trainer_state.json

{
  "best_metric": 0.8872,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0315/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": 149.78822326660156,
      "learning_rate": 8.728616793536588e-05,
      "loss": 2.9795,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6874666666666667,
      "eval_loss": 1.2588895559310913,
      "eval_runtime": 9.3587,
      "eval_samples_per_second": 400.698,
      "eval_steps_per_second": 6.304,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 101.07476806640625,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.8783,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8128,
      "eval_loss": 0.6587761640548706,
      "eval_runtime": 9.3492,
      "eval_samples_per_second": 401.103,
      "eval_steps_per_second": 6.311,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 79.67926788330078,
      "learning_rate": 6.75e-05,
      "loss": 0.5539,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8586666666666667,
      "eval_loss": 0.5099525451660156,
      "eval_runtime": 9.3663,
      "eval_samples_per_second": 400.371,
      "eval_steps_per_second": 6.299,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 138.72402954101562,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.4131,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8650666666666667,
      "eval_loss": 0.4616686701774597,
      "eval_runtime": 9.3467,
      "eval_samples_per_second": 401.209,
      "eval_steps_per_second": 6.312,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 203.91285705566406,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.3335,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.41663503646850586,
      "eval_runtime": 9.3721,
      "eval_samples_per_second": 400.123,
      "eval_steps_per_second": 6.295,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 63.299415588378906,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.2824,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.40055161714553833,
      "eval_runtime": 9.3566,
      "eval_samples_per_second": 400.786,
      "eval_steps_per_second": 6.306,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 56.5100212097168,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.2611,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.3829886019229889,
      "eval_runtime": 9.4147,
      "eval_samples_per_second": 398.312,
      "eval_steps_per_second": 6.267,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}