Merge release to develop branch

[kprokofi]

Summary

How to test

Checklist

• The PR title and description are clear and descriptive
• I have manually tested the changes
• All changes are covered by automated tests
• All related issues are linked to this PR (if applicable)
• Documentation has been updated (if applicable)

[Copilot]

Pull Request Overview

A routine merge from release to develop branch containing various bug fixes and improvements across multiple components including batch size adaptations, transform pipeline updates, keypoint detection enhancements, and version updates.

• Updated default batch sizes from 8 to 4 for training and validation subsets
• Added keypoint detection support with new test cases and improved pre-filtering logic
• Modified adaptive batch size algorithms with stricter memory bounds and XPU-specific optimizations

Reviewed Changes

Copilot reviewed 28 out of 28 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
library/tests/unit/tools/test_converter.py	Updated test assertions for new default batch size (4 instead of 8)
library/tests/unit/data/transform_libs/test_torchvision.py	Added segmentation data fixture and transform test for RandomAffine
library/tests/unit/data/test_pre_filtering.py	Enhanced test coverage for keypoint detection validation with Points annotations
library/tests/unit/backend/native/tools/adaptive_bs/test_bs_search_algo.py	Updated memory thresholds and expected batch size assertions
library/tests/assets/geti/model_configs/detection.yaml	Added batch_size parameter to hyperparameters section
library/src/otx/tools/converter.py	Added update_batch_size function for configuration management
library/src/otx/recipe/detection/*.yaml	Reordered transform pipelines and enabled RandomFlip transforms
library/src/otx/data/utils/pre_filtering.py	Enhanced validation logic for keypoint detection with Points annotations
library/src/otx/data/transform_libs/torchvision.py	Improved RandomAffine transform handling for masks and polygons
library/src/otx/data/module.py	Simplified pre-filtering condition for keypoint detection
library/src/otx/backend/native/tools/adaptive_bs/*.py	Updated memory bounds and XPU-specific batch size handling
library/src/otx/backend/native/models/detection/*.py	Enhanced DFINE model configuration and checkpoint loading
library/src/otx/backend/native/engine.py	Added CPU precision configuration
library/src/otx/backend/native/callbacks/batchsize_finder.py	Increased steps per trial and adjusted XPU epochs
library/src/otx/init.py	Version bump to 2.7.0dev
library/pyproject.toml	Updated setuptools and onnxconverter-common versions
CHANGELOG.md	Updated changelog for v2.6.0 release
.github/workflows/publish.yaml	Fixed build paths and added build tools installation

Comments suppressed due to low confidence (1)

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[github-actions]

Docker Image Sizes

Image	Size
geti-tune-pr-4903	1.5G
geti-tune-sha-38da436	1.5G

[github-actions]

📊 Test coverage report

Metric	Coverage
Lines	39.5%
Functions	36.2%
Branches	84.5%
Statements	39.5%

[leoll2]

LGTM, could you please summarize the changes (or link the release PRs) for the record?