trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.4,
  "eval_steps": 500,
  "global_step": 9,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 69.07160949707031,
      "learning_rate": 5e-06,
      "loss": 1.9082,
      "step": 1
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 28.931533813476562,
      "learning_rate": 5e-06,
      "loss": 1.829,
      "step": 2
    },
    {
      "epoch": 0.8,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 1.7257,
      "step": 3
    },
    {
      "epoch": 0.8,
      "eval_loss": NaN,
      "eval_runtime": 2.4891,
      "eval_samples_per_second": 40.577,
      "eval_steps_per_second": 0.804,
      "step": 3
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 4
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 5
    },
    {
      "epoch": 1.6,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 6
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 7
    },
    {
      "epoch": 1.8666666666666667,
      "eval_loss": NaN,
      "eval_runtime": 2.4704,
      "eval_samples_per_second": 40.884,
      "eval_steps_per_second": 0.81,
      "step": 7
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 8
    },
    {
      "epoch": 2.4,
      "grad_norm": NaN,
      "learning_rate": 5e-06,
      "loss": 0.0,
      "step": 9
    },
    {
      "epoch": 2.4,
      "eval_loss": NaN,
      "eval_runtime": 1.9225,
      "eval_samples_per_second": 52.537,
      "eval_steps_per_second": 1.04,
      "step": 9
    },
    {
      "epoch": 2.4,
      "step": 9,
      "total_flos": 4.249521302148219e+17,
      "train_loss": 0.606980217827691,
      "train_runtime": 621.1198,
      "train_samples_per_second": 9.269,
      "train_steps_per_second": 0.014
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 9,
  "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": 4.249521302148219e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}