trainer_state.json

{
  "best_metric": 0.9408,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0598/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": NaN,
      "learning_rate": 5e-05,
      "loss": 0.8684,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9314666666666667,
      "eval_loss": 0.24375179409980774,
      "eval_runtime": 13.3961,
      "eval_samples_per_second": 279.933,
      "eval_steps_per_second": 4.404,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3410260677337646,
      "learning_rate": 5e-05,
      "loss": 0.1465,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9290666666666667,
      "eval_loss": 0.24055688083171844,
      "eval_runtime": 13.2645,
      "eval_samples_per_second": 282.709,
      "eval_steps_per_second": 4.448,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.09504510462284088,
      "learning_rate": 5e-05,
      "loss": 0.0707,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9248,
      "eval_loss": 0.27425527572631836,
      "eval_runtime": 13.3118,
      "eval_samples_per_second": 281.705,
      "eval_steps_per_second": 4.432,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 71.33308410644531,
      "learning_rate": 5e-05,
      "loss": 0.047,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9352,
      "eval_loss": 0.24885211884975433,
      "eval_runtime": 13.2736,
      "eval_samples_per_second": 282.517,
      "eval_steps_per_second": 4.445,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.03257712349295616,
      "learning_rate": 5e-05,
      "loss": 0.0276,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9309333333333333,
      "eval_loss": 0.2906922996044159,
      "eval_runtime": 13.2889,
      "eval_samples_per_second": 282.191,
      "eval_steps_per_second": 4.44,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.00939291063696146,
      "learning_rate": 5e-05,
      "loss": 0.0237,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9306666666666666,
      "eval_loss": 0.2664060592651367,
      "eval_runtime": 13.5372,
      "eval_samples_per_second": 277.015,
      "eval_steps_per_second": 4.358,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.005621448624879122,
      "learning_rate": 5e-05,
      "loss": 0.0147,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9408,
      "eval_loss": 0.2817372977733612,
      "eval_runtime": 13.3031,
      "eval_samples_per_second": 281.89,
      "eval_steps_per_second": 4.435,
      "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
}