GitLab

* initial commit

* formatting

* minor change

* replaced pdf link with webpage link

* replaced pdf link with webpage link

* minor

* minor
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* add convnext.rst

* refactor code

* add convnext
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* [FEAT] Add efficientnetv2 doc.

* [FIX] Precommits fix.

* Apply suggestions from code review
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* [DOC] Add new EfficientNet doc.

* [DOC] Add missing file.

* [FIX] Precommits fix.

* [FIX] Precommits fix.

* [ENH] Integrate few documentation comments from code review

* Update docs/source/models/efficientnet.rst
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* [proto] Added crop_bounding_box op

* Added tests for resized_crop_bounding_box

* Fixed code formatting

Not sure what changed in 61.2.0, but it causes build failures

* test: add functional vertical flip tests on segmentation mask

* refactor: improve test readibility

* Update test_prototype_transforms_functional.py
Co-authored-by: Federico Pozzi <federico.pozzi@argo.vision>
Co-authored-by: vfdev <vfdev.5@gmail.com>

* Restructuring metrics meta-data for detection, segmentation and optical flow.

* Renaming acc to pixel_acc for segmentation

* Restructure video meta-data.

* Restructure classification and quantization meta-data.

* Fix tests.

* Fix documentation

Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Added `crop_segmentation_mask` op

* Fixed failed mypy

* Add docs for regnet, still need to update the comment docs on models

* Fix a little typo on .rst file

* Update regnet docstring

* Add vision_transformer docs, and fix typo on regnet docs

* Update docstring to make sure it does not exceed 120 chars per line

* Improve formatting

* Change the new line location for vision_transformer docstring

Co-authored-by: Nikita Shulga <nshulga@fb.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Removing `task`, `architecture` and `quantization`

* Fix mypy

* Remove size field

* Remove unused import.

* Fix mypy

* Remove size from schema list.

* update todo

* Simplify with assert

* Adding min_size to all models.

* Update RAFT min size to 128

Co-authored-by: Bruno Korbar <bkorbar@quansight.com>
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* Remove publication_year and interpolation meta

* Add type to _COMMON_META and _COMMON_SWAG_META to prevent error from mypy check

* Remove test to check interpolation and publication_year meta
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Convert weights only if `old_key` is in `state_dict`

* Fix linter

* Minor updates to optical flow ref for consistency

* Actually put back name

* linting

* First PR for model doc revamp

* Deactivating fail on warning, temporarily

* Remove commnet

* Minor changes

* Typos

* Added TODO in Makefile

* Keep old models.rst file intact, move new docs into new models_new.rst file

* Added docs LOL

* First PR for model doc revamp

* Deactivating fail on warning, temporarily

* Remove commnet

* Minor changes

* Typos

* Added TODO in Makefile

* Keep old models.rst file intact, move new docs into new models_new.rst file

the assert msg should be same
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* add warning if font is not set

* added quotes

* added changes after review

* Apply suggestions from code review
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* escaped warning string

* Apply suggestions from code review
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* updated warning message in test

* refactored code

* Apply suggestions from code review
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

To resolve issue #5815
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Update focal_loss.py

updated docstring for the `sigmoid_focal_loss`

* edited docstring of focal_loss

* Update focal_loss.py

* fix doc style.

* formatting

* formatting

* minor edits

* move attribution link
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

There is a mistake in FGVCAircraft dataset documentation. If you load dataset and sum up 'trainval' and 'test' splits, there are not 10 200, but 10 000 images. Also mistake is in the number of classes: if we choose `annotation_level` to be `variant`, then unique number of classes is going to be not 102, but 100, and with `manufacturer` level it is not 41, but 30.
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* Add test for giou_loss

* Use standard assert for integer comparison
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Fix ufmt format and add log after assert
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

As Ubuntu-1604 runners will no longer be available in early May 2022
Update `ubuntu-1604:201903-01` to `ubuntu-2004:202104-01`
and `ubuntu-1604-cuda-10.2:202012-01` to `ubuntu-2004-cuda-11.4:202110-01`
Per [CircleCI Configuration reference](https://circleci.com/docs/2.0/configuration-reference/)

* Add SWAG model that only the linear classifier head is finetuned with frozen trunk weight

* Add accuracy from experiments

* Change name from SWAG_LC to SWAG_LINEAR

* Add comment on SWAG_LINEAR weight

* Remove the comment docs (moved to PR description), and add the PR url as recipe. Also change name of previous swag model to SWAG_E2E_V1

* Update _register_onnx_ops.py

* Update misc.py

* Add test for sigmoid_focal_loss

* Update test/test_ops.py

Improve code by using torch.testing.make_tensor
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Update test/test_ops.py

Remove unnecessary assert
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Update test/test_ops.py

Refactor code for generating inputs and targets
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Improve focal_loss test code suggested on comment by Philip

* Use fuser2 to prevent fuser bug

* Combine function to generate input, dont set the fuser when device is cpu

* Add github issue for the fuser problem
Co-authored-by: Philip Meier <github.pmeier@posteo.de>