:param X: 3d-tensor, similar to the input for predict :return: 3d array (num_obs, max_words max_sentences) containing the attention weights for each ...
確定! 回上一頁