trainer_state.json

{
  "best_metric": 0.8482666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0681/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": 331.52972412109375,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 3.6233,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5210666666666667,
      "eval_loss": 2.9816372394561768,
      "eval_runtime": 8.9398,
      "eval_samples_per_second": 419.471,
      "eval_steps_per_second": 6.6,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 223.42098999023438,
      "learning_rate": 3.571428571428572e-05,
      "loss": 1.8291,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7450666666666667,
      "eval_loss": 1.0185669660568237,
      "eval_runtime": 8.8643,
      "eval_samples_per_second": 423.045,
      "eval_steps_per_second": 6.656,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 158.19406127929688,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.8363,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8013333333333333,
      "eval_loss": 0.6908562779426575,
      "eval_runtime": 9.0164,
      "eval_samples_per_second": 415.91,
      "eval_steps_per_second": 6.544,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 331.5224609375,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.6147,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8338666666666666,
      "eval_loss": 0.5660020112991333,
      "eval_runtime": 8.8612,
      "eval_samples_per_second": 423.195,
      "eval_steps_per_second": 6.658,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 273.75042724609375,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.513,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.5299258232116699,
      "eval_runtime": 9.0107,
      "eval_samples_per_second": 416.172,
      "eval_steps_per_second": 6.548,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 90.43047332763672,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.4668,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5228175520896912,
      "eval_runtime": 9.0898,
      "eval_samples_per_second": 412.549,
      "eval_steps_per_second": 6.491,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 344.5986328125,
      "learning_rate": 0.0,
      "loss": 0.4249,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8482666666666666,
      "eval_loss": 0.5048677921295166,
      "eval_runtime": 8.8603,
      "eval_samples_per_second": 423.236,
      "eval_steps_per_second": 6.659,
      "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
}