{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9570141159582104,
  "eval_steps": 500,
  "global_step": 750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01276018821277614,
      "grad_norm": 37637.11328125,
      "learning_rate": 9e-06,
      "loss": 0.6589,
      "step": 10
    },
    {
      "epoch": 0.02552037642555228,
      "grad_norm": 28946.435546875,
      "learning_rate": 1.9e-05,
      "loss": 0.6001,
      "step": 20
    },
    {
      "epoch": 0.038280564638328414,
      "grad_norm": 23968.130859375,
      "learning_rate": 2.9e-05,
      "loss": 0.478,
      "step": 30
    },
    {
      "epoch": 0.05104075285110456,
      "grad_norm": 29737.384765625,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.3571,
      "step": 40
    },
    {
      "epoch": 0.06380094106388069,
      "grad_norm": 51756.8203125,
      "learning_rate": 4.9e-05,
      "loss": 0.2617,
      "step": 50
    },
    {
      "epoch": 0.07656112927665683,
      "grad_norm": 47605.94921875,
      "learning_rate": 5.9e-05,
      "loss": 0.3241,
      "step": 60
    },
    {
      "epoch": 0.08932131748943296,
      "grad_norm": 12073.0615234375,
      "learning_rate": 6.9e-05,
      "loss": 0.1533,
      "step": 70
    },
    {
      "epoch": 0.10208150570220911,
      "grad_norm": 5818.654296875,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.1012,
      "step": 80
    },
    {
      "epoch": 0.11484169391498525,
      "grad_norm": 4877.6806640625,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.0995,
      "step": 90
    },
    {
      "epoch": 0.12760188212776138,
      "grad_norm": 4604.2353515625,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.1039,
      "step": 100
    },
    {
      "epoch": 0.14036207034053752,
      "grad_norm": 4012.2919921875,
      "learning_rate": 9.995728791936504e-05,
      "loss": 0.0861,
      "step": 110
    },
    {
      "epoch": 0.15312225855331366,
      "grad_norm": 3052.91357421875,
      "learning_rate": 9.980973490458728e-05,
      "loss": 0.0783,
      "step": 120
    },
    {
      "epoch": 0.1658824467660898,
      "grad_norm": 3616.212158203125,
      "learning_rate": 9.955712478597579e-05,
      "loss": 0.092,
      "step": 130
    },
    {
      "epoch": 0.17864263497886593,
      "grad_norm": 4166.802734375,
      "learning_rate": 9.9199990360734e-05,
      "loss": 0.0955,
      "step": 140
    },
    {
      "epoch": 0.19140282319164206,
      "grad_norm": 4267.69677734375,
      "learning_rate": 9.873908488539836e-05,
      "loss": 0.1084,
      "step": 150
    },
    {
      "epoch": 0.20416301140441823,
      "grad_norm": 3430.815185546875,
      "learning_rate": 9.817538048709367e-05,
      "loss": 0.0884,
      "step": 160
    },
    {
      "epoch": 0.21692319961719436,
      "grad_norm": 3015.459716796875,
      "learning_rate": 9.751006611315356e-05,
      "loss": 0.0793,
      "step": 170
    },
    {
      "epoch": 0.2296833878299705,
      "grad_norm": 3575.607666015625,
      "learning_rate": 9.674454502343076e-05,
      "loss": 0.0902,
      "step": 180
    },
    {
      "epoch": 0.24244357604274663,
      "grad_norm": 3760.454833984375,
      "learning_rate": 9.588043183058606e-05,
      "loss": 0.0936,
      "step": 190
    },
    {
      "epoch": 0.25520376425552277,
      "grad_norm": 3904.21240234375,
      "learning_rate": 9.491954909459895e-05,
      "loss": 0.0995,
      "step": 200
    },
    {
      "epoch": 0.2679639524682989,
      "grad_norm": 3201.078857421875,
      "learning_rate": 9.386392347868215e-05,
      "loss": 0.0858,
      "step": 210
    },
    {
      "epoch": 0.28072414068107504,
      "grad_norm": 3236.521728515625,
      "learning_rate": 9.271578147470825e-05,
      "loss": 0.079,
      "step": 220
    },
    {
      "epoch": 0.2934843288938512,
      "grad_norm": 3096.511962890625,
      "learning_rate": 9.147754470716408e-05,
      "loss": 0.0878,
      "step": 230
    },
    {
      "epoch": 0.3062445171066273,
      "grad_norm": 3840.806396484375,
      "learning_rate": 9.01518248255376e-05,
      "loss": 0.0945,
      "step": 240
    },
    {
      "epoch": 0.31900470531940345,
      "grad_norm": 4038.00732421875,
      "learning_rate": 8.874141799591007e-05,
      "loss": 0.1045,
      "step": 250
    },
    {
      "epoch": 0.3317648935321796,
