trainer_state.json

{
  "best_metric": 0.8752,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0912/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": 371.94537353515625,
      "learning_rate": 6.733578363789503e-05,
      "loss": 3.355,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6029333333333333,
      "eval_loss": 1.889343500137329,
      "eval_runtime": 9.167,
      "eval_samples_per_second": 409.078,
      "eval_steps_per_second": 6.436,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 78.37898254394531,
      "learning_rate": 5.974873734152915e-05,
      "loss": 1.1092,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.808,
      "eval_loss": 0.7165209650993347,
      "eval_runtime": 8.149,
      "eval_samples_per_second": 460.177,
      "eval_steps_per_second": 7.24,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 155.7366180419922,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.6219,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.5720428228378296,
      "eval_runtime": 8.1191,
      "eval_samples_per_second": 461.876,
      "eval_steps_per_second": 7.267,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 46.028385162353516,
      "learning_rate": 3.5e-05,
      "loss": 0.464,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.5032939910888672,
      "eval_runtime": 8.1402,
      "eval_samples_per_second": 460.676,
      "eval_steps_per_second": 7.248,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 128.08802795410156,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.382,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.856,
      "eval_loss": 0.49264124035835266,
      "eval_runtime": 8.4286,
      "eval_samples_per_second": 444.912,
      "eval_steps_per_second": 7.0,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 72.91575622558594,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.3311,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.4504142701625824,
      "eval_runtime": 8.3652,
      "eval_samples_per_second": 448.283,
      "eval_steps_per_second": 7.053,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 83.11953735351562,
      "learning_rate": 2.664216362104964e-06,
      "loss": 0.3092,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4409746825695038,
      "eval_runtime": 8.1131,
      "eval_samples_per_second": 462.218,
      "eval_steps_per_second": 7.272,
      "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
}