You can deploy this model and start earning money today!

RISys-Lab/
ReasonSigLIP2-So14-384-S0-Rea

Name: ReasonSigLIP2-So14-384-S0-Rea
Availability: InStock

transformers

safetensors

siglip

zero-shot-image-classification

2606.26794

cc-by-nc-sa-4.0

endpoints_compatible

Views

No views yet

Model Details

Model: ReasonSigLIP2-So14-384-S0-Rea
Base model: google/siglip2-so400m-patch14-384
Architecture: SigLIP2 ViT-So400M/14
Image resolution: 384
Training stage: Stage 0 - Reasoning
Training data: Only reasoning caption-image pairs from ReasonLite-42M and ReasonPro-16M

Method

Resources

GitHub: RISys-Lab/ReasonCLIP
Paper: arXiv:2606.26794
Playground: ReasonCLIP Playground

Usage

python

from transformers import AutoModel, AutoProcessor

model_id = "RISys-Lab/ReasonSigLIP2-So14-384-S0-Rea"
model = AutoModel.from_pretrained(model_id)
processor = AutoProcessor.from_pretrained(model_id)

For the full checkpoint list, see the ReasonCLIP model card.