trainer_state.json

{
  "best_metric": 0.9232,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0814/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.06472177803516388,
      "learning_rate": 0.0002625,
      "loss": 0.5966,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4314057230949402,
      "eval_runtime": 11.3555,
      "eval_samples_per_second": 330.237,
      "eval_steps_per_second": 5.196,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.09871386736631393,
      "learning_rate": 0.000225,
      "loss": 0.242,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4600445628166199,
      "eval_runtime": 10.7166,
      "eval_samples_per_second": 349.923,
      "eval_steps_per_second": 5.505,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.02054961584508419,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.1337,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8898666666666667,
      "eval_loss": 0.4202483296394348,
      "eval_runtime": 10.4185,
      "eval_samples_per_second": 359.937,
      "eval_steps_per_second": 5.663,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.11374368518590927,
      "learning_rate": 0.00015,
      "loss": 0.0778,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.4009053409099579,
      "eval_runtime": 10.5462,
      "eval_samples_per_second": 355.578,
      "eval_steps_per_second": 5.594,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.010078508406877518,
      "learning_rate": 0.0001125,
      "loss": 0.0391,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9082666666666667,
      "eval_loss": 0.37463992834091187,
      "eval_runtime": 10.6785,
      "eval_samples_per_second": 351.175,
      "eval_steps_per_second": 5.525,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.004321741871535778,
      "learning_rate": 7.5e-05,
      "loss": 0.0135,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.3593113422393799,
      "eval_runtime": 10.3353,
      "eval_samples_per_second": 362.836,
      "eval_steps_per_second": 5.709,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.7612612612612614e-05,
      "loss": 0.0027,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9221333333333334,
      "eval_loss": 0.344949871301651,
      "eval_runtime": 10.5447,
      "eval_samples_per_second": 355.629,
      "eval_steps_per_second": 5.595,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.0035311137326061726,
      "learning_rate": 1.1261261261261261e-07,
      "loss": 0.0009,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9232,
      "eval_loss": 0.3251197934150696,
      "eval_runtime": 10.3831,
      "eval_samples_per_second": 361.165,
      "eval_steps_per_second": 5.682,
      "step": 2664
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}