{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 10,
  "global_step": 625,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008,
      "grad_norm": 15.360989570617676,
      "learning_rate": 8.000000000000001e-06,
      "loss": 5.9088,
      "step": 5
    },
    {
      "epoch": 0.016,
      "grad_norm": 7.644533634185791,
      "learning_rate": 1.8e-05,
      "loss": 4.5305,
      "step": 10
    },
    {
      "epoch": 0.016,
      "eval_loss": 4.309304714202881,
      "eval_runtime": 110.5075,
      "eval_samples_per_second": 4.525,
      "eval_steps_per_second": 4.525,
      "step": 10
    },
    {
      "epoch": 0.024,
      "grad_norm": 13.015691757202148,
      "learning_rate": 1.999791251006346e-05,
      "loss": 4.2406,
      "step": 15
    },
    {
      "epoch": 0.032,
      "grad_norm": 5.692902088165283,
      "learning_rate": 1.9991650911776697e-05,
      "loss": 4.0798,
      "step": 20
    },
    {
      "epoch": 0.032,
      "eval_loss": 4.059751033782959,
      "eval_runtime": 112.8939,
      "eval_samples_per_second": 4.429,
      "eval_steps_per_second": 4.429,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.954951763153076,
      "learning_rate": 1.9977958222476374e-05,
      "loss": 4.0475,
      "step": 25
    },
    {
      "epoch": 0.048,
      "grad_norm": 5.692638874053955,
      "learning_rate": 1.9957756633258264e-05,
      "loss": 3.9848,
      "step": 30
    },
    {
      "epoch": 0.048,
      "eval_loss": 3.947784900665283,
      "eval_runtime": 110.4433,
      "eval_samples_per_second": 4.527,
      "eval_steps_per_second": 4.527,
      "step": 30
    },
    {
      "epoch": 0.056,
      "grad_norm": 6.3086256980896,
      "learning_rate": 1.9931059322181396e-05,
      "loss": 3.9191,
      "step": 35
    },
    {
      "epoch": 0.064,
      "grad_norm": 4.923604965209961,
      "learning_rate": 1.9897883704644982e-05,
      "loss": 3.9104,
      "step": 40
    },
    {
      "epoch": 0.064,
      "eval_loss": 3.9222826957702637,
      "eval_runtime": 112.332,
      "eval_samples_per_second": 4.451,
      "eval_steps_per_second": 4.451,
      "step": 40
    },
    {
      "epoch": 0.072,
      "grad_norm": 5.413952350616455,
      "learning_rate": 1.9858251422027903e-05,
      "loss": 3.8996,
      "step": 45
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.031176567077637,
      "learning_rate": 1.98121883275714e-05,
      "loss": 3.8933,
      "step": 50
    },
    {
      "epoch": 0.08,
      "eval_loss": 3.8840794563293457,
      "eval_runtime": 111.9613,
      "eval_samples_per_second": 4.466,
      "eval_steps_per_second": 4.466,
      "step": 50
    },
    {
      "epoch": 0.088,
      "grad_norm": 5.418991565704346,
      "learning_rate": 1.9759724469514302e-05,
      "loss": 3.8838,
      "step": 55
    },
    {
      "epoch": 0.096,
      "grad_norm": 5.6602864265441895,
      "learning_rate": 1.9700894071491736e-05,
      "loss": 3.8928,
      "step": 60
    },
    {
      "epoch": 0.096,
      "eval_loss": 3.859248638153076,
      "eval_runtime": 111.6304,
      "eval_samples_per_second": 4.479,
      "eval_steps_per_second": 4.479,
      "step": 60
    },
    {
      "epoch": 0.104,
      "grad_norm": 4.850898265838623,
      "learning_rate": 1.9635735510210053e-05,
      "loss": 3.8717,
      "step": 65
    },
    {
      "epoch": 0.112,
      "grad_norm": 4.956837177276611,
      "learning_rate": 1.956429129041269e-05,
      "loss": 3.8501,
      "step": 70
    },
    {
      "epoch": 0.112,
      "eval_loss": 3.862858533859253,
      "eval_runtime": 111.0424,
      "eval_samples_per_second": 4.503,
      "eval_steps_per_second": 4.503,
      "step": 70
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.260834693908691,
      "learning_rate": 1.94866080171531e-05,
      "loss": 3.8313,
      "step": 75
    },
    {
      "epoch": 0.128,
      "grad_norm": 5.144981384277344,
      "learning_rate": 1.940273636539301e-05,
      "loss": 3.8422,
      "step": 80
    },
    {
      "epoch": 0.128,
      "eval_loss": 3.8454525470733643,
