This model is a SigLIP2-trained Vision Transformer (ViT) with 400 million parameters and 384px resolution, hosted in the timm library. It excels at zero-shot image classification and producing high-quality image embeddings for downstream computer vision tasks. The openly available weights allow researchers and engineers to leverage state-of-the-art contrastive vision capabilities.
ViT SO400M 16 SigLIP2 384 is an Other AI project. It focuses on performing zero-shot image classification and retrieval without task-specific training data. ViT SO400M 16 SigLIP2 384 is an open-source project aimed at developers. The project is open source (Apache-2.0). ViT SO400M 16 SigLIP2 384 is available on the web and API.
Behind ViT SO400M 16 SigLIP2 384 is timm, and it first shipped in 2022. Development happens publicly on GitHub with 3.5k stars. Among its 3 catalogued features are zero-shot classification, image embeddings, and sigLIP2 training.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do