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

[Fix] Unloaded weights will not be initialized when using PretrainedIinit #764

Merged
merged 10 commits into from
Jan 9, 2023
Merged

[Fix] Unloaded weights will not be initialized when using PretrainedIinit #764

merged 10 commits into from
Jan 9, 2023

Conversation

HAOCHENYE
Copy link
Collaborator

@HAOCHENYE HAOCHENYE commented Nov 24, 2022 •
edited
Loading

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Resolve #691 .

Modification

Please briefly describe what modification is made in this PR.

BC-breaking (Optional)

Does the modification introduce changes that break the backward-compatibility of the downstream repos?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Checklist

  1. Pre-commit or other linting tools are used to fix the potential lint issues.
  2. The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
  3. If the modification has potential influence on downstream projects, this PR should be tested with downstream projects, like MMDet or MMCls.
  4. The documentation has been modified accordingly, like docstring or example tutorials.

@HAOCHENYE HAOCHENYE force-pushed the HAOCHENYE/fix_PretrainedInit branch from 4153bac to cb4b70d Compare November 30, 2022 06:59
@HAOCHENYE HAOCHENYE changed the title [Draft] Initialize the module not loaded pretrained weights [Fix] Initialize the module not loaded pretrained weights Nov 30, 2022
@HAOCHENYE HAOCHENYE marked this pull request as ready for review November 30, 2022 06:59
@zhouzaida zhouzaida requested a review from RangiLyu December 12, 2022 03:30
@HAOCHENYE HAOCHENYE requested a review from MeowZheng December 26, 2022 03:25
MeowZheng
MeowZheng previously approved these changes Jan 6, 2023
View reviewed changes
@HAOCHENYE HAOCHENYE force-pushed the HAOCHENYE/fix_PretrainedInit branch from 763470a to 98ac467 Compare January 9, 2023 04:23
@codecov
Copy link

codecov bot commented Jan 9, 2023 •
edited
Loading

Codecov Report

Base: 78.66% // Head: 78.14% // Decreases project coverage by -0.52% ⚠️

Coverage data is based on head (11f769e) compared to base (fe26c65).
Patch coverage: 50.72% of modified lines in pull request are covered.

❗ Current head 11f769e differs from pull request most recent head 611eec3. Consider uploading reports for the commit 611eec3 to get more accurate results

Additional details and impacted files 
@@            Coverage Diff             @@
##             main     #764      +/-   ##
==========================================
- Coverage   78.66%   78.14%   -0.53%     
==========================================
  Files         128      130       +2     
  Lines        9348     9431      +83     
  Branches     1848     1887      +39     
==========================================
+ Hits         7354     7370      +16     
- Misses       1679     1741      +62     
- Partials      315      320       +5
Flag Coverage Δ
unittests 78.14% <50.72%> (-0.53%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Δ
mmengine/model/wrappers/__init__.py 66.66% <ø> (-3.34%) ⬇️
mmengine/optim/optimizer/zero_optimizer.py 52.17% <0.00%> (-4.97%) ⬇️
mmengine/registry/build_functions.py 80.20% <ø> (ø)
mmengine/registry/default_scope.py 100.00% <ø> (ø)
mmengine/runner/runner.py 84.91% <ø> (-0.24%) ⬇️
mmengine/structures/base_data_element.py 93.61% <ø> (-0.04%) ⬇️
mmengine/structures/label_data.py 100.00% <ø> (ø)
mmengine/visualization/vis_backend.py 83.79% <ø> (-0.70%) ⬇️
mmengine/visualization/visualizer.py 93.03% <ø> (-0.17%) ⬇️
mmengine/runner/checkpoint.py 52.07% <14.28%> (+8.11%) ⬆️
... and 54 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

zhouzaida
zhouzaida previously approved these changes Jan 9, 2023
View reviewed changes
zhouzaida
zhouzaida previously approved these changes Jan 9, 2023
View reviewed changes
@HAOCHENYE HAOCHENYE changed the title [Fix] Initialize the module not loaded pretrained weights [Fix] Unloaded weights will not be initialized when using PretrainedIinit Jan 9, 2023
@zhouzaida zhouzaida merged commit 925ac87 into open-mmlab:main Jan 9, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@C1rN09 C1rN09 C1rN09 left review comments

@zhouzaida zhouzaida zhouzaida approved these changes

@MeowZheng MeowZheng MeowZheng left review comments

@RangiLyu RangiLyu Awaiting requested review from RangiLyu

Assignees

@ice-tong ice-tong

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

Unexpected weight initialization
5 participants
@HAOCHENYE @zhouzaida @MeowZheng @C1rN09 @ice-tong