      "eval_runtime": 112.0635,
      "eval_samples_per_second": 4.462,
      "eval_steps_per_second": 4.462,
      "step": 80
    },
    {
      "epoch": 0.136,
      "grad_norm": 5.56984806060791,
      "learning_rate": 1.9312731046945694e-05,
      "loss": 3.836,
      "step": 85
    },
    {
      "epoch": 0.144,
      "grad_norm": 4.828178882598877,
      "learning_rate": 1.9216650774785975e-05,
      "loss": 3.8071,
      "step": 90
    },
    {
      "epoch": 0.144,
      "eval_loss": 3.831526041030884,
      "eval_runtime": 111.1507,
      "eval_samples_per_second": 4.498,
      "eval_steps_per_second": 4.498,
      "step": 90
    },
    {
      "epoch": 0.152,
      "grad_norm": 5.023971080780029,
      "learning_rate": 1.911455822475009e-05,
      "loss": 3.8,
      "step": 95
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.5282793045043945,
      "learning_rate": 1.9006519994650516e-05,
      "loss": 3.7826,
      "step": 100
    },
    {
      "epoch": 0.16,
      "eval_loss": 3.8195042610168457,
      "eval_runtime": 109.8545,
      "eval_samples_per_second": 4.551,
      "eval_steps_per_second": 4.551,
      "step": 100
    },
    {
      "epoch": 0.168,
      "grad_norm": 5.027553081512451,
      "learning_rate": 1.8892606560832335e-05,
      "loss": 3.7953,
      "step": 105
    },
    {
      "epoch": 0.176,
      "grad_norm": 4.897686004638672,
      "learning_rate": 1.8772892232199594e-05,
      "loss": 3.8194,
      "step": 110
    },
    {
      "epoch": 0.176,
      "eval_loss": 3.784461498260498,
      "eval_runtime": 112.129,
      "eval_samples_per_second": 4.459,
      "eval_steps_per_second": 4.459,
      "step": 110
    },
    {
      "epoch": 0.184,
      "grad_norm": 4.60252046585083,
      "learning_rate": 1.8647455101741492e-05,
      "loss": 3.774,
      "step": 115
    },
    {
      "epoch": 0.192,
      "grad_norm": 4.88078498840332,
      "learning_rate": 1.8516376995590185e-05,
      "loss": 3.7932,
      "step": 120
    },
    {
      "epoch": 0.192,
      "eval_loss": 3.8124334812164307,
      "eval_runtime": 111.9015,
      "eval_samples_per_second": 4.468,
      "eval_steps_per_second": 4.468,
      "step": 120
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.119283676147461,
      "learning_rate": 1.837974341964331e-05,
      "loss": 3.7665,
      "step": 125
    },
    {
      "epoch": 0.208,
      "grad_norm": 4.381007671356201,
      "learning_rate": 1.8237643503786097e-05,
      "loss": 3.7776,
      "step": 130
    },
    {
      "epoch": 0.208,
      "eval_loss": 3.7836484909057617,
      "eval_runtime": 111.6098,
      "eval_samples_per_second": 4.48,
      "eval_steps_per_second": 4.48,
      "step": 130
    },
    {
      "epoch": 0.216,
      "grad_norm": 4.681989669799805,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 3.7758,
      "step": 135
    },
    {
      "epoch": 0.224,
      "grad_norm": 4.945594310760498,
      "learning_rate": 1.7937418940642076e-05,
      "loss": 3.7704,
      "step": 140
    },
    {
      "epoch": 0.224,
      "eval_loss": 3.781540632247925,
      "eval_runtime": 111.7063,
      "eval_samples_per_second": 4.476,
      "eval_steps_per_second": 4.476,
      "step": 140
    },
    {
      "epoch": 0.232,
      "grad_norm": 4.896081924438477,
      "learning_rate": 1.7779490138195544e-05,
      "loss": 3.7896,
      "step": 145
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.702622890472412,
      "learning_rate": 1.7616486557764187e-05,
      "loss": 3.7802,
      "step": 150
    },
    {
      "epoch": 0.24,
      "eval_loss": 3.7781624794006348,
      "eval_runtime": 112.1095,
      "eval_samples_per_second": 4.46,
      "eval_steps_per_second": 4.46,
      "step": 150
    },
    {
      "epoch": 0.248,
      "grad_norm": 4.960013389587402,
      "learning_rate": 1.7448514531121265e-05,
      "loss": 3.7478,
      "step": 155
    },
    {
      "epoch": 0.256,
      "grad_norm": 4.85302734375,
      "learning_rate": 1.72756836310958e-05,
      "loss": 3.7776,
      "step": 160
