自然语言查询、分析数据(NL-TO-SQL)的尝试

让查数据这件事，不再是高高在上、遥不可及的技能，而是人人都能玩得转，妙趣横生的小技巧。比如“我的店上周赚了多少钱？哪个商品即将售罄？这个月卖的最多的商品是啥？”，下一秒，答案就像变魔术一样蹦出来。直接看效果

流程如下，

• 前端使用gradio，采集麦克风声音
• 后端收到声音，使用openai的speech-to-text
• 文字通过dbgpt（链接各种自定义模型），转换为sql进行数据库查询
• 返回对应数据

关键代码

# -*- coding: utf-8 -*-
import gradio as gr
from openai import OpenAI
import os
import requests
import json

os.environ["OPENAI_API_KEY"] = "sk-3zxxxxxx"  # Replace with your key

client = OpenAI()

def processAudio(audio):
    #print(audio)
    if audio is None:
        return "No audio inputs were provided."
    

    au = open(audio, "rb")
    transcript = client.audio.transcriptions.create(
        model="whisper-1", 
        file=au,
        response_format="text"
    )
    print(transcript)
    #return transcript
    url = "https://demo-dbgpt-backend.sunmi.com/api/v1/chat/completions"
    headers = {"Content-Type": "application/json; charset=utf-8"}
    data = {
        "select_param": "bmw",
        "chat_mode": "chat_with_db_execute",
        "model_name": "chatgpt_proxyllm",
        "user_input": transcript
    }
    response = requests.post(url, headers=headers, data=json.dumps(data))
    response.encoding='utf-8'
    if response.status_code == 200:
        print(response.text)
        return response.text
    else:
        return "Error making request."
demo2 = gr.Interface(
    fn=processAudio,
    inputs=[
        gr.Audio(type="filepath", label="Upload Audio")
    ],
    outputs="textbox",
    title="语音查询数据库 - 融合openai&DBGPT",
    description="Please ask database questions in natural language; the data set is fabricated and includes information such as orders, products, users, etc."
)


if __name__ == "__main__":
    demo2.launch(server_name='0.0.0.0', server_port=1088)

注意事项

• DBGPT官方部署文档写的不够直观，部署稍费事
• gradio采集麦克风需要使用https，否则出发浏览器安全限制