Tl;dr - Have you tried NVIDIA Nemotron Nano 2 yet? Check out the features, try it out, and come to office hours on Thursday, September 4 at 5:30pm PDT to ask questions live.
NVIDIA came out with the Nemotron Nano 2 9B model last week - a groundbreaking 9B parameter open, multilingual reasoning model that earned a spot on the Artificial Analysis Intelligence Index leaderboard, at release, among open models within the same parameter range.
Key highlights of the release:
-
Leading Accuracy and Performance: By combining the strengths of Transformer and Mamba architectures, achieves up to 6X faster throughput compared to the next best 8B open model and highest reasoning accuracy
-
Thinking Budget: Set
max_thinking_tokensto meet response-time targets giving you fine-grained control. -
Open Model and Datasets: The training datasets of this model are fully open, giving maximum transparency in using the model for enterprise applications
Getting Started
We encourage you to try it out and let us know what you think. Your feedback helps us continue to build tools that empower the open-source community.
-
Hugging Face: https://nvda.ws/4oH5DCP
-
NVIDIA NIM: https://nvda.ws/4mKuILd
-
Dataset: https://nvda.ws/3USA7nQ
-
Research Paper: https://nvda.ws/420kTAP
Office Hours
We’re also hosting Office Hours on Thursday September 4 at 5:30pm PDT on the NVIDIA Developer YouTube channel. Drop your questions/feedback here in the forum or on Discord.
See you there!
Stay up to date on NVIDIA Nemotron by subscribing to NVIDIA news and following NVIDIA AI on LinkedIn, X, Discord and YouTube.
Access open Nemotron Models on Hugging Face and a collection of NIM microservices and Developer Examples on build.nvidia.com