{
  "best_global_step": 50,
  "best_metric": 0.5247398614883423,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15037593984962405,
      "grad_norm": 1994.7098388671875,
      "learning_rate": 8.571428571428571e-06,
      "loss": 34.788,
      "step": 10
    },
    {
      "epoch": 0.3007518796992481,
      "grad_norm": 478.9690246582031,
      "learning_rate": 1.8095238095238097e-05,
      "loss": 16.7412,
      "step": 20
    },
    {
      "epoch": 0.45112781954887216,
      "grad_norm": 247.04168701171875,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 10.0861,
      "step": 30
    },
    {
      "epoch": 0.6015037593984962,
      "grad_norm": 130.6533203125,
      "learning_rate": 1.8e-05,
      "loss": 7.6831,
      "step": 40
    },
    {
      "epoch": 0.7518796992481203,
      "grad_norm": 158.5700225830078,
      "learning_rate": 1.688888888888889e-05,
      "loss": 8.4341,
      "step": 50
    },
    {
      "epoch": 0.7518796992481203,
      "eval_accuracy": 0.71982421875,
      "eval_loss": 0.5247398614883423,
      "eval_runtime": 1.3258,
      "eval_samples_per_second": 565.679,
      "eval_steps_per_second": 18.102,
      "step": 50
    },
    {
      "epoch": 0.9022556390977443,
      "grad_norm": 302.4800720214844,
      "learning_rate": 1.577777777777778e-05,
      "loss": 8.2982,
      "step": 60
    },
    {
      "epoch": 1.045112781954887,
      "grad_norm": 147.75057983398438,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 8.1503,
      "step": 70
    },
    {
      "epoch": 1.1954887218045114,
      "grad_norm": 37.97437286376953,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 7.4996,
      "step": 80
    },
    {
      "epoch": 1.3458646616541352,
      "grad_norm": 217.54083251953125,
      "learning_rate": 1.2444444444444446e-05,
      "loss": 7.1527,
      "step": 90
    },
    {
      "epoch": 1.4962406015037595,
      "grad_norm": 198.07823181152344,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 6.4484,
      "step": 100
    },
    {
      "epoch": 1.4962406015037595,
      "eval_accuracy": 0.749609375,
      "eval_loss": 0.539933979511261,
      "eval_runtime": 1.3139,
      "eval_samples_per_second": 570.817,
      "eval_steps_per_second": 18.266,
      "step": 100
    },
    {
      "epoch": 1.6466165413533833,
      "grad_norm": 267.6050109863281,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 7.155,
      "step": 110
    },
    {
      "epoch": 1.7969924812030076,
      "grad_norm": 105.75324249267578,
      "learning_rate": 9.111111111111112e-06,
      "loss": 6.9728,
      "step": 120
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 242.2601318359375,
      "learning_rate": 8.000000000000001e-06,
      "loss": 6.3463,
      "step": 130
    },
    {
      "epoch": 2.090225563909774,
      "grad_norm": 29.438447952270508,
      "learning_rate": 6.88888888888889e-06,
      "loss": 4.3959,
      "step": 140
    },
    {
      "epoch": 2.2406015037593985,
      "grad_norm": 170.81954956054688,
      "learning_rate": 5.777777777777778e-06,
      "loss": 3.5158,
      "step": 150
    },
    {
      "epoch": 2.2406015037593985,
      "eval_accuracy": 0.76240234375,
      "eval_loss": 0.5935441851615906,
      "eval_runtime": 1.3059,
      "eval_samples_per_second": 574.323,
      "eval_steps_per_second": 18.378,
      "step": 150
    },
    {
      "epoch": 2.3909774436090228,
      "grad_norm": 115.6646957397461,
      "learning_rate": 4.666666666666667e-06,
      "loss": 3.3307,
      "step": 160
    },
    {
      "epoch": 2.5413533834586466,
      "grad_norm": 102.5296630859375,
      "learning_rate": 3.555555555555556e-06,
      "loss": 2.8094,
      "step": 170
    },
    {
      "epoch": 2.6917293233082704,
      "grad_norm": 70.7061767578125,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 2.7766,
      "step": 180
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 146.64231872558594,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 3.0117,
      "step": 190
    },
    {
      "epoch": 2.992481203007519,
      "grad_norm": 122.64640045166016,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 2.7322,
      "step": 200
    },
    {
      "epoch": 2.992481203007519,
      "eval_accuracy": 0.759375,
      "eval_loss": 0.5873631238937378,
      "eval_runtime": 1.3022,
      "eval_samples_per_second": 575.956,
      "eval_steps_per_second": 18.431,
      "step": 200
    }
  ],
  "logging_steps": 10,
  "max_steps": 201,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0258179678850253e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}