    },
    {
      "epoch": 0.256,
      "eval_loss": 3.767767906188965,
      "eval_runtime": 111.5609,
      "eval_samples_per_second": 4.482,
      "eval_steps_per_second": 4.482,
      "step": 160
    },
    {
      "epoch": 0.264,
      "grad_norm": 4.286273002624512,
      "learning_rate": 1.7098106600095204e-05,
      "loss": 3.7231,
      "step": 165
    },
    {
      "epoch": 0.272,
      "grad_norm": 5.215882778167725,
      "learning_rate": 1.6952705502317577e-05,
      "loss": 3.777,
      "step": 170
    },
    {
      "epoch": 0.272,
      "eval_loss": 3.758565902709961,
      "eval_runtime": 112.4381,
      "eval_samples_per_second": 4.447,
      "eval_steps_per_second": 4.447,
      "step": 170
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.810149669647217,
      "learning_rate": 1.6766879380776983e-05,
      "loss": 3.7346,
      "step": 175
    },
    {
      "epoch": 0.288,
      "grad_norm": 4.446249961853027,
      "learning_rate": 1.6576639035437975e-05,
      "loss": 3.7554,
      "step": 180
    },
    {
      "epoch": 0.288,
      "eval_loss": 3.7531182765960693,
      "eval_runtime": 111.8589,
      "eval_samples_per_second": 4.47,
      "eval_steps_per_second": 4.47,
      "step": 180
    },
    {
      "epoch": 0.296,
      "grad_norm": 4.627930164337158,
      "learning_rate": 1.6382108565373783e-05,
      "loss": 3.7298,
      "step": 185
    },
    {
      "epoch": 0.304,
      "grad_norm": 4.859839916229248,
      "learning_rate": 1.6183414868225434e-05,
      "loss": 3.7494,
      "step": 190
    },
    {
      "epoch": 0.304,
      "eval_loss": 3.754574775695801,
      "eval_runtime": 111.1646,
      "eval_samples_per_second": 4.498,
      "eval_steps_per_second": 4.498,
      "step": 190
    },
    {
      "epoch": 0.312,
      "grad_norm": 4.594749927520752,
      "learning_rate": 1.5980687557422854e-05,
      "loss": 3.7233,
      "step": 195
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.1947479248046875,
      "learning_rate": 1.5774058877634473e-05,
      "loss": 3.7535,
      "step": 200
    },
    {
      "epoch": 0.32,
      "eval_loss": 3.740143299102783,
      "eval_runtime": 110.7913,
      "eval_samples_per_second": 4.513,
      "eval_steps_per_second": 4.513,
      "step": 200
    },
    {
      "epoch": 0.328,
      "grad_norm": 5.630258560180664,
      "learning_rate": 1.55636636185003e-05,
      "loss": 3.7192,
      "step": 205
    },
    {
      "epoch": 0.336,
      "grad_norm": 5.389300346374512,
      "learning_rate": 1.5349639026704916e-05,
      "loss": 3.7408,
      "step": 210
    },
    {
      "epoch": 0.336,
      "eval_loss": 3.7239396572113037,
      "eval_runtime": 113.2108,
      "eval_samples_per_second": 4.417,
      "eval_steps_per_second": 4.417,
      "step": 210
    },
    {
      "epoch": 0.344,
      "grad_norm": 4.697848796844482,
      "learning_rate": 1.5132124716447627e-05,
      "loss": 3.7209,
      "step": 215
    },
    {
      "epoch": 0.352,
      "grad_norm": 4.92866325378418,
      "learning_rate": 1.4911262578368233e-05,
      "loss": 3.7015,
      "step": 220
    },
    {
      "epoch": 0.352,
      "eval_loss": 3.724281072616577,
      "eval_runtime": 112.1344,
      "eval_samples_per_second": 4.459,
      "eval_steps_per_second": 4.459,
      "step": 220
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.7383036613464355,
      "learning_rate": 1.4687196686987832e-05,
      "loss": 3.7033,
      "step": 225
    },
    {
      "epoch": 0.368,
      "grad_norm": 5.267530918121338,
      "learning_rate": 1.4460073206724984e-05,
      "loss": 3.7526,
      "step": 230
    },
    {
      "epoch": 0.368,
      "eval_loss": 3.7359769344329834,
      "eval_runtime": 111.4037,
      "eval_samples_per_second": 4.488,
      "eval_steps_per_second": 4.488,
      "step": 230
    },
    {
      "epoch": 0.376,
      "grad_norm": 4.392041206359863,
      "learning_rate": 1.4230040296548588e-05,
      "loss": 3.7569,
      "step": 235
    },
    {
      "epoch": 0.384,
      "grad_norm": 4.574108600616455,
      "learning_rate": 1.3997248013329681e-05,
      "loss": 3.7129,
      "step": 240
    },
    {
      "epoch": 0.384,
      "eval_loss": 3.7244505882263184,
      "eval_runtime": 111.6979,
      "eval_samples_per_second": 4.476,
      "eval_steps_per_second": 4.476,
      "step": 240
    },
    {
      "epoch": 0.392,
      "grad_norm": 4.590445041656494,
      "learning_rate": 1.3761848213955136e-05,
      "loss": 3.6811,
      "step": 245
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.465660095214844,
      "learning_rate": 1.352399445626722e-05,
      "loss": 3.7496,
      "step": 250
    },
    {
      "epoch": 0.4,
      "eval_loss": 3.721489191055298,
      "eval_runtime": 112.1994,
      "eval_samples_per_second": 4.456,
      "eval_steps_per_second": 4.456,
      "step": 250
    },
    {
      "epoch": 0.408,
      "grad_norm": 4.747437953948975,
      "learning_rate": 1.328384189889351e-05,
      "loss": 3.7119,
      "step": 255
    },
    {
      "epoch": 0.416,
      "grad_norm": 5.685433387756348,
      "learning_rate": 1.3041547200032618e-05,
      "loss": 3.6913,
      "step": 260
    },
    {
      "epoch": 0.416,
      "eval_loss": 3.7043662071228027,
      "eval_runtime": 111.183,
      "eval_samples_per_second": 4.497,
      "eval_steps_per_second": 4.497,
      "step": 260
    },
    {
      "epoch": 0.424,
      "grad_norm": 4.718033313751221,
      "learning_rate": 1.2797268415261681e-05,
      "loss": 3.6865,
      "step": 265
    },
    {
      "epoch": 0.432,
      "grad_norm": 4.735610485076904,
      "learning_rate": 1.2551164894432305e-05,
      "loss": 3.6844,
      "step": 270
    },
    {
      "epoch": 0.432,
      "eval_loss": 3.7100021839141846,
      "eval_runtime": 111.4098,
      "eval_samples_per_second": 4.488,
      "eval_steps_per_second": 4.488,
      "step": 270
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.772237777709961,
      "learning_rate": 1.2303397177722234e-05,
      "loss": 3.6997,
      "step": 275
    },
    {
      "epoch": 0.448,
      "grad_norm": 5.533233642578125,
      "learning_rate": 1.2054126890910499e-05,
      "loss": 3.7172,
      "step": 280
    },
    {
      "epoch": 0.448,
      "eval_loss": 3.7194995880126953,
      "eval_runtime": 112.4891,
      "eval_samples_per_second": 4.445,
      "eval_steps_per_second": 4.445,
      "step": 280
    },
    {
      "epoch": 0.456,
      "grad_norm": 4.787185192108154,
      "learning_rate": 1.1803516639944452e-05,
      "loss": 3.7097,
      "step": 285
    },
    {
      "epoch": 0.464,
      "grad_norm": 4.525595664978027,
      "learning_rate": 1.1551729904867353e-05,
      "loss": 3.739,
      "step": 290
    },
    {
      "epoch": 0.464,
      "eval_loss": 3.70326828956604,
      "eval_runtime": 111.8761,
      "eval_samples_per_second": 4.469,
      "eval_steps_per_second": 4.469,
      "step": 290
    },
    {
      "epoch": 0.472,
      "grad_norm": 4.284120082855225,
      "learning_rate": 1.1298930933175805e-05,
      "loss": 3.684,
      "step": 295
    },
    {
      "epoch": 0.48,
      "grad_norm": 5.107572555541992,
      "learning_rate": 1.1045284632676535e-05,
      "loss": 3.6814,
      "step": 300
    },
    {
      "epoch": 0.48,
      "eval_loss": 3.701869249343872,
      "eval_runtime": 110.8874,
      "eval_samples_per_second": 4.509,
      "eval_steps_per_second": 4.509,
      "step": 300
    },
    {
      "epoch": 0.488,
      "grad_norm": 4.888582706451416,
      "learning_rate": 1.0790956463912439e-05,
      "loss": 3.6695,
      "step": 305
    },
    {
      "epoch": 0.496,
      "grad_norm": 5.606681823730469,
      "learning_rate": 1.0536112332228057e-05,
      "loss": 3.6809,
      "step": 310
    },
    {
      "epoch": 0.496,
      "eval_loss": 3.6872060298919678,
      "eval_runtime": 112.8373,
      "eval_samples_per_second": 4.431,
      "eval_steps_per_second": 4.431,
      "step": 310
    },
    {
      "epoch": 0.504,
      "grad_norm": 4.867734909057617,
      "learning_rate": 1.0280918479544914e-05,
      "loss": 3.707,
