{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.906542056074766,
  "eval_steps": 500,
  "global_step": 530,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018691588785046728,
      "grad_norm": 456.0,
      "learning_rate": 3.7735849056603773e-06,
      "loss": 46.1216,
      "step": 1
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 438.0,
      "learning_rate": 1.8867924528301888e-05,
      "loss": 43.7935,
      "step": 5
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 302.0,
      "learning_rate": 3.7735849056603776e-05,
      "loss": 37.6584,
      "step": 10
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 41.0,
      "learning_rate": 5.660377358490566e-05,
      "loss": 21.0859,
      "step": 15
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 30.125,
      "learning_rate": 7.547169811320755e-05,
      "loss": 17.8237,
      "step": 20
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 14.5,
      "learning_rate": 9.433962264150944e-05,
      "loss": 15.4732,
      "step": 25
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 6.1875,
      "learning_rate": 0.00011320754716981132,
      "loss": 14.2199,
      "step": 30
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 6.8125,
      "learning_rate": 0.0001320754716981132,
      "loss": 13.4105,
      "step": 35
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 18.75,
      "learning_rate": 0.0001509433962264151,
      "loss": 12.0055,
      "step": 40
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 42.0,
      "learning_rate": 0.000169811320754717,
      "loss": 9.0376,
      "step": 45
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 5.125,
      "learning_rate": 0.00018867924528301889,
      "loss": 2.8367,
      "step": 50
    },
    {
      "epoch": 0.9906542056074766,
      "eval_loss": 2.7326531410217285,
      "eval_runtime": 1.0199,
      "eval_samples_per_second": 4.902,
      "eval_steps_per_second": 1.961,
      "step": 53
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 2.90625,
      "learning_rate": 0.00019999132465602527,
      "loss": 1.6892,
      "step": 55
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 1.875,
      "learning_rate": 0.0001998937443221316,
      "loss": 1.4752,
      "step": 60
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00019968784563700586,
      "loss": 1.3666,
      "step": 65
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 1.171875,
      "learning_rate": 0.00019937385186393888,
      "loss": 1.251,
      "step": 70
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0001989521034775823,
      "loss": 1.2121,
      "step": 75
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00019842305779475968,
      "loss": 1.1493,
      "step": 80
    },
    {
      "epoch": 1.588785046728972,
      "grad_norm": 6.40625,
      "learning_rate": 0.0001977872884785815,
      "loss": 1.149,
      "step": 85
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00019704548491640192,
      "loss": 1.0847,
      "step": 90
    },
    {
      "epoch": 1.7757009345794392,
      "grad_norm": 2.75,
      "learning_rate": 0.00019619845147229138,
      "loss": 1.0808,
      "step": 95
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 10.75,
      "learning_rate": 0.00019524710661483592,
      "loss": 1.011,
      "step": 100
    },
    {
      "epoch": 1.9626168224299065,
      "grad_norm": 5.25,
      "learning_rate": 0.00019419248192120898,
      "loss": 1.0189,
      "step": 105
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.324605703353882,
      "eval_runtime": 1.0183,
      "eval_samples_per_second": 4.91,
      "eval_steps_per_second": 1.964,
      "step": 107
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00019303572095859546,
      "loss": 0.964,
      "step": 110
    },
    {
