Created
November 3, 2019 17:41
-
-
Save pvva/869a85ba1f9c62a8332a025830cd72ec to your computer and use it in GitHub Desktop.
rnn1_2
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| def detach_from_history(h): | |
| if type(h) == torch.Tensor: | |
| return h.detach() | |
| return tuple(detach_from_history(v) for v in h) | |
| class CharRnn(nn.Module): | |
| def __init__(self, vocab_size, n_fac, n_hidden, batch_size): | |
| super().__init__() | |
| self.e = nn.Embedding(vocab_size, n_fac) | |
| self.rnn = nn.RNN(n_fac, n_hidden) | |
| self.l_out = nn.Linear(n_hidden, vocab_size) | |
| self.n_hidden = n_hidden | |
| self.init_hidden_state(batch_size) | |
| def init_hidden_state(self, batch_size): | |
| self.h = torch.zeros(1, batch_size, self.n_hidden).cuda() | |
| def forward(self, inp): | |
| inp = self.e(inp) | |
| b_size = inp[0].size(0) | |
| if self.h[0].size(1) != b_size: | |
| self.init_hidden_state(b_size) | |
| outp, h = self.rnn(inp, self.h) | |
| self.h = detach_from_history(h) | |
| return F.log_softmax(self.l_out(outp[-1]), dim=-1) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment