Where do I put multiple models like 1B, 3B and 7B?

[Winnougan]

I have downloaded all of the Llasa TTS models from Huggingface. Can I put them all in the models folder? How do I do it?

[nivibilla]

It should auto download. But if you downloaded them separately then you might need to change the model paths in the code

[Winnougan]

Ok thanks

[Winnougan]

It should auto download. But if you downloaded them separately then you might need to change the model paths in the code

It auto downloads the 3B and 7B, but I would like to add the GGUF quantized versions and the 1B transformers folders. Where specifically does it place them?

[rmcc3]

It should auto download. But if you downloaded them separately then you might need to change the model paths in the code

It auto downloads the 3B and 7B, but I would like to add the GGUF quantized versions and the 1B transformers folders. Where specifically does it place them?

#13 adds 1B model support, but not GGUF. I'll look into GGUF.

[Winnougan]

It should auto download. But if you downloaded them separately then you might need to change the model paths in the code

It auto downloads the 3B and 7B, but I would like to add the GGUF quantized versions and the 1B transformers folders. Where specifically does it place them?

#13 adds 1B model support, but not GGUF. I'll look into GGUF.

Thanks mate.

[kadirnar]

@rmcc3 , @nivibilla Have you tested this model https://huggingface.co/HKUSTAudio/Llasa-1B-Multilingual? I input a long text but it produces only a 1-second output. How can I fix this?