Comment by smusamashah

Comment by smusamashah a day ago

5 replies

Tried it (used gibberish email address). It answers immediately/instantly/while you are still talking. But those are just filler sentences (cached answers?). Actual thing that you asked for is answered much later down the line, if it doesn't get stuck in a loop.

swyx a day ago

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...

  • l-m-z a day ago

    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).

    • swyx a day ago

      thanks Laurent! also congrats on releasing + fully believe you. just offering first impressions.

  • huac a day ago

    One guess is that the live demo is quantized to run fast on cheaper GPUs, and that degraded the performance a lot.

  • imjonse a day ago

    They are too prestigious to try shumering it.