Seamless M4T on ASR and TTS

[SanderGi]

Task: https://app.asana.com/0/1203067047205953/1205563311799030

ASR:

TTS:

Text translation: will be added in the translation workflow.
Speech to speech translation: the gpu pr supports it, so we need to find a workflow to add this feature to. Maybe ASR?

Relies on GooeyAI/gooey-gpu#4

Q/A checklist

• Do a code review of the changes
• Add any new dependencies to poetry & export to requirementst.txt (poetry export -o requirements.txt)
• Carefully think about the stuff that might break because of this change
• The relevant pages still run when you press submit
• If you added new settings / knobs, the values get saved if you save it on the UI
• The API for those pages still work (API tab)
• The public API interface doesn't change if you didn't want it to (check API tab > docs page)
• Do your UI changes (if applicable) look acceptable on mobile?