Azure Open AI Assistant using Code interpreter and Functions — Plot charts

Balamurugan Balakreshnan

·

Follow

5 min read

·

Apr 13, 2024

--

Share

Use Case

• Using assistant api to create charts using python code
• Only for dev purposes
• Using GPT 4 turbo
• using api version = 2024–02–15-preview

Pre-requisites

• Azure Open AI GPT 4 turbo
• Azure subscription
• Azure machine learning workspace
• Azure open ai key

Code

• First install libraries needed

%pip install openai
%pip install python-dotenv

• Import open ai library

import openai
import json
from typing import Optional

• Load the environment variables

from dotenv import dotenv_values

config = dotenv_values("env.env")

• Configure open ai

import os
from openai import AzureOpenAI

client = AzureOpenAI(
  azure_endpoint = config["AZURE_OPENAI_ENDPOINT"], 
  api_key=config["AZURE_OPENAI_KEY"],  
  api_version="2024-02-15-preview"
)

• json display function

import json

def show_json(obj):
    display(json.loads(obj.model_dump_json()))

• Create a function to plot the graph and save as image