trainer_state.json

{
  "best_metric": 0.9141333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0947/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": 50.85430908203125,
      "learning_rate": 8.554359905560886e-05,
      "loss": 2.8081,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7810666666666667,
      "eval_loss": 0.9082329869270325,
      "eval_runtime": 5.4125,
      "eval_samples_per_second": 692.836,
      "eval_steps_per_second": 10.901,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 55.9850959777832,
      "learning_rate": 7.305704108364301e-05,
      "loss": 0.6343,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.45142531394958496,
      "eval_runtime": 4.9286,
      "eval_samples_per_second": 760.865,
      "eval_steps_per_second": 11.971,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 65.28044891357422,
      "learning_rate": 5.501344202803415e-05,
      "loss": 0.3903,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9010666666666667,
      "eval_loss": 0.347413033246994,
      "eval_runtime": 4.9383,
      "eval_samples_per_second": 759.373,
      "eval_steps_per_second": 11.947,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 134.03805541992188,
      "learning_rate": 3.498655797196586e-05,
      "loss": 0.2826,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9024,
      "eval_loss": 0.3259020149707794,
      "eval_runtime": 5.7365,
      "eval_samples_per_second": 653.713,
      "eval_steps_per_second": 10.285,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 47.07770919799805,
      "learning_rate": 1.6942958916356995e-05,
      "loss": 0.2185,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9112,
      "eval_loss": 0.29677507281303406,
      "eval_runtime": 4.9632,
      "eval_samples_per_second": 755.567,
      "eval_steps_per_second": 11.888,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 84.27220916748047,
      "learning_rate": 4.456400944391144e-06,
      "loss": 0.1998,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9114666666666666,
      "eval_loss": 0.30070847272872925,
      "eval_runtime": 4.9945,
      "eval_samples_per_second": 750.831,
      "eval_steps_per_second": 11.813,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 58.21573257446289,
      "learning_rate": 0.0,
      "loss": 0.1819,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9141333333333334,
      "eval_loss": 0.2893497943878174,
      "eval_runtime": 5.0347,
      "eval_samples_per_second": 744.836,
      "eval_steps_per_second": 11.719,
      "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
}