trainer_state.json

{
  "best_metric": 0.96,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0346/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.611292362213135,
      "learning_rate": 3e-05,
      "loss": 1.128,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9469333333333333,
      "eval_loss": 0.24842382967472076,
      "eval_runtime": 4.3187,
      "eval_samples_per_second": 868.315,
      "eval_steps_per_second": 13.661,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.1394,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9578666666666666,
      "eval_loss": 0.16906480491161346,
      "eval_runtime": 4.1067,
      "eval_samples_per_second": 913.132,
      "eval_steps_per_second": 14.367,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.15933063626289368,
      "learning_rate": 3e-05,
      "loss": 0.0512,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9509333333333333,
      "eval_loss": 0.18644753098487854,
      "eval_runtime": 4.2361,
      "eval_samples_per_second": 885.239,
      "eval_steps_per_second": 13.928,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1135998964309692,
      "learning_rate": 3e-05,
      "loss": 0.0205,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9538666666666666,
      "eval_loss": 0.18307766318321228,
      "eval_runtime": 4.6611,
      "eval_samples_per_second": 804.528,
      "eval_steps_per_second": 12.658,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.11185336112976074,
      "learning_rate": 3e-05,
      "loss": 0.0106,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9578666666666666,
      "eval_loss": 0.1693788766860962,
      "eval_runtime": 4.2374,
      "eval_samples_per_second": 884.976,
      "eval_steps_per_second": 13.924,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.20613664388656616,
      "learning_rate": 3e-05,
      "loss": 0.0067,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.96,
      "eval_loss": 0.170301616191864,
      "eval_runtime": 4.0783,
      "eval_samples_per_second": 919.51,
      "eval_steps_per_second": 14.467,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}