{
  "best_global_step": 400,
  "best_metric": 0.19629092514514923,
  "best_model_checkpoint": "/workspace/andy-D1-andyft/checkpoint-400",
  "epoch": 0.7358013336399172,
  "eval_steps": 100,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.358754662796855,
      "epoch": 0.03679006668199586,
      "grad_norm": 0.322265625,
      "learning_rate": 6.785714285714286e-05,
      "loss": 1.9387958526611329,
      "mean_token_accuracy": 0.6328379280865193,
      "num_tokens": 552605.0,
      "step": 20
    },
    {
      "entropy": 1.2090021014213561,
      "epoch": 0.07358013336399172,
      "grad_norm": 0.3359375,
      "learning_rate": 9.988791089826816e-05,
      "loss": 1.0881237983703613,
      "mean_token_accuracy": 0.7621329009532929,
      "num_tokens": 1118812.0,
      "step": 40
    },
    {
      "entropy": 0.5365577314049006,
      "epoch": 0.11037020004598758,
      "grad_norm": 0.265625,
      "learning_rate": 9.91120793909987e-05,
      "loss": 0.5044384002685547,
      "mean_token_accuracy": 0.8858215417712927,
      "num_tokens": 1693574.0,
      "step": 60
    },
    {
      "entropy": 0.4165896608028561,
      "epoch": 0.14716026672798344,
      "grad_norm": 0.2451171875,
      "learning_rate": 9.760895058620235e-05,
      "loss": 0.38671829700469973,
      "mean_token_accuracy": 0.9098233781754971,
      "num_tokens": 2267338.0,
      "step": 80
    },
    {
      "entropy": 0.3438017311505973,
      "epoch": 0.1839503334099793,
      "grad_norm": 0.2197265625,
      "learning_rate": 9.540078421165121e-05,
      "loss": 0.3093204736709595,
      "mean_token_accuracy": 0.9240799810737371,
      "num_tokens": 2850973.0,
      "step": 100
    },
    {
      "epoch": 0.1839503334099793,
      "eval_entropy": 0.33466541236795194,
      "eval_loss": 0.3281489610671997,
      "eval_mean_token_accuracy": 0.9268425211115175,
      "eval_num_tokens": 2850973.0,
      "eval_runtime": 156.4612,
      "eval_samples_per_second": 2.921,
      "eval_steps_per_second": 1.464,
      "step": 100
    },
    {
      "entropy": 0.3134843219071627,
      "epoch": 0.22074040009197515,
      "grad_norm": 0.189453125,
      "learning_rate": 9.252028084644137e-05,
      "loss": 0.28039026260375977,
      "mean_token_accuracy": 0.92984459400177,
      "num_tokens": 3429327.0,
      "step": 120
    },
    {
      "entropy": 0.3144511237274855,
      "epoch": 0.257530466773971,
      "grad_norm": 0.22265625,
      "learning_rate": 8.901009766043847e-05,
      "loss": 0.27316670417785643,
      "mean_token_accuracy": 0.9302996207028628,
      "num_tokens": 3995502.0,
      "step": 140
    },
    {
      "entropy": 0.28906950131058695,
      "epoch": 0.29432053345596687,
      "grad_norm": 0.267578125,
      "learning_rate": 8.492221670727694e-05,
      "loss": 0.2562049150466919,
      "mean_token_accuracy": 0.9345210112631321,
      "num_tokens": 4561185.0,
      "step": 160
    },
    {
      "entropy": 0.309068296616897,
      "epoch": 0.33111060013796273,
      "grad_norm": 0.255859375,
      "learning_rate": 8.031717512581784e-05,
      "loss": 0.2678884267807007,
      "mean_token_accuracy": 0.9299660053104162,
      "num_tokens": 5133715.0,
      "step": 180
    },
    {
      "entropy": 0.27301658713258803,
      "epoch": 0.3679006668199586,
      "grad_norm": 0.35546875,
      "learning_rate": 7.526316864995647e-05,
      "loss": 0.23800337314605713,
      "mean_token_accuracy": 0.938116031512618,
      "num_tokens": 5698771.0,
      "step": 200
    },
    {
      "epoch": 0.3679006668199586,
      "eval_entropy": 0.261124916667501,
      "eval_loss": 0.24931415915489197,
      "eval_mean_token_accuracy": 0.9429609785954505,
      "eval_num_tokens": 5698771.0,
      "eval_runtime": 156.272,
      "eval_samples_per_second": 2.924,
      "eval_steps_per_second": 1.465,
