trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 25,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 56.660274505615234,
      "learning_rate": 2e-05,
      "loss": 6.2872,
      "step": 1
    },
    {
      "epoch": 2.0,
      "grad_norm": 56.60550308227539,
      "learning_rate": 2e-05,
      "loss": 6.2876,
      "step": 2
    },
    {
      "epoch": 3.0,
      "grad_norm": 57.12910461425781,
      "learning_rate": 2e-05,
      "loss": 6.2093,
      "step": 3
    },
    {
      "epoch": 4.0,
      "grad_norm": 57.13336181640625,
      "learning_rate": 2e-05,
      "loss": 6.2127,
      "step": 4
    },
    {
      "epoch": 5.0,
      "grad_norm": 58.30503845214844,
      "learning_rate": 2e-05,
      "loss": 6.1902,
      "step": 5
    },
    {
      "epoch": 6.0,
      "grad_norm": NaN,
      "learning_rate": 2e-05,
      "loss": 6.1393,
      "step": 6
    },
    {
      "epoch": 7.0,
      "grad_norm": 58.863895416259766,
      "learning_rate": 2e-05,
      "loss": 6.1219,
      "step": 7
    },
    {
      "epoch": 8.0,
      "grad_norm": 59.03289031982422,
      "learning_rate": 2e-05,
      "loss": 6.1135,
      "step": 8
    },
    {
      "epoch": 9.0,
      "grad_norm": 60.73344039916992,
      "learning_rate": 2e-05,
      "loss": 6.0639,
      "step": 9
    },
    {
      "epoch": 10.0,
      "grad_norm": 60.50625991821289,
      "learning_rate": 2e-05,
      "loss": 6.0119,
      "step": 10
    },
    {
      "epoch": 10.0,
      "step": 10,
      "total_flos": 20710135851600.0,
      "train_loss": 6.163743114471435,
      "train_runtime": 42.0404,
      "train_samples_per_second": 2.379,
      "train_steps_per_second": 0.238
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 20710135851600.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}