{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12,
  "eval_steps": 1000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 1.2400550842285156,
      "learning_rate": 7.920000000000001e-05,
      "loss": 8.362694702148438,
      "step": 100
    },
    {
      "epoch": 0.004,
      "grad_norm": 2.1569149494171143,
      "learning_rate": 0.00015920000000000002,
      "loss": 6.592593383789063,
      "step": 200
    },
    {
      "epoch": 0.006,
      "grad_norm": 4.070979595184326,
      "learning_rate": 0.00023920000000000001,
      "loss": 6.208985595703125,
      "step": 300
    },
    {
      "epoch": 0.008,
      "grad_norm": 4.392622470855713,
      "learning_rate": 0.0003192,
      "loss": 5.921519775390625,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.653286457061768,
      "learning_rate": 0.0003992,
      "loss": 5.658856201171875,
      "step": 500
    },
    {
      "epoch": 0.012,
      "grad_norm": 2.6610119342803955,
      "learning_rate": 0.00047920000000000005,
      "loss": 5.450134887695312,
      "step": 600
    },
    {
      "epoch": 0.014,
      "grad_norm": 2.6298203468322754,
      "learning_rate": 0.0005592,
      "loss": 5.260469360351562,
      "step": 700
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.681970238685608,
      "learning_rate": 0.0006392,
      "loss": 5.112535095214843,
      "step": 800
    },
    {
      "epoch": 0.018,
      "grad_norm": 1.5673184394836426,
      "learning_rate": 0.0007191999999999999,
      "loss": 5.0076907348632815,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6654455661773682,
      "learning_rate": 0.0007992,
      "loss": 4.9159228515625,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "eval_accuracy": 0.2274637964774951,
      "eval_loss": 4.8702616691589355,
      "eval_runtime": 5.1375,
      "eval_samples_per_second": 194.646,
      "eval_steps_per_second": 1.557,
      "step": 1000
    },
    {
      "epoch": 0.022,
      "grad_norm": 1.3562606573104858,
      "learning_rate": 0.0008792,
      "loss": 4.811459045410157,
      "step": 1100
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.1230770349502563,
      "learning_rate": 0.0009592000000000001,
      "loss": 4.759424438476563,
      "step": 1200
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.9026045203208923,
      "learning_rate": 0.0010391999999999999,
      "loss": 4.691336669921875,
      "step": 1300
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.8914104104042053,
      "learning_rate": 0.0011192,
      "loss": 4.6415283203125,
      "step": 1400
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8752658367156982,
      "learning_rate": 0.0011992,
      "loss": 4.560531311035156,
      "step": 1500
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.7435582280158997,
      "learning_rate": 0.0012791999999999999,
      "loss": 4.516304016113281,
      "step": 1600
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.7041499018669128,
      "learning_rate": 0.0013592,
      "loss": 4.48115234375,
      "step": 1700
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.5327283143997192,
      "learning_rate": 0.0014392,
      "loss": 4.429443969726562,
      "step": 1800
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.45207348465919495,
      "learning_rate": 0.0015192,
      "loss": 4.37362060546875,
      "step": 1900
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5152547359466553,
      "learning_rate": 0.0015992,
      "loss": 4.377236938476562,
      "step": 2000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.2694227005870842,
      "eval_loss": 4.3582682609558105,
      "eval_runtime": 5.3878,
      "eval_samples_per_second": 185.606,
      "eval_steps_per_second": 1.485,
      "step": 2000
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.45765218138694763,
      "learning_rate": 0.0016792,
      "loss": 4.341053161621094,
      "step": 2100
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.3776390254497528,
      "learning_rate": 0.0017592,
      "loss": 4.304255676269531,
      "step": 2200
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.48194995522499084,
      "learning_rate": 0.0018392,
      "loss": 4.308780822753906,
      "step": 2300
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.40603286027908325,
      "learning_rate": 0.0019192,
      "loss": 4.272806396484375,
      "step": 2400
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.41386568546295166,
      "learning_rate": 0.0019992,
      "loss": 4.267073974609375,
      "step": 2500
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.4916844666004181,
      "learning_rate": 0.001999978563623903,
      "loss": 4.220490417480469,
      "step": 2600
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.37477609515190125,
      "learning_rate": 0.0019999133871331223,
      "loss": 4.224794006347656,
      "step": 2700
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.405086874961853,
      "learning_rate": 0.0019998044711915177,
      "loss": 4.185610961914063,
      "step": 2800
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.334581196308136,
      "learning_rate": 0.0019996518205634257,
      "loss": 4.165193176269531,
      "step": 2900
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.3760662376880646,
      "learning_rate": 0.0019994554419262797,
      "loss": 4.161107177734375,
      "step": 3000
    },
    {
      "epoch": 0.06,
      "eval_accuracy": 0.28686888454011744,
      "eval_loss": 4.15764856338501,
      "eval_runtime": 5.1036,
      "eval_samples_per_second": 195.938,
      "eval_steps_per_second": 1.568,
      "step": 3000
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.38909581303596497,
      "learning_rate": 0.001999215343870317,
      "loss": 4.1489105224609375,
