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

[dist] revert communicator patch (#66)

Browse Source 
### What this PR does / why we need it?
Revert communicator patch as
https://github.com/vllm-project/vllm/pull/13208 has been merged.

### Does this PR introduce _any_ user-facing change?
N/A

### How was this patch tested?
test locally by
https://github.com/vllm-project/vllm-ascend/pull/30#issuecomment-2650251266

Signed-off-by: MengqingCao <cmq0113@163.com>
This commit is contained in:
Mengqing Cao
2025-02-17 11:42:33 +08:00
committed by GitHub
parent bfbfbce184
commit 4544e99d88
4 changed files with 14 additions and 145 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm_ascend/worker.py
View File

@@ -457,8 +457,6 @@ def init_worker_distributed_environment(
backend: str = "hccl") -> None:
"""Initialize the distributed environment."""
set_custom_all_reduce(not parallel_config.disable_custom_all_reduce)
# register communicator patch before init dist env
from vllm_ascend import patch # noqa: F401
init_distributed_environment(parallel_config.world_size, rank,
distributed_init_method, local_rank, backend)