How to Fine-Tune an LLM on NVIDIA GPUs With Unsloth
Fine-tune popular AI models faster in Unsloth with NVIDIA RTX AI PCs and DGX Spark to build personalized assistants for studying, work, creative proje...
NVIDIA Nemotron Research
research.nvidia.comSaved December 15, 2025
Summary
NVIDIA Nemotron is a family of open-source models, datasets, and technologies including Nano, Super, and Ultra variants for agentic AI tasks like reasoning, coding, math, RAG, and safety.
These models use a hybrid Mamba-Transformer Mixture-of-Experts architecture, NVFP4 training, and LatentMoE for leading accuracy and up to 4x higher throughput than prior generations. The project provides over 9T tokens of open synthetic datasets and contributes techniques openly via Hugging Face and NVIDIA NIM APIs.
