trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 105,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 9.201884414807651,
      "learning_rate": 5e-06,
      "loss": 0.9991,
      "step": 10
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.5083267309317445,
      "learning_rate": 5e-06,
      "loss": 0.8974,
      "step": 20
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.4531839790824428,
      "learning_rate": 5e-06,
      "loss": 0.8534,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8311977386474609,
      "eval_runtime": 25.0496,
      "eval_samples_per_second": 37.685,
      "eval_steps_per_second": 0.599,
      "step": 35
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 2.251516118880847,
      "learning_rate": 5e-06,
      "loss": 0.8148,
      "step": 40
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.9005237142151287,
      "learning_rate": 5e-06,
      "loss": 0.7837,
      "step": 50
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.8989515401550041,
      "learning_rate": 5e-06,
      "loss": 0.7796,
      "step": 60
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7329080348225255,
      "learning_rate": 5e-06,
      "loss": 0.7696,
      "step": 70
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.803409218788147,
      "eval_runtime": 24.4026,
      "eval_samples_per_second": 38.684,
      "eval_steps_per_second": 0.615,
      "step": 70
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 1.107006832030335,
      "learning_rate": 5e-06,
      "loss": 0.7186,
      "step": 80
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.9752023188417169,
      "learning_rate": 5e-06,
      "loss": 0.7148,
      "step": 90
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6496713319096532,
      "learning_rate": 5e-06,
      "loss": 0.7087,
      "step": 100
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7998417019844055,
      "eval_runtime": 23.1343,
      "eval_samples_per_second": 40.805,
      "eval_steps_per_second": 0.648,
      "step": 105
    },
    {
      "epoch": 3.0,
      "step": 105,
      "total_flos": 175669531115520.0,
      "train_loss": 0.7996833460671561,
      "train_runtime": 3733.4078,
      "train_samples_per_second": 14.397,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 10,
  "max_steps": 105,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 175669531115520.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}