Results 8 issues of Vinnam Kim

### Changes - Generate HTML report for ONNX e2e test - Add model accuracy degradation acceptance criteria for ONNX e2e test ### Reason for changes To keep alignment with other...

ready for reviews

- Weighted loss function is needed for the dataset which has patially annotated keypoints Signed-off-by: Vinnam Kim

- There was a bug when obtaining padding of the cropped bounding box Signed-off-by: Vinnam Kim

### Changes Add a new mixed-precision QAT algorithm, FracBits [[paper]](https://arxiv.org/abs/2007.02017) and [[code]](https://github.com/deJQK/FracBits) as an experimental feature. ### Reason for changes To expand the choice of mixed-precision QAT algorithms for users....

documentation
NNCF PT
NNCF Common
experimental

### Summary - Ticket no. 128949 - Stack images as possible for all tasks - Pad images during instance segmentation data pipeline as possible - Add non_blocking option to instance...

OTX 2.0

This is a minimal reproduction. If I have this Python code which implements `jsonargparse.ArgumentParser` ```python # test.py from lightning.pytorch.cli import LRSchedulerCallable, ReduceLROnPlateau from jsonargparse import ArgumentParser, ActionConfigFile class Model: def...

bug

### 🐛 Describe the bug Please see this line on the main branch. https://github.com/pytorch/vision/blob/5181a854d8b127cf465cd22a67c1b5aaf6ccae05/torchvision/transforms/v2/functional/_geometry.py#L200 It decides whether the image resize can be done for uint8 datatype with native CPU. However,...

### Summary - Revert https://github.com/openvinotoolkit/training_extensions/pull/3291 and upgrade `jsonargparse` to `4.28.0`. ### How to test ### Checklist - [ ] I have added unit tests to cover my changes.​ - [...

BUILD
OTX 2.0