GitLab

* add ufmt as code formatter

* cleanup

* quote ufmt requirement

* split imports into more groups

* regenerate circleci config

* fix CI

* clarify local testing utils section

* use ufmt pre-commit hook

* split relative imports into local category

* Revert "split relative imports into local category"

This reverts commit f2e224cde2008c56c9347c1f69746d39065cdd51.

* pin black and usort dependencies

* fix local test utils detection

* fix ufmt rev

* add reference utils to local category

* fix usort config

* remove custom categories sorting

* Run pre-commit without fixing flake8

* got a double import in merge
Co-authored-by: Nicolas Hug <nicolashug@fb.com>

* draft commit

* Polish and add corresponding test

* Update docs

* Update torchvision/models/feature_extraction.py

* Update docs/source/feature_extraction.rst
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Moving _make_divisible to utils.

* Replace the old ConvBNReLU and ConvBNActivation layers

* Fix minor bug.

* Moving SE layer to ops.

* Adding deprecation warnings on old layers.

* Apply changes to regnets.

* initial code

* add SqueezeExcitation

* initial code

* add SqueezeExcitation

* add SqueezeExcitation

* regnet blocks, stems and model definition

* nit

* add fc layer

* use Callable instead of Enum for block, stem and activation

* add regnet_x and regnet_y model build functions, add docs

* remove unused depth

* use BN/activation constructor and ConvBNActivation

* add expected test pkl files

* allow custom activation in SqueezeExcitation

* use ReLU as the default activation

* initial code

* add SqueezeExcitation

* initial code

* add SqueezeExcitation

* add SqueezeExcitation

* regnet blocks, stems and model definition

* nit

* add fc layer

* use Callable instead of Enum for block, stem and activation

* add regnet_x and regnet_y model build functions, add docs

* remove unused depth

* use BN/activation constructor and ConvBNActivation

* reuse SqueezeExcitation from efficientnet

* refactor RegNetParams into BlockParams

* use nn.init, replace np with torch

* update README

* construct model with stem, block, classifier instances

* Revert "construct model with stem, block, classifier instances"

This reverts commit 850f5f3ed01a2a9b36fcbf8405afd6e41d2e58ef.

* remove unused blocks

* support scaled model

* fuse into ConvBNActivation

* make reset_parameters private

* fix type errors

* fix for unit test

* add pretrained weights for 6 variant models, update docs

* Reuse EfficientNet SE layer.

* Deprecating the mobilenetv3.SqueezeExcitation layer.

* Passing the right activation on quantization.

* Making strict named param.

* Set default params if missing.

* Fixing typos.

* allow custom activation in SqueezeExcitation

* use ReLU as the default activation

* make scale activation parameterizable
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Allow gradient backpropagation through GeneralizedRCNNTransform to inputs
Signed-off-by: Beat Buesser <beat.buesser@ie.ibm.com>

* Add unit tests for gradient backpropagation to inputs
Signed-off-by: Beat Buesser <beat.buesser@ie.ibm.com>

* Update torchvision/models/detection/transform.py
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Update _check_input_backprop
Signed-off-by: Beat Buesser <beat.buesser@ie.ibm.com>

* Account for tests requiring cuda
Signed-off-by: Beat Buesser <beat.buesser@ie.ibm.com>
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

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

* Added paper references to detection models

* Ignore linter warning

* Break long line into two

* 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>

* 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>

* fix

* add typings

* fixup some more types

* Type more

* remove mypy ignore

* add missing typings

* fix a few mypy errors

* fix mypy errors

* fix mypy

* ignore types

* fixup annotation

* fix remaining types

* cleanup #TODO comments
Co-authored-by: Philip Meier <github.pmeier@posteo.de>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Adding code skeleton

* Adding MBConvConfig.

* Extend SqueezeExcitation to support custom min_value and activation.

* Implement MBConv.

* Replace stochastic_depth with operator.

* Adding the rest of the EfficientNet implementation

* Update torchvision/models/efficientnet.py

* Replacing 1st activation of SE with SiLU.

* Adding efficientnet_b3.

* Replace mobilenetv3 assets with custom.

* Switch to standard sigmoid and reconfiguring BN.

* Reconfiguration of efficientnet.

* Add repr

* Add weights.

* Update weights.

* Adding B5-B7 weights.

* Update docs and hubconf.

* Fix doc link.

* Fix typo on comment.

* style: Added typing to models/video

* style: Fixed typing

* style: Fixed typing

* style: Fixed typing

* refactor: Removed default value for stem

* docs: Fixed docstring of VideoResNet

* style: Refactored typing

* docs: Fixed docstring

* style: Fixed typing

* docs: Specified docstring

* typing: Fixed tying

* docs: Fixed docstring

* Undoing change.