      "step": 3100
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.4108506143093109,
      "learning_rate": 0.0019989315368982054,
      "loss": 4.123897705078125,
      "step": 3200
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.27484413981437683,
      "learning_rate": 0.0019986040334245797,
      "loss": 4.101358032226562,
      "step": 3300
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.29958581924438477,
      "learning_rate": 0.001998232847775504,
      "loss": 4.10423583984375,
      "step": 3400
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.29483968019485474,
      "learning_rate": 0.0019978179961878404,
      "loss": 4.078011169433593,
      "step": 3500
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.36317598819732666,
      "learning_rate": 0.001997359496808541,
      "loss": 4.076428527832031,
      "step": 3600
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.34998658299446106,
      "learning_rate": 0.001996857369693855,
      "loss": 4.052371826171875,
      "step": 3700
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.31070172786712646,
      "learning_rate": 0.0019963116368084486,
      "loss": 4.059148559570312,
      "step": 3800
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.2655385732650757,
      "learning_rate": 0.001995722322024446,
      "loss": 4.039622802734375,
      "step": 3900
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4650065302848816,
      "learning_rate": 0.001995089451120385,
      "loss": 4.011789855957031,
      "step": 4000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.29829745596868884,
      "eval_loss": 4.0341691970825195,
      "eval_runtime": 5.1511,
      "eval_samples_per_second": 194.133,
      "eval_steps_per_second": 1.553,
      "step": 4000
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.43114492297172546,
      "learning_rate": 0.0019944130517800893,
      "loss": 4.0238558959960935,
      "step": 4100
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.26026225090026855,
      "learning_rate": 0.001993693153591457,
      "loss": 4.017567443847656,
      "step": 4200
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.23712828755378723,
      "learning_rate": 0.0019929297880451674,
      "loss": 3.9972265625,
      "step": 4300
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.44415482878685,
      "learning_rate": 0.0019921229885333023,
      "loss": 3.9874578857421876,
      "step": 4400
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.27471208572387695,
      "learning_rate": 0.001991272790347886,
      "loss": 3.998799133300781,
      "step": 4500
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.2248983383178711,
      "learning_rate": 0.0019903792306793415,
      "loss": 3.9638967895507813,
      "step": 4600
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.36029258370399475,
      "learning_rate": 0.001989442348614863,
      "loss": 3.972870178222656,
      "step": 4700
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.26471713185310364,
      "learning_rate": 0.0019884621851367075,
      "loss": 3.959658508300781,
      "step": 4800
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.23001010715961456,
      "learning_rate": 0.001987438783120401,
      "loss": 3.9595660400390624,
      "step": 4900
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.23091939091682434,
      "learning_rate": 0.001986372187332862,
      "loss": 3.9660198974609373,
      "step": 5000
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.3069041095890411,
      "eval_loss": 3.95263409614563,
      "eval_runtime": 5.269,
      "eval_samples_per_second": 189.788,
      "eval_steps_per_second": 1.518,
      "step": 5000
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.28329843282699585,
      "learning_rate": 0.0019852624444304467,
      "loss": 3.9279766845703126,
      "step": 5100
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.4390493333339691,
      "learning_rate": 0.0019841096029569044,
      "loss": 3.940526123046875,
      "step": 5200
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.22625097632408142,
      "learning_rate": 0.0019829137133412556,
      "loss": 3.9342303466796875,
      "step": 5300
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.21681281924247742,
      "learning_rate": 0.001981674827895587,
      "loss": 3.9172735595703125,
      "step": 5400
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.3054690957069397,
      "learning_rate": 0.00198039300081276,
      "loss": 3.920041198730469,
      "step": 5500
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.24492567777633667,
      "learning_rate": 0.0019790682881640448,
      "loss": 3.918515319824219,
      "step": 5600
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.20615942776203156,
      "learning_rate": 0.001977700747896664,
      "loss": 3.902435302734375,
      "step": 5700
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.2998456358909607,
      "learning_rate": 0.001976290439831259,
      "loss": 3.9016647338867188,
      "step": 5800
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.22004957497119904,
      "learning_rate": 0.0019748374256592736,
      "loss": 3.897244567871094,
      "step": 5900
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.2709659934043884,
      "learning_rate": 0.0019733417689402543,
      "loss": 3.892081604003906,
      "step": 6000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.31208414872798435,
      "eval_loss": 3.8932552337646484,
      "eval_runtime": 5.3089,
      "eval_samples_per_second": 188.362,
      "eval_steps_per_second": 1.507,
      "step": 6000
    }
  ],
  "logging_steps": 100,
  "max_steps": 50000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}