checkpoint-2500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.928571428571429,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.571428571428571,
      "grad_norm": 1.0717273950576782,
      "learning_rate": 0.00032727272727272726,
      "loss": 1.1979,
      "step": 1000
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 1.0056803226470947,
      "learning_rate": 0.00014545454545454546,
      "loss": 0.4925,
      "step": 2000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 2800,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6240558120960000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}