-
Notifications
You must be signed in to change notification settings - Fork 3.4k
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
[fix] Enable manual optimization DeepSpeed #7970
Conversation
Hello @tchaton! Thanks for updating this PR. There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻 Comment last updated at 2021-06-16 09:02:29 UTC |
for more information, see https://pre-commit.ci
Codecov Report
@@ Coverage Diff @@
## master #7970 +/- ##
======================================
- Coverage 92% 91% -0%
======================================
Files 207 207
Lines 13378 13380 +2
======================================
- Hits 12248 12214 -34
- Misses 1130 1166 +36 |
Let's get #7234 merged, then I'll rebase this PR and address comments for thomas! |
# Conflicts: # dockers/base-cuda/Dockerfile # pytorch_lightning/plugins/training_type/deepspeed.py
The test failing is associated to #7790 so should be good to go |
Cherry picking this for 1.3.x will be difficult as it relies on the later DeepSpeed Infinity API, so we'll wait for 1.4. |
What does this PR do?
This PR enables support for manual optimization for DeepSpeed.
However, DeepSpeed only supports 1 optimizer.
Fixes #7957 #6503
Before submitting
PR review
Anyone in the community is free to review the PR once the tests have passed.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:
Did you have fun?
Make sure you had fun coding 🙃