wukong100m / README.md
wanng's picture
Update README.md
2d102e6
---
annotations_creators:
- crowdsourced
language_creators:
- crowdsourced
language:
- zh
license:
- cc-by-nc-sa-4.0
multilinguality:
- monolingual
pretty_name: Wukong100M
task_categories:
- feature-extraction
---
# wukong100m
## 简介 Brief Introduction
取自Noah-Wukong多语言多模态数据集中的中文部分,一共100M个图文对。
A subset from Noah-Wukong (a multimodal dataset), around 100M image-text pairs (only Chinese).
## 数据集信息 Dataset Information
大约一共100M个中文图文对。大约占用16GB空间(仅仅是url等文本信息,不包含图片)。下载成功率在80%左右。(虽然我没有统计下载之后会占用多少空间,但是,可以说非常非常大)
- Homepage: [Noah-Wukong](https://wukong-dataset.github.io/wukong-dataset/index.html)
## 下载 Download
```bash
mkdir wukong100m && cd wukong100m
for i in {00000..00031}; do wget https://huggingface.co/datasets/wanng/wukong100m/resolve/main/data/train-$i-of-00032.parquet; done
cd ..
```
## Lisence
CC BY-NC-SA 4.0