app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# 加载模型和 tokenizer
model_name = "LilithHu/mbert-manipulative-detector"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

# 设置为评估模式
model.eval()

# 设置运行设备
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

# 标签名
labels = ["Non-manipulative / 非操纵性", "Manipulative / 操纵性"]

# 推理函数
def classify(text):
    inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
    with torch.no_grad():
        outputs = model(**inputs)
        probs = torch.softmax(outputs.logits, dim=1)
        pred = torch.argmax(probs, dim=1).item()
        confidence = probs[0][pred].item()
        return f"🧠 预测 / Prediction: {labels[pred]}\n🔢 置信度 / Confidence: {confidence*100:.2f}%"

# Gradio 界面
interface = gr.Interface(
    fn=classify,
    inputs=gr.Textbox(lines=4, placeholder="Enter text in English or Chinese... / 输入中文或英文句子"),
    outputs="text",
    title="🔍 Manipulative Language Detector / 操纵性语言识别器",
    description="🧪 输入英文或中文句子，系统将判断其是否包含操纵性语言。\nEnter a sentence in English or Chinese to detect if it's manipulative.",
)

interface.launch()