File size: 1,157 Bytes
56943c6
 
 
 
 
 
4041dfe
56943c6
f1d3933
56943c6
 
4914695
56943c6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
---
title: GraphGen Demo
emoji: ๐Ÿ“Š
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: "5.44.0"
python_version: "3.10"
app_file: app.py
suggested_hardware: cpu-basic
pinned: false
short_description: "Knowledge-driven synthetic data generation demo"
tags:
  - synthetic-data
  - knowledge-graph
  - gradio-demo
---

# GraphGen Space ๐Ÿค–๐Ÿ“Š

This is the **official Hugging Face Space** for [GraphGen](https://github.com/open-sciencelab/GraphGen) โ€“ a framework that leverages knowledge graphs to generate high-quality synthetic questionโ€“answer pairs for supervised fine-tuning of LLMs.

๐Ÿ”— Paper: [arXiv 2505.20416](https://arxiv.org/abs/2505.20416)  
๐Ÿ™ GitHub: [open-sciencelab/GraphGen](https://github.com/open-sciencelab/GraphGen)

---

## How to use (๐Ÿ–ฑ๏ธ 3 clicks)

1. Open the **Gradio app** above.  
2. Upload or paste your source text โ†’ click **Generate KG**.  
3. Download the generated QA pairs directly.

---

## Local quick start (optional)

```bash
git clone https://github.com/open-sciencelab/GraphGen
cd GraphGen
uv venv --python 3.10 && uv pip install -r requirements.txt
uv run webui/app.py   # http://localhost:7860
```