This tutorial will walk you through the steps to use Ollama server to download the Llama model and set up an inference server. Instead of Llama models that is just used as an example, you can download any open source AI model. Ollama can be installed on the kiosk computer or a server on the local network where the kiosks are installed.
Step 1: Installing Ollama
- Go to the Ollama website (https://ollama.com/) and click on the "Download" button to download the latest version of Ollama Server.
- Follow the installation instructions for your operating system.
Step 2: Launching Ollama App
- Once Ollama is installed, launch the application.
Step 3: Downloading the AI model
- Download Llama recommanded model or any other open source model.
Step 4: Set up HelloKiosk
- Go to Settings of HelloKiosk and select Ollama as the main AI provider
- Enter this URL as endoint: http://localhost:11434/api/
- Note: you can replace localhost with the IP address of another machine if you installed the AI engine on another machine.
- Click on Refresh the list of model. Select the downloaded AI model.
- NOTE: Make sure that the computers have fixed IP addresses if you are using an IP address
That's it! You are done. You can have conversations using the Ollama engine.