[Refactor] 1/N Refactor attention_v1 & extract attention_cp (#4628)

Browse Source

RFC: https://github.com/vllm-project/vllm-ascend/issues/4629
Reason：
The functions related to Cp differ significantly from those of normal
Attention, but the coupling is quite severe.

Steps：
Isolate PCP and DCP
(1) Forward class extraction (100%)
(2) Metadata coupling processing
(3) Builder processing

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

---------

Signed-off-by: weijinqian_v1 <weijinqian@huawei.com>
Co-authored-by: weijinqian_v1 <weijinqian@huawei.com>

...

This commit is contained in:

weijinqian0

2025-12-06 09:33:28 +08:00

committed by GitHub

parent 3740b3edfc

commit a78f49ea57

2 changed files with 1001 additions and 830 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

916

vllm_ascend/attention/attention_v1.py

File diff suppressed because it is too large Load Diff