trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 34,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 12.47676411128599,
      "learning_rate": 9.330127018922195e-06,
      "loss": 5.493,
      "step": 10
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 3.8434719643397486,
      "learning_rate": 5e-06,
      "loss": 0.1882,
      "step": 20
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 6.466585393095021,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.1184,
      "step": 30
    },
    {
      "epoch": 2.0,
      "step": 34,
      "total_flos": 4767598772224.0,
      "train_loss": 1.7220316529273987,
      "train_runtime": 350.6476,
      "train_samples_per_second": 3.091,
      "train_steps_per_second": 0.097
    }
  ],
  "logging_steps": 10,
  "max_steps": 34,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4767598772224.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}