GitLab

* Avoid freezing bn1 if all layers are trainable.

* Remove misleading comments.

* Adding TODO placeholders.

* More placeholders.

* Add MobileNetV3 small pre-trained weights.

* Remove placeholders.

* Refactoring mobilenetv3 to make code reusable.

* Adding quantizable MobileNetV3 architecture.

* Fix bug on reference script.

* Moving documentation of quantized models in the right place.

* Update documentation.

* Workaround for loading correct weights of quant model.

* Update weight URL and readme.

* Adding eval.

Disable pretrained backbone downloading if pretrained is True
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Document undodcumented parameters

* remove setup.cfg changes

* Properly pass normalize down instead of deprecating it

* Fix flake8

* Add new CI check

* Fix type spec

* Leave normalize be part of kwargs
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Making _segm_resnet() generic and reusable.

* Adding fcn and deeplabv3 directly on mobilenetv3 backbone.

* Adding tests for segmentation models.

* Rename is_strided with _is_cn.

* Add dilation support on MobileNetV3 for Segmentation.

* Add Lite R-ASPP with MobileNetV3 backbone.

* Add pretrained model weights.

* Removing model fcn_mobilenet_v3_large.

* Adding docs and imports.

* Fixing typo and readme.

Removes deprecation warning

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

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

* Fixed warning.

UserWarning: The default behavior for interpolate/upsample with float scale_factor changed in 1.6.0 to align with other frameworks/libraries, and now uses scale_factor directly, instead of relying on the computed output size. If you wish to restore the old behavior, please set recompute_scale_factor=True. See the documentation of nn.Upsample for details.

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

* Tag fasterrcnn mobilenetv3 model with 320, add new inference config that makes it 2x faster sacrificing a bit of mAP.

* Add a high resolution fasterrcnn mobilenetv3 model.

* Update tests and expected values.

* Simplify code and remove used vars.

* Simplify expressions and remove used parenthesis.

* Jit fixes.

* Making check more readable.

* fixing styles

* Minor refactoring of a private method to make it reusuable.

* Adding a FasterRCNN + MobileNetV3 with & w/o FPN models.

* Reducing Resolution to 320-640 and anchor sizes to 16-256.

* Increase anchor sizes.

* Adding rpn score threshold param on the train script.

* Adding trainable_backbone_layers param on the train script.

* Adding rpn_score_thresh param directly in fasterrcnn_mobilenet_v3_large_fpn.

* Remove fasterrcnn_mobilenet_v3_large prototype and update expected file.

* Update documentation and adding weights.

* Use buildin Identity.

* Fix spelling.

* Add MobileNetV3 Architecture in TorchVision (#3182)

* Adding implementation of network architecture

* Adding rmsprop support on the train.py

* Adding auto-augment and random-erase in the training scripts.

* Adding support for reduced tail on MobileNetV3.

* Tagging blocks with comments.

* Adding documentation, pre-trained model URL and a minor refactoring.

* Handling better untrained supported models.

* Introduce small score threshold on rpn

* Adding docs and fixing keypoint and mask.

* Making value 0.0 by default for BC.

* Fixing for onnx.

* Update threshold.

* Removing non-default threshold from reference scripts.
Co-authored-by: Francisco Massa <fvsmassa@gmail.com>

* Fixing trainable_layers bug.

* minor doc fixes

* remove unused imports after manual review

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

* remove two more instances
Co-authored-by: Ben Weinstein <benweinstein@Bens-MacBook-Pro.local>
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

* Patches required for FBCode merge.

* Patching quantization model imports.

* import QuantizableMobileNetV2

* Adding newline to avoid lit errors

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

* Moving mobilenet.py to mobilenetv2.py

* Adding mobilenet.py for BC.

* Extending ConvBNReLU for reuse.

* Reduce import scope on mobilenet to only the public and versioned classes and methods.

* Fixing incorrect doc example in MNASNet.

* Fixing incorrect output.

* Replacing all torch.jit.annotations with typing

* Replacing remaining typing

* Fix an issue that ShuffleNetV2 model is exported to a wrong ONNX file if dynamic_axes field was provided.

* Add a ut for the bug fix.

* Fix flake8 issue.

* Don't access each element in x.shape, use x.size() instead.
Co-authored-by: Vasilis Vryniotis <datumbox@users.noreply.github.com>

Replace tabs with spaces, add newlines to files and replace whitelist with allowlist

* Correcting incorrect types

* Add missing type statement

* Fix type annotations in unittest

* Fix TypeError

* Fix TypeError

* Fix type equality judgment

* Fix recursive compile

* Use string for class name annotation.
Co-authored-by: zhiqiang <zhiqwang@outlook.com>

* Enable support for images without annotations

* Ensuring gradient propagates to RegressionHead.

* Rewriting losses to remove branching.

* Fix the seed on DeformConv autocast test.

Adding Python type hints, correcting incorrect types, removing unnecessary vars and simplifying code. (#3045)

* refactor: Fixed typing exception throwing issues with JIT

* style: Added back mypy typing to forward

* chore: Added back densenet module as mypy exception

* Fix MNASNet docstrings so they are rendered correctly

* Add dot after url link in models docstrings for consistency

* Making quantized inception torchscriptable.

* Adding a test.

* Fix mypy warning.

* Remove model download from tests

* Refactor trainable_layers checks in detection models

* Bugfix

* Finish tests and fixes

* Fix lint

* style: Added annotation typing for shufflenet

* fix: Removed duplicate type hint

* refactor: Removed un-necessary import

* fix: Fixed constructor typing

* style: Added black formatting on depthwise_conv

* style: Fixed stage typing in shufflenet

* style: Added annotation typing for densenet

* fix: Fixed import

* refactor: Removed un-necessary import

* fix: Fixed constructor typing

* chore: Updated mypy.ini

* fix: Fixed tuple typing

* style: Ignored some mypy errors

* style: Fixed typing

* fix: Added missing constructor typing

* changes to value error

* fixes it :)

* simpler fix

* Overwriting FrozenBN eps=0.0 if pretrained=True for detection models.

* Moving the method to detection utils and adding comments.