SakanaAI/JA-Multi-Image-VQA

--- size_categories: - n<1K task_categories: - visual-question-answering dataset_info: features: - name: images sequence: image - name: page_urls sequence: string - name: image_urls sequence: string - name: question dtype: string - name: answer dtype: string splits: - name: test num_bytes: 102782224 num_examples: 55 download_size: 29527995 dataset_size: 102782224 configs: - config_name: default data_files: - split: test path: data/test-* license: apache-2.0 language: - ja --- # JA-Multi-Image-VQA ## Dataset Description **JA-Multi-Image-VQA** is a dataset for evaluating the question answering capabilities on multiple image inputs. We carefully collected a diverse set of 39 images with 55 questions in total. Some images contain Japanese culture and objects in Japan. The Japanese questions and answers were created manually. <!-- To evaluate Japanese VLMs, please go to [our Github repository](https://github.com/SakanaAI/evolutionary-model-merge). --> ## Usage ```python from datasets import load_dataset dataset = load_dataset("SakanaAI/JA-Multi-Image-VQA", split="test") ``` ## Uses The images in this dataset are sourced from Unsplash and are free to use under the Unsplash License. They cannot be sold without significant modification and cannot be used to replicate similar or competing services. All other parts of this dataset, excluding the images, are licensed under the Apache 2.0 License. ## Citation ```bibtex @misc{Llama-3-EvoVLM-JP-v2, url = {[https://huggingface.co/SakanaAI/Llama-3-EvoVLM-JP-v2](https://huggingface.co/SakanaAI/Llama-3-EvoVLM-JP-v2)}, title = {Llama-3-EvoVLM-JP-v2}, author = {Yuichi, Inoue and Takuya, Akiba and Shing, Makoto} } ```size_categories: - 样本量小于1000（n<1K） task_categories: - 视觉问答（visual-question-answering） dataset_info: features: - name: images sequence: 图像序列 - name: page_urls sequence: 字符串序列 - name: image_urls sequence: 字符串序列 - name: question dtype: 字符串 - name: answer dtype: 字符串 splits: - name: 测试集 num_bytes: 102782224 num_examples: 55 download_size: 29527995 dataset_size: 102782224 configs: - config_name: 默认配置 data_files: - split: test path: data/test-* license: Apache 2.0许可协议 language: - 日语 --- # JA-Multi-Image-VQA ## 数据集描述 **JA-Multi-Image-VQA** 是一款用于评估多图像输入场景下问答能力的数据集。我们精心采集了涵盖多样主题的39张图像，总计配套55个问答对。 部分图像包含日本文化元素与日本本土物品，配套的日语问题与答案均为人工撰写。 <!-- 若需评估日语视觉语言模型（Vision Language Model, VLM），请访问我们的GitHub仓库：[https://github.com/SakanaAI/evolutionary-model-merge](https://github.com/SakanaAI/evolutionary-model-merge)。 --> ## 使用方法 python from datasets import load_dataset dataset = load_dataset("SakanaAI/JA-Multi-Image-VQA", split="test") ## 数据集用途说明 本数据集的图像资源源自Unsplash平台，可依据Unsplash许可协议免费使用。未经大幅修改不得用于售卖，亦不得用于复刻同类或竞争性服务。 本数据集除图像外的其余内容均遵循Apache 2.0许可协议进行授权。 ## 引用格式 bibtex @misc{Llama-3-EvoVLM-JP-v2, url = {[https://huggingface.co/SakanaAI/Llama-3-EvoVLM-JP-v2](https://huggingface.co/SakanaAI/Llama-3-EvoVLM-JP-v2)}, title = {Llama-3-EvoVLM-JP-v2}, author = {Yuichi, Inoue and Takuya, Akiba and Shing, Makoto} }