{
  "best_global_step": 540,
  "best_metric": 0.43301182985305786,
  "best_model_checkpoint": "Mistral-7B-v0.1/r4/checkpoint-540",
  "epoch": 3.0,
  "eval_steps": 60,
  "global_step": 720,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.3212984272175365,
      "epoch": 0.25069637883008355,
      "grad_norm": 4.5548248291015625,
      "learning_rate": 7.638888888888889e-05,
      "loss": 2.8492,
      "mean_token_accuracy": 0.6425296268943284,
      "num_tokens": 1474560.0,
      "step": 60
    },
    {
      "epoch": 0.25069637883008355,
      "eval_entropy": 0.6633218830669081,
      "eval_loss": 0.5672379732131958,
      "eval_mean_token_accuracy": 0.8630788239178719,
      "eval_num_tokens": 1474560.0,
      "eval_runtime": 1043.9462,
      "eval_samples_per_second": 2.357,
      "eval_steps_per_second": 0.295,
      "step": 60
    },
    {
      "entropy": 0.5157420211368137,
      "epoch": 0.5013927576601671,
      "grad_norm": 1.0181090831756592,
      "learning_rate": 9.891743688752738e-05,
      "loss": 0.5017,
      "mean_token_accuracy": 0.8742851712637477,
      "num_tokens": 2949120.0,
      "step": 120
    },
    {
      "epoch": 0.5013927576601671,
      "eval_entropy": 0.4764551486965124,
      "eval_loss": 0.46269991993904114,
      "eval_mean_token_accuracy": 0.882140948400869,
      "eval_num_tokens": 2949120.0,
      "eval_runtime": 1038.9576,
      "eval_samples_per_second": 2.369,
      "eval_steps_per_second": 0.296,
      "step": 120
    },
    {
      "entropy": 0.46944635676013097,
      "epoch": 0.7520891364902507,
      "grad_norm": 1.4101026058197021,
      "learning_rate": 9.389450641873323e-05,
      "loss": 0.4597,
      "mean_token_accuracy": 0.8829208799534374,
      "num_tokens": 4423680.0,
      "step": 180
    },
    {
      "epoch": 0.7520891364902507,
      "eval_entropy": 0.4749697573579751,
      "eval_loss": 0.45166516304016113,
      "eval_mean_token_accuracy": 0.8840483388343414,
      "eval_num_tokens": 4423680.0,
      "eval_runtime": 1040.6612,
      "eval_samples_per_second": 2.365,
      "eval_steps_per_second": 0.296,
      "step": 180
    },
    {
      "entropy": 0.4655254686648926,
      "epoch": 1.0,
      "grad_norm": 1.558998465538025,
      "learning_rate": 8.518351670729529e-05,
      "loss": 0.4543,
      "mean_token_accuracy": 0.8834420842735955,
      "num_tokens": 5879808.0,
      "step": 240
    },
    {
      "epoch": 1.0,
      "eval_entropy": 0.4612268569407525,
      "eval_loss": 0.4448107182979584,
      "eval_mean_token_accuracy": 0.8855506770022503,
      "eval_num_tokens": 5879808.0,
      "eval_runtime": 1040.4289,
      "eval_samples_per_second": 2.365,
      "eval_steps_per_second": 0.296,
      "step": 240
    },
    {
      "entropy": 0.45157561596069073,
      "epoch": 1.2506963788300836,
      "grad_norm": 0.6592407822608948,
      "learning_rate": 7.351637360519813e-05,
      "loss": 0.4387,
      "mean_token_accuracy": 0.8871405469046698,
      "num_tokens": 7354368.0,
      "step": 300
    },
    {
      "epoch": 1.2506963788300836,
      "eval_entropy": 0.4438396056557631,
      "eval_loss": 0.44106850028038025,
      "eval_mean_token_accuracy": 0.8862724611898521,
      "eval_num_tokens": 7354368.0,
      "eval_runtime": 1037.3961,
      "eval_samples_per_second": 2.372,
      "eval_steps_per_second": 0.297,
      "step": 300
    },
    {
      "entropy": 0.4534259519229333,
      "epoch": 1.501392757660167,
      "grad_norm": 0.6076720356941223,
      "learning_rate": 5.9873361855649876e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8862164333462715,
      "num_tokens": 8828928.0,
      "step": 360
    },
    {
      "epoch": 1.501392757660167,
      "eval_entropy": 0.44579531716835963,
      "eval_loss": 0.43730634450912476,
      "eval_mean_token_accuracy": 0.8868549373242762,
      "eval_num_tokens": 8828928.0,
      "eval_runtime": 1044.4842,
      "eval_samples_per_second": 2.356,
      "eval_steps_per_second": 0.295,
      "step": 360
    },
    {
      "entropy": 0.44368693138621434,
      "epoch": 1.7520891364902507,
      "grad_norm": 0.6398904919624329,
      "learning_rate": 4.5400780612818626e-05,
      "loss": 0.4223,
