Spaces:

---

laion

/

CoCa

like 124

Sleeping

Demonstration of language-vision models for zero-shot classification, retrieval and captioning (openMaMMUT), made by LAION and open-sci. Following NeurIPS paper https://openreview.net/forum?id=cWnZLIdeKn https://arxiv.org/abs/2506.04598 . The model collection is available at HF https://huggingface.co/collections/laion/openmammut-openclip-models-and-scaling-laws, source code for reproduction https://github.com/LAION-AI/scaling-laws-for-comparison