trainer_state.json

{
  "best_metric": 0.8829333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0350/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": 198.1342315673828,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 3.3083,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6554666666666666,
      "eval_loss": 1.6205730438232422,
      "eval_runtime": 7.38,
      "eval_samples_per_second": 508.132,
      "eval_steps_per_second": 7.995,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 227.05104064941406,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 0.899,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8245333333333333,
      "eval_loss": 0.6284153461456299,
      "eval_runtime": 7.4564,
      "eval_samples_per_second": 502.921,
      "eval_steps_per_second": 7.913,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 86.43342590332031,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 0.4825,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.46674153208732605,
      "eval_runtime": 7.4056,
      "eval_samples_per_second": 506.374,
      "eval_steps_per_second": 7.967,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 250.90768432617188,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 0.334,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.43646275997161865,
      "eval_runtime": 7.7474,
      "eval_samples_per_second": 484.034,
      "eval_steps_per_second": 7.615,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 124.33551025390625,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.2571,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4266894459724426,
      "eval_runtime": 7.473,
      "eval_samples_per_second": 501.806,
      "eval_steps_per_second": 7.895,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 120.90621948242188,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.2051,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.42480984330177307,
      "eval_runtime": 7.4727,
      "eval_samples_per_second": 501.824,
      "eval_steps_per_second": 7.895,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 114.56005859375,
      "learning_rate": 0.0,
      "loss": 0.1844,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.3937285840511322,
      "eval_runtime": 7.4561,
      "eval_samples_per_second": 502.943,
      "eval_steps_per_second": 7.913,
      "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
}