---
title: Leopard-Instruct
canonical_url: "https://www.modelscope.cn/datasets/AI-ModelScope/Leopard-Instruct"
md_url: "https://www.modelscope.cn/datasets/AI-ModelScope/Leopard-Instruct.md"
repository: AI-ModelScope/Leopard-Instruct
last_updated: 2024-11-08
license: "Apache License 2.0"
storage_size: "564 GB"
downloads: 2432
stars: 0
---

# Leopard-Instruct

> Leopard-Instruct - AI-ModelScope 在 ModelScope 开源的数据集。Paper | Github | Models-LLaVA | Models-Idefics2

AI-ModelScope/Leopard-Instruct 是 ModelScope 魔搭社区上的数据集，存储大小 564 GB，采用 Apache License 2.0 许可。

- **Repository**: AI-ModelScope/Leopard-Instruct
- **License**: Apache License 2.0
- **Storage size**: 564 GB
- **Downloads**: 2432
- **Stars**: 0
- **Last updated**: 2024-11-08

Source: https://www.modelscope.cn/datasets/AI-ModelScope/Leopard-Instruct

---

# Leopard-Instruct

[Paper](https://arxiv.org/abs/2410.01744) | [Github](https://github.com/tencent-ailab/Leopard) | [Models-LLaVA](https://huggingface.co/wyu1/Leopard-LLaVA) | [Models-Idefics2](https://huggingface.co/wyu1/Leopard-Idefics2)

## Summaries

Leopard-Instruct is a large instruction-tuning dataset, comprising 925K instances, with 739K specifically designed for text-rich, multiimage scenarios. It's been used to train **Leopard-LLaVA** [\[checkpoint\]](https://huggingface.co/wyu1/Leopard-LLaVA)  and **Leopard-Idefics2** [\[checkpoint\]](https://huggingface.co/wyu1/Leopard-Idefics2).

## Loading dataset

- to load the dataset without automatically downloading and process the images (Please run the following codes with datasets==2.18.0)
```python
import datasets
dataset = datasets.load_dataset("wyu1/Leopard-Instruct", "webvision") 
# print(dataset['train'][0]['images'], dataset['train'][0]['texts'])
```


- to load all the subsets of the images
```python
from datasets import get_dataset_config_names, load_dataset
config_dataset = {}
for config_name in get_dataset_config_names():
    config_dataset[config_name] = load_dataset("wyu1/Leopard-Instruct", config_name)
```

## Citation
```
@article{jia2024leopard,
  title={LEOPARD: A Vision Language Model For Text-Rich Multi-Image Tasks},
  author={Jia, Mengzhao and Yu, Wenhao and Ma, Kaixin and Fang, Tianqing and Zhang, Zhihan and Ouyang, Siru and Zhang, Hongming and Jiang, Meng and Yu, Dong},
  journal={arXiv preprint arXiv:2410.01744},
  year={2024}
}
```