Multilingual, multimodal reranker for text and visual documents across 29+ languages via Qwen2-VL backbone
View on HuggingFace
View on ModelScope