我已经在Tensorflow（Python）中尝试使用Seq2Seq模型几个星期了，我有一个工作模型，它使用双向编码器和基于注意力的解码器，之前运行得很好。今天我添加了Beam Search，但我注意到当Beam宽度为1或更大时，推理过程现在需要很长时间，而之前仅使用双向编码器和注意力解码器时，推理只需要几秒钟。
环境详情：TensorFlow版本：1.3.0，MacOS 10.12.4
以下是我代码的相关部分：
def decoding_layer(dec_input, encoder_state, target_sequence_length, max_target_sequence_length, rnn_size, num_layers, target_vocab_to_int, target_vocab_size, batch_size, keep_prob, decoding_embedding_size , encoder_outputs): """ 创建解码层 :param dec_input: 解码器输入 :param encoder_state: 编码器状态 :param target_sequence_length: 目标批次中每个序列的长度 :param max_target_sequence_length: 目标序列的最大长度 :param rnn_size: RNN大小 :param num_layers: 层数 :param target_vocab_to_int: 从目标词到ID的字典 :param target_vocab_size: 目标词汇大小 :param batch_size: 批次大小 :param keep_prob: Dropout保留概率 :param decoding_embedding_size: 解码嵌入大小 :encoder_outputs : 编码器的输出 :return: 包含（训练BasicDecoderOutput，推理BasicDecoderOutput）的元组 """ encoder_outputs_tr =encoder_outputs #tf.transpose(encoder_outputs,[1,0,2]) # 1. 解码器嵌入 dec_embeddings = tf.Variable(tf.random_uniform([target_vocab_size, decoding_embedding_size])) dec_embed_input = tf.nn.embedding_lookup(dec_embeddings, dec_input) # 2. 构建解码器单元 def create_cell(rnn_size): lstm_cell = tf.contrib.rnn.LSTMCell(rnn_size, initializer=tf.random_uniform_initializer(-0.1,0.1,seed=2)) drop = tf.contrib.rnn.DropoutWrapper(lstm_cell, output_keep_prob=keep_prob) return drop def create_complete_cell(rnn_size,num_layers,encoder_outputs_tr,batch_size,encoder_state , infer ): if infer and beam_width >0: encoder_outputs_tr = tf.contrib.seq2seq.tile_batch(encoder_outputs_tr, multiplier=beam_width) encoder_state = tf.contrib.seq2seq.tile_batch(encoder_state, multiplier=beam_width) batch_size = batch_size * beam_width dec_cell = tf.contrib.rnn.MultiRNNCell([create_cell(rnn_size) for _ in range(num_layers)]) attention_mechanism = tf.contrib.seq2seq.BahdanauAttention(num_units=rnn_size, memory=encoder_outputs_tr) attn_cell = tf.contrib.seq2seq.AttentionWrapper(dec_cell, attention_mechanism , attention_layer_size=rnn_size , output_attention=False) attn_zero = attn_cell.zero_state(batch_size , tf.float32 ) attn_zero = attn_zero.clone(cell_state = encoder_state) return attn_zero , attn_cell intial_train_state , train_cell = create_complete_cell(rnn_size,num_layers,encoder_outputs_tr,batch_size,encoder_state , False ) intial_infer_state , infer_cell = create_complete_cell(rnn_size,num_layers,encoder_outputs_tr,batch_size,encoder_state , True ) output_layer = Dense(target_vocab_size, kernel_initializer = tf.truncated_normal_initializer(mean = 0.0, stddev=0.1)) with tf.variable_scope("decode"): train_decoder_out = decoding_layer_train(intial_train_state, train_cell, dec_embed_input, target_sequence_length, max_target_sequence_length, output_layer, keep_prob) with tf.variable_scope("decode", reuse=True): if beam_width == 0 : infer_decoder_out = decoding_layer_infer(intial_infer_state, infer_cell, dec_embeddings, target_vocab_to_int['<GO>'], target_vocab_to_int['<EOS>'], max_target_sequence_length, target_vocab_size, output_layer, batch_size, keep_prob) else : infer_decoder_out = decoding_layer_infer_with_Beam(intial_infer_state, infer_cell, dec_embeddings, target_vocab_to_int['<GO>'], target_vocab_to_int['<EOS>'], max_target_sequence_length, target_vocab_size, output_layer, batch_size, keep_prob) print('beam search') return (train_decoder_out, infer_decoder_out)"""DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE"""#tests.test_decoding_layer(decoding_layer)def decoding_layer_infer_with_Beam(encoder_state, dec_cell, dec_embeddings, start_of_sequence_id, end_of_sequence_id, max_target_sequence_length, vocab_size, output_layer, batch_size, keep_prob): """ 为推理创建解码层 :param encoder_state: 编码器状态 :param dec_cell: 解码器RNN单元 :param dec_embeddings: 解码器嵌入 :param start_of_sequence_id: GO ID :param end_of_sequence_id: EOS Id :param max_target_sequence_length: 目标序列的最大长度 :param vocab_size: 解码器/目标词汇大小 :param decoding_scope: TensorFlow变量作用域用于解码 :param output_layer: 应用输出层的函数 :param batch_size: 批次大小 :param keep_prob: Dropout保留概率 :return: 包含推理logits和sample_id的BasicDecoderOutput """ start_tokens = tf.tile(tf.constant([start_of_sequence_id], dtype=tf.int32), [batch_size], name='start_tokens') inference_decoder = tf.contrib.seq2seq.BeamSearchDecoder( cell=dec_cell, embedding=dec_embeddings, start_tokens=start_tokens, end_token=end_of_sequence_id, initial_state=encoder_state, beam_width=beam_width, output_layer=output_layer) inference_decoder_output = tf.contrib.seq2seq.dynamic_decode(inference_decoder, impute_finished=False )[0] return inference_decoder_output"""DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE"""#tests.test_decoding_layer_infer(decoding_layer_infer)以下是模型参数：
# 轮数epochs = 200# 批次大小batch_size = 30# RNN大小rnn_size = 512# 层数num_layers = 2# 嵌入大小encoding_embedding_size = 100decoding_embedding_size = 100# 学习率learning_rate = 0.001# Dropout保留概率keep_probability = 0.55display_step = 10beam_width=1我非常希望得到您的帮助，我不确定具体哪里出了问题。
谢谢您
回答：
好的，我刚刚发现了我做错的地方。
我只需要在动态解码函数中设置最大迭代值如下：
inference_decoder_output = tf.contrib.seq2seq.dynamic_decode(inference_decoder, impute_finished=False, maximum_iterations=max_target_sequence_length)[0]