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

Update llama.py

Browse source 
update n_batch default to 512 to match upstream llama.cpp
This commit is contained in:
eiery 2023-04-23 20:56:40 -04:00 committed by GitHub
parent e99caedbbd
commit aa12d8a81f
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
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/llama.py
View file

@ -37,7 +37,7 @@ class Llama:
use_mlock: bool = False, use_mlock: bool = False,
embedding: bool = False, embedding: bool = False,
n_threads: Optional[int] = None, n_threads: Optional[int] = None,
n_batch: int = 8, n_batch: int = 512,
last_n_tokens_size: int = 64, last_n_tokens_size: int = 64,
lora_base: Optional[str] = None, lora_base: Optional[str] = None,
lora_path: Optional[str] = None, lora_path: Optional[str] = None,