trainer_state.json

{
  "best_metric": 0.9112,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0938/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": 11.145367622375488,
      "learning_rate": 0.0004809698831278217,
      "loss": 0.7738,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8317333333333333,
      "eval_loss": 0.56048983335495,
      "eval_runtime": 6.057,
      "eval_samples_per_second": 619.113,
      "eval_steps_per_second": 9.741,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0004269850409335979,
      "loss": 0.4081,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5557873845100403,
      "eval_runtime": 5.812,
      "eval_samples_per_second": 645.221,
      "eval_steps_per_second": 10.151,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.883262634277344,
      "learning_rate": 0.0003459431688641196,
      "loss": 0.2525,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8272,
      "eval_loss": 0.6146223545074463,
      "eval_runtime": 5.9209,
      "eval_samples_per_second": 633.348,
      "eval_steps_per_second": 9.965,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 41.10908508300781,
      "learning_rate": 0.00025029481906207064,
      "loss": 0.1438,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8653333333333333,
      "eval_loss": 0.5049653053283691,
      "eval_runtime": 5.8125,
      "eval_samples_per_second": 645.163,
      "eval_steps_per_second": 10.151,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.36567759513855,
      "learning_rate": 0.00015460158573036287,
      "loss": 0.0609,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8933333333333333,
      "eval_loss": 0.40818119049072266,
      "eval_runtime": 5.7901,
      "eval_samples_per_second": 647.662,
      "eval_steps_per_second": 10.19,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.005491859279572964,
      "learning_rate": 7.343189618242854e-05,
      "loss": 0.0131,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9066666666666666,
      "eval_loss": 0.40785181522369385,
      "eval_runtime": 5.9238,
      "eval_samples_per_second": 633.045,
      "eval_steps_per_second": 9.96,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0023938706144690514,
      "learning_rate": 1.9143099846872597e-05,
      "loss": 0.0023,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9112,
      "eval_loss": 0.3915362060070038,
      "eval_runtime": 5.799,
      "eval_samples_per_second": 646.666,
      "eval_steps_per_second": 10.174,
      "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
}