migrate to transformers v5

[omkar-334]

What does this PR do?

This PR migrates the backbone to transformers v5

Resolves #730

Type of Change

• Refactoring (no functional changes)

Testing

All `not gpu` tests pass

• I have tested this change locally

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code where necessary, particularly in hard-to-understand areas
• My changes generate no new warnings or errors
• I have updated the documentation accordingly (if applicable)

[codecov]

Codecov Report

❌ Patch coverage is 96.42857% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 69%. Comparing base (6ee2b3e) to head (0c49b15).
⚠️ Report is 81 commits behind head on develop.

❌ Your project check has failed because the head coverage (69%) is below the target coverage (95%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@          Coverage Diff           @@
##           develop   #760   +/-   ##
======================================
  Coverage       69%    69%           
======================================
  Files           56     56           
  Lines         6966   6983   +17     
======================================
+ Hits          4807   4823   +16     
- Misses        2159   2160    +1     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull request overview

Migrates RF-DETR’s DINOv2 windowed-attention backbone integration to be compatible with Transformers v5, aligning the project’s dependency constraints with the requested transformers>=5.0.0 support (Issue #730).

Changes:

• Updated imports/backbone mixin initialization to match Transformers v5 APIs.
• Removed head_mask plumbing throughout the windowed-attention DINOv2 implementation.
• Updated package dependency constraint to require Transformers v5.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.

File	Description
src/rfdetr/models/backbone/dinov2_with_windowed_attn.py	Refactors the DINOv2 windowed-attention backbone implementation for Transformers v5 API compatibility.
pyproject.toml	Updates the Transformers dependency to >=5.0.0.

Comments suppressed due to low confidence (1)

src/rfdetr/models/backbone/dinov2_with_windowed_attn.py:864

• PR description says this is a refactor with no functional changes, but head_mask was removed from the model forward(...) API and its docstrings. If this is an intentional breaking change required by Transformers v5, consider updating the PR description (and/or release notes) to reflect the public API change.

    def forward(
        self,
        pixel_values: Optional[torch.Tensor] = None,
        bool_masked_pos: Optional[torch.Tensor] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
    ) -> Union[Tuple, BaseModelOutputWithPooling]: