GitLab

* Windows use size_t

* Temporarily pin jpeg <=9b

* Temporarily pin jpeg ==9c

* Remove pinning.

* Warmup on Classficiation references.

* Adjust epochs for cosine.

* Warmup on Segmentation references.

* Warmup on Video classification references.

* Adding support of both types of warmup in segmentation.

* Use LinearLR in detection.

* Fix deprecation warning.

* Enable building torchvision with ffmpeg regardless of version on Linux

* Skip tests with python 3.9

* Removed PY39_SKIP

* Cleaned up a bit

* Skip building torchvision with ffmpeg when python==3.9

* Add FIXME to comments

Co-authored-by: Julien RIPOCHE <ripoche@magic-lemp.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Add RandomMixupCutmix.

* Add test with real data.

* Use dataloader and collate in the test.

* Making RandomMixupCutmix JIT scriptable.

* Move out label_smoothing and try roll instead of flip

* Adding mixup/cutmix in references script.

* Handle one-hot encoded target in accuracy.

* Add support of devices on tests.

* Separate Mixup from Cutmix.

* Add check for floats.

* Adding device on expect value.

* Remove hardcoded weights.

* One-hot only when necessary.

* Fix linter.

* Moving mixup and cutmix to references.

* Final code clean up.

* Updated annotation to be a Union of Tensor and List

* Updated check_roi_boxes_shape.

* Fix ffmpeg version

* Ignore platforms other than Linux for ffmpeg

* Removed unnecessary changes

* Update log message.

* Update fstring.

* Added LFW Dataset

* Added dataset to list in __init__.py

* Updated lfw.py
* Created a common superclass for people and pairs type datatsets
* corrected the .download() method

* Added docstrings and updated datasets.rst

* Wrote tests for LFWPeople and LFWPairs

* Resolved mypy error: Need type annotation for "data"

* Updated inject_fake_data method for LFWPeople

* Updated tests for LFW

* Updated LFW tests and minor changes in lfw.py

* Updated LFW
* Added functionality for 10-fold validation view
* Optimized the code so to replace repeated lines by method in super
  class
* Updated LFWPeople to get classes from all lfw-names.txt rather than
  just the classes fron trainset

* Updated lfw.py and tests
* Updated inject_fake_data method to create 10fold fake data
* Minor changes in docstring and extra_repr

* resolved py lint errors

* Added checksums for annotation files

* Minor changes in test

* Updated docstrings, defaults and minor changes in test

* Removed 'os.path.exists' check
Co-authored-by: ABD-01 <myac931@gmai.com>
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* if there's dll, overwrite it

* temp change

* revert last
Co-authored-by: Nikita Shulga <nshulga@fb.com>

* rm install_cuda_compatible_cmath

* Remove unused script
Co-authored-by: Nicolas Hug <nicolashug@fb.com>
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* add pre-commit hooks

* ignore yamls in packaging/*

* add pre-commit to contributing guide lines

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

* remove some hooks

* fix docstrings

* fix end of files
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* Add deprecation warnings on vision::models

* Change the C++ example.

* Chage readme.

* Update deprecation warning.

* Fix training resuming in references/segmentation

* Clarification for training resnext101_32x8d

* Update references/classification/README.md
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* Added Exponential Moving Average support to classification reference script

* Addressed review comments

* Updated model argument

* Added paper references to detection models

* Ignore linter warning

* Break long line into two

Co-authored-by: Vassilis Nicodemou <nikodim@ics.forth.gr>
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/63414



Misuse of raw pointer in here where stack is never nullable.
ghstack-source-id: 136938318

Reviewed By: ejguan

Differential Revision: D30375410

fbshipit-source-id: 9d65b620bb76d90d886c800f54308520095d58ee
Co-authored-by: Zhengxu Chen <zhxchen17@fb.com>

* add fx feature extraction util

* Make it possible to use train and eval mode

* FX feature extraction - Tweaks and small bug fixes

* FX feature extraction - add tests

* move to feature_extraction.py, add LeafModuleAwareTracer, add docs

* Tweaks to docs

* addressing latest round of feedback

* undo line spacing changes

* change type hints in docstrings

* fix sphinx indentation

* expose feature_extraction

* add maskrcnn example

* add api refernce subheading

* address latest review notes, refactor names, fix regex, cosmetics

* Add back efficientnet to models

* fix tests for effnet

* fix linting issue

* fix test tracer kwargs
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Add types in transform.

* Trace on eval mode.
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* WIP on fixing index overflow issue

* Fixed backward pass for large num_kernels

* Fixed clang formatting

* Fixed GET_BLOCKS int/int64_t types issue
Co-authored-by: vfdev-5 <vfdev-5@gmail.com>
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Fix RA bugs.

* Fix bins for TA.

* Initial Proposal

* Tensor Save Test + Test Name Fix

* Formatting + removing unused argument

* fix old argument

* fix isnan check error + indexing error with jit

* Fix Flake8 error.

* Fix MyPy error.

* Fix Flake8 error.

* Fix PyTorch JIT error in UnitTests due to type annotation.

* Fixing tests.

* Removing type ignore.

* Adding support of ta_wide in references.

* Move methods in classes.

* Moving new classes to the bottom.

* Specialize to TA to TAwide

* Add missing type

* Fixing lint

* Fix doc

* Fix search space of TrivialAugment.
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>
Co-authored-by: Vasilis Vryniotis <vvryniotis@fb.com>

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

* Adding docs for RandAugment.

* Fix docs.

* Adding randaugment implementation

* Refactoring.

* Adding num_magnitude_bins.

* Adding FIXME.

* Adding label smoothing on classification reference.

* Replace underscore with dash.

* Explicitely store a distance value that is reused 

I don't see a reason to calculate the value twice for each distance. 

Knowing that this code is going to be called at every epoch and that probably there is not a compiler that optimizes this (+ also code clarity), I think is best to store the value for x and y.

* Update torchvision/models/detection/_utils.py
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* removing spaces
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Fixed a couple of typos and removed unnecessary bracket

* Fixed clang error