trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.997686030934113,
  "eval_steps": 500,
  "global_step": 3078,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.48715138229204724,
      "grad_norm": 0.16452117264270782,
      "learning_rate": 0.00019316938294480988,
      "loss": 1.498,
      "step": 500
    },
    {
      "epoch": 0.9743027645840945,
      "grad_norm": 0.1559494435787201,
      "learning_rate": 0.00016145696457272202,
      "loss": 1.3549,
      "step": 1000
    },
    {
      "epoch": 1.4608452076482767,
      "grad_norm": 0.18282005190849304,
      "learning_rate": 0.00011243098608364038,
      "loss": 1.1913,
      "step": 1500
    },
    {
      "epoch": 1.947996589940324,
      "grad_norm": 0.21383380889892578,
      "learning_rate": 5.990296960148566e-05,
      "loss": 1.1556,
      "step": 2000
    },
    {
      "epoch": 2.4345390330045062,
      "grad_norm": 0.29522448778152466,
      "learning_rate": 1.8671025926443465e-05,
      "loss": 0.9622,
      "step": 2500
    },
    {
      "epoch": 2.9216904152965535,
      "grad_norm": 0.2826710343360901,
      "learning_rate": 3.509538621403463e-07,
      "loss": 0.9327,
      "step": 3000
    }
  ],
  "logging_steps": 500,
  "max_steps": 3078,
  "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": 5.296618368751632e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}