      "epoch": 2.149532710280374,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000191778078044181,
      "loss": 0.9563,
      "step": 115
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 2.5625,
      "learning_rate": 0.00019042091688505105,
      "loss": 0.9415,
      "step": 120
    },
    {
      "epoch": 2.336448598130841,
      "grad_norm": 3.265625,
      "learning_rate": 0.00018896570909947475,
      "loss": 0.9123,
      "step": 125
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 4.75,
      "learning_rate": 0.00018741403262117657,
      "loss": 0.9691,
      "step": 130
    },
    {
      "epoch": 2.5233644859813085,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00018576756998832665,
      "loss": 0.8991,
      "step": 135
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 2.71875,
      "learning_rate": 0.00018402810651910442,
      "loss": 0.8936,
      "step": 140
    },
    {
      "epoch": 2.710280373831776,
      "grad_norm": 3.671875,
      "learning_rate": 0.00018219752837581466,
      "loss": 0.8844,
      "step": 145
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 1.859375,
      "learning_rate": 0.00018027782051965407,
      "loss": 0.9217,
      "step": 150
    },
    {
      "epoch": 2.897196261682243,
      "grad_norm": 0.875,
      "learning_rate": 0.00017827106455834733,
      "loss": 0.9125,
      "step": 155
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0001761794364889855,
      "loss": 0.8788,
      "step": 160
    },
    {
      "epoch": 2.9906542056074765,
      "eval_loss": 2.2786262035369873,
      "eval_runtime": 1.0188,
      "eval_samples_per_second": 4.908,
      "eval_steps_per_second": 1.963,
      "step": 160
    },
    {
      "epoch": 3.0841121495327104,
      "grad_norm": 0.609375,
      "learning_rate": 0.00017400520433851456,
      "loss": 0.8796,
      "step": 165
    },
    {
      "epoch": 3.177570093457944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00017175072570443312,
      "loss": 0.8224,
      "step": 170
    },
    {
      "epoch": 3.2710280373831777,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00016941844519836508,
      "loss": 0.8814,
      "step": 175
    },
    {
      "epoch": 3.364485981308411,
      "grad_norm": 0.703125,
      "learning_rate": 0.00016701089179528032,
      "loss": 0.8648,
      "step": 180
    },
    {
      "epoch": 3.457943925233645,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00016453067609123654,
      "loss": 0.804,
      "step": 185
    },
    {
      "epoch": 3.5514018691588785,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00016198048747261709,
      "loss": 0.8115,
      "step": 190
    },
    {
      "epoch": 3.6448598130841123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0001593630911999333,
      "loss": 0.7971,
      "step": 195
    },
    {
      "epoch": 3.7383177570093458,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0001566813254093538,
      "loss": 0.8242,
      "step": 200
    },
    {
      "epoch": 3.831775700934579,
      "grad_norm": 2.359375,
      "learning_rate": 0.00015393809803521214,
      "loss": 0.8288,
      "step": 205
    },
    {
      "epoch": 3.925233644859813,
      "grad_norm": 2.0,
      "learning_rate": 0.00015113638365682995,
      "loss": 0.8104,
      "step": 210
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.2753615379333496,
      "eval_runtime": 1.0175,
      "eval_samples_per_second": 4.914,
      "eval_steps_per_second": 1.966,
      "step": 214
    },
    {
      "epoch": 4.018691588785047,
      "grad_norm": 1.078125,
      "learning_rate": 0.00014827922027307451,
      "loss": 0.82,
      "step": 215
    },
    {
      "epoch": 4.11214953271028,
      "grad_norm": 0.8125,
      "learning_rate": 0.00014536970600814787,
      "loss": 0.7523,
      "step": 220
    },
    {
      "epoch": 4.205607476635514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001424109957521806,
