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.

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

Release 0.7 (round 2) #142

Merged
merged 13 commits into from
Apr 8, 2022
Merged

Release 0.7 (round 2) #142

merged 13 commits into from
Apr 8, 2022

Conversation

darsnack
Copy link
Member

@darsnack darsnack commented Apr 4, 2022

I rebased #140 on #132 by accident, so #132 is already merged. No problem, it was ready. But this PR is to try and fix some of the memory pressure issues in the tests. Also an opportunity to suggest changes that should have been in #132.

@theabhirath
Copy link
Member

Docs CI still seems to be erroring out for some reason 😬 Also, this error in the CI is one I've encountered occasionally locally as well (particularly when I update Flux and load it for the first time post precompilation). Not sure what the underlying package is, though.

@ToucheSir
Copy link
Member

Unsure if the warnings have anything to do with the fatal error, but the first is JuliaMath/BFloat16s.jl#37. The second I haven't seen before.

@theabhirath
Copy link
Member

theabhirath commented Apr 5, 2022

The second as in the rpad one? That's just because Flux 0.12.9 doesn't have MLUtils yet, so the rpad definition from Flux and the one from MLUtils (because Metalhead has MLUtils as a dep) are conflicting - should be solved when Flux 0.13 lands. The BFloat16s has never caused an error before though, it's only ever a warning - so I don't think that's the one causing CI failure

@darsnack
Copy link
Member Author

darsnack commented Apr 5, 2022

The CI failure might be completely unrelated and due to Publish.jl.

It's cause our quickstart uses a larger batch size than our CI can handle: https://github.com/FluxML/Metalhead.jl/blob/master/docs/tutorials/quickstart.md

@darsnack
Copy link
Member Author

darsnack commented Apr 8, 2022

Okay I think the release looks ready. @ToucheSir anything else? @theabhirath I replaced uses of Diagonal with the new and improved Scale. Are there any replacements possible with Flux 0.13 that I missed?

@ToucheSir
Copy link
Member

Nope, I think this is good to go.

@theabhirath
Copy link
Member

Seems perfect!

@darsnack darsnack merged commit 5386046 into FluxML:master Apr 8, 2022
@darsnack darsnack deleted the release-0.7 branch April 8, 2022 18:36
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.

3 participants