trainer_state.json

{
  "best_metric": 0.8562666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0095/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": 234.65670776367188,
      "learning_rate": 4.752422169756048e-05,
      "loss": 3.6027,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5885333333333334,
      "eval_loss": 2.9084270000457764,
      "eval_runtime": 14.1215,
      "eval_samples_per_second": 265.553,
      "eval_steps_per_second": 4.178,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 56.211448669433594,
      "learning_rate": 4.058724504646834e-05,
      "loss": 1.6145,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7776,
      "eval_loss": 0.8594679832458496,
      "eval_runtime": 13.9588,
      "eval_samples_per_second": 268.647,
      "eval_steps_per_second": 4.227,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 102.34169006347656,
      "learning_rate": 3.056302334890786e-05,
      "loss": 0.71,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8288,
      "eval_loss": 0.5973041653633118,
      "eval_runtime": 14.6072,
      "eval_samples_per_second": 256.723,
      "eval_steps_per_second": 4.039,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 158.91070556640625,
      "learning_rate": 1.9436976651092144e-05,
      "loss": 0.5227,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8421333333333333,
      "eval_loss": 0.5301016569137573,
      "eval_runtime": 14.0856,
      "eval_samples_per_second": 266.228,
      "eval_steps_per_second": 4.189,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 127.1190185546875,
      "learning_rate": 9.412754953531663e-06,
      "loss": 0.4359,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.4951532483100891,
      "eval_runtime": 14.1026,
      "eval_samples_per_second": 265.909,
      "eval_steps_per_second": 4.184,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 121.82305908203125,
      "learning_rate": 2.475778302439524e-06,
      "loss": 0.3916,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.4894227981567383,
      "eval_runtime": 13.9837,
      "eval_samples_per_second": 268.17,
      "eval_steps_per_second": 4.219,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 109.89532470703125,
      "learning_rate": 0.0,
      "loss": 0.3755,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8562666666666666,
      "eval_loss": 0.48643195629119873,
      "eval_runtime": 14.1017,
      "eval_samples_per_second": 265.926,
      "eval_steps_per_second": 4.184,
      "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
}