GitLab

* Refactor model builder

* Add 3 more convnext variants.

* Adding weights for convnext_small.

* Fix minor bug.

* Fix number of parameters for small model.

* Adding weights for the base variant.

* Adding weights for the large variant.

* Simplify LayerNorm2d implementation.

* Optimize speed of CNBlock.

* Repackage weights.

* Adding conv_stem support

* fix lint

* bug fix

* address comments

* fix after merge

* adding back checking lines

* fix failing tests

* fix iignore

* add unittest & address comments

* fix memory issue

* address comments

* Some Qs

* Some modifications

* don't need _loader in __init__

* list_names -> list_datasets

* Update torchvision/prototype/datasets/utils/_resource.py
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Remove unsued import

* fix tests

* Some missing renames
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Add PCAM dataset to prototype area

* use BytesIO instead of writing file to disk

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

* remove noqa

* typos

* Use _Resource namedtuple

* Add h5py to unittest_prototype job

* use .item() on target

* Forgot to call regenerate.py
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Start stuff

* Start adding some implementation

* Add implementation

* Use torch for colorwheel

* Add simple test

* Fix small numpy rename

* Adapt the changes

* Fix stuff

* Add suggestions

* Minor fixes for float

* Remove idx logic

* Simplify

* Fix test

* Update code and add test

* Add expected flow asset

* Fix expected flow path

* Doc nits
Co-authored-by: Nicolas Hug <nicolashug@fb.com>
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* fix bug when the target is empty

* Add unittest for empty instance training
Co-authored-by: Zhiqiang Wang <zhiqwang@foxmail.com>
Co-authored-by: Joao Gomes <joaopsgomes@gmail.com>

* expect the CI to fail

* retrigger tests

* rewriting to close the containers and avoid potential clashes

* addressing Prabhat's commetns
Co-authored-by: Bruno Korbar <bkorbar@quansight.com>
Co-authored-by: Prabhat Roy <prabhatroy@fb.com>

* replace assert with valueerror

* pytest should raise ValueError not AssertionError

* minor edit

* raise assert changed to raise valueerror in test

* Update torchvision/models/detection/backbone_utils.py
Co-authored-by: Aditya Oke <47158509+oke-aditya@users.noreply.github.com>

* Update torchvision/models/detection/backbone_utils.py
Co-authored-by: Aditya Oke <47158509+oke-aditya@users.noreply.github.com>

* minor edits

* minor edits

* added one test

* added another test

* added another test

* test for mobilenet

* ufmt formatting

* cant have unused variables

* suggested changes

* minor edit

* corrected bug pointed out by datumbox

* corrected bug pointed out by datumbox

* bug correction and shorten msg

* ufmt stuff

* resolved last comment
Co-authored-by: Abhijit Deo <72816663+abhi-glitchhg@users.noreply.github.com>
Co-authored-by: Aditya Oke <47158509+oke-aditya@users.noreply.github.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* use fixture for dataset root in tests

* fix home dir generation

* remove caching
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Added center option to F.affine and RandomAffine ops

* Updates according to the review

* Add seek in GPU decoder

* Merge the two tests

* Refine unit test

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

* renamed ImageNet weights

* renamed COCO weights

* renamed COCO with VOC labels weights

* renamed Kinetics 400 weights

* rename default with DEFAULT

* update test

* fix typos

* update test

* update test

* update test

* indent as w was weight_enum

* revert

* Adding back the capitalization test
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* add fcos

* update fcos

* add giou_loss

* add BoxLinearCoder for FCOS

* add full code for FCOS

* add giou loss

* add fcos

* add __all__

* Fixing lint

* Fixing lint in giou_loss.py

* Add typing annotation to fcos

* Add trained checkpoints

* Use partial to replace lambda

* Minor fixes to docstrings

* Apply ufmt format

* Fixing docstrings

* Fixing jit scripting

* Minor fixes to docstrings

* Fixing jit scripting

* Ignore mypy in fcos

* Fixing trained checkpoints

* Fixing unit-test of jit script

* Fixing docstrings

* Add test/expect/ModelTester.test_fcos_resnet50_fpn_expect.pkl

* Fixing test_detection_model_trainable_backbone_layers

* Update test_fcos_resnet50_fpn_expect.pkl

* rename stride to box size

* remove TODO and fix some typo

* merge some code for better

* impove the comments

* remove decode and encode of BoxLinearCoder

* remove some unnecessary hints

* use the default value in detectron2.

