preprocess.py 文件源码

def vectorize_stories(stories, vocab : Vocabulary):
    X = []
    Q = []
    C = []
    A = []

    for s, q, a, c in stories:
        x = vocab.convert2idx(s)
        xq = vocab.convert2idx(q)
        xc = vocab.convert2idx(c)
        X.append(x)
        Q.append(xq)
        C.append(xc)
        A.append(vocab.getIdx(a))

    X = X
    Q = Q
    C = C
    A = torch.LongTensor(A)
    return X, Q, A, C