How to integrate more benchmarks in your pipeline

[StupidBuluchacha]

Hi ViDoRAG team,

Thanks for your good work, and congrats for the publication. Our team is trying to integrate some latest visual document retrieval benchmarks with your codebase pipeline. Do you have some detailed instruction for the community, or would you plan to integrate them in the near future?

It would be great if the following example benchmarks (multi-hop query or reasoning-intensive) can be tested in your evaluation pipeline as well as ViDoRAG framework:

• MRMR: A Realistic and Expert-Level Multidisciplinary Benchmark for Reasoning-Intensive Multimodal Retrieval [huggingface: https://huggingface.co/MRMRbenchmark]
• MR2-Bench: Going Beyond Matching to Reasoning in Multimodal Retrieval [huggingface: https://huggingface.co/datasets/JUNJIE99/MR2-Bench]
• Are We on the Right Way for Assessing Document Retrieval-Augmented Generation? [huggingface: https://huggingface.co/datasets/Episoode/Double-Bench]

Look forward to your reply sincerely.