trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 14.82353919212971,
      "learning_rate": 5e-06,
      "loss": 1.1242,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.0261839628219604,
      "eval_runtime": 9.8591,
      "eval_samples_per_second": 32.559,
      "eval_steps_per_second": 0.609,
      "step": 12
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 7.332552262162801,
      "learning_rate": 5e-06,
      "loss": 1.0017,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.9571768641471863,
      "eval_runtime": 9.6257,
      "eval_samples_per_second": 33.348,
      "eval_steps_per_second": 0.623,
      "step": 24
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.0127017936556957,
      "learning_rate": 5e-06,
      "loss": 0.9348,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.9238602519035339,
      "eval_runtime": 9.0015,
      "eval_samples_per_second": 35.661,
      "eval_steps_per_second": 0.667,
      "step": 36
    },
    {
      "epoch": 3.0,
      "step": 36,
      "total_flos": 60091961180160.0,
      "train_loss": 0.99822465578715,
      "train_runtime": 1482.7704,
      "train_samples_per_second": 12.336,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 10,
  "max_steps": 36,
  "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": 60091961180160.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}