trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.010036130068245684,
  "eval_steps": 500,
  "global_step": 75,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0026763013515321826,
      "grad_norm": 3.8896071910858154,
      "learning_rate": 7.733333333333333e-05,
      "loss": 1.7014,
      "step": 20
    },
    {
      "epoch": 0.005352602703064365,
      "grad_norm": 4.4182610511779785,
      "learning_rate": 5.0666666666666674e-05,
      "loss": 1.3569,
      "step": 40
    },
    {
      "epoch": 0.008028904054596548,
      "grad_norm": 4.7788567543029785,
      "learning_rate": 2.4e-05,
      "loss": 1.1297,
      "step": 60
    },
    {
      "epoch": 0.010036130068245684,
      "step": 75,
      "total_flos": 649967869886464.0,
      "train_loss": 1.3115534464518228,
      "train_runtime": 102.821,
      "train_samples_per_second": 0.727,
      "train_steps_per_second": 0.729
    }
  ],
  "logging_steps": 20,
  "max_steps": 75,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 649967869886464.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}