Skip to content
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.

Already on GitHub? Sign in to your account

Jump to bottom

Fix VAE training stability #14

Open
0xtristan opened this issue Jan 20, 2021 · 0 comments
Open

Fix VAE training stability #14

0xtristan opened this issue Jan 20, 2021 · 0 comments
Assignees
@0xtristan

Comments

@0xtristan
Copy link
Collaborator

0xtristan commented Jan 20, 2021 •
edited
Loading

Currently @sholtodouglas uses dynamic gradient clipping (gradient clipped if 4x the previous norm). This still seems to cause some issues with loss spikes.

Other options are static clipping, gradient penalties, weight regularisation/normalisation/decay

A personal favourite of mine that worked very well (and is also leveraged in NVAE) is SpectralNorm. TFA has an implementation here:
https://www.tensorflow.org/addons/api_docs/python/tfa/layers/SpectralNormalization
@0xtristan 0xtristan self-assigned this Jan 20, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@0xtristan 0xtristan

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@0xtristan