      "loss": 0.772,
      "step": 225
    },
    {
      "epoch": 4.299065420560748,
      "grad_norm": 1.515625,
      "learning_rate": 0.0001394062977402717,
      "loss": 0.7393,
      "step": 230
    },
    {
      "epoch": 4.392523364485982,
      "grad_norm": 4.0625,
      "learning_rate": 0.00013635887007368467,
      "loss": 0.7492,
      "step": 235
    },
    {
      "epoch": 4.485981308411215,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0001332720171869723,
      "loss": 0.7598,
      "step": 240
    },
    {
      "epoch": 4.579439252336448,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0001301490862648603,
      "loss": 0.7433,
      "step": 245
    },
    {
      "epoch": 4.672897196261682,
      "grad_norm": 1.5,
      "learning_rate": 0.0001269934636127754,
      "loss": 0.7615,
      "step": 250
    },
    {
      "epoch": 4.766355140186916,
      "grad_norm": 1.1875,
      "learning_rate": 0.00012380857098495356,
      "loss": 0.7309,
      "step": 255
    },
    {
      "epoch": 4.859813084112149,
      "grad_norm": 0.921875,
      "learning_rate": 0.00012059786187410984,
      "loss": 0.7343,
      "step": 260
    },
    {
      "epoch": 4.953271028037383,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00011736481776669306,
      "loss": 0.7626,
      "step": 265
    },
    {
      "epoch": 4.990654205607477,
      "eval_loss": 2.320525884628296,
      "eval_runtime": 1.018,
      "eval_samples_per_second": 4.912,
      "eval_steps_per_second": 1.965,
      "step": 267
    },
    {
      "epoch": 5.046728971962617,
      "grad_norm": 0.671875,
      "learning_rate": 0.00011411294436778561,
      "loss": 0.6804,
      "step": 270
    },
    {
      "epoch": 5.140186915887851,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00011084576779974257,
      "loss": 0.662,
      "step": 275
    },
    {
      "epoch": 5.233644859813084,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00010756683077869132,
      "loss": 0.6635,
      "step": 280
    },
    {
      "epoch": 5.327102803738318,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00010427968877303808,
      "loss": 0.6852,
      "step": 285
    },
    {
      "epoch": 5.420560747663552,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00010098790614814658,
      "loss": 0.6885,
      "step": 290
    },
    {
      "epoch": 5.5140186915887845,
      "grad_norm": 1.046875,
      "learning_rate": 9.769505230136962e-05,
      "loss": 0.6665,
      "step": 295
    },
    {
      "epoch": 5.607476635514018,
      "grad_norm": 0.9453125,
      "learning_rate": 9.440469779162407e-05,
      "loss": 0.6704,
      "step": 300
    },
    {
      "epoch": 5.700934579439252,
      "grad_norm": 0.7890625,
      "learning_rate": 9.112041046770653e-05,
      "loss": 0.7145,
      "step": 305
    },
    {
      "epoch": 5.794392523364486,
      "grad_norm": 0.8203125,
      "learning_rate": 8.784575159954748e-05,
      "loss": 0.6567,
      "step": 310
    },
    {
      "epoch": 5.88785046728972,
      "grad_norm": 0.79296875,
      "learning_rate": 8.458427201659927e-05,
      "loss": 0.6752,
      "step": 315
    },
    {
      "epoch": 5.981308411214953,
      "grad_norm": 1.046875,
      "learning_rate": 8.13395082575451e-05,
      "loss": 0.6665,
      "step": 320
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.3902792930603027,
      "eval_runtime": 1.0171,
      "eval_samples_per_second": 4.916,
      "eval_steps_per_second": 1.966,
      "step": 321
    },
    {
      "epoch": 6.074766355140187,
      "grad_norm": 0.859375,
      "learning_rate": 7.81149787355039e-05,
      "loss": 0.6076,
      "step": 325
    },
    {
      "epoch": 6.168224299065421,
      "grad_norm": 1.0625,
      "learning_rate": 7.491417992288927e-05,
      "loss": 0.5788,
      "step": 330
    },
    {
