trainer_state.json

{
  "best_metric": 0.8912,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0372/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": 121.99246215820312,
      "learning_rate": 7.875e-05,
      "loss": 2.9624,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7104,
      "eval_loss": 1.1521905660629272,
      "eval_runtime": 13.6665,
      "eval_samples_per_second": 274.394,
      "eval_steps_per_second": 4.317,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 49.45231246948242,
      "learning_rate": 6.75e-05,
      "loss": 0.7643,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8352,
      "eval_loss": 0.5465917587280273,
      "eval_runtime": 14.1576,
      "eval_samples_per_second": 264.875,
      "eval_steps_per_second": 4.167,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 44.094818115234375,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 0.4675,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.44856172800064087,
      "eval_runtime": 12.6482,
      "eval_samples_per_second": 296.486,
      "eval_steps_per_second": 4.665,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 97.32904052734375,
      "learning_rate": 4.5e-05,
      "loss": 0.3426,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.4073420763015747,
      "eval_runtime": 12.62,
      "eval_samples_per_second": 297.148,
      "eval_steps_per_second": 4.675,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 42.35023880004883,
      "learning_rate": 3.375e-05,
      "loss": 0.2774,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.3763943910598755,
      "eval_runtime": 12.63,
      "eval_samples_per_second": 296.912,
      "eval_steps_per_second": 4.671,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 91.82551574707031,
      "learning_rate": 2.25e-05,
      "loss": 0.223,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8853333333333333,
      "eval_loss": 0.3675535023212433,
      "eval_runtime": 12.7518,
      "eval_samples_per_second": 294.075,
      "eval_steps_per_second": 4.627,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 299.60699462890625,
      "learning_rate": 1.125e-05,
      "loss": 0.1905,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.35570651292800903,
      "eval_runtime": 12.9745,
      "eval_samples_per_second": 289.029,
      "eval_steps_per_second": 4.547,
      "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
}