JoePenna/Dreambooth-Stable-Diffusion

Implementation of Dreambooth (https://arxiv.org/abs/2208.12242) by way of Textual Inversion (https://arxiv.org/abs/2208.01618) for Stable Diffusion (https://arxiv.org/abs/2112.10752). Tweaks focused on training faces, objects, and styles.

49
/ 100
Emerging

Combines textual inversion and class-preservation techniques to fine-tune Stable Diffusion on limited GPU memory (24GB VRAM), supporting multi-concept training with optional captions and class regularization images to prevent semantic shift. Deployable across RunPod, Vast.AI, Google Colab, and local environments via Jupyter notebooks with built-in model pruning to compress trained weights from 11-12GB to ~2GB. Includes debugging guidance and recommends using celebrity token names rather than artist identities to mitigate training-data bias and respect creator attribution.

3,222 stars. No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 23 / 25

How are scores calculated?

Stars

3,222

Forks

536

Language

Jupyter Notebook

License

MIT

Last pushed

Jan 08, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/JoePenna/Dreambooth-Stable-Diffusion"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.