ilsp/MMBench_Greek
收藏资源简介:
该数据集是MMBench的希腊语翻译版本,MMBench是一个用于评估多模态和视觉语言模型的多项选择基准。此版本的目标是使MMBench可作为评估希腊语能力视觉语言模型的基准。数据集保持了与原始MMBench_dev数据集相同的结构和列名,文本字段已翻译成希腊语,而元数据字段保持不变。每个示例包含一个图像、一个问题、可选提示文本、多项选择答案选项、正确答案标签以及类别/来源元数据。
This dataset is a Greek translation of MMBench, a multiple-choice benchmark for evaluating multimodal and Vision-Language Models. The goal of this version is to make MMBench usable as an evaluation benchmark for Greek-capable VLMs. The dataset keeps the same structure and column names as the original MMBench_dev dataset. The text fields have been translated into Greek, while metadata fields are kept unchanged. Each example contains an image, a question, optional hint text, multiple-choice answer options, the correct answer label, and category/source metadata.




