trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 10.068081855773926,
      "learning_rate": 0.00018333333333333334,
      "loss": 76.6032,
      "step": 2
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.153128147125244,
      "learning_rate": 0.00015000000000000001,
      "loss": 51.6375,
      "step": 4
    },
    {
      "epoch": 1.5925925925925926,
      "grad_norm": 14.278098106384277,
      "learning_rate": 0.00011666666666666668,
      "loss": 71.4472,
      "step": 6
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.574069023132324,
      "learning_rate": 0.0001,
      "loss": 46.9068,
      "step": 8
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 16.224937438964844,
      "learning_rate": 6.666666666666667e-05,
      "loss": 65.4849,
      "step": 10
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.509884834289551,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 43.5394,
      "step": 12
    },
    {
      "epoch": 3.0,
      "step": 12,
      "total_flos": 45454155461280.0,
      "train_loss": 59.26982498168945,
      "train_runtime": 282.0933,
      "train_samples_per_second": 0.766,
      "train_steps_per_second": 0.043
    }
  ],
  "logging_steps": 2,
  "max_steps": 12,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 45454155461280.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}