trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.742857142857143,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9142857142857143,
      "eval_loss": 1.0328664779663086,
      "eval_runtime": 3.7236,
      "eval_samples_per_second": 30.884,
      "eval_steps_per_second": 0.537,
      "step": 4
    },
    {
      "epoch": 1.8285714285714287,
      "eval_loss": 0.9912787079811096,
      "eval_runtime": 3.7287,
      "eval_samples_per_second": 30.842,
      "eval_steps_per_second": 0.536,
      "step": 8
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 4.032851257413885,
      "learning_rate": 5e-06,
      "loss": 1.1915,
      "step": 10
    },
    {
      "epoch": 2.742857142857143,
      "eval_loss": 0.974350094795227,
      "eval_runtime": 3.1997,
      "eval_samples_per_second": 35.941,
      "eval_steps_per_second": 0.625,
      "step": 12
    },
    {
      "epoch": 2.742857142857143,
      "step": 12,
      "total_flos": 19891067289600.0,
      "train_loss": 1.1598685383796692,
      "train_runtime": 630.3631,
      "train_samples_per_second": 10.38,
      "train_steps_per_second": 0.019
    }
  ],
  "logging_steps": 10,
  "max_steps": 12,
  "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": 19891067289600.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}