trainer_state.json

{
  "best_metric": 0.9184,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0673/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": 0.23114794492721558,
      "learning_rate": 0.00042857142857142855,
      "loss": 0.7471,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8357333333333333,
      "eval_loss": 0.5567205548286438,
      "eval_runtime": 10.8645,
      "eval_samples_per_second": 345.16,
      "eval_steps_per_second": 5.431,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00035735735735735734,
      "loss": 0.3679,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.5077999234199524,
      "eval_runtime": 10.7801,
      "eval_samples_per_second": 347.864,
      "eval_steps_per_second": 5.473,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.025174502283334732,
      "learning_rate": 0.00028592878592878594,
      "loss": 0.2063,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.46254974603652954,
      "eval_runtime": 10.7941,
      "eval_samples_per_second": 347.411,
      "eval_steps_per_second": 5.466,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.381617307662964,
      "learning_rate": 0.0002145002145002145,
      "loss": 0.1006,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8818666666666667,
      "eval_loss": 0.4799429178237915,
      "eval_runtime": 10.7748,
      "eval_samples_per_second": 348.035,
      "eval_steps_per_second": 5.476,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.031321343034505844,
      "learning_rate": 0.00014307164307164307,
      "loss": 0.0418,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8986666666666666,
      "eval_loss": 0.4223364591598511,
      "eval_runtime": 10.762,
      "eval_samples_per_second": 348.449,
      "eval_steps_per_second": 5.482,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.004208560101687908,
      "learning_rate": 7.164307164307165e-05,
      "loss": 0.0098,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9141333333333334,
      "eval_loss": 0.36887598037719727,
      "eval_runtime": 14.6571,
      "eval_samples_per_second": 255.849,
      "eval_steps_per_second": 4.025,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.008635880425572395,
      "learning_rate": 2.1450021450021452e-07,
      "loss": 0.0015,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9184,
      "eval_loss": 0.3570334017276764,
      "eval_runtime": 10.8045,
      "eval_samples_per_second": 347.077,
      "eval_steps_per_second": 5.461,
      "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
}