      "grad_norm": 3151.041015625,
      "learning_rate": 8.724929900337186e-05,
      "loss": 0.0817,
      "step": 260
    },
    {
      "epoch": 0.3445250817449557,
      "grad_norm": 2820.68505859375,
      "learning_rate": 8.567861497770053e-05,
      "loss": 0.0802,
      "step": 270
    },
    {
      "epoch": 0.35728526995773185,
      "grad_norm": 2988.68359375,
      "learning_rate": 8.403267875553519e-05,
      "loss": 0.0904,
      "step": 280
    },
    {
      "epoch": 0.370045458170508,
      "grad_norm": 3647.082275390625,
      "learning_rate": 8.231496189304704e-05,
      "loss": 0.0939,
      "step": 290
    },
    {
      "epoch": 0.3828056463832841,
      "grad_norm": 3880.467529296875,
      "learning_rate": 8.052908734384399e-05,
      "loss": 0.1022,
      "step": 300
    },
    {
      "epoch": 0.3955658345960603,
      "grad_norm": 2817.3056640625,
      "learning_rate": 7.86788218175523e-05,
      "loss": 0.0798,
      "step": 310
    },
    {
      "epoch": 0.40832602280883645,
      "grad_norm": 2925.299560546875,
      "learning_rate": 7.676806783519304e-05,
      "loss": 0.0804,
      "step": 320
    },
    {
      "epoch": 0.4210862110216126,
      "grad_norm": 3108.75146484375,
      "learning_rate": 7.480085549810897e-05,
      "loss": 0.0905,
      "step": 330
    },
    {
      "epoch": 0.4338463992343887,
      "grad_norm": 3450.92529296875,
      "learning_rate": 7.278133398780366e-05,
      "loss": 0.0919,
      "step": 340
    },
    {
      "epoch": 0.44660658744716486,
      "grad_norm": 4121.8212890625,
      "learning_rate": 7.071376281461994e-05,
      "loss": 0.1026,
      "step": 350
    },
    {
      "epoch": 0.459366775659941,
      "grad_norm": 2675.00390625,
      "learning_rate": 6.860250283371681e-05,
      "loss": 0.0776,
      "step": 360
    },
    {
      "epoch": 0.47212696387271713,
      "grad_norm": 3107.235595703125,
      "learning_rate": 6.645200704729266e-05,
      "loss": 0.0811,
      "step": 370
    },
    {
      "epoch": 0.48488715208549327,
      "grad_norm": 3101.115478515625,
      "learning_rate": 6.426681121245527e-05,
      "loss": 0.0849,
      "step": 380
    },
    {
      "epoch": 0.4976473402982694,
      "grad_norm": 3454.42578125,
      "learning_rate": 6.205152427454747e-05,
      "loss": 0.0911,
      "step": 390
    },
    {
      "epoch": 0.5104075285110455,
      "grad_norm": 3896.09619140625,
      "learning_rate": 5.98108186461068e-05,
      "loss": 0.103,
      "step": 400
    },
    {
      "epoch": 0.5231677167238217,
      "grad_norm": 2808.469970703125,
      "learning_rate": 5.7549420351961844e-05,
      "loss": 0.0856,
      "step": 410
    },
    {
      "epoch": 0.5359279049365978,
      "grad_norm": 2859.574951171875,
      "learning_rate": 5.527209906125169e-05,
      "loss": 0.083,
      "step": 420
    },
    {
      "epoch": 0.548688093149374,
      "grad_norm": 3273.62646484375,
      "learning_rate": 5.29836580273917e-05,
      "loss": 0.0856,
      "step": 430
    },
    {
      "epoch": 0.5614482813621501,
      "grad_norm": 3307.667236328125,
      "learning_rate": 5.068892395720483e-05,
      "loss": 0.0892,
      "step": 440
    },
    {
      "epoch": 0.5742084695749262,
      "grad_norm": 3913.605224609375,
      "learning_rate": 4.839273683058542e-05,
      "loss": 0.0994,
      "step": 450
    },
    {
      "epoch": 0.5869686577877024,
      "grad_norm": 2523.923583984375,
      "learning_rate": 4.6099939692168107e-05,
      "loss": 0.0773,
      "step": 460
    },
    {
      "epoch": 0.5997288460004785,
      "grad_norm": 2690.344482421875,
      "learning_rate": 4.381536843653262e-05,
      "loss": 0.0787,
      "step": 470
    },
    {
      "epoch": 0.6124890342132546,
      "grad_norm": 3163.53369140625,
      "learning_rate": 4.1543841608489075e-05,
      "loss": 0.0869,
      "step": 480
    },
    {
      "epoch": 0.6252492224260308,
      "grad_norm": 3396.5908203125,
      "learning_rate": 3.9290150239957057e-05,
      "loss": 0.0928,
      "step": 490
    },
    {
      "epoch": 0.6380094106388069,
      "grad_norm": 3951.349365234375,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.0986,
      "step": 500
    },
    {
      "epoch": 0.650769598851583,
      "grad_norm": 2700.265625,
      "learning_rate": 3.4855239893445946e-05,