* update doc

* Add unittest for BoxLinearCoder

* Add types in FCOS

* Add docstring for BoxLinearCoder

* Minor fix for the docstring

* update doc

* Update fcos_resnet50_fpn_coco pretained weights url

* Update torchvision/models/detection/fcos.py
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Update torchvision/models/detection/fcos.py
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Update torchvision/models/detection/fcos.py
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Update torchvision/models/detection/fcos.py
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Add FCOS model documentation

* Fix typo in FCOS documentation

* Add fcos to the prototype builder

* Capitalize COCO_V1

* Fix params of fcos

* fix bug for partial

* Fixing docs indentation

* Fixing docs format in giou_loss

* Adopt Reference for GIoU Loss

* Rename giou_loss to generalized_box_iou_loss

* remove overwrite_eps

* Update AP test values

* Minor fixes for the docs

* Minor fixes for the docs

* Update torchvision/models/detection/fcos.py
Co-authored-by: Zhiqiang Wang <zhiqwang@foxmail.com>

* Update torchvision/prototype/models/detection/fcos.py
Co-authored-by: Zhiqiang Wang <zhiqwang@foxmail.com>
Co-authored-by: zhiqiang <zhiqwang@foxmail.com>
Co-authored-by: Joao Gomes <jdsgomes@fb.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>
Co-authored-by: Joao Gomes <joaopsgomes@gmail.com>

* Adding CNBlock and skeleton architecture

* Completed implementation.

* Adding model in prototypes.

* Add test and minor refactor for JIT.

* Fix mypy.

* Fixing naming conventions.

* Fixing tests.

* Fix stochastic depth percentages.

* Adding stochastic depth to tiny variant.

* Minor refactoring and adding comments.

* Adding weights.

* Update default weights.

* Fix transforms issue

* Move convnext to prototype.

* linter fix

* fix docs

* Addressing code review comments.

* add SVHN prototype dataset

* add test

* refactor prototype dataset tests to use public API for loading

* add explanation

* use loop alternative

* allow single extension as str in make_dataset

* remove test class

* remove regex

* revert collection to tuple

* cleanup

* Adding min_size as a required field.

* Adding min_size to classification models (quantized and not)

* Adding min_size to video models meta.

* Moving min_size to _COMMON_META

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

* Change default of download for Food101 and DTD

* Set download default to False and put it at the end

* Keep stuff private

* GTSRB: train -> split. Also use pathlib

* mypy

* Remove split and partition for SUN397

* mypy

* mypy

* move download param for SST2

* Use make_dataset in SST2

* Use a base URL for GTSRB

* Let's make this code more complictaed than it needs to be because why not

* improve error message for prototype datasets without options

* remove early exit

* add test

* Adding multiweight support for shufflenetv2 prototype models

* Revert "Adding multiweight support for shufflenetv2 prototype models"

This reverts commit 31fadbee7d1a65cd73ae43dfd4ac6e97e7ca7b01.

* Adding multiweight support for shufflenetv2 prototype models

* Revert "Adding multiweight support for shufflenetv2 prototype models"

This reverts commit 4e3d900f796c1e3e667312087e77956ca4a4c017.

