trainer_state.json

{
  "best_metric": 0.8824,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0169/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": 48.75693893432617,
      "learning_rate": 8.728616793536588e-05,
      "loss": 2.9888,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6968,
      "eval_loss": 1.1764038801193237,
      "eval_runtime": 13.5514,
      "eval_samples_per_second": 276.724,
      "eval_steps_per_second": 4.354,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 54.80712127685547,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.8355,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8154666666666667,
      "eval_loss": 0.6381157040596008,
      "eval_runtime": 13.0008,
      "eval_samples_per_second": 288.445,
      "eval_steps_per_second": 4.538,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 123.06652069091797,
      "learning_rate": 6.75e-05,
      "loss": 0.5158,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8522666666666666,
      "eval_loss": 0.5004444122314453,
      "eval_runtime": 13.2114,
      "eval_samples_per_second": 283.845,
      "eval_steps_per_second": 4.466,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 96.99659729003906,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.3768,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8616,
      "eval_loss": 0.4430471360683441,
      "eval_runtime": 13.332,
      "eval_samples_per_second": 281.279,
      "eval_steps_per_second": 4.425,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 158.4842529296875,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.298,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.44203805923461914,
      "eval_runtime": 12.9703,
      "eval_samples_per_second": 289.122,
      "eval_steps_per_second": 4.549,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 248.22616577148438,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.2469,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.43266671895980835,
      "eval_runtime": 12.9938,
      "eval_samples_per_second": 288.6,
      "eval_steps_per_second": 4.541,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 156.6045684814453,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.2203,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.4075658917427063,
      "eval_runtime": 13.0178,
      "eval_samples_per_second": 288.067,
      "eval_steps_per_second": 4.532,
      "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
}