[Feature] Enable inference support for Deepseekr1-w8a8-MTP

[Irving11-BKN]

Support the inference of the Deepseekr1-w8a8-mtp model with statically-quantized shared_head in MTP layers.

cherry-pick from #1584

Signed-off-by: curryliu [email protected]

• vLLM version: v0.9.2
• vLLM main: vllm-project/vllm@ca4eb82

[codecov]

Codecov Report

Attention: Patch coverage is 33.33333% with 12 lines in your changes missing coverage. Please review.

Project coverage is 60.40%. Comparing base (c30ddb8) to head (a67bc73).
Report is 163 commits behind head on main.

Files with missing lines	Patch %	Lines
vllm_ascend/quantization/quant_config.py	33.33%	6 Missing ⚠️
vllm_ascend/models/deepseek_mtp.py	42.85%	4 Missing ⚠️
vllm_ascend/models/deepseek_v2.py	0.00%	2 Missing ⚠️

❌ Your patch check has failed because the patch coverage (33.33%) is below the target coverage (100.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #1834       +/-   ##
===========================================
+ Coverage   27.39%   60.40%   +33.00%     
===========================================
  Files          56       72       +16     
  Lines        6191     8117     +1926     
===========================================
+ Hits         1696     4903     +3207     
+ Misses       4495     3214     -1281     

Flag	Coverage Δ
unittests	60.40% <33.33%> (+33.00%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[wangxiyuan]

why only we need prefix here？The behavior of w8a8 weight is different with others?

[Irving11-BKN]

prefix is used for the function is_layer_skipped in quant_config.py file, the behavior is the same