CancerPath-170K-v1 is a multimodal dataset containing 170,000 cancer pathology image-text pairs, specifically built for the training and evaluation of multimodal models in cancer pathology.
# SigLIP (shape-optimized model) SigLIP model pre-trained on WebLi at resolution 384x384. It was introduced in the paper [Sigmoid Loss for Language Image Pre-Training](https://arxiv.org/abs/2303.15
