Top Answers

1. Understanding accumulated gradients in PyTorch
   44 votes
2. Understanding input shape to PyTorch LSTM
   22 votes
3. net.zero_grad() vs optim.zero_grad() pytorch
   22 votes
4. No matching distribution found for torch==1.5.0+cpu on Heroku
   17 votes
5. TransformerEncoder with a padding mask
   15 votes