trainer_state.json

{
  "best_metric": 0.8114666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0099/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": 100.15577697753906,
      "learning_rate": 2.8858192987669303e-05,
      "loss": 3.8236,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4,
      "eval_loss": 3.6561756134033203,
      "eval_runtime": 7.4972,
      "eval_samples_per_second": 500.188,
      "eval_steps_per_second": 7.87,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 100.78751373291016,
      "learning_rate": 2.5606601717798212e-05,
      "loss": 3.1276,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5992,
      "eval_loss": 2.4974286556243896,
      "eval_runtime": 8.3475,
      "eval_samples_per_second": 449.238,
      "eval_steps_per_second": 7.068,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 99.02912139892578,
      "learning_rate": 2.074025148547635e-05,
      "loss": 1.7502,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7154666666666667,
      "eval_loss": 1.284273386001587,
      "eval_runtime": 7.8417,
      "eval_samples_per_second": 478.215,
      "eval_steps_per_second": 7.524,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 202.98892211914062,
      "learning_rate": 1.5e-05,
      "loss": 1.0422,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.772,
      "eval_loss": 0.8980584740638733,
      "eval_runtime": 8.6398,
      "eval_samples_per_second": 434.04,
      "eval_steps_per_second": 6.829,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 584.0546875,
      "learning_rate": 9.259748514523654e-06,
      "loss": 0.8002,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7968,
      "eval_loss": 0.7743034958839417,
      "eval_runtime": 7.4677,
      "eval_samples_per_second": 502.164,
      "eval_steps_per_second": 7.901,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 151.7611846923828,
      "learning_rate": 4.393398282201788e-06,
      "loss": 0.6925,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8106666666666666,
      "eval_loss": 0.7218307852745056,
      "eval_runtime": 7.719,
      "eval_samples_per_second": 485.817,
      "eval_steps_per_second": 7.644,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 268.1744079589844,
      "learning_rate": 1.1418070123306989e-06,
      "loss": 0.6527,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8114666666666667,
      "eval_loss": 0.7029733061790466,
      "eval_runtime": 7.2158,
      "eval_samples_per_second": 519.695,
      "eval_steps_per_second": 8.177,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}