Pyannote AuthTokens Not Working

If you’re trying to work with parsing of audio files (diarization) you’re going to begin working with Pyannote at some point. And you’re going to have to authenticate a token on huggingspace.co to access the models… AND … it’s going to suck cause you can’t figure out why it won’t authenticate. 24 hours later (literal 24 hours). Here’s your solution to why you can’t authenticate.

But.. real quick:
You need this link: https://huggingface.co/pyannote/speaker-diarization-3.1/blob/main/README.md

Right now here’s your solution:

You have to run your command prompt in Administrator mode.

~~ awkard silence ~~

That’s it. The following code will now work.

pipeline = Pipeline.from_pretrained(
    "pyannote/speaker-diarization-3.1",
    use_auth_token="<your token>"
)
pipeline.to(torch.device("cuda"))