Spaces:

---

small-models-for-glam

/

README

like 1

Running

Hi @wjbmattingly — opening as a PR rather than a direct commit since you're co-maintainer and this changes the landing-page framing meaningfully. Veto/edit anything you want.

Current README is the placeholder ("Lightweight AI models for cultural heritage institutions / More coming soon"). With 14 models + 9 datasets in the org now (plus a couple of book chapters explaining the recipes), it felt time to flesh it out.

Changes:

• Kept the hero gif and frontmatter unchanged.
• Replaced the "more coming soon" stub with a landing-page body.
• Framing: the org as a community relay — open foundations + community datasets → small task-specific models → extensible by the next institution. Named YOLO/DETR/BERT/Qwen-VL, BigLAM, and the index-card-detector-v5 extension of NLS's detector as the worked example.
• Linked AI Patterns for GLAM (incl. Boring AI + Beyond Chatbots) for the why.
• CTA: a "share a model, or suggest one" section pointing at Discussions, with the idea of curating contributed models into an org collection.
• Maintainer line names both of us + flags wider community contributions.

Happy to take this in, partially merge, or pass on it.