Created by: wu8320175

版本、环境信息：    1）PaddlePaddle版本：1.8-静态图 LSTM文档：https://www.paddlepaddle.org.cn/documentation/docs/zh/api_cn/layers_cn/lstm_cn.html#lstm

文档中说了配置好相应参数后应该返回2*hidden_size，但是我按下面的代码配置为啥得不到正确的返回呢？

import paddle.fluid as fluid
import paddle.fluid.layers as layers

emb_dim = 256
vocab_size = 10000
data = fluid.layers.data(name='x1', shape=[-1, 100, 1],
               dtype='int64')
emb = fluid.layers.embedding(input=data, size=[vocab_size, emb_dim], is_sparse=True)
batch_size = 20
max_len = 100
dropout_prob = 0.2
hidden_size = 150
num_layers = 1
init_h = layers.fill_constant( [num_layers*2, batch_size, hidden_size], 'float32', 0.0 )
init_c = layers.fill_constant( [num_layers*2, batch_size, hidden_size], 'float32', 0.0 )
rnn_out, last_h, last_c = layers.lstm(emb,init_h, init_c, max_len, hidden_size, num_layers,is_bidirec=True, dropout_prob=dropout_prob)
rnn_out.shape,last_h.shape,last_c.shape
# rnn_out.shape  # (-1, 100, 150)
# last_h.shape  # (2, 20, 150)
# last_c.shape  # (2, 20, 150)

lastc和lasth能得到正确结果但是rnn_out明显不对呀 [batch_size,seq_len,hidden_size∗2]应该是(-1, 100, 300)呀？