基于gemini SQL问答

采用 gemini 模型实现语言学习模型（llm）的 SQL 问答功能，将自然语言转化为 SQL 语句（即 nl2sql）。该过程利用 llm 的 SQL 生成能力，将用户复杂的自然语言请求解析后映射到相应的 SQL 查询，从而更精准的获取数据库信息。这种技术在数据库管理、信息检索等方面具有重要应用价值。

代码来自于 ，原仓库是使用 streamlit 的，简单试用下就不要 streamlit 了。

1. gemini 生成 sql

import os
import pymysql
import streamlit as st
import google.generativeai as genai
 
genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
 
## Function To Load Google Gemini Model and provide queries as response
def get_gemini_response(question,prompt):
    model=genai.GenerativeModel('gemini-pro')
    response=model.generate_content([prompt[0], question])
    return response.text

2. pysql 执行 sql 语句

import pymysql
 
def read_sql_query(sql, db_config):
    # db_config is a dictionary containing your MySQL connection parameters
    # Example: {'host': 'your_host', 'user': 'your_user', 'password': 'your_password', 'database': 'your_database'}
 
    # Connect to the MySQL database
    conn = pymysql.connect(**db_config)
 
    try:
        # Create a cursor object
        with conn.cursor() as cur:
            # Execute the SQL query
            cur.execute(sql)
 
            # Fetch all the rows
            rows = cur.fetchall()
 
            Print the rows
            for row in rows:
                print(row)
 
            return rows
    finally:
        # Close the connection when done
        conn.close()
 
 
#测试
sql_query = "SELECT * FROM STUDENT;"
database_config = {'host': 'localhost', 'port': 61106, 'user': 'root', 'password': 'rn', 'database': 'gemini'}
result = read_sql_query(sql_query, database_config)

3. 实际应用

prompt=[
    """
    You are an expert in converting English questions to SQL query!
    The SQL database has the name STUDENT and has the following columns - NAME, CLASS, 
    SECTION \n\nFor example,\nExample 1 - How many entries of records are present?, 
    the SQL command will be something like this SELECT COUNT(*) FROM STUDENT ;
    \nExample 2 - Tell me all the students studying in Data Science class?, 
    the SQL command will be something like this SELECT * FROM STUDENT 
    where CLASS="Data Science"; 
    also the sql code should not have ``` in beginning or end and sql word in output
 
    """
 
]
 
question="Tell who is the highest MARKS in Data Science class?"
 
sql_cmd = get_gemini_response(question,prompt)
print("Gemini generate SQL:", sql_cmd)
response = read_sql_query(sql_cmd, database_config)
print("SQL query result:", response)

执行效果如下，也有可能 sql 语句不对的，要是用 nl2sql 的数据微调下会比较好。

我使用的是 pymysql, 这是插入 mysql 部分代码。

也可以建表后直接命令行插入：

INSERT INTO STUDENT (NAME, CLASS, SECTION, MARKS) VALUES
    ('Krish', 'Data Science', 'A', 90),
    ('Sudhanshu', 'Data Science', 'B', 100),
    ('Darius', 'Data Science', 'A', 86),
    ('Vikash', 'DEVOPS', 'A', 50),
    ('Dipesh', 'DEVOPS', 'A', 35);

create_db.py

import pymysql
 
# MySQL 连接
mysql_connection = pymysql.connect(
    host="145.22.xx.xx",
    port=6306,
    user="root",
    password="xxxx",
    database="gemini",
    cursorclass=pymysql.cursors.DictCursor  # 设置字典游标以获取字典格式的结果
)
 
 
mysql_cursor = mysql_connection.cursor()
 
# 创建 MySQL 表结构
mysql_cursor.execute("""
CREATE TABLE STUDENT (NAME VARCHAR(25),
    CLASS VARCHAR(25),
    SECTION VARCHAR(25),
    MARKS INT
)
""")
 
# 插入数据
data_to_insert = [
    ('Krish', 'Data Science', 'A', 90),
    ('Sudhanshu', 'Data Science', 'B', 100),
    ('Darius', 'Data Science', 'A', 86),
    ('Vikash', 'DEVOPS', 'A', 50),
    ('Dipesh', 'DEVOPS', 'A', 35)
]
 
mysql_cursor.executemany('''INSERT INTO STUDENT VALUES (%s, %s, %s, %s)''', data_to_insert)
 
# 提交更改
mysql_connection.commit()
 
# 显示插入的数据
mysql_cursor.execute('''SELECT * FROM STUDENT''')
result = mysql_cursor.fetchall()
for row in result:
    print(row)
 
# 关闭连接
mysql_connection.close()