emm, thank you for your share. here is an error. when i run this code.
in decode part, such as follows:
shapes = self.p_char_emb.get_shape().as_list() _, self.sep_p_char_encodes = rnn( 'bi-gru', inputs=tf.reshape(self.p_char_emb, (shapes[0] * shapes[1], shapes[2], -1)), length=tf.reshape(self.p_char_length, (shapes[0] * shapes[1],)), hidden_size=self.char_hidden_size )
the shapes return [None,None,None,300], so 'shapes[0] * shapes[1]' is not fit. what do u think
emm, thank you for your share. here is an error. when i run this code. in decode part, such as follows:
shapes = self.p_char_emb.get_shape().as_list()
_, self.sep_p_char_encodes = rnn( 'bi-gru', inputs=tf.reshape(self.p_char_emb, (shapes[0] * shapes[1], shapes[2], -1)), length=tf.reshape(self.p_char_length, (shapes[0] * shapes[1],)), hidden_size=self.char_hidden_size )
the shapes return [None,None,None,300], so 'shapes[0] * shapes[1]' is not fit. what do u think