January 2025
Intermediate to advanced
312 pages
6h 8m
English
In this chapter, you’ll deploy and manage a multi-container AI chatbot using Docker Compose. Along the way, we’ll look at how the chatbot app works.
The app is a multi-container app with three services — a web interface, a backend API, and a model server. The model server runs Ollama to execute a local instance of a large language model (LLM) that’s pre-trained to answer developer-related questions and keep answers short. LLMs require a lot of memory and run best GPUs. However, not everyone has access to GPUs, so I’ve configured the app to work on systems with CPUs and a minimum of 8GB of RAM. I’ll show you how to configure the app for GPUs if your Docker host has NVIDIA GPUs.
When referring to Docker Compose, ...
Read now
Unlock full access