trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.88,
  "eval_steps": 500,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "grad_norm": 1.6399102210998535,
      "learning_rate": 4.1069690242163484e-05,
      "loss": 2.9087,
      "step": 5
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.1874444484710693,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 2.7345,
      "step": 10
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.500931978225708,
      "learning_rate": 3.3493649053890326e-06,
      "loss": 2.4515,
      "step": 15
    },
    {
      "epoch": 2.88,
      "step": 18,
      "total_flos": 711163669118976.0,
      "train_loss": 2.6407934294806585,
      "train_runtime": 59.852,
      "train_samples_per_second": 5.012,
      "train_steps_per_second": 0.301
    }
  ],
  "logging_steps": 5,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 711163669118976.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}