      "loss": 0.0822,
      "step": 510
    },
    {
      "epoch": 0.6635297870643592,
      "grad_norm": 2672.4716796875,
      "learning_rate": 3.268337488688748e-05,
      "loss": 0.0768,
      "step": 520
    },
    {
      "epoch": 0.6762899752771353,
      "grad_norm": 3514.844970703125,
      "learning_rate": 3.0548033553583705e-05,
      "loss": 0.0865,
      "step": 530
    },
    {
      "epoch": 0.6890501634899114,
      "grad_norm": 3326.14501953125,
      "learning_rate": 2.8453719687353285e-05,
      "loss": 0.0916,
      "step": 540
    },
    {
      "epoch": 0.7018103517026876,
      "grad_norm": 4190.75732421875,
      "learning_rate": 2.6404850548190235e-05,
      "loss": 0.1027,
      "step": 550
    },
    {
      "epoch": 0.7145705399154637,
      "grad_norm": 2811.360595703125,
      "learning_rate": 2.4405747545519963e-05,
      "loss": 0.0796,
      "step": 560
    },
    {
      "epoch": 0.7273307281282398,
      "grad_norm": 3221.398193359375,
      "learning_rate": 2.246062712362043e-05,
      "loss": 0.0796,
      "step": 570
    },
    {
      "epoch": 0.740090916341016,
      "grad_norm": 3041.626220703125,
      "learning_rate": 2.0573591868432406e-05,
      "loss": 0.0858,
      "step": 580
    },
    {
      "epoch": 0.7528511045537921,
      "grad_norm": 3378.84375,
      "learning_rate": 1.874862185451608e-05,
      "loss": 0.0862,
      "step": 590
    },
    {
      "epoch": 0.7656112927665683,
      "grad_norm": 3626.47265625,
      "learning_rate": 1.6989566250404702e-05,
      "loss": 0.097,
      "step": 600
    },
    {
      "epoch": 0.7783714809793445,
      "grad_norm": 2655.00146484375,
      "learning_rate": 1.5300135200061168e-05,
      "loss": 0.0786,
      "step": 610
    },
    {
      "epoch": 0.7911316691921206,
      "grad_norm": 2849.29931640625,
      "learning_rate": 1.368389199756075e-05,
      "loss": 0.081,
      "step": 620
    },
    {
      "epoch": 0.8038918574048968,
      "grad_norm": 3150.29931640625,
      "learning_rate": 1.2144245571504898e-05,
      "loss": 0.0824,
      "step": 630
    },
    {
      "epoch": 0.8166520456176729,
      "grad_norm": 3491.138427734375,
      "learning_rate": 1.0684443295017721e-05,
      "loss": 0.0897,
      "step": 640
    },
    {
      "epoch": 0.829412233830449,
      "grad_norm": 4051.920166015625,
      "learning_rate": 9.307564136490254e-06,
      "loss": 0.1012,
      "step": 650
    },
    {
      "epoch": 0.8421724220432252,
      "grad_norm": 2838.83544921875,
      "learning_rate": 8.016512165518569e-06,
      "loss": 0.0764,
      "step": 660
    },
    {
      "epoch": 0.8549326102560013,
      "grad_norm": 3116.049072265625,
      "learning_rate": 6.814010427732786e-06,
      "loss": 0.0795,
      "step": 670
    },
    {
      "epoch": 0.8676927984687774,
      "grad_norm": 3072.74365234375,
      "learning_rate": 5.702595201436101e-06,
      "loss": 0.083,
      "step": 680
    },
    {
      "epoch": 0.8804529866815536,
      "grad_norm": 3425.1845703125,
      "learning_rate": 4.684610648167503e-06,
      "loss": 0.0856,
      "step": 690
    },
    {
      "epoch": 0.8932131748943297,
      "grad_norm": 4003.127197265625,
      "learning_rate": 3.762203868471087e-06,
      "loss": 0.1022,
      "step": 700
    },
    {
      "epoch": 0.9059733631071059,
      "grad_norm": 2913.515380859375,
      "learning_rate": 2.9373203733000232e-06,
      "loss": 0.0802,
      "step": 710
    },
    {
      "epoch": 0.918733551319882,
      "grad_norm": 2771.475341796875,
      "learning_rate": 2.211699980606946e-06,
      "loss": 0.0775,
      "step": 720
    },
    {
      "epoch": 0.9314937395326581,
      "grad_norm": 2954.710205078125,
      "learning_rate": 1.5868731457754139e-06,
      "loss": 0.0845,
      "step": 730
    },
    {
      "epoch": 0.9442539277454343,
      "grad_norm": 3425.241455078125,
      "learning_rate": 1.064157733632276e-06,
      "loss": 0.0877,
      "step": 740
    },
    {
      "epoch": 0.9570141159582104,
      "grad_norm": 3987.344970703125,
      "learning_rate": 6.44656238849195e-07,
      "loss": 0.1024,
      "step": 750
    }
  ],
  "logging_steps": 10,
  "max_steps": 784,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.009329291264e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}