pipeline_tag: audio-to-audio tags: - roformer - speech
Specially trained to separate the speaker's voice from background music that contains vocals.
