Multimodel RAG with Ollama:llava and LlamaIndex on Jetson

Here I use reComputer J4012 powered by NVIDIA Jetson Orin NX 16GB.

Prepare environment

# install the container tools
git clone https://github.com/dusty-nv/jetson-containers
bash jetson-containers/install.sh

# Ollam:llava will be our multimodel model
jetson-containers run --name ollama $(autotag ollama)
ollama run llava

The result is as follows:

Open a new terminal(Ctrl+Alt+T) and run:

cd jetson-containers
cd data
git clone https://github.com/Seeed-Projects/Multimodal-RAG-on-Jetson
cd ..
jetson-containers run $(autotag l4t-pytorch)
cd data/Multimodal-RAG-on-Jetson

Run Multimodel-RAG-on-Jetson

bash run.sh

Result

you can check here in hackster for more information.