trainer_state.json

{
  "best_metric": 0.9466666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0228/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": 2.784193277359009,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.6226,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9176,
      "eval_loss": 0.2747783064842224,
      "eval_runtime": 5.721,
      "eval_samples_per_second": 655.477,
      "eval_steps_per_second": 10.313,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.03293422982096672,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.1354,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.23778438568115234,
      "eval_runtime": 6.88,
      "eval_samples_per_second": 545.055,
      "eval_steps_per_second": 8.576,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0514802485704422,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0552,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9349333333333333,
      "eval_loss": 0.24058791995048523,
      "eval_runtime": 6.1283,
      "eval_samples_per_second": 611.914,
      "eval_steps_per_second": 9.627,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.18970708549022675,
      "learning_rate": 5e-05,
      "loss": 0.0194,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9373333333333334,
      "eval_loss": 0.2399214804172516,
      "eval_runtime": 5.3385,
      "eval_samples_per_second": 702.44,
      "eval_steps_per_second": 11.052,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.02834269218146801,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0084,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9432,
      "eval_loss": 0.23917651176452637,
      "eval_runtime": 6.5652,
      "eval_samples_per_second": 571.191,
      "eval_steps_per_second": 8.987,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.08431679010391235,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.0027,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9429333333333333,
      "eval_loss": 0.24367468059062958,
      "eval_runtime": 7.5105,
      "eval_samples_per_second": 499.301,
      "eval_steps_per_second": 7.856,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.82861996937452e-06,
      "loss": 0.0016,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9466666666666667,
      "eval_loss": 0.23360523581504822,
      "eval_runtime": 6.6992,
      "eval_samples_per_second": 559.767,
      "eval_steps_per_second": 8.807,
      "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
}