hugpadhugpad

google / siglip-so400m-patch14-384

Vision878M694.5K704Updated Sep 26, 2024
Hugging Face

Downloads

694.5K

Likes

704

Last updated

Sep 26, 2024

License

apache-2.0

SigLIP model pre-trained on WebLi at resolution 384x384. It was introduced in the paper Sigmoid Loss for Language Image Pre-Training by Zhai et al. and first released in this repository.

Tasks & library

  • zero shot image classification
  • transformers

Tags

  • transformers
  • siglip
  • vision

Technical specifications

Parameters878M
Librarytransformers
LanguagesNot available
Licenseapache-2.0