A large SigLIP vision model (ViT-SO400M) with 384px resolution, available through the timm library and OpenCLIP. It supports zero-shot image classification using natural language prompts and is trained on the WebLI dataset. The model weights are openly available on Hugging Face.
In the Other AI space, ViT SO400M 14 SigLIP 384 takes a focused approach. It focuses on performing zero-shot image classification and retrieval without task-specific training data. It is built as an open-source project for developers. ViT SO400M 14 SigLIP 384 is open source under the Apache-2.0 license. ViT SO400M 14 SigLIP 384 is available on the web and API.
It is developed by timm, and it first shipped in 2022. Development happens publicly on GitHub with 3.5k stars. PulseGate's similarity index places it among 5 comparable projects. Among its 3 catalogued features are Zero-shot Classification, Image Embeddings, and Vision Transformer. It exposes integrations via a public API.
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