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

[DP][V1] Fix rank set in DP scenario & Bump torch-npu version to 2.5.1.post1.dev20250528 (#1235)

Browse Source 
### What this PR does / why we need it?
1. Fix rank set in DP scenario. The new poc version of torch-npu support
setting `ASCEND_RT_VISIBLE_DEVICES` dynamically, thus we could use the
rank set in `DPEngineCoreProc` directly instead of calculating local
rank across dp by hand in the patched `_init_data_parallel`

Closes: https://github.com/vllm-project/vllm-ascend/issues/1170

2. Bump torch-npu version to 2.5.1.post1.dev20250528

Closes: https://github.com/vllm-project/vllm-ascend/pull/1242
Closes: https://github.com/vllm-project/vllm-ascend/issues/1232


### How was this patch tested?
CI passed with new added test.

---------

Signed-off-by: MengqingCao <cmq0113@163.com>
Signed-off-by: Icey <1790571317@qq.com>
Co-authored-by: Icey <1790571317@qq.com>
This commit is contained in:
Mengqing Cao
2025-06-16 23:09:53 +08:00
committed by GitHub
parent f5404dc650
commit 96fa7ff63b
19 changed files with 114 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
vllm_ascend/patch/__init__.py
View File

@@ -47,16 +47,7 @@
# Related PR (if no, explain why):
# Future Plan:
# Remove those patch when vllm merged them
# 2. `vllm.v1.engine.core.DPEngineCoreProc._init_data_parallel`
# Why:
# There is some bug for ASCEND_RT_VISIBLE_DEVICES usage.
# How：
# The ASCEND_RT_VISIBLE_DEVICES related code is dropped.
# Related PR (if no, explain why):
# No, this is a bug for vllm ascend
# Future Plan:
# Remove this patch once ASCEND_RT_VISIBLE_DEVICES bug is fixed.
# 3. `vllm.config.ParallelConfig.get_next_dp_init_port`
# 2. `vllm.config.ParallelConfig.get_next_dp_init_port`
# Why:
# vllm doesn't support get port from environment.
# How：
@@ -65,7 +56,7 @@
# Need a PR to vllm to support get port from environment.
# Future Plan:
# Remove those patch when vllm merged them
# 4. `vllm.config.ParallelConfig.ParallelConfig.stateless_init_dp_group`
# 3. `vllm.config.ParallelConfig.ParallelConfig.stateless_init_dp_group`
# Why:
# vLLM use gloo backend by default to initialize stateless dp process gourp, but we want to use hccl here to
# get better performance