eyehole/VisChainBench download history
eyehole/VisChainBench is a visual question answering dataset on the Hugging Face Hub. In the last 30 days it was downloaded 663 times (180 in the last 7 days), and 2,911 times in total. It ranks #26,464 among datasets by monthly downloads.
VisChainBench VisChainBench is a benchmark for multi-turn image-centric reasoning, designed to evaluate and test visual language models (VLMs) across various tasks and scenarios. Existing benchmarks increasingly emphasise multi-image understanding, they largely focus on static, language-h