* style: Added typing annotations to segmentation/_utils

* style: Added typing annotations to segmentation/segmentation

* style: Added typing annotations to remaining segmentation models

* style: Fixed typing of DeepLab

* style: Fixed typing

* fix: Fixed typing annotations & default values

* Fixing python_type_check

using nn.init.trunc_normal_ instead of scipy.stats.truncnorm
Co-authored-by: Vincent Moens <vmoens@fb.com>

* [doc] add minimum input size for alexnet builder

* [doc] add minimum input size for vgg builder

* [doc] add minimum input size for squeezenet builder

* [doc] add minimum input size for densenet builder

* [doc] add minimum input size for inception_v3 builder

* [doc] add minimum input size for googlenet builder
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Fix a bug when trainable_layers == 0

* Fix same issue on ssd.

* Moving tensors to the right device.

* Switch to gpu.medium

* Remove incorrect params from doc and add references to the paper.

* Add paper links in doc.

* Converting private parameters to public.

* Add kwargs to handle extra params.

* Add another kwargs.

* Add arguments in _mobilenet_extractor.

* Partial implementation of SSDlite.

* Add normal init and BN hyperparams.

* Refactor to keep JIT happy

* Completed SSDlite.

* Fix lint

* Update todos

* Add expected file in repo.

* Use C4 expansion instead of C4 output.

* Change scales formula for Default Boxes.

* Add cosine annealing on trainer.

* Make T_max count epochs.

* Fix test and handle corner-case.

* Add support of support width_mult

* Add ssdlite presets.

* Change ReLU6, [-1,1] rescaling, backbone init & no pretraining.

* Use _reduced_tail=True.

* Add sync BN support.

* Adding the best config along with its weights and documentation.

* Make mean/std configurable.

* Fix not implemented for half exception

* Refactor grid default boxes with torch.meshgrid

* Fix torch jit tracing

* Only doing the list multiplication once
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Make grid_default_box private as suggested
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Replace list multiplication with torch.repeat

* Move the clipping into _grid_default_boxes to accelerate
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Early skeleton of API.

* Adding MultiFeatureMap and vgg16 backbone.

* Making vgg16 backbone same as paper.

* Making code generic to support all vggs.

* Moving vgg's extra layers a separate class + L2 scaling.

* Adding header vgg layers.

* Fix maxpool patching.

* Refactoring code to allow for support of different backbones & sizes:
- Skeleton for Default Boxes generator class
- Dynamic estimation of configuration when possible
- Addition of types

* Complete the implementation of DefaultBox generator.

* Replace randn with empty.

* Minor refactoring

* Making clamping between 0 and 1 optional.

* Change xywh to xyxy encoding.

* Adding parameters and reusing objects in constructor.

* Temporarily inherit from Retina to avoid dup code.

* Implement forward methods + temp workarounds to inherit from retina.

* Inherit more methods from retinanet.

* Fix type error.

* Add Regression loss.

* Fixing JIT issues.

* Change JIT workaround to minimize new code.

* Fixing initialization bug.

* Add classification loss.

* Update todos.

* Add weight loading support.

* Support SSD512.

* Change kernel_size to get output size 1x1

* Add xavier init and refactoring.

* Adding unit-tests and fixing JIT issues.

* Add a test for dbox generator.

* Remove unnecessary import.

* Workaround on GeneralizedRCNNTransform to support fixed size input.

* Remove unnecessary random calls from the test.

* Remove more rand calls from the test.

* change mapping and handling of empty labels

* Fix JIT warnings.

* Speed up loss.

* Convert 0-1 dboxes to original size.

* Fix warning.

* Fix tests.

* Update comments.

* Fixing minor bugs.

* Introduce a custom DBoxMatcher.

* Minor refactoring

* Move extra layer definition inside feature extractor.

* handle no bias on init.

* Remove fixed image size limitation

* Change initialization values for bias of classification head.

* Refactoring and update test file.

* Adding ResNet backbone.

* Minor refactoring.

* Remove inheritance of retina and general refactoring.

* SSD should fix the input size.

* Fixing messages and comments.

* Silently ignoring exception if test-only.

* Update comments.

* Update regression loss.

* Restore Xavier init everywhere, update the negative sampling method, change the clipping approach.

* Fixing tests.

* Refactor to move the losses from the Head to the SSD.

* Removing resnet50 ssd version.

* Adding support for best performing backbone and its config.

* Refactor and clean up the API.

* Fix lint

* Update todos and comments.

* Adding RandomHorizontalFlip and RandomIoUCrop transforms.

* Adding necessary checks to our tranforms.

* Adding RandomZoomOut.

* Adding RandomPhotometricDistort.

* Moving Detection transforms to references.

* Update presets

* fix lint

* leave compose and object

* Adding scaling for completeness.

* Adding params in the repr

* Remove unnecessary import.

* minor refactoring

* Remove unnecessary call.

* Give better names to DBox* classes

* Port num_anchors estimation in generator

* Remove rescaling and fix presets

* Add the ability to pass a custom head and refactoring.

* fix lint

* Fix unit-test

* Update todos.

* Change mean values.

* Change the default parameter of SSD to train the full VGG16 and remove the catch of exception for eval only.

* Adding documentation

* Adding weights and updating readmes.

* Update the model weights with a more performing model.

* Adding doc for head.

* Restore import.

* Refactored set_cell_anchors() in AnchorGenerator

* Addressed review comment

* Fixed test failure