demo.py 文件源码

def get_question_features(question):
    ''' For a given question, a unicode string, returns the timeseris vector
    with each word (token) transformed into a 300 dimension representation
    calculated using Glove Vector '''
    word_embeddings = spacy.load('en', vectors='en_glove_cc_300_1m_vectors')
#    word_embeddings = spacy.load('en')#, vectors='en_glove_cc_300_1m_vectors')

#    nlp = English()
#    n_dimensions = nlp.vocab.load_vectors('glove.840B.300d.txt.bz2')
#    print n_dimensions
#    tokens = n_dimensions

#    embeddings_index = {}
#    f = open('glove.6B.300d.txt')
#    for line in f:
#        values = line.split()
#        word = values[0]
#        coefs = np.asarray(values[1:], dtype='float32')
#        embeddings_index[word] = coefs
#    f.close()
#
#    print('Found %s word vectors.' % len(embeddings_index))
#    
#    word_embeddings = spacy.load('en', vectors='glove.6B.30d.txt')

    tokens = word_embeddings(question)
    question_tensor = np.zeros((1, 30, 300))
    for j in xrange(len(tokens)):
            question_tensor[0,j,:] = tokens[j].vector
    return question_tensor