Skip to content

NOMRG Comments on prototype weights and model builders #4937

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

Closed
wants to merge 3 commits into from

Conversation

NicolasHug
Copy link
Member

Here are a bunch of comments on the new weights prototype. I'm opening this as a PR (not to be merged) instead of an issue as I think it will better help localize each discussion.

Overall I don't have any major comment or concern, except maybe for the somewhat BC breaking changes of some model building functions. I think we can work around BC for some of them, but not all.

@facebook-github-bot
Copy link

facebook-github-bot commented Nov 15, 2021

💊 CI failures summary and remediations

As of commit 500f826 (more details on the Dr. CI page):


  • 1/1 failures introduced in this PR

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

See CircleCI build python_type_check (1/1)

Step: "sudo apt-get update -y
sudo apt install -y libturbojpeg-dev
pip install --user --progress-bar off mypy
pip install --user --progress-bar off types-requests
pip install --user --progress-bar off --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
pip install --user --progress-bar off git+https://github.com/pytorch/data.git
pip install --user --progress-bar off --no-build-isolation --editable .
mypy --config-file mypy.ini
" (full log | diagnosis details | 🔁 rerun)

Found 2 errors in 1 file (checked 190 source files)
�[33mWARNING: You are using pip version 21.2.4; however, version 21.3.1 is available.
You should consider upgrading via the '/usr/local/bin/python -m pip install --upgrade pip' command.�[0m
torchvision/prototype/models/resnet.py:52: error: Name "blah" is not defined 
[name-defined]
                    f"Oops, you specified num_classes={blah} but this is i...
                    ^
torchvision/prototype/models/resnet.py:52: error: Name "blop" is not defined 
[name-defined]
                    f"Oops, you specified num_classes={blah} but this is i...
                    ^
Found 2 errors in 1 file (checked 190 source files)


Exited with code exit status 1


This comment was automatically generated by Dr. CI (expand for details).

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

Copy link
Contributor

@datumbox datumbox left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the comments @NicolasHug.

I'm going to go ahead and implement many of the proposals and send a PR (I'll add you as reviewer).

I also added more comments/context to those that I require more clarifications. Let me know your thoughts.

@datumbox
Copy link
Contributor

@NicolasHug I believe we addressed all the comments on this PR and we can close it. If there are more, happy to make the needed changes. Let me know

@NicolasHug
Copy link
Member Author

Thanks for all the discussions @datumbox @fmassa , we can safely close this one now. Latest discussions were tracked in #4652 (comment)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Multi pretrained weights: Cleanups and Refactoring
4 participants