We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
GRUStep in lib/ops.py does implement a bit untraditional GRU unit: https://github.com/soroushmehr/sampleRNN_ICLR2017/blob/master/lib/ops.py#L372
return (update * candidate) + ((one - update) * last_hidden)
which is supposed to return GRU as we know: S_t = (1 - z).h + z.S_{t-1}
S_t = (1 - z).h + z.S_{t-1}
Instead the code does: S_t = (1 - z).S_{t-1} + z.h
S_t = (1 - z).S_{t-1} + z.h
Logically it seems not too far from the original variant except z is inverted. But in practice does it affect any figures presented in the paper?
The text was updated successfully, but these errors were encountered:
No branches or pull requests
GRUStep in lib/ops.py does implement a bit untraditional GRU unit:
https://github.com/soroushmehr/sampleRNN_ICLR2017/blob/master/lib/ops.py#L372
return (update * candidate) + ((one - update) * last_hidden)which is supposed to return GRU as we know:
S_t = (1 - z).h + z.S_{t-1}Instead the code does:
S_t = (1 - z).S_{t-1} + z.hLogically it seems not too far from the original variant except z is inverted.
But in practice does it affect any figures presented in the paper?
The text was updated successfully, but these errors were encountered: