trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 64,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.625,
      "grad_norm": 1.1638071537017822,
      "learning_rate": 0.00016875,
      "loss": 0.8303,
      "step": 10
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.2804111242294312,
      "learning_rate": 0.0001375,
      "loss": 0.5166,
      "step": 20
    },
    {
      "epoch": 1.875,
      "grad_norm": 3.433356523513794,
      "learning_rate": 0.00010625000000000001,
      "loss": 0.503,
      "step": 30
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.124768018722534,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.3796,
      "step": 40
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.8484621047973633,
      "learning_rate": 4.375e-05,
      "loss": 0.2372,
      "step": 50
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.7986748218536377,
      "learning_rate": 1.25e-05,
      "loss": 0.2453,
      "step": 60
    },
    {
      "epoch": 4.0,
      "step": 64,
      "total_flos": 7.780265488738714e+16,
      "train_loss": 0.4349986780434847,
      "train_runtime": 34.4492,
      "train_samples_per_second": 29.144,
      "train_steps_per_second": 1.858
    }
  ],
  "logging_steps": 10,
  "max_steps": 64,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.780265488738714e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}