pytorch torch.nn.RNN

应用

rnn = nn.RNN(10, 20, 2)
input = torch.randn(5, 3, 10)
h0 = torch.randn(2, 3, 20)
output, hn = rnn(input, h0)

API

CLASS torch.nn.RNN(*args, **kwargs)

$h_t=tanh(W_{ih}{x}_t+b_{ih}+W_{hh}{h}_{t-1}+b_{hh}$
$h_t$：the hidden state at time t
$x_t$：the input at time t
$h_{t-1}$：the hidden state of the previous layer at time t-1
如果nonlinearity是relu，则会替换tanh

类

参数	描述
input_size	The number of expected features in the input x
hidden_size	The number of features in the hidden state h
num_layers	Number of recurrent layers，Default: 1
nonlinearity	The non-linearity to use.Default: ‘tanh’
bias	If False, then the layer does not use bias weights b_ih and b_hh. Default: True
batch_first	If True, then the input and output tensors are provided as (batch, seq, feature). Default: False
bidirectional	If True, becomes a bidirectional RNN. Default: False

input_size：是RNN的维度，注意不是句子或序列的长度，而是句子的一个词，或序列的一个元素的维度。比如，词向量的维度。比如说NLP中你需要把一个单词输入到RNN中，这个单词的编码是300维的，那么这个input_size就是300.
hidden_size：每个RNN的节点实际上就是一个BP网络，包含输入层，隐含层，输出层。这里就是指隐藏层的节点个数。
num_layers：如果num_layer=2的话，表示两个RNN堆叠在一起。

参考：https://www.cnblogs.com/dhName/p/11760610.html

对象

输入：

参数	描述
input of shape (seq_len, batch, input_size)	The input can also be a packed variable length sequence. See torch.nn.utils.rnn.pack_padded_sequence() or torch.nn.utils.rnn.pack_sequence() for details.
h_0 of shape (num_layers * num_directions, batch, hidden_size)	tensor containing the initial hidden state for each element in the batch. Defaults to zero if not provided. If the RNN is bidirectional, num_directions should be 2, else it should be 1.

输出：

参数	描述
output of shape (seq_len, batch, num_directions * hidden_size):
h_n of shape (num_layers * num_directions, batch, hidden_size):

参考：

https://www.icode9.com/content-4-622959.html
https://zhuanlan.zhihu.com/p/59772104
只是batch长度要求相同，但不同batch则不需要相同？
https://zhuanlan.zhihu.com/p/97378498
https://www.cnblogs.com/lindaxin/p/8052043.html
https://www.jianshu.com/p/f5b816750839
https://www.jianshu.com/p/efe045c24a93
https://zhuanlan.zhihu.com/p/161972223
https://zhuanlan.zhihu.com/p/34418001?edition=yidianzixun&utm_source=yidianzixun&yidian_docid=0IVwLf60
https://www.cnblogs.com/jiangkejie/p/13141664.html
https://zhuanlan.zhihu.com/p/64527432