error of function finetune of class LGBModel

🐛 Bug Description

Using function of finetune of class LGBModel raised AttributeError: 'tuple' object has no attribute 'empty'.

To Reproduce

Steps to reproduce the behavior:

1. Training LGBModel and saving params to 'LGB_params.pkl'.
2. Init model from 'LGB_params.pkl'.
3. Using function finetune.

Expected Behavior

Screenshot

Environment

Note: User could run cd scripts && python collect_info.py all under project directory to get system information
and paste them here directly.

• Qlib version: 0.8.4
• Python version: 3.7
• OS (Windows, Linux, MacOS): Windows
• Commit number (optional, please provide it if you are using the dev version):

Additional Notes

I read the raw code, the error located at line 111 of file : anaconda3\Lib\site-packages\qlib\contrib\model\gbdt.py.
Here is the raw code:

    def finetune(self, dataset: DatasetH, num_boost_round=10, verbose_eval=20, reweighter=None):
        """
        finetune model

        Parameters
        ----------
        dataset : DatasetH
            dataset for finetuning
        num_boost_round : int
            number of round to finetune model
        verbose_eval : int
            verbose level
        """
        # Based on existing model and finetune by train more rounds
        dtrain, _ = self._prepare_data(dataset, reweighter)  # pylint: disable=W0632
        if dtrain.empty:
            raise ValueError("Empty data from dataset, please check your dataset config.")
        self.model = lgb.train(
            self.params,
            dtrain,
            num_boost_round=num_boost_round,
            init_model=self.model,
            valid_sets=[dtrain],
            valid_names=["train"],
            verbose_eval=verbose_eval,
        )

dtrain is a tuple and function _prepare_data has already checked whether the dataset is empty, so I think the code could be modified to:

    def finetune(self, dataset: DatasetH, num_boost_round=10, verbose_eval=20, reweighter=None):
        """
        finetune model

        Parameters
        ----------
        dataset : DatasetH
            dataset for finetuning
        num_boost_round : int
            number of round to finetune model
        verbose_eval : int
            verbose level
        """
        # Based on existing model and finetune by train more rounds
        dtrain, _ = self._prepare_data(dataset, reweighter)  # pylint: disable=W0632
        self.model = lgb.train(
            self.params,
            dtrain[0],
            num_boost_round=num_boost_round,
            init_model=self.model,
            valid_sets=[dtrain[0]],
            valid_names=["train"],
            verbose_eval=verbose_eval,
        )