{
  "best_global_step": 6000,
  "best_metric": 0.034662000834941864,
  "best_model_checkpoint": "./training_output/checkpoint-6000",
  "epoch": 0.75,
  "eval_steps": 1000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0625,
      "grad_norm": 1.8642356395721436,
      "learning_rate": 1.8752500000000004e-05,
      "loss": 0.1279,
      "step": 500
    },
    {
      "epoch": 0.125,
      "grad_norm": 3.2498440742492676,
      "learning_rate": 1.7502500000000004e-05,
      "loss": 0.0926,
      "step": 1000
    },
    {
      "epoch": 0.125,
      "eval_accuracy": 0.944625,
      "eval_loss": 0.04429839178919792,
      "eval_runtime": 500.554,
      "eval_samples_per_second": 15.982,
      "eval_steps_per_second": 0.999,
      "step": 1000
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.005530397407710552,
      "learning_rate": 1.62525e-05,
      "loss": 0.0612,
      "step": 1500
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.27407366037368774,
      "learning_rate": 1.5002500000000002e-05,
      "loss": 0.0565,
      "step": 2000
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.925125,
      "eval_loss": 0.04796064645051956,
      "eval_runtime": 500.334,
      "eval_samples_per_second": 15.989,
      "eval_steps_per_second": 0.999,
      "step": 2000
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.02653448097407818,
      "learning_rate": 1.3752500000000003e-05,
      "loss": 0.0527,
      "step": 2500
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.2703651785850525,
      "learning_rate": 1.2502500000000003e-05,
      "loss": 0.0442,
      "step": 3000
    },
    {
      "epoch": 0.375,
      "eval_accuracy": 0.887125,
      "eval_loss": 0.21262305974960327,
      "eval_runtime": 500.0079,
      "eval_samples_per_second": 16.0,
      "eval_steps_per_second": 1.0,
      "step": 3000
    },
    {
      "epoch": 0.4375,
      "grad_norm": 0.08504923433065414,
      "learning_rate": 1.1252500000000001e-05,
      "loss": 0.0369,
      "step": 3500
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.14933931827545166,
      "learning_rate": 1.0002500000000001e-05,
      "loss": 0.0383,
      "step": 4000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.947875,
      "eval_loss": 0.071560338139534,
      "eval_runtime": 499.5693,
      "eval_samples_per_second": 16.014,
      "eval_steps_per_second": 1.001,
      "step": 4000
    },
    {
      "epoch": 0.5625,
      "grad_norm": 1.824542760848999,
      "learning_rate": 8.7525e-06,
      "loss": 0.0309,
      "step": 4500
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.012093538418412209,
      "learning_rate": 7.502500000000001e-06,
      "loss": 0.0332,
      "step": 5000
    },
    {
      "epoch": 0.625,
      "eval_accuracy": 0.877375,
      "eval_loss": 0.23829498887062073,
      "eval_runtime": 499.6395,
      "eval_samples_per_second": 16.012,
      "eval_steps_per_second": 1.001,
      "step": 5000
    },
    {
      "epoch": 0.6875,
      "grad_norm": 0.006525672040879726,
      "learning_rate": 6.2525e-06,
      "loss": 0.0281,
      "step": 5500
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.004029456060379744,
      "learning_rate": 5.0025e-06,
      "loss": 0.0253,
      "step": 6000
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.970625,
      "eval_loss": 0.034662000834941864,
      "eval_runtime": 499.746,
      "eval_samples_per_second": 16.008,
      "eval_steps_per_second": 1.001,
      "step": 6000
    }
  ],
  "logging_steps": 500,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2366352719872e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}