Mirrors/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-09-20 10:35:10 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

allow quantized model to be loaded from model dir (#760)

Browse source
This commit is contained in:
catalpaaa 2023-04-04 19:19:38 -07:00 committed by GitHub
parent ae1fe45bc0
commit 4ab679480e
WARNING! Although there is a key with this ID in the database it does not verify this commit! This commit is SUSPICIOUS.
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 5 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
modules/GPTQ_loader.py
View file

@ -74,7 +74,7 @@ def load_quantized(model_name):
exit()
# Now we are going to try to locate the quantized model file.
path_to_model = Path(f'models/{model_name}')
path_to_model = Path(f'{shared.args.model_dir}/{model_name}')
found_pts = list(path_to_model.glob("*.pt"))
found_safetensors = list(path_to_model.glob("*.safetensors"))
pt_path = None
@ -95,8 +95,8 @@ def load_quantized(model_name):
else:
pt_model = f'{model_name}-{shared.args.wbits}bit'
# Try to find the .safetensors or .pt both in models/ and in the subfolder
for path in [Path(p+ext) for ext in ['.safetensors', '.pt'] for p in [f"models/{pt_model}", f"{path_to_model}/{pt_model}"]]:
# Try to find the .safetensors or .pt both in the model dir and in the subfolder
for path in [Path(p+ext) for ext in ['.safetensors', '.pt'] for p in [f"{shared.args.model_dir}/{pt_model}", f"{path_to_model}/{pt_model}"]]:
if path.exists():
print(f"Found {path}")
pt_path = path

4
modules/models.py
View file

@ -42,7 +42,7 @@ def load_model(model_name):
t0 = time.time()
shared.is_RWKV = 'rwkv-' in model_name.lower()
shared.is_llamacpp = len(list(Path(f'models/{model_name}').glob('ggml*.bin'))) > 0
shared.is_llamacpp = len(list(Path(f'{shared.args.model_dir}/{model_name}').glob('ggml*.bin'))) > 0
# Default settings
if not any([shared.args.cpu, shared.args.load_in_8bit, shared.args.wbits, shared.args.auto_devices, shared.args.disk, shared.args.gpu_memory is not None, shared.args.cpu_memory is not None, shared.args.deepspeed, shared.args.flexgen, shared.is_RWKV, shared.is_llamacpp]):
@ -105,7 +105,7 @@ def load_model(model_name):
elif shared.is_llamacpp:
from modules.llamacpp_model import LlamaCppModel
model_file = list(Path(f'models/{model_name}').glob('ggml*.bin'))[0]
model_file = list(Path(f'{shared.args.model_dir}/{model_name}').glob('ggml*.bin'))[0]
print(f"llama.cpp weights detected: {model_file}\n")
model, tokenizer = LlamaCppModel.from_pretrained(model_file)