Ai Chatbot Tutorials #1 – BASIC OPEN AI API CALLS

Knowledgebase Article: https://www.libraryofcelsus.com/research/basic-open-ai-api-calls/

Chat-GPT

This guide walks you through making calls using the Open Ai Api.

1. Pre-requisites and Initial Setup

Before interacting with the OpenAI API, you must prepare your environment.

• Libraries: Import the required Python libraries.

import os
import json
import openai
import time
from time import time, sleep

• API Key Configuration: Load your OpenAI API key from a file.

with open('key_openai.txt', 'r', encoding='utf-8') as file:
    openai.api_key = file.read().strip()

2. OpenAI API Request Configuration

• Define a function to communicate with the OpenAI API. This function will send a user’s query and receive a response.

def chatgpt_completion(query):
    max_counter = 7
    counter = 0
    while True:
        try:
            completion = openai.ChatCompletion.create(
              model="gpt-3.5-turbo",
              max_tokens=800,
              temperature=0.5,
              messages=query
            )
            response = (completion.choices[0].message.content)
            return response
        except Exception as e:
            counter +=1
            if counter >= max_counter:
                print(f"Exiting with error: {e}")
                exit()
            print(f"Retrying with error: {e} in 20 seconds...")
            sleep(20)

3. Initializing the Main Loop

• Set up the main loop for chatbot interaction. This also involves initializing an empty list to store conversation details.

if __name__ == '__main__':

    conversation = []

4. Define Chatbot & User Details

• Define user and chatbot names, and create uppercase versions for display purposes.

    bot_name = open_file('./Prompts/bot_name.txt')
    user_name = open_file('./Prompts/user_name.txt')

    usernameupper = user_name.upper()
    botnameupper = bot_name.upper()

5. Interaction and Conversation Flow

Guide the user through the chat experience:

• User Input Capture: Prompt the user and get their input.

    while True:
        try:
            user_input = input(f'\n\n{usernameupper}: ')

• System Prompt: Introduce the assistant’s role.

            conversation.append({'role': 'system', 'content': f"You are {bot_name}, a helpful Ai assistant.  Give an answer to the user's inquiry in a natural, but verbose manner."})

• Chatbot Greeting: Start the interaction with a friendly message.

            conversation.append({'role': 'assistant', 'content': f"Hello!  I'm {bot_name}, your loyal personal assistant!  How can I help you today?"})

• User Message: Add the user’s input to the conversation.

            conversation.append({'role': 'user', 'content': user_input})

6. Response Generation

Utilize the configured function to generate a response using the OpenAI API.

            conv_completion = chatgpt_completion(conversation)
            print(f"{botnameupper}: conv_completion)

7. Clearing for Next Interaction

Reset the conversation list to prepare for the next user interaction. If any errors occur, they will be caught and displayed.

            conversation.clear()
        except Exception as e:
            print(e)