+      "learning_rate": 0.00015706039076376555,
+      "loss": 0.1913,
+      "step": 976
+    },
+    {
+      "epoch": 0.6503577966383758,
+      "grad_norm": 64.96234893798828,
+      "learning_rate": 0.00015701598579040852,
+      "loss": 0.2994,
+      "step": 977
+    },
+    {
+      "epoch": 0.6510234648027958,
+      "grad_norm": 33.34348678588867,
+      "learning_rate": 0.00015697158081705152,
+      "loss": 0.2654,
+      "step": 978
+    },
+    {
+      "epoch": 0.6516891329672159,
+      "grad_norm": 74.68543243408203,
+      "learning_rate": 0.0001569271758436945,
+      "loss": 0.481,
+      "step": 979
+    },
+    {
+      "epoch": 0.6523548011316359,
+      "grad_norm": 70.37760162353516,
+      "learning_rate": 0.00015688277087033749,
+      "loss": 0.6861,
+      "step": 980
+    },
+    {
+      "epoch": 0.653020469296056,
+      "grad_norm": 50.27323913574219,
+      "learning_rate": 0.00015683836589698046,
+      "loss": 0.5713,
+      "step": 981
+    },
+    {
+      "epoch": 0.653686137460476,
+      "grad_norm": 28.98605728149414,
+      "learning_rate": 0.00015679396092362346,
+      "loss": 0.4162,
+      "step": 982
+    },
+    {
+      "epoch": 0.654351805624896,
+      "grad_norm": 93.48116302490234,
+      "learning_rate": 0.00015674955595026643,
+      "loss": 0.6315,
+      "step": 983
+    },
+    {
+      "epoch": 0.6550174737893161,
+      "grad_norm": 79.4161148071289,
+      "learning_rate": 0.00015670515097690943,
+      "loss": 0.3387,
+      "step": 984
+    },
+    {
+      "epoch": 0.6556831419537361,
+      "grad_norm": 39.957183837890625,
+      "learning_rate": 0.0001566607460035524,
+      "loss": 0.2098,
+      "step": 985
+    },
+    {
+      "epoch": 0.656348810118156,
+      "grad_norm": 52.71916961669922,
+      "learning_rate": 0.00015661634103019537,
+      "loss": 0.4032,
+      "step": 986
+    },
+    {
+      "epoch": 0.6570144782825761,
+      "grad_norm": 74.70369720458984,
+      "learning_rate": 0.00015657193605683837,
+      "loss": 0.9639,
+      "step": 987
+    },
+    {
+      "epoch": 0.6576801464469961,
+      "grad_norm": 78.83412170410156,
+      "learning_rate": 0.00015652753108348134,
+      "loss": 0.4124,
+      "step": 988
+    },
+    {
+      "epoch": 0.6583458146114162,
+      "grad_norm": 56.441986083984375,
+      "learning_rate": 0.00015648312611012434,
+      "loss": 0.2709,
+      "step": 989
+    },
+    {
+      "epoch": 0.6590114827758362,
+      "grad_norm": 50.2053337097168,
+      "learning_rate": 0.0001564387211367673,
+      "loss": 0.5839,
+      "step": 990
+    },
+    {
+      "epoch": 0.6596771509402563,
+      "grad_norm": 46.05778884887695,
+      "learning_rate": 0.0001563943161634103,
+      "loss": 0.1879,
+      "step": 991
+    },
+    {
+      "epoch": 0.6603428191046763,
+      "grad_norm": 89.58129119873047,
+      "learning_rate": 0.00015634991119005328,
+      "loss": 0.4644,
+      "step": 992
+    },
+    {
+      "epoch": 0.6610084872690963,
+      "grad_norm": 72.7092514038086,
+      "learning_rate": 0.00015630550621669628,
+      "loss": 0.747,
+      "step": 993
+    },
+    {
+      "epoch": 0.6616741554335164,
+      "grad_norm": 26.638051986694336,
+      "learning_rate": 0.00015626110124333925,
+      "loss": 0.3813,
+      "step": 994
+    },
+    {
+      "epoch": 0.6623398235979364,
+      "grad_norm": 52.18305587768555,
+      "learning_rate": 0.00015621669626998225,
+      "loss": 0.2625,
+      "step": 995
+    },
+    {
+      "epoch": 0.6630054917623565,
+      "grad_norm": 79.29065704345703,
+      "learning_rate": 0.00015617229129662522,
+      "loss": 0.4931,
+      "step": 996
+    },
+    {
+      "epoch": 0.6636711599267765,
+      "grad_norm": 51.57868576049805,
+      "learning_rate": 0.00015612788632326821,
+      "loss": 0.3116,
+      "step": 997
+    },
+    {
+      "epoch": 0.6643368280911965,
+      "grad_norm": 36.20134353637695,
+      "learning_rate": 0.00015608348134991119,
+      "loss": 0.4365,
+      "step": 998
+    },
+    {
+      "epoch": 0.6650024962556166,
+      "grad_norm": 48.62169647216797,
+      "learning_rate": 0.00015603907637655418,
+      "loss": 0.3637,
+      "step": 999
+    },
+    {
+      "epoch": 0.6656681644200366,
+      "grad_norm": 91.92649841308594,
+      "learning_rate": 0.00015599467140319716,
+      "loss": 0.4506,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4506,
+  "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": 1.8679348395658445e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}