trainer_state.json

{
  "best_metric": 0.9314666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0833/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": 9.147697448730469,
      "learning_rate": 0.0004849231551964771,
      "loss": 0.6929,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8530666666666666,
      "eval_loss": 0.5027872323989868,
      "eval_runtime": 10.486,
      "eval_samples_per_second": 357.619,
      "eval_steps_per_second": 5.627,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0004416794553834106,
      "loss": 0.3766,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.43058347702026367,
      "eval_runtime": 10.4651,
      "eval_samples_per_second": 358.334,
      "eval_steps_per_second": 5.638,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0343607664108276,
      "learning_rate": 0.00037522688315464164,
      "loss": 0.2486,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.4993470013141632,
      "eval_runtime": 10.8188,
      "eval_samples_per_second": 346.62,
      "eval_steps_per_second": 5.453,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.48378264904022217,
      "learning_rate": 0.0002936701006654613,
      "loss": 0.157,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.43769678473472595,
      "eval_runtime": 10.4428,
      "eval_samples_per_second": 359.099,
      "eval_steps_per_second": 5.65,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 87.90697479248047,
      "learning_rate": 0.00020684605953398183,
      "loss": 0.0816,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.41821959614753723,
      "eval_runtime": 10.7076,
      "eval_samples_per_second": 350.218,
      "eval_steps_per_second": 5.51,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.007104875985532999,
      "learning_rate": 0.00012522702050703539,
      "loss": 0.0258,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9181333333333334,
      "eval_loss": 0.34937727451324463,
      "eval_runtime": 10.3625,
      "eval_samples_per_second": 361.88,
      "eval_steps_per_second": 5.694,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.22116513550281525,
      "learning_rate": 5.8657444259997555e-05,
      "loss": 0.0064,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9314666666666667,
      "eval_loss": 0.286752313375473,
      "eval_runtime": 10.2766,
      "eval_samples_per_second": 364.906,
      "eval_steps_per_second": 5.741,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}