Re:Dessa: the most authentic full-duplex speech model ever b...
by bm-rf on Friday October 23, 2026 @08:53PM
(#50217342)
Do you think that, much like how OpenAI scaled up GPT-2 to GPT-3, someone else can train a larger model like Dessa on, say, multiple GB200s?
[ Parent
| on HN ]
Re:Dessa: the most authentic full-duplex speech model ever b...
by WhitneyLand on Saturday October 24, 2026 @12:12PM
(#50222681)
Yes, most likely. The problem is even if it’s better it has to be super fast/real time to be convincing. That’s the part I think could take a few more hardware generations.