five

lmms-lab/COCO-Caption2017

收藏
Hugging Face2024-03-08 更新2024-04-19 收录
下载链接:
https://hf-mirror.com/datasets/lmms-lab/COCO-Caption2017
下载链接
链接失效反馈
官方服务:
资源简介:
--- dataset_info: features: - name: question_id dtype: string - name: image dtype: image - name: question dtype: string - name: answer sequence: string - name: id dtype: int64 - name: license dtype: int8 - name: file_name dtype: string - name: coco_url dtype: string - name: height dtype: int32 - name: width dtype: int32 - name: date_captured dtype: string splits: - name: val num_bytes: 788752747.0 num_examples: 5000 - name: test num_bytes: 6649116198.0 num_examples: 40670 download_size: 7444321699 dataset_size: 7437868945.0 configs: - config_name: default data_files: - split: val path: data/val-* - split: test path: data/test-* --- <p align="center" width="100%"> <img src="https://i.postimg.cc/g0QRgMVv/WX20240228-113337-2x.png" width="100%" height="80%"> </p> # Large-scale Multi-modality Models Evaluation Suite > Accelerating the development of large-scale multi-modality models (LMMs) with `lmms-eval` 🏠 [Homepage](https://lmms-lab.github.io/) | 📚 [Documentation](docs/README.md) | 🤗 [Huggingface Datasets](https://huggingface.co/lmms-lab) # This Dataset This is a formatted version of [COCO-Caption-2017-version](https://cocodataset.org/#home). It is used in our `lmms-eval` pipeline to allow for one-click evaluations of large multi-modality models. ``` @misc{lin2015microsoft, title={Microsoft COCO: Common Objects in Context}, author={Tsung-Yi Lin and Michael Maire and Serge Belongie and Lubomir Bourdev and Ross Girshick and James Hays and Pietro Perona and Deva Ramanan and C. Lawrence Zitnick and Piotr Dollár}, year={2015}, eprint={1405.0312}, archivePrefix={arXiv}, primaryClass={cs.CV} } ```

数据集信息: 特征字段: - 字段名:question_id,数据类型:字符串(string) - 字段名:image,数据类型:图像(image) - 字段名:question,数据类型:字符串(string) - 字段名:answer,数据类型:字符串序列(sequence: string) - 字段名:id,数据类型:64位整型(int64) - 字段名:license,数据类型:8位整型(int8) - 字段名:file_name,数据类型:字符串(string) - 字段名:coco_url,数据类型:字符串(string) - 字段名:height,数据类型:32位整型(int32) - 字段名:width,数据类型:32位整型(int32) - 字段名:date_captured,数据类型:字符串(string) 数据集划分: - 划分名称:val(验证集),占用字节数:788752747.0,样本量:5000 - 划分名称:test(测试集),占用字节数:6649116198.0,样本量:40670 下载总大小:7444321699,数据集总存储大小:7437868945.0 配置项: - 配置名称:default(默认配置),数据文件路径: - 划分:val(验证集),路径:data/val-* - 划分:test(测试集),路径:data/test-* <p align="center" width="100%"> <img src="https://i.postimg.cc/g0QRgMVv/WX20240228-113337-2x.png" width="100%" height="80%"> </p> # 大规模多模态模型评测套件(Large-scale Multi-modality Models Evaluation Suite) > 借助`lmms-eval`加速大规模多模态模型(Large-scale Multi-modality Models, LMMs)的研发 🏠 [项目主页](https://lmms-lab.github.io/) | 📚 [文档](docs/README.md) | 🤗 [Huggingface 数据集仓库](https://huggingface.co/lmms-lab) # 本数据集 本数据集为[COCO-Caption-2017版本](https://cocodataset.org/#home)的格式化版本,被应用于我们的`lmms-eval`评测流程中,可实现大规模多模态模型的一键式评测。 @misc{lin2015microsoft, title={Microsoft COCO: Common Objects in Context}, author={Tsung-Yi Lin and Michael Maire and Serge Belongie and Lubomir Bourdev and Ross Girshick and James Hays and Pietro Perona and Deva Ramanan and C. Lawrence Zitnick and Piotr Dollár}, year={2015}, eprint={1405.0312}, archivePrefix={arXiv}, primaryClass={cs.CV} }
提供机构:
lmms-lab
原始信息汇总

数据集概述

数据集信息

特征

  • question_id: 字符串类型
  • image: 图像类型
  • question: 字符串类型
  • answer: 字符串序列
  • id: 64位整数类型
  • license: 8位整数类型
  • file_name: 字符串类型
  • coco_url: 字符串类型
  • height: 32位整数类型
  • width: 32位整数类型
  • date_captured: 字符串类型

数据分割

  • val:
    • 字节数: 788752747.0
    • 样本数: 5000
  • test:
    • 字节数: 6649116198.0
    • 样本数: 40670

数据大小

  • 下载大小: 7444321699
  • 数据集大小: 7437868945.0

配置

  • default:
    • val: data/val-*
    • test: data/test-*

数据集来源

  • 该数据集是COCO-Caption-2017-version的格式化版本,用于lmms-eval流水线中,以便一键评估大型多模态模型。
搜集汇总
数据集介绍
main_image_url
背景与挑战
背景概述
This dataset is a curated collection of image-caption pairs from COCO-Caption-2017, optimized for evaluating multi-modality models. It includes diverse images each paired with multiple descriptive captions, facilitating comprehensive model assessment in image understanding and caption generation tasks.
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务