trainer_state.json

{
  "best_metric": 0.8973333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0613/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": 133.36058044433594,
      "learning_rate": 6.733578363789503e-05,
      "loss": 3.2533,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6677333333333333,
      "eval_loss": 1.5108211040496826,
      "eval_runtime": 7.406,
      "eval_samples_per_second": 506.345,
      "eval_steps_per_second": 7.966,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 151.68869018554688,
      "learning_rate": 5.974873734152915e-05,
      "loss": 0.8574,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8522666666666666,
      "eval_loss": 0.5300312638282776,
      "eval_runtime": 7.3901,
      "eval_samples_per_second": 507.433,
      "eval_steps_per_second": 7.984,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 156.96775817871094,
      "learning_rate": 4.839392013277814e-05,
      "loss": 0.4758,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.4159897267818451,
      "eval_runtime": 7.4019,
      "eval_samples_per_second": 506.627,
      "eval_steps_per_second": 7.971,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 46.31423568725586,
      "learning_rate": 3.5e-05,
      "loss": 0.3587,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.3890632092952728,
      "eval_runtime": 7.2769,
      "eval_samples_per_second": 515.33,
      "eval_steps_per_second": 8.108,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 146.5362548828125,
      "learning_rate": 2.1606079867221858e-05,
      "loss": 0.2866,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.3639943301677704,
      "eval_runtime": 7.3633,
      "eval_samples_per_second": 509.285,
      "eval_steps_per_second": 8.013,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 119.10710144042969,
      "learning_rate": 1.0251262658470837e-05,
      "loss": 0.2399,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8952,
      "eval_loss": 0.35094043612480164,
      "eval_runtime": 7.3478,
      "eval_samples_per_second": 510.357,
      "eval_steps_per_second": 8.03,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 142.36903381347656,
      "learning_rate": 2.664216362104964e-06,
      "loss": 0.2154,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8973333333333333,
      "eval_loss": 0.34734511375427246,
      "eval_runtime": 7.3036,
      "eval_samples_per_second": 513.445,
      "eval_steps_per_second": 8.078,
      "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
}