trainer_state.json

{
  "best_metric": 0.8538666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0981/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": 198.19772338867188,
      "learning_rate": 5e-05,
      "loss": 3.7433,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4056,
      "eval_loss": 3.308870315551758,
      "eval_runtime": 7.5792,
      "eval_samples_per_second": 494.773,
      "eval_steps_per_second": 7.784,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 61.61637496948242,
      "learning_rate": 5e-05,
      "loss": 2.0036,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7125333333333334,
      "eval_loss": 1.0829179286956787,
      "eval_runtime": 7.4441,
      "eval_samples_per_second": 503.752,
      "eval_steps_per_second": 7.926,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 75.48828887939453,
      "learning_rate": 5e-05,
      "loss": 0.9011,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7922666666666667,
      "eval_loss": 0.7145320177078247,
      "eval_runtime": 7.4699,
      "eval_samples_per_second": 502.015,
      "eval_steps_per_second": 7.898,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 257.66455078125,
      "learning_rate": 5e-05,
      "loss": 0.6517,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8082666666666667,
      "eval_loss": 0.6349654197692871,
      "eval_runtime": 7.9288,
      "eval_samples_per_second": 472.956,
      "eval_steps_per_second": 7.441,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 52.081966400146484,
      "learning_rate": 5e-05,
      "loss": 0.5281,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8269333333333333,
      "eval_loss": 0.5380356311798096,
      "eval_runtime": 7.4894,
      "eval_samples_per_second": 500.711,
      "eval_steps_per_second": 7.878,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 81.83316040039062,
      "learning_rate": 5e-05,
      "loss": 0.445,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8469333333333333,
      "eval_loss": 0.4900718927383423,
      "eval_runtime": 7.4151,
      "eval_samples_per_second": 505.722,
      "eval_steps_per_second": 7.957,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 156.2613983154297,
      "learning_rate": 5e-05,
      "loss": 0.3804,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.4655342996120453,
      "eval_runtime": 7.3555,
      "eval_samples_per_second": 509.821,
      "eval_steps_per_second": 8.021,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}