News for hackers. Stuff that matters.

A comment on GPT-4o. Parent comment

Re:GPT-4o by jacobsimon on Monday May 13, 2024 @06:30PM (#40349320)
I couldn’t quite tell from the announcement, but is there still a separate TTS step, where GPT is generating tones/pitches that are to be used, or is it completely end to end where GPT is generating the output sounds directly?
[ Parent | on HN ]
Re:GPT-4o by derac on Monday May 13, 2024 @06:46PM (#40349460)
It's one model with text/audio/image input and output.
[ Parent | on HN ]
Re:GPT-4o by jacobsimon on Tuesday May 14, 2024 @12:12AM (#40351544)
Very exciting, would love to read more about how the architecture of the image generation works. Is it still a diffusion model that has been integrated with a transformer somehow, or an entirely new architecture that is not diffusion based?
[ Parent | on HN ]