config doesn't have any attribute called WORD_EMBEDDING_DIM or MAX_NB_WORDS

[hadifar]

When use decomposable_attention for training model it throws exception because there is no attribute in config called WORD_EMBEDDING_DIM. I think you should change this to following:

nb_words = min(TrainConfig.MAX_NB_WORDS, len(word_index)) + 1
    embedding_matrix = np.zeros((nb_words, TrainConfig.WORD_EMBEDDING_DIM))
    for word, i in word_index.items():
        if word in word2vec.vocab:
            embedding_matrix[i] = word2vec.word_vec(word)
    print('Null word embeddings: %d' % np.sum(np.sum(embedding_matrix, axis=1) == 0))