LilithHu commited on
Commit
55c0410
·
verified ·
1 Parent(s): 636f520

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +11 -12
app.py CHANGED
@@ -45,31 +45,30 @@ interface = gr.Interface(
45
  fn=classify,
46
  inputs=gr.Textbox(
47
  lines=4,
48
- placeholder="Enter text in English or Chinese... / 输入中文或英文句子",
49
- label="📝 Input Text / 输入文本"
50
  ),
51
- outputs=gr.Markdown(label="📊 Prediction / 预测结果"),
52
- title="🔍 Manipulative Language Detector / 操纵性语言识别器",
53
  description="""
54
- 🧪 输入英文或中文句子,系统将判断其是否包含操纵性语言。
55
- Enter a sentence in English or Chinese to detect if it's manipulative.
56
 
57
- 📌 **Disclaimer / 免责声明**
58
  This system is for **research and educational purposes only**.
59
  It **does not guarantee accuracy** and **should not be used as legal or clinical evidence**.
60
- 本工具仅用于**学术研究与教学演示**,不构成法律、医疗或其他正式用途的依据。
61
 
62
- 🤖 **Model Info / 模型信息**
63
  - Model: `LilithHu/mbert-manipulative-detector`
64
  - Base: `mDeBERTa-v3` multilingual pre-trained model
65
  - Fine-tuned using HuggingFace Transformers on 10,000 labeled Chinese data
66
 
67
- 🌐 **Built with Gradio and hosted on HuggingFace Spaces**
68
-
69
- ⚠️ **About Examples / 关于例子**
70
  The examples provided below are those **cited in the paper**, including implicit moral coercion, polite masking and false positives.
71
 
72
 
 
 
 
73
 
74
  """,
75
  examples=[
 
45
  fn=classify,
46
  inputs=gr.Textbox(
47
  lines=4,
48
+ placeholder="Enter text in English or Chinese... ",
49
+ label="📝 Input Text"
50
  ),
51
+ outputs=gr.Markdown(label="📊 Prediction"),
52
+ title="🔍 Manipulative Language Detector",
53
  description="""
54
+ 🧪 Enter a sentence in English or Chinese to detect if it's manipulative.
 
55
 
56
+ 📌 **Disclaimer**
57
  This system is for **research and educational purposes only**.
58
  It **does not guarantee accuracy** and **should not be used as legal or clinical evidence**.
 
59
 
60
+ 🤖 **Model Info**
61
  - Model: `LilithHu/mbert-manipulative-detector`
62
  - Base: `mDeBERTa-v3` multilingual pre-trained model
63
  - Fine-tuned using HuggingFace Transformers on 10,000 labeled Chinese data
64
 
65
+ ⚠️ **About Examples**
 
 
66
  The examples provided below are those **cited in the paper**, including implicit moral coercion, polite masking and false positives.
67
 
68
 
69
+ 🌐 **Built with Gradio and hosted on HuggingFace Spaces**
70
+
71
+
72
 
73
  """,
74
  examples=[