trainer_state.json

{
  "best_metric": 0.8829333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0498/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": 132.84197998046875,
      "learning_rate": 7e-05,
      "loss": 3.4475,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5186666666666667,
      "eval_loss": 2.199477195739746,
      "eval_runtime": 9.2029,
      "eval_samples_per_second": 407.482,
      "eval_steps_per_second": 6.411,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 141.55218505859375,
      "learning_rate": 7e-05,
      "loss": 1.1834,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7909333333333334,
      "eval_loss": 0.7098614573478699,
      "eval_runtime": 8.3206,
      "eval_samples_per_second": 450.691,
      "eval_steps_per_second": 7.091,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 78.99305725097656,
      "learning_rate": 7e-05,
      "loss": 0.608,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8426666666666667,
      "eval_loss": 0.5311636328697205,
      "eval_runtime": 8.3361,
      "eval_samples_per_second": 449.851,
      "eval_steps_per_second": 7.078,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 68.00151824951172,
      "learning_rate": 7e-05,
      "loss": 0.4702,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.4521065950393677,
      "eval_runtime": 8.3181,
      "eval_samples_per_second": 450.826,
      "eval_steps_per_second": 7.093,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 133.61672973632812,
      "learning_rate": 7e-05,
      "loss": 0.3749,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8781333333333333,
      "eval_loss": 0.4139246344566345,
      "eval_runtime": 8.3319,
      "eval_samples_per_second": 450.078,
      "eval_steps_per_second": 7.081,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 69.53807067871094,
      "learning_rate": 7e-05,
      "loss": 0.3115,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.4030216932296753,
      "eval_runtime": 8.3079,
      "eval_samples_per_second": 451.378,
      "eval_steps_per_second": 7.102,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 174.940185546875,
      "learning_rate": 7e-05,
      "loss": 0.2605,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.3939755856990814,
      "eval_runtime": 8.3564,
      "eval_samples_per_second": 448.758,
      "eval_steps_per_second": 7.06,
      "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
}