krutrim-ai-labs/IndicLLaVABench

IndicLLaVABench是一个用于评估大型多模态模型在现实世界场景中的能力的基准数据集。它包含24张来自不同领域的多样化和具有挑战性的图像，包括室内和室外场景、绘画、表情包和草图等。每张图像都配有一个详细描述和60个精心设计的问题，这些问题旨在评估模型在不同领域和更复杂任务上的理解、推理和泛化能力。IndicLLaVABench is a benchmark designed to evaluate the capabilities of Large Multimodal Models (LMMs) in real-world scenarios. It consists of 24 diverse and challenging images from various domains, including indoor and outdoor scenes, paintings, memes, and sketches. Each image is accompanied by a detailed description and a set of 60 carefully curated questions designed to assess the models understanding, reasoning, and generalization across different domains and more complex tasks.