      "step": 200
    },
    {
      "entropy": 0.2801984048448503,
      "epoch": 0.40469073350195445,
      "grad_norm": 0.294921875,
      "learning_rate": 6.983504170283999e-05,
      "loss": 0.2422886610031128,
      "mean_token_accuracy": 0.9372635453939437,
      "num_tokens": 6263699.0,
      "step": 220
    },
    {
      "entropy": 0.2785345103126019,
      "epoch": 0.4414808001839503,
      "grad_norm": 0.2578125,
      "learning_rate": 6.411317903111779e-05,
      "loss": 0.24179139137268066,
      "mean_token_accuracy": 0.9373910058289766,
      "num_tokens": 6825294.0,
      "step": 240
    },
    {
      "entropy": 0.2691320342477411,
      "epoch": 0.4782708668659462,
      "grad_norm": 0.2578125,
      "learning_rate": 5.8182315292934406e-05,
      "loss": 0.23189406394958495,
      "mean_token_accuracy": 0.9388320956379175,
      "num_tokens": 7391294.0,
      "step": 260
    },
    {
      "entropy": 0.24280655251350253,
      "epoch": 0.515060933547942,
      "grad_norm": 0.29296875,
      "learning_rate": 5.213028022839188e-05,
      "loss": 0.20993587970733643,
      "mean_token_accuracy": 0.9451322961598635,
      "num_tokens": 7949107.0,
      "step": 280
    },
    {
      "entropy": 0.24249689073767514,
      "epoch": 0.5518510002299379,
      "grad_norm": 0.275390625,
      "learning_rate": 4.6046697995163003e-05,
      "loss": 0.21177098751068116,
      "mean_token_accuracy": 0.9459892265498638,
      "num_tokens": 8526495.0,
      "step": 300
    },
    {
      "epoch": 0.5518510002299379,
      "eval_entropy": 0.22384919074880505,
      "eval_loss": 0.2132847160100937,
      "eval_mean_token_accuracy": 0.9506369665200013,
      "eval_num_tokens": 8526495.0,
      "eval_runtime": 156.3249,
      "eval_samples_per_second": 2.923,
      "eval_steps_per_second": 1.465,
      "step": 300
    },
    {
      "entropy": 0.21746529480442406,
      "epoch": 0.5886410669119337,
      "grad_norm": 0.27734375,
      "learning_rate": 4.002165993070237e-05,
      "loss": 0.1864999294281006,
      "mean_token_accuracy": 0.9523148186504841,
      "num_tokens": 9103591.0,
      "step": 320
    },
    {
      "entropy": 0.22349709873087703,
      "epoch": 0.6254311335939297,
      "grad_norm": 0.26953125,
      "learning_rate": 3.414439039602495e-05,
      "loss": 0.18752355575561525,
      "mean_token_accuracy": 0.9500059768557548,
      "num_tokens": 9677492.0,
      "step": 340
    },
    {
      "entropy": 0.20582567660603673,
      "epoch": 0.6622212002759255,
      "grad_norm": 0.3203125,
      "learning_rate": 2.8501925458477353e-05,
      "loss": 0.1842264413833618,
      "mean_token_accuracy": 0.9537915319204331,
      "num_tokens": 10255825.0,
      "step": 360
    },
    {
      "entropy": 0.2267866414040327,
      "epoch": 0.6990112669579214,
      "grad_norm": 0.27734375,
      "learning_rate": 2.3177823980794727e-05,
      "loss": 0.19315229654312133,
      "mean_token_accuracy": 0.9484861396253109,
      "num_tokens": 10829988.0,
      "step": 380
    },
    {
      "entropy": 0.23870108206756413,
      "epoch": 0.7358013336399172,
      "grad_norm": 0.279296875,
      "learning_rate": 1.8250930203833732e-05,
      "loss": 0.2072918176651001,
      "mean_token_accuracy": 0.946313539147377,
      "num_tokens": 11399434.0,
      "step": 400
    },
    {
      "epoch": 0.7358013336399172,
      "eval_entropy": 0.20661315126710583,
      "eval_loss": 0.19629092514514923,
      "eval_mean_token_accuracy": 0.9545749706992936,
      "eval_num_tokens": 11399434.0,
      "eval_runtime": 155.9996,
      "eval_samples_per_second": 2.929,
      "eval_steps_per_second": 1.468,
      "step": 400
    }
  ],
  "logging_steps": 20,
  "max_steps": 544,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.45469995696554e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}