trainer_state.json

{
  "best_metric": 0.9258666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0592/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": 0.9174463748931885,
      "learning_rate": 0.00047524221697560476,
      "loss": 0.7602,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8266666666666667,
      "eval_loss": 0.6198466420173645,
      "eval_runtime": 10.7026,
      "eval_samples_per_second": 350.382,
      "eval_steps_per_second": 5.513,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.13336363434791565,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.3933,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.5244952440261841,
      "eval_runtime": 10.5086,
      "eval_samples_per_second": 356.85,
      "eval_steps_per_second": 5.614,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.8552050590515137,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2302,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.4408089816570282,
      "eval_runtime": 10.6253,
      "eval_samples_per_second": 352.931,
      "eval_steps_per_second": 5.553,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0897539854049683,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.105,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.44657278060913086,
      "eval_runtime": 10.6986,
      "eval_samples_per_second": 350.513,
      "eval_steps_per_second": 5.515,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0042423983104527,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0303,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9146666666666666,
      "eval_loss": 0.3529181182384491,
      "eval_runtime": 10.5063,
      "eval_samples_per_second": 356.928,
      "eval_steps_per_second": 5.616,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.023781634867191315,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0046,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.3179607391357422,
      "eval_runtime": 10.2706,
      "eval_samples_per_second": 365.119,
      "eval_steps_per_second": 5.745,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0012,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9258666666666666,
      "eval_loss": 0.3139592111110687,
      "eval_runtime": 10.3206,
      "eval_samples_per_second": 363.352,
      "eval_steps_per_second": 5.717,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}