This snap installs a hardware-optimized engine for inference with the Nemotron 3 Nano large language model with reasoning capabilities built with a hybrid Mixture-of-Experts (MoE) architecture.
📚 Documentation, learn how to use inference snaps
💬 Discussions, ask questions and share ideas
🐛 Issues, report bugs and request features
Clone this repo with its submodules:
git clone https://github.com/canonical/nemotron-3-nanoPrepare the required models by following the instructions for each model, under the components directory.
Build the snap and its component:
snapcraft pack -vRefer to the ./dev directory for additional development tools.