      "epoch": 6.261682242990654,
      "grad_norm": 0.8359375,
      "learning_rate": 7.174058256006012e-05,
      "loss": 0.6007,
      "step": 335
    },
    {
      "epoch": 6.355140186915888,
      "grad_norm": 0.79296875,
      "learning_rate": 6.859762789187259e-05,
      "loss": 0.6096,
      "step": 340
    },
    {
      "epoch": 6.4485981308411215,
      "grad_norm": 0.8671875,
      "learning_rate": 6.548872393621578e-05,
      "loss": 0.5978,
      "step": 345
    },
    {
      "epoch": 6.542056074766355,
      "grad_norm": 0.79296875,
      "learning_rate": 6.24172417885762e-05,
      "loss": 0.6354,
      "step": 350
    },
    {
      "epoch": 6.635514018691588,
      "grad_norm": 0.76171875,
      "learning_rate": 5.938651196663865e-05,
      "loss": 0.6094,
      "step": 355
    },
    {
      "epoch": 6.728971962616822,
      "grad_norm": 0.859375,
      "learning_rate": 5.639982079888726e-05,
      "loss": 0.5985,
      "step": 360
    },
    {
      "epoch": 6.822429906542056,
      "grad_norm": 0.765625,
      "learning_rate": 5.3460406861121894e-05,
      "loss": 0.5846,
      "step": 365
    },
    {
      "epoch": 6.91588785046729,
      "grad_norm": 0.9140625,
      "learning_rate": 5.0571457464755226e-05,
      "loss": 0.6015,
      "step": 370
    },
    {
      "epoch": 6.990654205607477,
      "eval_loss": 2.4630489349365234,
      "eval_runtime": 1.0189,
      "eval_samples_per_second": 4.907,
      "eval_steps_per_second": 1.963,
      "step": 374
    },
    {
      "epoch": 7.009345794392523,
      "grad_norm": 0.87109375,
      "learning_rate": 4.7736105200697056e-05,
      "loss": 0.6055,
      "step": 375
    },
    {
      "epoch": 7.102803738317757,
      "grad_norm": 0.93359375,
      "learning_rate": 4.495742454257418e-05,
      "loss": 0.5464,
      "step": 380
    },
    {
      "epoch": 7.196261682242991,
      "grad_norm": 0.93359375,
      "learning_rate": 4.2238428512969064e-05,
      "loss": 0.5556,
      "step": 385
    },
    {
      "epoch": 7.289719626168225,
      "grad_norm": 0.83984375,
      "learning_rate": 3.9582065416291926e-05,
      "loss": 0.548,
      "step": 390
    },
    {
      "epoch": 7.383177570093458,
      "grad_norm": 1.0390625,
      "learning_rate": 3.69912156418289e-05,
      "loss": 0.5074,
      "step": 395
    },
    {
      "epoch": 7.4766355140186915,
      "grad_norm": 0.80859375,
      "learning_rate": 3.4468688540433425e-05,
      "loss": 0.5497,
      "step": 400
    },
    {
      "epoch": 7.570093457943925,
      "grad_norm": 0.8359375,
      "learning_rate": 3.201721937824673e-05,
      "loss": 0.5436,
      "step": 405
    },
    {
      "epoch": 7.663551401869158,
      "grad_norm": 0.86328125,
      "learning_rate": 2.9639466370751068e-05,
      "loss": 0.5627,
      "step": 410
    },
    {
      "epoch": 7.757009345794392,
      "grad_norm": 0.890625,
      "learning_rate": 2.7338007800372024e-05,
      "loss": 0.566,
      "step": 415
    },
    {
      "epoch": 7.850467289719626,
      "grad_norm": 0.9296875,
      "learning_rate": 2.5115339220754797e-05,
      "loss": 0.557,
      "step": 420
    },
    {
      "epoch": 7.94392523364486,
      "grad_norm": 0.84765625,
      "learning_rate": 2.297387075074625e-05,
      "loss": 0.5699,
      "step": 425
    },
    {
      "epoch": 8.0,
      "eval_loss": 2.583679676055908,
      "eval_runtime": 1.0173,
      "eval_samples_per_second": 4.915,
      "eval_steps_per_second": 1.966,
      "step": 428
    },
    {
      "epoch": 8.037383177570094,
      "grad_norm": 0.796875,
      "learning_rate": 2.09159244610172e-05,
      "loss": 0.5156,
      "step": 430
    },
    {
      "epoch": 8.130841121495328,
      "grad_norm": 0.828125,
      "learning_rate": 1.89437318561583e-05,
      "loss": 0.509,
      "step": 435
    },
    {
      "epoch": 8.22429906542056,
