app.py

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

# 加载模型和分词器
model_name = "Qwen2-0.5B-Instruct"
model = AutoModelForCausalLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

def generate_response(input_text):
    inputs = tokenizer.encode(input_text, return_tensors="pt")
    outputs = model.generate(inputs, max_length=50, num_return_sequences=1)
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response

def chatbot(input_text):
    response = generate_response(input_text)
    return response

iface = gr.Interface(
    fn=chatbot,
    inputs=gr.Textbox(lines=2, placeholder="输入你的消息..."),
    outputs=gr.Textbox(lines=2),
    title="Qwen2-0.5B-Instruct 聊天机器人",
    description="基于Qwen2-0.5B-Instruct模型的聊天机器人。"
)

iface.launch()