Add MNv4 medium recipe

[sovrasov]

Summary

• Add recipes for 3 types of classification task. Training parameters are the same as for MNv3_large model
• H-cls task type was moved to datumaro data type in auto configurator. That didn't help to solve the issue with h-cls datumaro datasets: they are still considered as multilabel, but at least autoconfig triggers a warning\
• Bump timm to 1.0.10

How to test

Checklist

• I have added unit tests to cover my changes.​
• I have added integration tests to cover my changes.​
• I have ran e2e tests and there is no issues.
• I have added the description of my changes into CHANGELOG in my target branch (e.g., CHANGELOG in develop).​
• I have updated the documentation in my target branch accordingly (e.g., documentation in develop).
• I have linked related issues.

License

• I submit my code changes under the same Apache License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below).

# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0

🚀 Integration Run Summary

• Last Commit: Move h-cls to datumaro format
• Result: ❓
• Workflow Run: View Run

[eugene123tw]

/run multi_class_cls

[sovrasov]

Anomalib is incompatible with timm version containing MNv4. PR is on hold until anomaly engine refactoring is done