2028dffb2a29570c72f10bcb3e35e33fa7fb1c03,finetune/base.py,BaseModel,_inference,#BaseModel#Any#Any#,214

Before Change

        input_func = self.input_pipeline.get_predict_input_fn(Xs)
        length = len(Xs) if not callable(Xs) else None

        pred_gen = list(
            map(
                lambda y: y[mode] if mode else y,
                tqdm.tqdm(
                    estimator.predict(
                        input_fn=input_func, predict_keys=mode
                    ),
                    total=length,
                    desc="Inference"
                )
            )
        )
        return pred_gen

    def fit(self, *args, **kwargs):
         An alias for finetune. 

After Change

        self._data = self.input_pipeline._format_for_inference(Xs)
        self._iter_idx = 0

        n = len(self._data)

        if not getattr(self, "estimator", None):
            self.estimator = self.get_estimator()
            self._input_fn = lambda: self.input_pipeline._dataset_without_targets(
                self._data_generator, train=None
            ).batch(self.config.batch_size)
            self._predictions = self.estimator.predict(input_fn=self._input_fn)

        predictions = [None] * n
        for i in range(n):
            y = next(self._predictions)
            y = y[mode] if mode else y
            predictions[i] = y
        return predictions
        
    def fit(self, *args, **kwargs):
         An alias for finetune. 

Instances