trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 100.0,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 10.0,
      "grad_norm": 73.0,
      "learning_rate": 0.0001,
      "loss": 6.6386,
      "step": 10
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.890625,
      "learning_rate": 0.0001,
      "loss": 4.25,
      "step": 20
    },
    {
      "epoch": 30.0,
      "grad_norm": 7.5625,
      "learning_rate": 0.0001,
      "loss": 3.2984,
      "step": 30
    },
    {
      "epoch": 40.0,
      "grad_norm": 3.359375,
      "learning_rate": 0.0001,
      "loss": 2.9021,
      "step": 40
    },
    {
      "epoch": 50.0,
      "grad_norm": 2.734375,
      "learning_rate": 0.0001,
      "loss": 2.2109,
      "step": 50
    },
    {
      "epoch": 60.0,
      "grad_norm": 7.21875,
      "learning_rate": 0.0001,
      "loss": 1.1565,
      "step": 60
    },
    {
      "epoch": 70.0,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0001,
      "loss": 0.1664,
      "step": 70
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.0001,
      "loss": 0.0017,
      "step": 80
    },
    {
      "epoch": 90.0,
      "grad_norm": 0.00592041015625,
      "learning_rate": 0.0001,
      "loss": 0.0004,
      "step": 90
    },
    {
      "epoch": 100.0,
      "grad_norm": 0.0036468505859375,
      "learning_rate": 0.0001,
      "loss": 0.0002,
      "step": 100
    }
  ],
  "logging_steps": 10,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1168359222067200.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}