* Add RenderedSST2 dataset

* Address PR comments

* Fix bug in dataset verification
Co-authored-by: Nicolas Hug <contact@nicolas-hug.com>

* adding vit_h_14

* prototype and docs

* bug fix

* adding curl check

* fix and add test for sequence_to_str

* remove manual ids

* Return RGB frames as output of GPU decoder

* Move clamp to the conversion function

* Cleaned up a bit

* Remove utility functions from test

* Use data member width directly

* Fix linter error

* refactor prototype datasets tests

* skip tests with insufficient third party dependencies

* cleanup

* add tests for SBD prototype dataset

* add tests for SEMEION prototype dataset

* add tests for VOC prototype dataset

* add tests for CelebA prototype dataset

* add tests for DTD prototype dataset

* add tests for FER2013 prototype dataset

* add tests for CLEVR prototype dataset

* add tests for oxford-iiit-pet prototype dataset

* enforce tests for new datasets

* add missing archive generation for oxford-iiit-pet tests

* add tests for CUB200 prototype datasets

* fix split generation

* add capability to mark parametrization and xfail cub200 traverse tests

* [WIP]
*added stanford_cars

* [WIP]
added stanfordCars to docs

* [WIP]
minor edits

* [WIP]
minor edits

* edited StanfordCars class

* Adding Testcase for stanford cars

* Added Testcase for stanford cars

* Added Testcase for stanford cars

* minor edit

* made changes as per the suggestions

* fixed typo in naming stanford_cars.py

* cars_meta.mat file will be created in test

* Some cleanups

* Sigh

* don't convert to strings
Co-authored-by: Nicolas Hug <nicolashug@fb.com>

* feat: Added EuroSAT dataset

* test: Added unittest

* docs: Improved comments

* docs: Updated the documentation

* docs: Removed unnecessary comments

* fix: Fixed class implementation

* test: Fixed unittest

* fix: Fixed magic method len

* test: Fixed unittest

* refactor: Refactored EuroSAT

* refactor: Applied modifications

* Apply suggestions from code review

* refactor: Applied request changes

* refactor: Made var explicit

* fix: Fixed attribute initialization order

* refactor: Removed name mapping
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Add support for PCAM dataset

* mypy

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

* Remove classes and class_to_idx attributes

* Use _decompress
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* relax split requirement for prototype datasets

* remove obsolete tests

* appease mypy

* fix failing test

* fix load config test with default config

* add fvgc_aircraft dataset

* add docstring & remove useless import

* resolve lint issue

* address comments

* adding more annotation level

* nit

* address comments

* Apply suggestions from code review

* unify format

* remove useless line
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* add regnet_y_128gf

* fix test

* add expected test file

* update regnet factory function, add to prototype as well

* write torchscript to temp file instead bytesio in model test

* docs

* clear GPU memory

* no_grad

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

* Add Flowers102 datasets

* Fix initialization of images and labels

* Fix _check_exists in Flowers102

* Add Flowers102 to datasets and docs

* Add Flowers102TestCase to unittest

* Fixing Python type statically

* Shuffle the fake labels

* Update test/test_datasets.py
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Apply the suggestions by pmeier

* Use check_integrity to check file existence

* Save the labels to base_folder

* Minor fixes

* Using a loop makes this more concise without reducing readability
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Using a loop makes this more concise without reducing readability
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* Remove self.labels and self.label_to_index attributes

* minor simplification

* Check the exitence of image folder

* Revert the check

* Check the existence of image folder

* valid -> val

* keep some stuff private

* minor doc arrangements

* remove default FEATURE_TYPES

* Simplify the datasets existence

* check if the image folder exists
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

* isdir -> is_dir
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Nicolas Hug <nicolashug@fb.com>

* Add Country211 dataset

To addresses issue #5108.

* Add Country211 dataset

To addresses issue #5108.

* Update country211.py

* Update country211.py

* Code review reflected

Reflect code review

* Update test_datasets.py

* Update with review

Update with review

* inherit from ImageFolder

* Update test/test_datasets.py

* Docstring + minor test update
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Nicolas Hug <nicolashug@fb.com>

* refactor prototype datasets tests

* skip tests with insufficient third party dependencies