Artificial Intelligence

Comprehensive guides for deploying and managing AI infrastructure, from image generation to large language models.

Overview

This section covers everything you need to run AI workloads on your self-hosted infrastructure. Whether you’re setting up image generation with ComfyUI, running local LLMs with Ollama, or creating user-friendly interfaces with Open WebUI, these guides will walk you through multiple deployment options.

Featured Tools

ComfyUI

A powerful and modular interface for Stable Diffusion and other image generation models. ComfyUI offers a node-based workflow system that gives you complete control over your image generation pipeline.

Deployment Options:

Proxmox LXC containers
Docker containers
Bare metal installation (like on Bob, your AI machine)

Ollama

Run large language models locally with ease. Ollama provides a simple interface for downloading, managing, and running models like Llama 3, Mistral, and more.

Deployment Options:

Proxmox VMs with GPU passthrough
Docker containers with NVIDIA runtime

Open WebUI

A feature-rich web interface for interacting with LLMs. Similar to ChatGPT’s interface but for your self-hosted models. Connect it to Ollama or other OpenAI-compatible APIs.

Deployment Options:

Docker containers
Integration guides for connecting to Ollama and other AI services

Hardware Considerations

Running AI workloads efficiently requires proper hardware:

GPU requirements for image generation and LLM inference
RAM considerations for different model sizes
Storage planning for models and generated content

Many guides in this section reference “Bob” - a dedicated bare metal AI machine optimized for these workloads.

Integration with Other Services

AI tools become even more powerful when integrated with automation platforms like n8n or Home Assistant. Check the Automation section for guides on connecting these services together.

Select an AI tool from the sidebar to get started with your deployment.