trainer_state.json

{
  "best_metric": 0.9432,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0244/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": 1.8547084331512451,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.6307,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9197333333333333,
      "eval_loss": 0.2847656011581421,
      "eval_runtime": 4.6367,
      "eval_samples_per_second": 808.77,
      "eval_steps_per_second": 12.725,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.465467691421509,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.1369,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9157333333333333,
      "eval_loss": 0.2987215518951416,
      "eval_runtime": 4.3469,
      "eval_samples_per_second": 862.679,
      "eval_steps_per_second": 13.573,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8155102133750916,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0637,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9224,
      "eval_loss": 0.2845192849636078,
      "eval_runtime": 4.3379,
      "eval_samples_per_second": 864.474,
      "eval_steps_per_second": 13.601,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.046499673277139664,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0227,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9344,
      "eval_loss": 0.26931315660476685,
      "eval_runtime": 4.1269,
      "eval_samples_per_second": 908.682,
      "eval_steps_per_second": 14.297,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.017725791782140732,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0072,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9386666666666666,
      "eval_loss": 0.260152667760849,
      "eval_runtime": 4.2674,
      "eval_samples_per_second": 878.751,
      "eval_steps_per_second": 13.826,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.009165707975625992,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0017,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9426666666666667,
      "eval_loss": 0.24518676102161407,
      "eval_runtime": 4.3539,
      "eval_samples_per_second": 861.292,
      "eval_steps_per_second": 13.551,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1731488851999511e-05,
      "loss": 0.001,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9432,
      "eval_loss": 0.24306204915046692,
      "eval_runtime": 4.2791,
      "eval_samples_per_second": 876.359,
      "eval_steps_per_second": 13.788,
      "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
}