biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[V1] Unify VLLM_ENABLE_V1_MULTIPROCESSING handling in RayExecutor (#11472)

Browse Source
This commit is contained in:
Rui Qiao
2024-12-24 19:49:46 -08:00
committed by GitHub
parent 3f3e92e1f2
commit 9832e5572a
3 changed files with 4 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/v1/engine/llm_engine.py
View File

@@ -21,7 +21,6 @@ from vllm.v1.engine.core_client import EngineCoreClient
from vllm.v1.engine.detokenizer import Detokenizer
from vllm.v1.engine.processor import Processor
from vllm.v1.executor.abstract import Executor
from vllm.v1.executor.ray_utils import initialize_ray_cluster
logger = init_logger(__name__)
@@ -112,7 +111,6 @@ class LLMEngine:
distributed_executor_backend = (
vllm_config.parallel_config.distributed_executor_backend)
if distributed_executor_backend == "ray":
initialize_ray_cluster(vllm_config.parallel_config)
from vllm.v1.executor.ray_executor import RayExecutor
executor_class = RayExecutor
elif distributed_executor_backend == "mp":