Fix param mapping for Qwen2

[ChingTsai]

Description

• Resolves param_mapping.py conflicts introduced in this commit.
  • Add back the mapping for Qwen2.5-7b and Qwen2.5-14b

Tests

end2end test for qwen2.5-7b

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

❌ Patch coverage is 18.18182% with 9 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
...xtext/checkpoint_conversion/utils/param_mapping.py	18.18%	9 Missing ⚠️

📢 Thoughts on this report? Let us know!

[github-actions]

🤖 Hi @RissyRan, I've received your request, and I'm working on it now! You can track my progress in the logs for more details.

[RissyRan]

Thanks!

[github-actions]

## 📋 Review Summary

This PR successfully fixes the parameter mapping for Qwen2 models by restoring support for qwen2.5-7b and qwen2.5-14b. It also introduces necessary Q/K/V bias hooks and cleanly renames the internal mapping functions to be more general (QWEN_ instead of QWEN3_).

🔍 General Feedback

• The addition of reshape_bias perfectly complements the new bias hooks for Qwen attention layers.
• Consider removing the redundant conditional in the reshape_bias function to keep the code DRY, as the underlying reshape operation inherently handles both directions correctly.