model.py 文件源码

def _make_embedding_layer(self, word_index):
        embeddings = self._get_embeddings()
        nb_words = min(self.max_nr_words, len(word_index))
        embedding_matrix = np.zeros((nb_words, self.embedding_dim))

        for word, i in word_index.items():
            if i >= self.max_nr_words:
                continue
            embedding_vector = embeddings.get(word)
            if embedding_vector is not None:
                embedding_matrix[i] = embedding_vector

        embedding_layer = Embedding(nb_words, self.embedding_dim, 
                                    weights=[embedding_matrix], 
                                    input_length=self.sequence_length, trainable=False)
        return embedding_layer