trainer_state.json

{
  "best_metric": 0.8866666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0459/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 9.507764069616767e-05,
      "loss": 1.5583,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7704,
      "eval_loss": 0.8083885908126831,
      "eval_runtime": 4.3175,
      "eval_samples_per_second": 868.564,
      "eval_steps_per_second": 13.665,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.127974762931168e-05,
      "loss": 0.5272,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8272,
      "eval_loss": 0.5642760396003723,
      "eval_runtime": 5.3988,
      "eval_samples_per_second": 694.595,
      "eval_steps_per_second": 10.928,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.135300636291504,
      "learning_rate": 6.125740150220697e-05,
      "loss": 0.2944,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.844,
      "eval_loss": 0.5166472792625427,
      "eval_runtime": 7.8812,
      "eval_samples_per_second": 475.817,
      "eval_steps_per_second": 7.486,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.6662163734436035,
      "learning_rate": 3.900538894462112e-05,
      "loss": 0.1508,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8698666666666667,
      "eval_loss": 0.4516353905200958,
      "eval_runtime": 12.1817,
      "eval_samples_per_second": 307.839,
      "eval_steps_per_second": 4.843,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.125276580452919,
      "learning_rate": 1.893099394660879e-05,
      "loss": 0.066,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.43940502405166626,
      "eval_runtime": 5.8316,
      "eval_samples_per_second": 643.052,
      "eval_steps_per_second": 10.117,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.08664432168006897,
      "learning_rate": 5.0101966338242764e-06,
      "loss": 0.0261,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8856,
      "eval_loss": 0.4216085970401764,
      "eval_runtime": 4.5254,
      "eval_samples_per_second": 828.647,
      "eval_steps_per_second": 13.037,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.20598919689655304,
      "learning_rate": 1.8164143966159507e-10,
      "loss": 0.0154,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.41499799489974976,
      "eval_runtime": 4.4205,
      "eval_samples_per_second": 848.323,
      "eval_steps_per_second": 13.347,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}