this is TangoFlux, a new text-to-audio model that can generate 30 seconds of 44.1kHz audio in just 3.7 seconds on a single A40 GPU project page: https://tangoflux.github.io/ code: https://github.com/declare-lab... demo: https://huggingface.co/spaces/...

@michieldoteth don’t think this is gonna hold up for it but report back!

@NimEshed not sure how it has evolved, but check out tortoise and t5-tts

@pureshimon enjoy