WangZeJun's picture
Update README.md
75d5614 verified
|
raw
history blame
774 Bytes
---
license: apache-2.0
base_model:
- uer/gpt2-chinese-cluecorpussmall
---
# gpt2-chinese-ai-text-classifier
使用 `uer/gpt2-chinese-cluecorpussmall` 作为基座模型,基于 `Hello-SimpleAI/HC3-Chinese` 数据集训练得到的 AI 文本分类器。
## 训练数据
[Hello-SimpleAI/HC3-Chinese](https://huggingface.co/datasets/Hello-SimpleAI/HC3-Chinese)
## Demo
- gradio==3.50.0
- easytokenizer
- transformers
安装 [easytokenizer](https://github.com/zejunwang1/easytokenizer):
```shell
pip install git+https://github.com/zejunwang1/easytokenizer
```
启动 demo:
```shell
python gradio_demo.py --model gpt2-chinese-ai-text-classifier/ --port 9000
```
打开 `http://[server_ip]:9000/`,查看效果:
![](image/demo_1.png)
![](image/demo_2.png)