EVisRAG-Train

paper: [2510.09733](https://arxiv.org/abs/2510.09733)

**Dataset Description**

This is a VQA Training dataset, collected from [ChartQA](https://arxiv.org/abs/2203.10244), [InfographicVQA](https://arxiv.org/abs/2104.12756), and [MMLongBench-Doc](https://arxiv.org/abs/2407.01523).

**Load the dataset**
```python
import pandas as pd
import os
import sys
data_name = sys.argv[1]
df = pd.read_parquet(f"data/{data_name}/images.parquet", engine="pyarrow")
output_dir = f"data/{data_name}"
os.makedirs(f"{output_dir}/imgs", exist_ok=True)
for idx, row in df.iterrows():
img_bytes = row['image']['bytes']
output_path = os.path.join(output_dir, row["path"])
with open(output_path, "wb") as f:
f.write(img_bytes)
```

论文: [2510.09733](https://arxiv.org/abs/2510.09733)

**数据集说明**

本数据集为视觉问答（Visual Question Answering, VQA）训练数据集，采集自[ChartQA](https://arxiv.org/abs/2203.10244)、[InfographicVQA](https://arxiv.org/abs/2104.12756)以及[MMLongBench-Doc](https://arxiv.org/abs/2407.01523)。

**数据集加载方式**
python
import pandas as pd
import os
import sys
data_name = sys.argv[1]
df = pd.read_parquet(f"data/{data_name}/images.parquet", engine="pyarrow")
output_dir = f"data/{data_name}"
os.makedirs(f"{output_dir}/imgs", exist_ok=True)
for idx, row in df.iterrows():
img_bytes = row['image']['bytes']
output_path = os.path.join(output_dir, row["path"])
with open(output_path, "wb") as f:
f.write(img_bytes)

搜集汇总

背景与挑战

背景概述

EVisRAG-Train是由OpenBMB发布的VQA训练数据集，整合了ChartQA、InfographicVQA和MMLongBench-Doc三个来源的数据。该数据集大小为4.27GB，采用Apache License 2.0许可，主要用于视觉问答任务的训练。