trainer_state.json

{
  "best_metric": 0.9146666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0522/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": 22.42638397216797,
      "learning_rate": 0.00047524221697560476,
      "loss": 1.3009,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.46364444494247437,
      "eval_runtime": 7.7041,
      "eval_samples_per_second": 486.754,
      "eval_steps_per_second": 7.658,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 25.34224510192871,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.371,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8826666666666667,
      "eval_loss": 0.39825254678726196,
      "eval_runtime": 7.6924,
      "eval_samples_per_second": 487.497,
      "eval_steps_per_second": 7.67,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.398799896240234,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2008,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.3613670766353607,
      "eval_runtime": 7.7207,
      "eval_samples_per_second": 485.704,
      "eval_steps_per_second": 7.642,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 37.0283203125,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.0904,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9016,
      "eval_loss": 0.37688493728637695,
      "eval_runtime": 7.7741,
      "eval_samples_per_second": 482.371,
      "eval_steps_per_second": 7.589,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 35.5752067565918,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0349,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.3778795897960663,
      "eval_runtime": 7.8078,
      "eval_samples_per_second": 480.291,
      "eval_steps_per_second": 7.557,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 39.45241928100586,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0213,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9138666666666667,
      "eval_loss": 0.3492687940597534,
      "eval_runtime": 7.6896,
      "eval_samples_per_second": 487.672,
      "eval_steps_per_second": 7.673,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 26.497146606445312,
      "learning_rate": 0.0,
      "loss": 0.0142,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9146666666666666,
      "eval_loss": 0.3619638979434967,
      "eval_runtime": 7.6701,
      "eval_samples_per_second": 488.913,
      "eval_steps_per_second": 7.692,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}