-
Notifications
You must be signed in to change notification settings - Fork 7k
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
Feature/layer freezing maskrcnn keypointrcnn #2242
Feature/layer freezing maskrcnn keypointrcnn #2242
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.
Changes look good to me, thanks a lot @muaz-urwa !
There seems to be some python lint failures, can you fix those? Here are the errors https://circleci.com/gh/pytorch/vision/142749?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link:
./test/test_models_detection_utils.py:37:1: W293 blank line contains whitespace
./test/test_models_detection_utils.py:46:45: E127 continuation line over-indented for visual indent
./test/test_models_detection_utils.py:47:45: E127 continuation line over-indented for visual indent
./test/test_models_detection_utils.py:61:45: E128 continuation line under-indented for visual indent
./test/test_models_detection_utils.py:62:45: E128 continuation line under-indented for visual indent
./torchvision/models/detection/mask_rcnn.py:318:37: E225 missing whitespace around operator
./torchvision/models/detection/mask_rcnn.py:318:73: E225 missing whitespace around operator
./torchvision/models/detection/keypoint_rcnn.py:322:37: E225 missing whitespace around operator
./torchvision/models/detection/keypoint_rcnn.py:322:73: E225 missing whitespace around operator
@fmassa my bad. I have corrected the linting errors but some checks are still failing. I have checked the logs, these changes don't appear to be the reason. Will be happy to look deeper if you think otherwise. |
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 a lot!
This pull request is a follow up pull reguest to PR, as requested by @fmassa .
This PR adds arguments to keypointrcnn_resnet50_fpn and maskrcnn_resnet50_fpn functions, which allow the user to control the number of frozen layers in the backbone. Visit the previous PR (link above) to learn more about it.
I have also added respective test cases.
Let me know if any changes are required.