finds.dev

camenduru/controlnet-colab

View on GitHub

★ 533Jupyter NotebookUnlicenseupdated Jun 2023

// the take

A collection of Google Colab notebooks for running ControlNet with various Stable Diffusion 1.5-based fine-tunes, all without needing local GPU hardware. It's a quick-start toolkit for anyone who wants to experiment with pose/edge/depth-conditioned image generation on free Colab compute.

Zero local setup — click a badge, get a running notebook in 30 seconds. Covers a wide range of fine-tuned SD 1.5 checkpoints so you can test ControlNet conditioning across very different aesthetics without hunting down each model separately. The per-notebook structure means you don't need to understand the broader codebase to run a single experiment.

Abandoned since June 2023 — the SD ecosystem has moved well past 1.5 and SDXL/Flux are the current baselines, so most of these notebooks are two generations stale. A significant chunk of entries are marked 🚧 (broken/untested), and there's no indication of which actually work today. No abstraction at all — 60+ near-identical notebooks with copy-pasted boilerplate instead of a parameterized single notebook; if the install steps break upstream, every notebook breaks individually. Colab free tier GPU availability has degraded substantially since 2023, so the 'free compute' premise is much weaker in practice.

We read the README, the directory tree and a key source file in July 2026. Nobody pays to be listed. How we judge →

// at a glance

Stars
533
Forks
66
Language
Jupyter Notebook
License
Unlicense
Last commit
26 Jun 2023

// nearby

If controlnet-colab isn't quite it.

All repos →