trainer_state.json

{
  "best_metric": 0.9168,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0709/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": 40.27912902832031,
      "learning_rate": 0.00047524221697560476,
      "loss": 1.3098,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8341333333333333,
      "eval_loss": 0.5541201829910278,
      "eval_runtime": 8.9691,
      "eval_samples_per_second": 418.104,
      "eval_steps_per_second": 6.578,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 82.39599609375,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.435,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.3993593156337738,
      "eval_runtime": 8.973,
      "eval_samples_per_second": 417.918,
      "eval_steps_per_second": 6.575,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 30.44105339050293,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2661,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8925333333333333,
      "eval_loss": 0.3913598954677582,
      "eval_runtime": 8.7234,
      "eval_samples_per_second": 429.877,
      "eval_steps_per_second": 6.763,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.052017211914062,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.1577,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9098666666666667,
      "eval_loss": 0.32333359122276306,
      "eval_runtime": 8.7842,
      "eval_samples_per_second": 426.903,
      "eval_steps_per_second": 6.717,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 34.94878005981445,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.091,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9149333333333334,
      "eval_loss": 0.3177701234817505,
      "eval_runtime": 8.5624,
      "eval_samples_per_second": 437.962,
      "eval_steps_per_second": 6.891,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 28.133563995361328,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0607,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9133333333333333,
      "eval_loss": 0.31960639357566833,
      "eval_runtime": 8.5937,
      "eval_samples_per_second": 436.365,
      "eval_steps_per_second": 6.865,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 26.874265670776367,
      "learning_rate": 0.0,
      "loss": 0.0444,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9168,
      "eval_loss": 0.3242078721523285,
      "eval_runtime": 8.8164,
      "eval_samples_per_second": 425.345,
      "eval_steps_per_second": 6.692,
      "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
}