8000 [PyTorch] Add Bi-RNN by AnirudhDagar · Pull Request #1351 · d2l-ai/d2l-en · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

[PyTorch] Add Bi-RNN #1351

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

8000

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 14, 2020
Merged

[PyTorch] Add Bi-RNN #1351

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 21 additions & 1 deletion chapter_recurrent-modern/bi-rnn.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,6 +131,27 @@ num_epochs, lr = 500, 1
d2l.train_ch8(model, train_iter, vocab, lr, num_epochs, device)
```

```{.python .input}
#@tab pytorch
from d2l import torch as d2l
import torch
from torch import nn

# Load data
batch_size, num_steps, device = 32, 35, d2l.try_gpu()
train_iter, vocab = d2l.load_data_time_machine(batch_size, num_steps)
# Define the model
vocab_size, num_hiddens, num_layers = len(vocab), 256, 2
num_inputs = vocab_size
lstm_layer = nn.LSTM(num_inputs, num_hiddens, num_layers, bidirectional=True)
lstm_layer.bidirectional
model = d2l.RNNModel(lstm_layer, len(vocab))
model = model.to(device)
# Train the model
num_epochs, lr = 500, 1
d2l.train_ch8(model, train_iter, vocab, lr, num_epochs, device)
```

The output is clearly unsatisfactory for the reasons described above. For a
discussion of more effective uses of bidirectional models, please see the sentiment
classification in :numref:`sec_sentiment_rnn`.
Expand All @@ -148,7 +169,6 @@ classification in :numref:`sec_sentiment_rnn`.
1. Design a bidirectional recurrent neural network with multiple hidden layers.
1. Implement a sequence classification algorithm using bidirectional RNNs. Hint: use the RNN to embed each word and then aggregate (average) all embedded outputs before sending the output into an MLP for classification. For instance, if we have $(\mathbf{o}_1, \mathbf{o}_2, \mathbf{o}_3)$, we compute $\bar{\mathbf{o}} = \frac{1}{3} \sum_i \mathbf{o}_i$ first and then use the latter for sentiment classification.


:begin_tab:`mxnet`
[Discussions](https://discuss.d2l.ai/t/339)
:end_tab:
0