chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

[FEAT] Shared mem pool based cuda ipc for multi-modal data transport (#11917)

Browse Source 
Co-authored-by: kousakawang <wanghanpei@bytedance.com>
Co-authored-by: Yuan Luo <4908075+yuan-luo@users.noreply.github.com>
This commit is contained in:
kousakawang
2025-11-02 16:46:37 +08:00
committed by GitHub
parent a920b9dace
commit 7efd8b3d1f
8 changed files with 424 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
python/sglang/srt/server_args.py
View File

@@ -3889,6 +3889,11 @@ def set_global_server_args_for_scheduler(server_args: ServerArgs):
_global_server_args = server_args
def set_global_server_args_for_tokenizer(server_args: ServerArgs):
global _global_server_args
_global_server_args = server_args
def get_global_server_args() -> ServerArgs:
if _global_server_args is None:
raise ValueError("Global server args is not set yet!")