We’ve open-sourced our 4B full-duplex speech model under Apache 2.0 !!
Our model was trained on 40M hours and uses continuous tokens.
Full-duplex AI for "Human like AI" is still largely unexplored.
Let’s push the boundaries together!
Check out our demos! 🚀
Website: https://t.co/hOlupgh4bM
Code: https://t.co/Qx4oFJPVj0
Model: https://t.co/CBslmM4yr2
Technical blog: https://t.co/R6l26UCD6q
Real-world noise makes speech tokens unstable.
We train a neural codec to recover clean speech from degraded audio through augment-and-denoise training.
Our new blog is here!!
https://t.co/HASC9TX5Ze
DubGuild is looking for exceptional talent based in Japan.
You don’t need to speak Japanese — English is enough. We also support visa and permanent residency applications, housing, and other essentials for living in Japan.
If you’re interested in building the future of Speech AI with us, please apply here:
https://t.co/gjZn8hzW2w