EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Fusion] [Graph] Add qknorm rope fusion operator (#4711)

Browse Source 
### What this PR does / why we need it?
This PR add `qkv_rmsnorm_rope` operator and introduces a graph fusion
pass for `qknorm_rope` operations. The implementation includes a new
configuration flag, a pattern matching pass using
`torch._inductor.pattern_matcher`, and a custom Triton kernel for the
fused operation.

Co-authored-by: Angazenn
[supperccell@163.com](mailto:supperccell@163.com)

### Does this PR introduce _any_ user-facing change?
Yes, add new additional_config

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

---------

Signed-off-by: wxsIcey <1790571317@qq.com>
This commit is contained in:
Icey
2025-12-17 08:53:44 +08:00
committed by GitHub
parent b1a853b0f6
commit cadfa5ddc1
14 changed files with 754 additions and 71 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/e2e/singlecard/test_aclgraph_accuracy.py
View File

@@ -210,4 +210,4 @@ def test_aclgraph_enable():
# after check_and_update_config, mode should be VLLM_COMPILE and piecewise cudagraph
NPUPlatform.check_and_update_config(VllmConfig)
assert VllmConfig.compilation_config.mode == CompilationMode.VLLM_COMPILE
assert VllmConfig.compilation_config.cudagraph_mode == CUDAGraphMode.PIECEWISE
assert VllmConfig.compilation_config.cudagraph_mode == CUDAGraphMode.PIECEWISE