trainer_state.json

{
  "best_metric": 0.8752,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0554/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": 130.5053253173828,
      "learning_rate": 4.849231551964771e-05,
      "loss": 3.6033,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5714666666666667,
      "eval_loss": 2.815537452697754,
      "eval_runtime": 8.1243,
      "eval_samples_per_second": 461.576,
      "eval_steps_per_second": 7.262,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 138.881103515625,
      "learning_rate": 4.415111107797445e-05,
      "loss": 1.5501,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7858666666666667,
      "eval_loss": 0.8150323033332825,
      "eval_runtime": 7.4748,
      "eval_samples_per_second": 501.683,
      "eval_steps_per_second": 7.893,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 123.15934753417969,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.6966,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8392,
      "eval_loss": 0.5543141961097717,
      "eval_runtime": 7.6815,
      "eval_samples_per_second": 488.185,
      "eval_steps_per_second": 7.681,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 295.48095703125,
      "learning_rate": 2.9341204441673266e-05,
      "loss": 0.5089,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.48516231775283813,
      "eval_runtime": 8.0029,
      "eval_samples_per_second": 468.581,
      "eval_steps_per_second": 7.372,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 99.43540954589844,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.4075,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.4392818212509155,
      "eval_runtime": 7.8771,
      "eval_samples_per_second": 476.063,
      "eval_steps_per_second": 7.49,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 213.40008544921875,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.3548,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.42601078748703003,
      "eval_runtime": 7.5081,
      "eval_samples_per_second": 499.458,
      "eval_steps_per_second": 7.858,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 176.50839233398438,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.3255,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4141671359539032,
      "eval_runtime": 7.4433,
      "eval_samples_per_second": 503.81,
      "eval_steps_per_second": 7.927,
      "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
}