      "step": 315
    },
    {
      "epoch": 0.512,
      "grad_norm": 7.736087322235107,
      "learning_rate": 1.0025541375917274e-05,
      "loss": 3.7031,
      "step": 320
    },
    {
      "epoch": 0.512,
      "eval_loss": 3.6830782890319824,
      "eval_runtime": 110.3497,
      "eval_samples_per_second": 4.531,
      "eval_steps_per_second": 4.531,
      "step": 320
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.444705486297607,
      "learning_rate": 9.770147610939098e-06,
      "loss": 3.7152,
      "step": 325
    },
    {
      "epoch": 0.528,
      "grad_norm": 15.434615135192871,
      "learning_rate": 9.514903785072998e-06,
      "loss": 3.7322,
      "step": 330
    },
    {
      "epoch": 0.528,
      "eval_loss": 3.698810577392578,
      "eval_runtime": 111.3388,
      "eval_samples_per_second": 4.491,
      "eval_steps_per_second": 4.491,
      "step": 330
    },
    {
      "epoch": 0.536,
      "grad_norm": 5.347226142883301,
      "learning_rate": 9.259976400972147e-06,
      "loss": 3.6663,
      "step": 335
    },
    {
      "epoch": 0.544,
      "grad_norm": 4.670411109924316,
      "learning_rate": 9.005531754865929e-06,
      "loss": 3.6813,
      "step": 340
    },
    {
      "epoch": 0.544,
      "eval_loss": 3.7001984119415283,
      "eval_runtime": 113.107,
      "eval_samples_per_second": 4.421,
      "eval_steps_per_second": 4.421,
      "step": 340
    },
    {
      "epoch": 0.552,
      "grad_norm": 5.2611589431762695,
      "learning_rate": 8.751735828080308e-06,
      "loss": 3.6737,
      "step": 345
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.978248596191406,
      "learning_rate": 8.498754178763564e-06,
      "loss": 3.6661,
      "step": 350
    },
    {
      "epoch": 0.56,
      "eval_loss": 3.6846611499786377,
      "eval_runtime": 111.743,
      "eval_samples_per_second": 4.475,
      "eval_steps_per_second": 4.475,
      "step": 350
    },
    {
      "epoch": 0.568,
      "grad_norm": 5.137356758117676,
      "learning_rate": 8.246751833888122e-06,
      "loss": 3.6944,
      "step": 355
    },
    {
      "epoch": 0.576,
      "grad_norm": 5.919885635375977,
      "learning_rate": 7.995893181598866e-06,
      "loss": 3.7133,
      "step": 360
    },
    {
      "epoch": 0.576,
      "eval_loss": 3.682840585708618,
      "eval_runtime": 110.2256,
      "eval_samples_per_second": 4.536,
      "eval_steps_per_second": 4.536,
      "step": 360
    },
    {
      "epoch": 0.584,
      "grad_norm": 4.792472839355469,
      "learning_rate": 7.746341863978151e-06,
      "loss": 3.6581,
      "step": 365
    },
    {
      "epoch": 0.592,
      "grad_norm": 4.619278430938721,
      "learning_rate": 7.4982606702975505e-06,
      "loss": 3.7072,
      "step": 370
    },
    {
      "epoch": 0.592,
      "eval_loss": 3.6833138465881348,
      "eval_runtime": 112.4846,
      "eval_samples_per_second": 4.445,
      "eval_steps_per_second": 4.445,
      "step": 370
    },
    {
      "epoch": 0.6,
      "grad_norm": 5.981200218200684,
      "learning_rate": 7.251811430825846e-06,
      "loss": 3.6818,
      "step": 375
    },
    {
      "epoch": 0.608,
      "grad_norm": 5.758677005767822,
      "learning_rate": 7.007154911262678e-06,
      "loss": 3.665,
      "step": 380
    },
    {
      "epoch": 0.608,
      "eval_loss": 3.6854116916656494,
      "eval_runtime": 112.1801,
      "eval_samples_per_second": 4.457,
      "eval_steps_per_second": 4.457,
      "step": 380
    },
    {
      "epoch": 0.616,
      "grad_norm": 4.582080841064453,
      "learning_rate": 6.764450707866577e-06,
      "loss": 3.667,
      "step": 385
    },
    {
      "epoch": 0.624,
      "grad_norm": 5.349440574645996,
      "learning_rate": 6.523857143345884e-06,
      "loss": 3.6779,
      "step": 390
    },
    {
      "epoch": 0.624,
      "eval_loss": 3.67708158493042,
      "eval_runtime": 111.3211,
      "eval_samples_per_second": 4.492,
      "eval_steps_per_second": 4.492,
      "step": 390
    },
    {
      "epoch": 0.632,
      "grad_norm": 4.663772106170654,
