Is the uv detector trained in the pretrained pitch detector?

[auspicious3000]

Hi,

Although the voice-unvoice detector part is removed from the pitch detector code, is the uv detector still pretrained in your provided JDCnet checkpoint? In other words, if I add back the code for uv detection and load from the checkpoint, can I use it as a uv detector?

Looks like the classifier output only has one dimension. Are there any ways to detect voiced or unvoiced frames using the provided pitch extractor from this repo?

Thanks