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

[Misc] Remove deprecated code (#12383)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-01-25 03:45:20 +08:00
committed by GitHub
parent ab5bbf5ae3
commit df5dafaa5b
6 changed files with 25 additions and 78 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
vllm/config.py
View File

@@ -1227,9 +1227,6 @@ class ParallelConfig:
pipeline_parallel_size: int = 1 # Number of pipeline parallel groups.
tensor_parallel_size: int = 1 # Number of tensor parallel groups.
# Deprecated, use distributed_executor_backend instead.
worker_use_ray: Optional[bool] = None
# Maximum number of multiple batches
# when load model sequentially. To avoid RAM OOM when using tensor
# parallel and large models.
@@ -1283,13 +1280,6 @@ class ParallelConfig:
self.world_size = self.pipeline_parallel_size * \
self.tensor_parallel_size
if self.worker_use_ray:
if self.distributed_executor_backend is None:
self.distributed_executor_backend = "ray"
elif not self.use_ray:
raise ValueError(f"worker-use-ray can't be used with "
f"distributed executor backend "
f"'{self.distributed_executor_backend}'.")
ray_only_devices = ["tpu"]
from vllm.platforms import current_platform
if (current_platform.device_type in ray_only_devices