wsy-w1k/YANAMI_Bench

YANAMI Benchmark

This Benchmark accompanies the paper YANAMI: A Large-Scale Multimodal Dataset and Benchmark for Multiturn Stateful Interaction. It provides a large-scale multimodal dialogue dataset (YANAMI-400K) and a benchmark (YANAMI-Bench) for stateful interaction across multiple turns.

For more details, please visit the project page.