Dataset.py 文件源码

def shuffle(self):
        data = list(zip(self.documents, self.querys, self.candidates, self.answers))
        self.documents, self.querys, self.candidates, self.answers = zip(*[data[i] for i in torch.randperm(len(data))])