They have a smaller model[1] that might be amenable to Whisper-ization. Much smaller language matrix though. [1]: https://github.com/facebookresearch/seamless

They have a smaller model[1] that might be amenable to Whisper-ization. Much smaller language matrix though. [1]: https://github.com/facebookresearch/seamless_communication/blob/main/docs/m4t/on_device_README.md https://github.com/facebookresearch/seamless_communication/b...