This paper shows how to train big language models faster and cheaper by using 4-bit numbers (NVFP4) without losing much accuracy.
Nemotron 3 is a new family of open AI models (Nano, Super, Ultra) built to think better while running faster and cheaper.