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

Use separate llama-cpp-python packages for GGML support

Browse source
This commit is contained in:
jllllll 2023-08-26 09:15:11 -05:00
parent 6e6431e73f
commit 4a999e3bcd
No known key found for this signature in database
GPG key ID: 7FCD00C417935797
7 changed files with 74 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

4
modules/models_settings.py
View file

@ -24,9 +24,9 @@ def infer_loader(model_name):
loader = None
elif Path(f'{shared.args.model_dir}/{model_name}/quantize_config.json').exists() or ('wbits' in model_settings and type(model_settings['wbits']) is int and model_settings['wbits'] > 0):
loader = 'AutoGPTQ'
elif len(list(path_to_model.glob('*.gguf*'))) > 0:
elif len(list(path_to_model.glob('*.gguf*')) + list(path_to_model.glob('*ggml*.bin'))) > 0:
loader = 'llama.cpp'
elif re.match(r'.*\.gguf', model_name.lower()):
elif re.match(r'.*\.gguf|.*ggml.*\.bin', model_name.lower()):
loader = 'llama.cpp'
elif re.match(r'.*rwkv.*\.pth', model_name.lower()):
loader = 'RWKV'