# WebLI Pretraining

A ViT image encoder based on SigLIP 2, utilizing global average pooling, with the attention pooling head removed, suitable for image feature extraction tasks.

A ViT image encoder based on SigLIP 2, utilizing global average pooling and removing the attention pooling head, suitable for image feature extraction tasks.

A vision Transformer model based on SigLIP 2, using Global Average Pooling (GAP) instead of attention pooling head for image encoding

SigLIP 2 vision-language model trained on WebLI dataset, suitable for zero-shot image classification tasks.

A SigLIP 2 vision-language model trained on the WebLI dataset, suitable for zero-shot image classification tasks.

A SigLIP 2 vision-language model trained on the WebLI dataset, suitable for zero-shot image classification tasks.

A SigLIP 2 vision-language model trained on the WebLI dataset, suitable for zero-shot image classification tasks.

SigLIP 2 is an improved model based on SigLIP, integrating multiple technologies to enhance semantic understanding, localization, and dense feature extraction capabilities.

SigLIP 2 is an improved multilingual vision-language encoder based on SigLIP, enhancing semantic understanding, localization, and dense feature extraction capabilities.

A multimodal model based on the SoViT backbone network, improved with the Sigmoid loss function, supporting zero-shot image classification and image-text retrieval

SigLIP (Sigmoid Loss for Language-Image Pretraining) model trained on the WebLI dataset, suitable for zero-shot image classification tasks.

SigLIP (Sigmoid Loss for Language-Image Pre-training) model trained on the WebLI dataset for zero-shot image classification tasks.

SigLIP (Sigmoid Loss Language-Image Pretraining) model trained on the WebLI dataset for zero-shot image classification tasks