-
Notifications
You must be signed in to change notification settings - Fork 27.4k
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
Add WSD scheduler #30231
Add WSD scheduler #30231
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nice! Seems handy. Could you rebase from main
and see if that fixes the failing tests?
b8e64ec
to
e8eadd2
Compare
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Looks like it fixed some tests but not all. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LG2M, no idea why that pytorch test is failing 😕
cc @amyeroberts for final review (and ideas(?) was seeing that on another PR of mine iirc)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for adding this!
@visheratin CI should be green on rebase, we found the issue :) |
e8eadd2
to
ee9f3c3
Compare
@muellerzr a couple of tests are still failing. |
@visheratin Looks like they were failing because of connection issues with the hub. I've re-run - which should hopefully resolve |
* Added WSD scheduler. * Added tests. * Fixed errors. * Fix formatting. * CI fixes.
What does this PR do?
This PR adds the Warmup-Stable-Decay learning rate scheduler. The scheduler and its benefits are described in detail in the MiniCPM blog post. In short, it allows continuous training and performs better than standard cosine scheduler.
Here is an example of learning rate dynamics with this schedule:
Before submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
@muellerzr, @pacman100