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

[Bugfix]v0.18.0 support FlashComm1 & DCP for Qwen (#7726)

Browse Source 
### What this PR does / why we need it?
This PR backports the changes from #7673 ([Bugfix] support FlashComm1 &
DCP for Qwen) to the releases/v0.18.0 branch.

--------
Signed-off-by: Yang Yuxi <907276627@qq.com>
This commit is contained in:
Yang Yuxi
2026-03-29 15:59:19 +08:00
committed by GitHub
parent 9cc41c9457
commit e776d5c0f1
2 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/e2e/multicard/4-cards/long_sequence/test_basic.py
View File

@@ -236,7 +236,8 @@ def test_dcp_basic():
decode_context_parallel_size=2,
max_num_batched_tokens=1024,
enable_expert_parallel=True,
block_size=128) as runner:
block_size=128,
compilation_config={"pass_config": {"enable_sp": True}}) as runner:
runner.model.generate(prompts, sampling_params)