trainer_state.json

{
  "best_metric": 0.9141333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0023/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": 18.314071655273438,
      "learning_rate": 7e-05,
      "loss": 0.7871,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.30739882588386536,
      "eval_runtime": 12.0936,
      "eval_samples_per_second": 310.083,
      "eval_steps_per_second": 4.879,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.650655210018158,
      "learning_rate": 7e-05,
      "loss": 0.184,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9016,
      "eval_loss": 0.318776398897171,
      "eval_runtime": 11.9085,
      "eval_samples_per_second": 314.9,
      "eval_steps_per_second": 4.954,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.8196659088134766,
      "learning_rate": 7e-05,
      "loss": 0.1049,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.3235853612422943,
      "eval_runtime": 12.5976,
      "eval_samples_per_second": 297.677,
      "eval_steps_per_second": 4.683,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.925463318824768,
      "learning_rate": 7e-05,
      "loss": 0.0654,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9101333333333333,
      "eval_loss": 0.3254566788673401,
      "eval_runtime": 12.3829,
      "eval_samples_per_second": 302.838,
      "eval_steps_per_second": 4.765,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.025853216648101807,
      "learning_rate": 7e-05,
      "loss": 0.0512,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9130666666666667,
      "eval_loss": 0.34731459617614746,
      "eval_runtime": 11.8576,
      "eval_samples_per_second": 316.253,
      "eval_steps_per_second": 4.976,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.01416233740746975,
      "learning_rate": 7e-05,
      "loss": 0.0374,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9125333333333333,
      "eval_loss": 0.3698119819164276,
      "eval_runtime": 11.918,
      "eval_samples_per_second": 314.65,
      "eval_steps_per_second": 4.95,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 7e-05,
      "loss": 0.0361,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9141333333333334,
      "eval_loss": 0.39841771125793457,
      "eval_runtime": 13.1141,
      "eval_samples_per_second": 285.951,
      "eval_steps_per_second": 4.499,
      "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
}