Re:Moshi: A speech-text foundation model for real time dialogue
by swyx on Wednesday September 18, 2024 @01:33PM
(#41582931)
yeah i tried this demo when it first came out and then again today. Not to be all Reflection 70B again but it just doesnt seem like the same weights was uploaded as was showed in their original demo from July https://the-decoder.com/french-ai-lab-kyutai-unveils-convers...
[ Parent
| on HN ]
Re:Moshi: A speech-text foundation model for real time dialogue
by imjonse on Wednesday September 18, 2024 @01:45PM
(#41583115)
They are too prestigious to try shumering it.
Re:Moshi: A speech-text foundation model for real time dialogue
by huac on Wednesday September 18, 2024 @04:21PM
(#41584987)
One guess is that the live demo is quantized to run fast on cheaper GPUs, and that degraded the performance a lot.
Re:Moshi: A speech-text foundation model for real time dialogue
by l-m-z on Wednesday September 18, 2024 @06:22PM
(#41586240)
Hi swyx, laurent from kyutai here. We actually used the online demo at moshi.chat for the live event (the original demo), so same quantization. We updated the weights on the online version since then to add support for more emotions but we haven't noticed it being worse.
One thing to point out is that it takes time to get used to interact with the model, what tends to work, how to make it speak. The live event was far from perfect but we certainly used this experience. I would encourage you to try a bit the same kind of interaction we add on the live event and you should get similar results (though the model is very unpredictable so hard to be sure, you can see that some part of the live events definitely didn't work as expected).