trainer_state.json

{
  "best_metric": 0.892,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0582/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": 106.08540344238281,
      "learning_rate": 8.728616793536588e-05,
      "loss": 2.7997,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7413333333333333,
      "eval_loss": 1.0126051902770996,
      "eval_runtime": 8.0478,
      "eval_samples_per_second": 465.966,
      "eval_steps_per_second": 7.331,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 77.95970916748047,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.6885,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.5353417992591858,
      "eval_runtime": 7.925,
      "eval_samples_per_second": 473.186,
      "eval_steps_per_second": 7.445,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 74.24547576904297,
      "learning_rate": 6.75e-05,
      "loss": 0.3973,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.4386550188064575,
      "eval_runtime": 8.0405,
      "eval_samples_per_second": 466.391,
      "eval_steps_per_second": 7.338,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 57.576656341552734,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.2529,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.397809237241745,
      "eval_runtime": 8.0156,
      "eval_samples_per_second": 467.838,
      "eval_steps_per_second": 7.361,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 48.43045425415039,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.176,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.3910093307495117,
      "eval_runtime": 7.9157,
      "eval_samples_per_second": 473.742,
      "eval_steps_per_second": 7.454,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 59.2003288269043,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.1307,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8904,
      "eval_loss": 0.3855428993701935,
      "eval_runtime": 7.8203,
      "eval_samples_per_second": 479.522,
      "eval_steps_per_second": 7.544,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 292.388427734375,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.1007,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.892,
      "eval_loss": 0.39858677983283997,
      "eval_runtime": 8.0359,
      "eval_samples_per_second": 466.656,
      "eval_steps_per_second": 7.342,
      "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
}