[0.11.0][Bugfix] fix e2e full test

[zhangxinyuehfad]

What this PR does / why we need it?

pin Transformer version to 4.57.1 fix 'dict' object has no attribute 'model_type'
https://github.com/vllm-project/vllm-ascend/actions/runs/19660859460/job/56306822464

picked from #4423

Does this PR introduce any user-facing change?

How was this patch tested?

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

[gemini-code-assist]

Code Review

This pull request fixes an AttributeError in the test_fused_moe E2E test by mocking is_vl_model. The approach is correct. I've suggested a small refactoring to make the patching cleaner by using string paths, which avoids unnecessary imports. Additionally, the similar test test_token_dispatcher_with_all_gather_quant appears to be unpatched and may suffer from the same issue. It would be good to apply the same fix to that test as well to ensure all tests pass reliably.