trainer_state.json

{
  "best_metric": 0.8634666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0204/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": 116.08509063720703,
      "learning_rate": 4.752422169756048e-05,
      "loss": 3.638,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5533333333333333,
      "eval_loss": 2.958641767501831,
      "eval_runtime": 8.0982,
      "eval_samples_per_second": 463.067,
      "eval_steps_per_second": 7.286,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 137.5364990234375,
      "learning_rate": 4.058724504646834e-05,
      "loss": 1.6958,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7714666666666666,
      "eval_loss": 0.8904324173927307,
      "eval_runtime": 8.3103,
      "eval_samples_per_second": 451.249,
      "eval_steps_per_second": 7.1,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 485.0519104003906,
      "learning_rate": 3.056302334890786e-05,
      "loss": 0.7323,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8290666666666666,
      "eval_loss": 0.6214796900749207,
      "eval_runtime": 8.0052,
      "eval_samples_per_second": 468.443,
      "eval_steps_per_second": 7.37,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 207.94081115722656,
      "learning_rate": 1.9436976651092144e-05,
      "loss": 0.5224,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.5285818576812744,
      "eval_runtime": 7.9173,
      "eval_samples_per_second": 473.644,
      "eval_steps_per_second": 7.452,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 325.723876953125,
      "learning_rate": 9.412754953531663e-06,
      "loss": 0.4292,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.4923974573612213,
      "eval_runtime": 7.9758,
      "eval_samples_per_second": 470.173,
      "eval_steps_per_second": 7.397,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 90.5422134399414,
      "learning_rate": 2.475778302439524e-06,
      "loss": 0.3823,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.47771885991096497,
      "eval_runtime": 7.8901,
      "eval_samples_per_second": 475.278,
      "eval_steps_per_second": 7.478,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 228.3003387451172,
      "learning_rate": 0.0,
      "loss": 0.3656,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8634666666666667,
      "eval_loss": 0.46754133701324463,
      "eval_runtime": 7.9003,
      "eval_samples_per_second": 474.668,
      "eval_steps_per_second": 7.468,
      "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
}