trainer_state.json

{
  "best_metric": 0.9498666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0048/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": 0.9382069706916809,
      "learning_rate": 8e-05,
      "loss": 0.6469,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.2545296847820282,
      "eval_runtime": 7.9414,
      "eval_samples_per_second": 472.209,
      "eval_steps_per_second": 7.429,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0730382204055786,
      "learning_rate": 7.000000000000001e-05,
      "loss": 0.1271,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9344,
      "eval_loss": 0.23175273835659027,
      "eval_runtime": 6.6162,
      "eval_samples_per_second": 566.795,
      "eval_steps_per_second": 8.918,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.30895015597343445,
      "learning_rate": 6e-05,
      "loss": 0.0534,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9336,
      "eval_loss": 0.2555180490016937,
      "eval_runtime": 9.0903,
      "eval_samples_per_second": 412.526,
      "eval_steps_per_second": 6.49,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.018605751916766167,
      "learning_rate": 5e-05,
      "loss": 0.0229,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9346666666666666,
      "eval_loss": 0.25183552503585815,
      "eval_runtime": 6.323,
      "eval_samples_per_second": 593.074,
      "eval_steps_per_second": 9.331,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.04055001959204674,
      "learning_rate": 4e-05,
      "loss": 0.0108,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9394666666666667,
      "eval_loss": 0.2511155307292938,
      "eval_runtime": 6.2647,
      "eval_samples_per_second": 598.589,
      "eval_steps_per_second": 9.418,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.011056169867515564,
      "learning_rate": 3e-05,
      "loss": 0.0043,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.24384605884552002,
      "eval_runtime": 6.4988,
      "eval_samples_per_second": 577.033,
      "eval_steps_per_second": 9.079,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.0030030030030034e-05,
      "loss": 0.0016,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9453333333333334,
      "eval_loss": 0.2452995777130127,
      "eval_runtime": 6.4915,
      "eval_samples_per_second": 577.681,
      "eval_steps_per_second": 9.089,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.008450817316770554,
      "learning_rate": 1.0030030030030031e-05,
      "loss": 0.001,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9498666666666666,
      "eval_loss": 0.23273681104183197,
      "eval_runtime": 6.1806,
      "eval_samples_per_second": 606.735,
      "eval_steps_per_second": 9.546,
      "step": 2664
    }
  ],
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}