Fetching the paper…

VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks · Around