-
Notifications
You must be signed in to change notification settings - Fork 3.6k
Add EMAWeightAveraging callback to weight_averaging.py
#21260
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
base: master
Are you sure you want to change the base?
Conversation
for more information, see https://pre-commit.ci
|
Hi @amorehead , can you please write tests for it? |
|
Hi, @deependujha. I've added a few unit tests as requested. |
Co-authored-by: GdoongMathew <[email protected]>
Thanks @GdoongMathew! Co-authored-by: GdoongMathew <[email protected]>
|
@Borda, (nearly) all tests are passing. Are we OK with merging this PR as is? |
Codecov Report✅ All modified and coverable lines are covered by tests.
Additional details and impacted files@@ Coverage Diff @@
## master #21260 +/- ##
=========================================
- Coverage 87% 79% -8%
=========================================
Files 269 266 -3
Lines 23745 23707 -38
=========================================
- Hits 20570 18689 -1881
- Misses 3175 5018 +1843 |
What does this PR do?
Adds a default
EMAWeightAveragingcallback toweight_averaging.pyfor convenience.Now, users of
weight_averaging.pycan directly instantiate an EMA callback for model training with safe default hyperparameters chosen.Before submitting
PR review
Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:
Reviewer checklist
📚 Documentation preview 📚: https://pytorch-lightning--21260.org.readthedocs.build/en/21260/