ilsp/vibeeval_greek

Vibe-Eval Greek数据集是一个包含269个示例的基准数据集，用于评估多模态聊天模型，特别是具有挑战性的示例。它是从VibeEval数据集手动翻译成希腊语的。每个示例包含媒体URL、示例ID、类别、用户提示（英语和希腊语）、参考答案（英语和希腊语）。The Vibe-Eval Greek dataset is a benchmark of 269 examples for evaluating multimodal chat models, especially challenging examples. It has been manually translated into Greek from the VibeEval dataset. Each example includes a media URL, example ID, category, user prompts (in English and Greek), and reference answers (in English and Greek).