      "learning_rate": 6.2855311635804495e-06,
      "loss": 3.6561,
      "step": 395
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.236244201660156,
      "learning_rate": 6.049628235241459e-06,
      "loss": 3.6556,
      "step": 400
    },
    {
      "epoch": 0.64,
      "eval_loss": 3.6733860969543457,
      "eval_runtime": 111.5456,
      "eval_samples_per_second": 4.482,
      "eval_steps_per_second": 4.482,
      "step": 400
    },
    {
      "epoch": 0.648,
      "grad_norm": 5.29210090637207,
      "learning_rate": 5.816302244376206e-06,
      "loss": 3.6791,
      "step": 405
    },
    {
      "epoch": 0.656,
      "grad_norm": 4.581356048583984,
      "learning_rate": 5.585705396023931e-06,
      "loss": 3.6803,
      "step": 410
    },
    {
      "epoch": 0.656,
      "eval_loss": 3.67871356010437,
      "eval_runtime": 110.9923,
      "eval_samples_per_second": 4.505,
      "eval_steps_per_second": 4.505,
      "step": 410
    },
    {
      "epoch": 0.664,
      "grad_norm": 5.390838623046875,
      "learning_rate": 5.357988114928221e-06,
      "loss": 3.6861,
      "step": 415
    },
    {
      "epoch": 0.672,
      "grad_norm": 4.902568340301514,
      "learning_rate": 5.133298947410812e-06,
      "loss": 3.6451,
      "step": 420
    },
    {
      "epoch": 0.672,
      "eval_loss": 3.666407585144043,
      "eval_runtime": 112.3761,
      "eval_samples_per_second": 4.449,
      "eval_steps_per_second": 4.449,
      "step": 420
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.62513542175293,
      "learning_rate": 4.911784464470624e-06,
      "loss": 3.69,
      "step": 425
    },
    {
      "epoch": 0.688,
      "grad_norm": 4.63794469833374,
      "learning_rate": 4.693589166171466e-06,
      "loss": 3.6554,
      "step": 430
    },
    {
      "epoch": 0.688,
      "eval_loss": 3.6700761318206787,
      "eval_runtime": 110.8149,
      "eval_samples_per_second": 4.512,
      "eval_steps_per_second": 4.512,
      "step": 430
    },
    {
      "epoch": 0.696,
      "grad_norm": 5.994736671447754,
      "learning_rate": 4.4788553873806054e-06,
      "loss": 3.6735,
      "step": 435
    },
    {
      "epoch": 0.704,
      "grad_norm": 4.715162754058838,
      "learning_rate": 4.267723204919764e-06,
      "loss": 3.6115,
      "step": 440
    },
    {
      "epoch": 0.704,
      "eval_loss": 3.660256862640381,
      "eval_runtime": 112.647,
      "eval_samples_per_second": 4.439,
      "eval_steps_per_second": 4.439,
      "step": 440
    },
    {
      "epoch": 0.712,
      "grad_norm": 8.556255340576172,
      "learning_rate": 4.060330346189125e-06,
      "loss": 3.6289,
      "step": 445
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.761578559875488,
      "learning_rate": 3.8568120993238936e-06,
      "loss": 3.6678,
      "step": 450
    },
    {
      "epoch": 0.72,
      "eval_loss": 3.67647385597229,
      "eval_runtime": 110.1504,
      "eval_samples_per_second": 4.539,
      "eval_steps_per_second": 4.539,
      "step": 450
    },
    {
      "epoch": 0.728,
      "grad_norm": 4.757038593292236,
      "learning_rate": 3.657301224942098e-06,
      "loss": 3.7008,
      "step": 455
    },
    {
      "epoch": 0.736,
      "grad_norm": 4.461075305938721,
      "learning_rate": 3.4619278695411495e-06,
      "loss": 3.651,
      "step": 460
    },
    {
      "epoch": 0.736,
      "eval_loss": 3.6577467918395996,
      "eval_runtime": 110.5987,
      "eval_samples_per_second": 4.521,
      "eval_steps_per_second": 4.521,
      "step": 460
    },
    {
      "epoch": 0.744,
      "grad_norm": 3.8970530033111572,
      "learning_rate": 3.2708194805996252e-06,
      "loss": 3.6794,
      "step": 465
    },
    {
      "epoch": 0.752,
      "grad_norm": 4.632080554962158,
      "learning_rate": 3.0841007234397655e-06,
      "loss": 3.6475,
      "step": 470
    },
    {
      "epoch": 0.752,
      "eval_loss": 3.661729574203491,
      "eval_runtime": 112.6803,
      "eval_samples_per_second": 4.437,
      "eval_steps_per_second": 4.437,