      "mean_token_accuracy": 0.888861709115218,
      "num_tokens": 11753472.0,
      "step": 420
    },
    {
      "epoch": 1.7520891364902507,
      "eval_entropy": 0.4433383915524978,
      "eval_loss": 0.4355394244194031,
      "eval_mean_token_accuracy": 0.8874243900373384,
      "eval_num_tokens": 11753472.0,
      "eval_runtime": 1037.2958,
      "eval_samples_per_second": 2.373,
      "eval_steps_per_second": 0.297,
      "step": 420
    },
    {
      "entropy": 0.4386935969919301,
      "epoch": 2.0,
      "grad_norm": 1.2995598316192627,
      "learning_rate": 3.131463026883449e-05,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8898014740997486,
      "num_tokens": 13209600.0,
      "step": 480
    },
    {
      "epoch": 2.0,
      "eval_entropy": 0.4436397281076227,
      "eval_loss": 0.4339684844017029,
      "eval_mean_token_accuracy": 0.8877705109196824,
      "eval_num_tokens": 13209600.0,
      "eval_runtime": 1041.4468,
      "eval_samples_per_second": 2.363,
      "eval_steps_per_second": 0.296,
      "step": 480
    },
    {
      "entropy": 0.43632681195934614,
      "epoch": 2.2506963788300833,
      "grad_norm": 0.5876320004463196,
      "learning_rate": 1.8798442914793663e-05,
      "loss": 0.4221,
      "mean_token_accuracy": 0.8905482163031896,
      "num_tokens": 14684160.0,
      "step": 540
    },
    {
      "epoch": 2.2506963788300833,
      "eval_entropy": 0.4378794138985021,
      "eval_loss": 0.43301182985305786,
      "eval_mean_token_accuracy": 0.8880488022968367,
      "eval_num_tokens": 14684160.0,
      "eval_runtime": 1046.5384,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.294,
      "step": 540
    },
    {
      "entropy": 0.4391553503357702,
      "epoch": 2.501392757660167,
      "grad_norm": 0.630813479423523,
      "learning_rate": 8.903840820084096e-06,
      "loss": 0.4249,
      "mean_token_accuracy": 0.8895656656887796,
      "num_tokens": 16158720.0,
      "step": 600
    },
    {
      "epoch": 2.501392757660167,
      "eval_entropy": 0.4278067753880055,
      "eval_loss": 0.4333657920360565,
      "eval_mean_token_accuracy": 0.8881023328799706,
      "eval_num_tokens": 16158720.0,
      "eval_runtime": 1045.4866,
      "eval_samples_per_second": 2.354,
      "eval_steps_per_second": 0.295,
      "step": 600
    },
    {
      "entropy": 0.42997388628621896,
      "epoch": 2.752089136490251,
      "grad_norm": 0.7219184637069702,
      "learning_rate": 2.462178103591678e-06,
      "loss": 0.4144,
      "mean_token_accuracy": 0.8923869328366385,
      "num_tokens": 17633280.0,
      "step": 660
    },
    {
      "epoch": 2.752089136490251,
      "eval_entropy": 0.42755064682720545,
      "eval_loss": 0.4343859851360321,
      "eval_mean_token_accuracy": 0.8881968432045603,
      "eval_num_tokens": 17633280.0,
      "eval_runtime": 1047.511,
      "eval_samples_per_second": 2.349,
      "eval_steps_per_second": 0.294,
      "step": 660
    },
    {
      "entropy": 0.43753848961564934,
      "epoch": 3.0,
      "grad_norm": 1.3657286167144775,
      "learning_rate": 1.4689549754337562e-08,
      "loss": 0.4228,
      "mean_token_accuracy": 0.8900845710481151,
      "num_tokens": 19089408.0,
      "step": 720
    },
    {
      "epoch": 3.0,
      "eval_entropy": 0.42660339708839146,
      "eval_loss": 0.4343104064464569,
      "eval_mean_token_accuracy": 0.8882718165586521,
      "eval_num_tokens": 19089408.0,
      "eval_runtime": 1048.422,
      "eval_samples_per_second": 2.347,
      "eval_steps_per_second": 0.294,
      "step": 720
    },
    {
      "epoch": 3.0,
      "step": 720,
      "total_flos": 8.146999757147996e+17,
      "train_loss": 0.035229322645399304,
      "train_runtime": 4770.3564,
      "train_samples_per_second": 7.222,
      "train_steps_per_second": 0.151
    }
  ],
  "logging_steps": 60,
  "max_steps": 720,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 60,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.01
      },
      "attributes": {
        "early_stopping_patience_counter": 13
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.146999757147996e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}