We have done a major update on distillation and speculative decoding, and release our Nemotron-Labs-Sigma model family on Huggingface. Check it out at huggingface.co/collections/nv…
Kudos to all collaborators and looking forward to the future of continuous diffusion LLMs!
Kudos to all collaborators and looking forward to the future of continuous diffusion LLMs!
📣 NVIDIA is releasing a collection of open model weights: Nemotron-Labs-Sigma — 14 checkpoints at 3B/8B.View quoted post on X
Language models have stayed discrete. We bring continuous diffusion to language as in vision.
Smooth ODE trajectories (16-D) unlock distillation and steering: 8 TPFs for math, 4 TPFs for code, with accuracy and diversity preserved. 🧵 1/