trainer_state.json

{
  "best_metric": 0.9597333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0848/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": 13.275174140930176,
      "learning_rate": 6.733578363789503e-05,
      "loss": 0.6826,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9370666666666667,
      "eval_loss": 0.20735614001750946,
      "eval_runtime": 6.5462,
      "eval_samples_per_second": 572.847,
      "eval_steps_per_second": 9.013,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.036105893552303314,
      "learning_rate": 5.974873734152915e-05,
      "loss": 0.1187,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9432,
      "eval_loss": 0.1828790158033371,
      "eval_runtime": 6.945,
      "eval_samples_per_second": 539.96,
      "eval_steps_per_second": 8.495,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05290098860859871,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.0487,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9528,
      "eval_loss": 0.16495433449745178,
      "eval_runtime": 6.9536,
      "eval_samples_per_second": 539.289,
      "eval_steps_per_second": 8.485,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.046987466514110565,
      "learning_rate": 3.5e-05,
      "loss": 0.0173,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9544,
      "eval_loss": 0.1794421523809433,
      "eval_runtime": 6.8865,
      "eval_samples_per_second": 544.54,
      "eval_steps_per_second": 8.567,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.011084523051977158,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.0062,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9576,
      "eval_loss": 0.17380423843860626,
      "eval_runtime": 6.9194,
      "eval_samples_per_second": 541.956,
      "eval_steps_per_second": 8.527,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.03804173320531845,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.0029,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9581333333333333,
      "eval_loss": 0.16106151044368744,
      "eval_runtime": 6.8548,
      "eval_samples_per_second": 547.059,
      "eval_steps_per_second": 8.607,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.6800339785621635e-06,
      "loss": 0.0018,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9597333333333333,
      "eval_loss": 0.1684000939130783,
      "eval_runtime": 6.893,
      "eval_samples_per_second": 544.032,
      "eval_steps_per_second": 8.559,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}