Mirrors/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-12-24 13:28:59 +01:00
Code Issues Packages Projects Releases Wiki Activity

Replace ggml occurences with gguf

Browse Source
This commit is contained in:
oobabooga 2023-08-26 01:06:59 -07:00
parent 1a642c12b5
commit 83640d6f43
5 changed files with 12 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -156,7 +156,7 @@ text-generation-webui
In the "Model" tab of the UI, those models can be automatically downloaded from Hugging Face. You can also download them via the command-line with `python download-model.py organization/model`.
* GGML models are a single file and should be placed directly into `models`. Example:
* GGUF models are a single file and should be placed directly into `models`. Example:
```
text-generation-webui
@ -258,7 +258,7 @@ Optionally, you can use the following command-line flags:
| `--quant_type QUANT_TYPE` | quant_type for 4-bit. Valid options: nf4, fp4. |
| `--use_double_quant` | use_double_quant for 4-bit. |
#### GGML (for llama.cpp and ctransformers)
#### GGUF (for llama.cpp and ctransformers)
| Flag | Description |
|-------------|-------------|

12
download-model.py
View File

@ -57,7 +57,7 @@ class ModelDownloader:
classifications = []
has_pytorch = False
has_pt = False
# has_ggml = False
# has_gguf = False
has_safetensors = False
is_lora = False
while True:
@ -78,10 +78,10 @@ class ModelDownloader:
is_pytorch = re.match(r"(pytorch|adapter|gptq)_model.*\.bin", fname)
is_safetensors = re.match(r".*\.safetensors", fname)
is_pt = re.match(r".*\.pt", fname)
is_ggml = re.match(r".*ggml.*\.bin", fname)
is_gguf = re.match(r'.*\.gguf', fname)
is_tokenizer = re.match(r"(tokenizer|ice|spiece).*\.model", fname)
is_text = re.match(r".*\.(txt|json|py|md)", fname) or is_tokenizer
if any((is_pytorch, is_safetensors, is_pt, is_ggml, is_tokenizer, is_text)):
if any((is_pytorch, is_safetensors, is_pt, is_gguf, is_tokenizer, is_text)):
if 'lfs' in dict[i]:
sha256.append([fname, dict[i]['lfs']['oid']])
@ -101,9 +101,9 @@ class ModelDownloader:
elif is_pt:
has_pt = True
classifications.append('pt')
elif is_ggml:
# has_ggml = True
classifications.append('ggml')
elif is_gguf:
# has_gguf = True
classifications.append('gguf')
cursor = base64.b64encode(f'{{"file_name":"{dict[-1]["path"]}"}}'.encode()) + b':50'
cursor = base64.b64encode(cursor)

2
modules/llamacpp_hf.py
View File

@ -165,7 +165,7 @@ class LlamacppHF(PreTrainedModel):
if path.is_file():
model_file = path
else:
model_file = list(path.glob('*ggml*.bin'))[0]
model_file = list(path.glob('*.gguf*'))[0]
logger.info(f"llama.cpp weights detected: {model_file}\n")

2
modules/models.py
View File

@ -241,7 +241,7 @@ def llamacpp_loader(model_name):
if path.is_file():
model_file = path
else:
model_file = list(Path(f'{shared.args.model_dir}/{model_name}').glob('*ggml*.bin'))[0]
model_file = list(Path(f'{shared.args.model_dir}/{model_name}').glob('*.gguf*'))[0]
logger.info(f"llama.cpp weights detected: {model_file}")
model, tokenizer = LlamaCppModel.from_pretrained(model_file)

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('*ggml*.bin'))) > 0:
elif len(list(path_to_model.glob('*.gguf*'))) > 0:
loader = 'llama.cpp'
elif re.match(r'.*ggml.*\.bin', model_name.lower()):
elif re.match(r'.*\.gguf', model_name.lower()):
loader = 'llama.cpp'
elif re.match(r'.*rwkv.*\.pth', model_name.lower()):
loader = 'RWKV'