Skip to content

Conversation

pyu1538
Copy link
Contributor

@pyu1538 pyu1538 commented Jun 5, 2020

…activation, so it matches the paper

@tracek
Copy link

tracek commented Jun 7, 2020

Doing so will prevent the kernel from using cuDNN version of GRU, defaulting to the generic implementation. That in turn will severely slow down the training (assuming you're using GPU for the training).

@pyu1538
Copy link
Contributor Author

pyu1538 commented Jun 10, 2020

Doing so will prevent the kernel from using cuDNN version of GRU, defaulting to the generic implementation. That in turn will severely slow down the training (assuming you're using GPU for the training).

Got it.
I trained it without GPU, so I didn't feel slow down. RNNoise demo is also trained by "relu", it's better to use exact same model for comparison.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants