baalajimaestro/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Update n_batch for server

Browse source
This commit is contained in:
Andrei Betlen 2023-04-25 09:11:32 -04:00
parent cc706fb944
commit 3cab3ef4cb
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
llama_cpp/server/__main__.py
View file

@ -28,7 +28,7 @@ from sse_starlette.sse import EventSourceResponse
class Settings(BaseSettings):
model: str
n_ctx: int = 2048
n_batch: int = 8
n_batch: int = 512
n_threads: int = max((os.cpu_count() or 2) // 2, 1)
f16_kv: bool = True
use_mlock: bool = False # This causes a silent failure on platforms that don't support mlock (e.g. Windows) took forever to figure out...