update to vllm 12-19

[leo-pony]

What this PR does / why we need it?

Does this PR introduce any user-facing change?

Fix vllm break:

1. [Enable cuda graph for deepepHT, 5.3% throughput improvement, 4.4% TTFT improvement] ([Perf] Enable cuda graph for deepepHT, 5.3% throughput improvement, 4.4% TTFT improvement vllm#29558)
   Fix Solution: Add the now-necessary all2all_backend parameter. The impact of this parameter on the original set_splitting_ops_for_v1 implementation is only that graph mode is disabled in vllm if deepep_high_throughput is enabled; it has no effect on the vllm-ascend logic.

2.[Migrate legacy ViT MultiHeadAttention to new MMEncoderAttention interface ] (vllm-project/vllm#30684)
Fix Solution: The reason why the GPU does not need to convert qkv to 3D is that the GPU's flash_attention operator is compatible with 3D and 4D (b s h d and s b ( h d)), but the NPU's flash_attention_unpad operator only supports 3D (s b ( h d)). Therefore, we need to introduce the reshape_qkv_to_3d operation.

4.Skip Tencent-Hunyuan/HunyuanOCR test case, as it has following issue in upgrade vllm code:
#5297

How was this patch tested?

Co-authored-by: zxwang 1476209578@qq.com

• vLLM version: release/v0.13.0
• vLLM main: vllm-project/vllm@ad32e3e

[gemini-code-assist]

Code Review

This pull request updates the versioning policy documentation to include a specific vLLM commit hash for compatibility with the main branch. While this adds precision, the pull request description is empty. More importantly, the corresponding Chinese translation file has not been updated, leading to inconsistent documentation. This should be addressed to ensure all users have correct information.

[gemini-code-assist]

The compatibility matrix has been updated, but the corresponding Chinese translation file (docs/source/locale/zh_CN/LC_MESSAGES/community/versioning_policy.po) was not updated. This will cause a discrepancy and provide outdated information to users relying on the Chinese documentation. Please update the translation files to ensure consistency across all supported languages.

[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.