trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 200.0,
  "eval_steps": 200,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 20.0,
      "grad_norm": 0.38066378235816956,
      "learning_rate": 8e-05,
      "loss": 0.1782,
      "step": 20
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.008850287646055222,
      "learning_rate": 0.00016,
      "loss": 0.0092,
      "step": 40
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.0006731236935593188,
      "learning_rate": 0.0001866666666666667,
      "loss": 0.0,
      "step": 60
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.00021651879069395363,
      "learning_rate": 0.00016,
      "loss": 0.0,
      "step": 80
    },
    {
      "epoch": 100.0,
      "grad_norm": 0.00013040985504630953,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 120.0,
      "grad_norm": 9.823329310165718e-05,
      "learning_rate": 0.00010666666666666667,
      "loss": 0.0,
      "step": 120
    },
    {
      "epoch": 140.0,
      "grad_norm": 8.462095138384029e-05,
      "learning_rate": 8e-05,
      "loss": 0.0,
      "step": 140
    },
    {
      "epoch": 160.0,
      "grad_norm": 7.46678197174333e-05,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.0,
      "step": 160
    },
    {
      "epoch": 180.0,
      "grad_norm": 6.831328937551007e-05,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.0,
      "step": 180
    },
    {
      "epoch": 200.0,
      "grad_norm": 6.943791231606156e-05,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 200
    },
    {
      "epoch": 200.0,
      "eval_loss": 1.2972288459422998e-05,
      "eval_runtime": 0.4099,
      "eval_samples_per_second": 2.439,
      "eval_steps_per_second": 2.439,
      "step": 200
    },
    {
      "epoch": 200.0,
      "step": 200,
      "total_flos": 1572670596710400.0,
      "train_loss": 0.018746382653935145,
      "train_runtime": 146.7988,
      "train_samples_per_second": 21.799,
      "train_steps_per_second": 1.362
    }
  ],
  "logging_steps": 20,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 200,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1572670596710400.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}