AI大模子情况—Vanna安装分享（text-2-sql）

Vanna 简介

Vanna 是一个基于 MIT 允许的开源 Python RAG（检索增强天生）框架，用于 SQL 天生和干系功能。它允许用户在数据上练习一个 RAG “模子”，然后提问标题，这将天生在数据库上运行的 SQL 查询语句，并将查询效果通过表格和图表的方式展示给用户。

安装情况
   CentOS 7.X
  1.安装 anaconda3 
   wget https://repo.anaconda.com/archive/Anaconda3-2024.02-1-Linux-x86_64.sh
  sh Anaconda3-2024.02-1-Linux-x86_64.sh
  source ./anaconda3/bin/activate
  
2.更新pip
   pip install --upgrade pip
  >Successfully installed pip-24.0
  
3.安装模块包
   pip install pandas -i Simple Index
  pip install openai -i Simple Index
  pip install pymysql -i Simple Index
  pip install 'vanna[chromadb,openai]' -i Simple Index
  
4. 通过conda情况启动jupyter（pyhton开辟情况），下令行实验
   jupyter-lab
  大概
  jupyter notebook --allow-root
  

5. 申请Vanna的api_key
打开网址Log in to Vanna.AI，输入邮箱，获取Vanna.AI for free Key，示比方下：
from vanna.remote import VannaDefault
vn = VannaDefault(model='vanna-dxs', api_key='57d72396****')

6. 运行python模子

1. ##初始化 Vanna 对象
2. from vanna.remote import VannaDefault
3. vn = VannaDefault(model='data', api_key='d6647*****')
5. ##连接数据库
6. import pandas as pd
7. import pymysql
8. # MySQL数据库连接参数
9. conn_details = {
10. "host": "192.168.1.56",
11. "port": 13306,
12. "user": "root",
13. "password": "123@123",
14. "database": "vanna",
15. "charset": "utf8mb4",
16. }
18. # 建立MySQL数据库连接
19. conn = pymysql.connect(**conn_details)
22. # 定义一个函数，用于执行SQL查询并返回一个Pandas DataFrame
23. def run_sql(sql: str) -> pd.DataFrame:
24.     df = pd.read_sql_query(sql, conn)
25.     return df
27. # 将函数设置到vn.run_sql中
28. vn.run_sql = run_sql
29. vn.run_sql_is_set = True
31. ##训练数据
32. vn.train(ddl="""
33. CREATE TABLE IF NOT EXISTS users (
34.     id INT PRIMARY KEY COMMENT '用户ID',
35.     username VARCHAR(50) COMMENT '用户名',
36.     email VARCHAR(100) COMMENT '电子邮件',
37.     age INT COMMENT '年龄',
38.     gender VARCHAR(10) COMMENT '性别（男/女）',
39.     city VARCHAR(50) COMMENT '城市'
40. ) COMMENT='用户信息表' CHARACTER SET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
41. """)
43. ##插入示例数据
44. INSERT INTO users (id,username,email,age,gender,city) VALUES
45. (1,'张三','[email protected]',30,'男','北京')，
46. (2,'李四','[email protected]', 25,'女','上海')，
47. (3,'王五','[email protected]',40,'男','广州')，
48. (4,'赵六','[email protected]',35,'女','深圳')，
49. (5,'小明','[email protected]',28,'男','成都')，
50. (6,'小红','[email protected]',45,'女','重庆')，
51. (7,'小华','[email protected]',32,'男','天津')，
52. (8,'小丽','[email protected]',27,'女','南京')，
53. (9,'小李','[email protected]',38,'男','武汉')，
54. (10,'小美','[email protected]',33,'女','西安');

复制代码

7. 打开web访问界面 

1. from vanna.flask import VannaFlaskApp
2. VannaFlaskApp(vn).run()

复制代码

1. 返回如下：
2. Your app is running at:
3. http://localhost:8084
4. * Serving Flask app 'vanna.flask'
5. * Debug mode: off

复制代码

8. 打开Vanna界面

别的安装参考：
安装Vanna模块
   [yangwx@localhost bin]$ ./pip install vanna
  Collecting vanna
  Downloading vanna-0.0.33-py3-none-any.whl.metadata (8.8 kB)
Requirement already satisfied: requests in /opt/anaconda3/lib/python3.7/site-packages (from vanna) (2.19.1)
......
Installing collected packages: kaleido, tenacity, tabulate, sqlparse, plotly, vanna
Successfully installed kaleido-0.2.1 plotly-5.18.0 sqlparse-0.4.4 tabulate-0.9.0 tenacity-8.2.3 vanna-0.0.33

更新安装pandas
   [yangwx@localhost bin]$ ./pip install --upgrade pandas
  Requirement already satisfied: pandas in /opt/anaconda3/lib/python3.7/site-packages (0.23.4)
Collecting pandas
  Downloading pandas-1.1.5-cp37-cp37m-manylinux1_x86_64.whl.metadata (4.7 kB)
Successfully installed numpy-1.21.6 pandas-1.1.5

安装openAI
   ./pip install openai
  >>openai-0.8.0

安装chromadb
   ./pip install chromadb   --ignore-installed
  Downloading filelock-3.12.2-py3-none-any.whl (10 kB)
Using cached fsspec-2023.1.0-py3-none-any.whl (143 kB)
Installing collected packages: pypika, mpmath, monotonic, flatbuffers, zipp, websockets, uvloop, urllib3, typing-extensions, tqdm, sympy, sniffio, six, pyyaml, python-dotenv, protobuf, packaging, overrides, numpy, idna, humanfriendly, httptools, grpcio, graphlib-backport, fsspec, filelock, exceptiongroup, charset-normalizer, certifi, bcrypt, backoff, requests, python-dateutil, pydantic-core, pulsar-client, importlib-resources, importlib-metadata, h11, coloredlogs, chroma-hnswlib, anyio, annotated-types, watchfiles, starlette, pydantic, posthog, onnxruntime, huggingface_hub, click, uvicorn, typer, tokenizers, fastapi, chromadb

---

开放页面IP访问设置：
   修改设置文件，让界面通过IP访问
  先实验如下下令，重新天生设置文件
  jupyter notebook  --generate-config
  
  将c.NotebookApp.ip =localhost修改为c.NotebookApp.ip = '*'
   vim /home/yangwx/.jupyter/jupyter_notebook_config.py
  
参考链接：
   https://pypi.org/project/vanna/