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

Change defaults for Layers to match Pytorch #808

Closed
avik-pal opened this issue Jul 30, 2024 · 1 comment · Fixed by #744
Closed

Change defaults for Layers to match Pytorch #808

avik-pal opened this issue Jul 30, 2024 · 1 comment · Fixed by #744
Labels
breaking
Milestone
v1.0

Comments

@avik-pal
Copy link
Member

avik-pal commented Jul 30, 2024 •
edited
Loading

Currently this is a WIP list.

Normalization

  • InstanceNorm: affine is false
  • LayerNorm: doesn't accept dims. We should use that as the default and use optimized kernels in LuxLib

Default epsilon values should fallback to the ones in LuxLib
@avik-pal avik-pal added this to the v1.0 milestone Jul 30, 2024
@avik-pal
Copy link
Member Author

avik-pal commented Aug 2, 2024

Other changes that we carried over from Flux also differ from Pytorch -- see FluxML/Flux.jl#2410

@avik-pal avik-pal mentioned this issue Sep 1, 2024
Merged
3 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
breaking
Projects
None yet
Milestone
v1.0
Development

Successfully merging a pull request may close this issue.

Aggregate changes for v1 LuxDL/Lux.jl
1 participant
@avik-pal