      "grad_norm": 0.87109375,
      "learning_rate": 1.7059431454979824e-05,
      "loss": 0.5226,
      "step": 440
    },
    {
      "epoch": 8.317757009345794,
      "grad_norm": 0.8046875,
      "learning_rate": 1.52650664716397e-05,
      "loss": 0.5088,
      "step": 445
    },
    {
      "epoch": 8.411214953271028,
      "grad_norm": 0.82421875,
      "learning_rate": 1.3562582600113294e-05,
      "loss": 0.4948,
      "step": 450
    },
    {
      "epoch": 8.504672897196262,
      "grad_norm": 0.8203125,
      "learning_rate": 1.1953825904408034e-05,
      "loss": 0.5357,
      "step": 455
    },
    {
      "epoch": 8.598130841121495,
      "grad_norm": 0.8203125,
      "learning_rate": 1.0440540816810395e-05,
      "loss": 0.551,
      "step": 460
    },
    {
      "epoch": 8.69158878504673,
      "grad_norm": 0.91796875,
      "learning_rate": 9.024368246335735e-06,
      "loss": 0.5092,
      "step": 465
    },
    {
      "epoch": 8.785046728971963,
      "grad_norm": 0.83203125,
      "learning_rate": 7.706843799431984e-06,
      "loss": 0.5244,
      "step": 470
    },
    {
      "epoch": 8.878504672897197,
      "grad_norm": 0.8671875,
      "learning_rate": 6.489396114866941e-06,
      "loss": 0.5235,
      "step": 475
    },
    {
      "epoch": 8.97196261682243,
      "grad_norm": 0.8203125,
      "learning_rate": 5.373345314604205e-06,
      "loss": 0.5017,
      "step": 480
    },
    {
      "epoch": 8.990654205607477,
      "eval_loss": 2.6300225257873535,
      "eval_runtime": 1.0187,
      "eval_samples_per_second": 4.908,
      "eval_steps_per_second": 1.963,
      "step": 481
    },
    {
      "epoch": 9.065420560747663,
      "grad_norm": 0.81640625,
      "learning_rate": 4.359901572347758e-06,
      "loss": 0.5185,
      "step": 485
    },
    {
      "epoch": 9.158878504672897,
      "grad_norm": 0.84765625,
      "learning_rate": 3.450163801307582e-06,
      "loss": 0.5366,
      "step": 490
    },
    {
      "epoch": 9.25233644859813,
      "grad_norm": 0.796875,
      "learning_rate": 2.6451184626087644e-06,
      "loss": 0.5114,
      "step": 495
    },
    {
      "epoch": 9.345794392523365,
      "grad_norm": 0.78125,
      "learning_rate": 1.945638495636515e-06,
      "loss": 0.5138,
      "step": 500
    },
    {
      "epoch": 9.439252336448599,
      "grad_norm": 0.8203125,
      "learning_rate": 1.3524823714768375e-06,
      "loss": 0.486,
      "step": 505
    },
    {
      "epoch": 9.532710280373832,
      "grad_norm": 0.84375,
      "learning_rate": 8.662932704792792e-07,
      "loss": 0.5139,
      "step": 510
    },
    {
      "epoch": 9.626168224299064,
      "grad_norm": 0.7890625,
      "learning_rate": 4.875983848335874e-07,
      "loss": 0.5092,
      "step": 515
    },
    {
      "epoch": 9.719626168224298,
      "grad_norm": 0.7890625,
      "learning_rate": 2.1680834691628626e-07,
      "loss": 0.4957,
      "step": 520
    },
    {
      "epoch": 9.813084112149532,
      "grad_norm": 0.8125,
      "learning_rate": 5.421678402741659e-08,
      "loss": 0.5253,
      "step": 525
    },
    {
      "epoch": 9.906542056074766,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0,
      "loss": 0.4969,
      "step": 530
    },
    {
      "epoch": 9.906542056074766,
      "eval_loss": 2.6219420433044434,
      "eval_runtime": 1.0178,
      "eval_samples_per_second": 4.912,
      "eval_steps_per_second": 1.965,
      "step": 530
    },
    {
      "epoch": 9.906542056074766,
      "step": 530,
      "total_flos": 8.09000149933097e+17,
      "train_loss": 2.438210007829486,
      "train_runtime": 3271.6973,
      "train_samples_per_second": 2.613,
      "train_steps_per_second": 0.162
    }
  ],
  "logging_steps": 5,
  "max_steps": 530,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 8.09000149933097e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}