Open jasonsu123 opened 10 months ago
The binary files with the models are available for free download on Hugging Face.
Compressed models in ZIP format with mlmodelc
in the file name are not supported.
As for the model size you're welcome to test, but in theory I believe large model is more accurate.
Download ggml-large.bin
and ggml-medium.bin
and compare.
Hello, where can I download Whisper desktop-specific models?
The link on the Hugging Face website seems to be down before I could access it.
Also, in the speech-to-text functionality, is the medium-sized model more accurate in recognition than the large-sized model?
Thank you.