OpenAI's Whisper models converted to ggml format
For more information, visit:
https://github.com/ggerganov/whisper.cpp/tree/master/models
| Model | Mem | SHA |
|---|---|---|
| ggml-small-q4_0.bin | ~142 MiB | f6099c17a4305c6d173ef90ab6f15150dcbb85332b7831313bf36a378d7c5a85 |
| large-v3-turbo-q5_0 | ~526 MiB | 19fea4b380c3a618ec4723c3eef2eb785ffba0d0538cf43f8f235e7b3b34220f |
| ggml-large-v3-turbo-q5_0.bin | ~560 MiB | 394221709cd5ad1f40c46e6031ca61bce88931e6e088c188294c6d5a55ffa7e2 |
Models are multilingual . Models ending in -q5_0 are quantized.
SDK下载
#安装ModelScope
pip install modelscope
#SDK模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('linguacine/whisper.cpp')
Git下载
#Git模型下载
git clone https://www.modelscope.cn/linguacine/whisper.cpp.git