trainer_state.json

{
  "best_metric": 0.8872,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0988/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": 291.8343811035156,
      "learning_rate": 8.728616793536588e-05,
      "loss": 2.9457,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6965333333333333,
      "eval_loss": 1.1611312627792358,
      "eval_runtime": 9.641,
      "eval_samples_per_second": 388.962,
      "eval_steps_per_second": 6.12,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 77.53011322021484,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.7643,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8130666666666667,
      "eval_loss": 0.5882033109664917,
      "eval_runtime": 9.301,
      "eval_samples_per_second": 403.182,
      "eval_steps_per_second": 6.343,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 88.40230560302734,
      "learning_rate": 6.75e-05,
      "loss": 0.497,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.47132354974746704,
      "eval_runtime": 9.2463,
      "eval_samples_per_second": 405.57,
      "eval_steps_per_second": 6.381,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 176.52503967285156,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.3785,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8672,
      "eval_loss": 0.40484538674354553,
      "eval_runtime": 9.3394,
      "eval_samples_per_second": 401.524,
      "eval_steps_per_second": 6.317,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 84.0312728881836,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.3117,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.4069541096687317,
      "eval_runtime": 9.3791,
      "eval_samples_per_second": 399.826,
      "eval_steps_per_second": 6.291,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 61.067840576171875,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.2655,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.3988337814807892,
      "eval_runtime": 9.1877,
      "eval_samples_per_second": 408.154,
      "eval_steps_per_second": 6.422,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 52.87115478515625,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.2409,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.35879382491111755,
      "eval_runtime": 9.038,
      "eval_samples_per_second": 414.915,
      "eval_steps_per_second": 6.528,
      "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
}