How to handle long audio files?

#6
by q00u - opened

I have a transcriber that uses this model and the 0.6b-v2/v3 models. For shorter audio files, they all work great. For longer files, 0.6b-v2/v3 chunk correctly, but THIS model somehow causes a [WinError 267] error.

Error during transcription processing: [WinError 267] The directory name is invalid: 'E:\\tmp\\tmp4f90fr0a\\manifest.json'

Now, that manifest file exists, but from the error it sounds like it's trying to access it like a directory.

What's the correct way to configure this 1.1b model for longer audio files? We know it's not the same as the 0.6b-v2/v3 models.

Sign up or log in to comment