Multimodal Question Answering for Unified Information Extraction - View it on GitHub