      "step": 470
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.098201751708984,
      "learning_rate": 2.901893399904797e-06,
      "loss": 3.6647,
      "step": 475
    },
    {
      "epoch": 0.768,
      "grad_norm": 4.249239921569824,
      "learning_rate": 2.724316368904201e-06,
      "loss": 3.694,
      "step": 480
    },
    {
      "epoch": 0.768,
      "eval_loss": 3.652348041534424,
      "eval_runtime": 111.7392,
      "eval_samples_per_second": 4.475,
      "eval_steps_per_second": 4.475,
      "step": 480
    },
    {
      "epoch": 0.776,
      "grad_norm": 4.952099800109863,
      "learning_rate": 2.5514854688787406e-06,
      "loss": 3.6771,
      "step": 485
    },
    {
      "epoch": 0.784,
      "grad_norm": 4.995110034942627,
      "learning_rate": 2.383513442235812e-06,
      "loss": 3.6666,
      "step": 490
    },
    {
      "epoch": 0.784,
      "eval_loss": 3.653099298477173,
      "eval_runtime": 111.5607,
      "eval_samples_per_second": 4.482,
      "eval_steps_per_second": 4.482,
      "step": 490
    },
    {
      "epoch": 0.792,
      "grad_norm": 5.493462085723877,
      "learning_rate": 2.2205098618044584e-06,
      "loss": 3.6716,
      "step": 495
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.578746795654297,
      "learning_rate": 2.0625810593579286e-06,
      "loss": 3.64,
      "step": 500
    },
    {
      "epoch": 0.8,
      "eval_loss": 3.6576831340789795,
      "eval_runtime": 109.6662,
      "eval_samples_per_second": 4.559,
      "eval_steps_per_second": 4.559,
      "step": 500
    },
    {
      "epoch": 0.808,
      "grad_norm": 4.769631385803223,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 3.6298,
      "step": 505
    },
    {
      "epoch": 0.816,
      "grad_norm": 5.270910263061523,
      "learning_rate": 1.7623564962139061e-06,
      "loss": 3.6375,
      "step": 510
    },
    {
      "epoch": 0.816,
      "eval_loss": 3.658618211746216,
      "eval_runtime": 112.7334,
      "eval_samples_per_second": 4.435,
      "eval_steps_per_second": 4.435,
      "step": 510
    },
    {
      "epoch": 0.824,
      "grad_norm": 4.440089225769043,
      "learning_rate": 1.6202565803566917e-06,
      "loss": 3.6672,
      "step": 515
    },
    {
      "epoch": 0.832,
      "grad_norm": 4.2689337730407715,
      "learning_rate": 1.4836230044098164e-06,
      "loss": 3.6544,
      "step": 520
    },
    {
      "epoch": 0.832,
      "eval_loss": 3.6577441692352295,
      "eval_runtime": 111.1537,
      "eval_samples_per_second": 4.498,
      "eval_steps_per_second": 4.498,
      "step": 520
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.282012939453125,
      "learning_rate": 1.352544898258511e-06,
      "loss": 3.6416,
      "step": 525
    },
    {
      "epoch": 0.848,
      "grad_norm": 5.834847927093506,
      "learning_rate": 1.2271077678004084e-06,
      "loss": 3.653,
      "step": 530
    },
    {
      "epoch": 0.848,
      "eval_loss": 3.6557505130767822,
      "eval_runtime": 111.4698,
      "eval_samples_per_second": 4.486,
      "eval_steps_per_second": 4.486,
      "step": 530
    },
    {
      "epoch": 0.856,
      "grad_norm": 4.881261825561523,
      "learning_rate": 1.1073934391676666e-06,
      "loss": 3.6859,
      "step": 535
    },
    {
      "epoch": 0.864,
      "grad_norm": 4.997443675994873,
      "learning_rate": 9.934800053494886e-07,
      "loss": 3.6678,
      "step": 540
    },
    {
      "epoch": 0.864,
      "eval_loss": 3.6547605991363525,
      "eval_runtime": 112.6768,
      "eval_samples_per_second": 4.437,
      "eval_steps_per_second": 4.437,
      "step": 540
    },
    {
      "epoch": 0.872,
      "grad_norm": 5.601071834564209,
      "learning_rate": 8.854417752499112e-07,
      "loss": 3.6535,
      "step": 545
    },
    {
      "epoch": 0.88,
      "grad_norm": 5.277810573577881,
      "learning_rate": 7.833492252140284e-07,
      "loss": 3.6455,
      "step": 550
    },
    {
      "epoch": 0.88,
      "eval_loss": 3.6565444469451904,
      "eval_runtime": 111.5954,
