mmarch/text-generation-webui
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Install exllama (#83)

Browse source 
* Install exllama

* Handle updating exllama
This commit is contained in:
jllllll 2023-06-17 17:10:36 -05:00 • committed by GitHub
parent 657049d7d0
commit b1d05cbbf6
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 10 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

11
webui.py
View file

@ -167,8 +167,17 @@ def update_dependencies():
if not os.path.exists("repositories/"): if not os.path.exists("repositories/"):
os.mkdir("repositories") os.mkdir("repositories")
# Install GPTQ-for-LLaMa which enables 4bit CUDA quantization
os.chdir("repositories") os.chdir("repositories")
# Install or update exllama as needed
if not os.path.exists("exllama/"):
run_cmd("git clone https://github.com/turboderp/exllama.git", environment=True)
else:
os.chdir("exllama")
run_cmd("git pull", environment=True)
os.chdir("..")
# Install GPTQ-for-LLaMa which enables 4bit CUDA quantization
if not os.path.exists("GPTQ-for-LLaMa/"): if not os.path.exists("GPTQ-for-LLaMa/"):
run_cmd("git clone https://github.com/oobabooga/GPTQ-for-LLaMa.git -b cuda", assert_success=True, environment=True) run_cmd("git clone https://github.com/oobabooga/GPTQ-for-LLaMa.git -b cuda", assert_success=True, environment=True)