      "eval_samples_per_second": 4.48,
      "eval_steps_per_second": 4.48,
      "step": 550
    },
    {
      "epoch": 0.888,
      "grad_norm": 4.55835485458374,
      "learning_rate": 6.872689530543087e-07,
      "loss": 3.6681,
      "step": 555
    },
    {
      "epoch": 0.896,
      "grad_norm": 4.985684871673584,
      "learning_rate": 5.972636346069949e-07,
      "loss": 3.6437,
      "step": 560
    },
    {
      "epoch": 0.896,
      "eval_loss": 3.656482696533203,
      "eval_runtime": 111.9622,
      "eval_samples_per_second": 4.466,
      "eval_steps_per_second": 4.466,
      "step": 560
    },
    {
      "epoch": 0.904,
      "grad_norm": 5.561159133911133,
      "learning_rate": 5.133919828468992e-07,
      "loss": 3.6635,
      "step": 565
    },
    {
      "epoch": 0.912,
      "grad_norm": 5.3784003257751465,
      "learning_rate": 4.357087095873136e-07,
      "loss": 3.6519,
      "step": 570
    },
    {
      "epoch": 0.912,
      "eval_loss": 3.656426429748535,
      "eval_runtime": 110.8291,
      "eval_samples_per_second": 4.511,
      "eval_steps_per_second": 4.511,
      "step": 570
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.003520965576172,
      "learning_rate": 3.6426448978995054e-07,
      "loss": 3.6271,
      "step": 575
    },
    {
      "epoch": 0.928,
      "grad_norm": 4.7706403732299805,
      "learning_rate": 2.9910592850826983e-07,
      "loss": 3.7089,
      "step": 580
    },
    {
      "epoch": 0.928,
      "eval_loss": 3.6552963256835938,
      "eval_runtime": 111.738,
      "eval_samples_per_second": 4.475,
      "eval_steps_per_second": 4.475,
      "step": 580
    },
    {
      "epoch": 0.936,
      "grad_norm": 5.243898391723633,
      "learning_rate": 2.402755304856974e-07,
      "loss": 3.6502,
      "step": 585
    },
    {
      "epoch": 0.944,
      "grad_norm": 5.9423298835754395,
      "learning_rate": 1.8781167242860276e-07,
      "loss": 3.6603,
      "step": 590
    },
    {
      "epoch": 0.944,
      "eval_loss": 3.656373977661133,
      "eval_runtime": 110.7007,
      "eval_samples_per_second": 4.517,
      "eval_steps_per_second": 4.517,
      "step": 590
    },
    {
      "epoch": 0.952,
      "grad_norm": 4.393427848815918,
      "learning_rate": 1.4174857797209951e-07,
      "loss": 3.6765,
      "step": 595
    },
    {
      "epoch": 0.96,
      "grad_norm": 4.435468673706055,
      "learning_rate": 1.0211629535501811e-07,
      "loss": 3.6715,
      "step": 600
    },
    {
      "epoch": 0.96,
      "eval_loss": 3.6557369232177734,
      "eval_runtime": 111.0884,
      "eval_samples_per_second": 4.501,
      "eval_steps_per_second": 4.501,
      "step": 600
    },
    {
      "epoch": 0.968,
      "grad_norm": 4.849489212036133,
      "learning_rate": 6.894067781860636e-08,
      "loss": 3.6554,
      "step": 605
    },
    {
      "epoch": 0.976,
      "grad_norm": 6.290756702423096,
      "learning_rate": 4.2243366741735457e-08,
      "loss": 3.6473,
      "step": 610
    },
    {
      "epoch": 0.976,
      "eval_loss": 3.6551318168640137,
      "eval_runtime": 111.5978,
      "eval_samples_per_second": 4.48,
      "eval_steps_per_second": 4.48,
      "step": 610
    },
    {
      "epoch": 0.984,
      "grad_norm": 5.660240650177002,
      "learning_rate": 2.2041777523627018e-08,
      "loss": 3.6221,
      "step": 615
    },
    {
      "epoch": 0.992,
      "grad_norm": 5.215294361114502,
      "learning_rate": 8.349088223306157e-09,
      "loss": 3.6463,
      "step": 620
    },
    {
      "epoch": 0.992,
      "eval_loss": 3.655787944793701,
      "eval_runtime": 111.9272,
      "eval_samples_per_second": 4.467,
      "eval_steps_per_second": 4.467,
      "step": 620
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.055933475494385,
      "learning_rate": 1.1742309631845861e-09,
      "loss": 3.6125,
      "step": 625
    }
  ],
  "logging_steps": 5,
  "max_steps": 625,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.8614575095808e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}