gpt-to-chatgpt-py

Convert a regular GPT call into a ChatGPT call

TYPESCRIPT VERSION HERE -> https://github.com/bramses/gpt-to-chatgpt-ts

Installation

pip install gpt-to-chatgpt==0.1.4

Functions

toChatML()

Converts a string message into a Chat Markup Language (ChatML) object.

Usage

toChatML(message:str, options:Optional[dict]=None) -> List[dict]

Arguments

• message: The string message to be converted to ChatML.
• options (optional): A dictionary that can contain the following keys:
  • system_messages: A list of strings that represent system messages to be added to the ChatML object.
  • role: The role of the message (either Role.USER or Role.ASSISTANT).

Examples

toChatML('hello')
# Output: [{'role': Role.USER, 'content': 'hello'}]

toChatML('hello', {'system_messages': ['hi'], 'role': Role.ASSISTANT})
# Output: [{'role': Role.SYSTEM, 'content': 'hi'}, {'role': Role.ASSISTANT, 'content': 'hello'}]

get_message()

Extracts the message content from a response object.

Usage

get_message(response: dict, options: Optional[dict] = None) -> Union[str, dict]

Full Usage

import openai
import os
from dotenv import load_dotenv
from gpt_to_chatgpt import toChatML, get_message

load_dotenv()

openai.api_key = os.getenv("OPENAI_API_KEY")

res = openai.ChatCompletion.create(
  model="gpt-3.5-turbo",
  messages=toChatML("this is a test"),
)

print(get_message(res))

# As an AI language model, I don't really take tests, but I'm always ready to respond to your prompts and queries. How can I assist you today?

Arguments

• response: The response object from which to extract the message content.
• options (optional): A dictionary that can contain the following keys:
  • usage: A boolean value that indicates whether to return the usage information of the response.
  • role: A boolean value that indicates whether to return the role of the message.
  • isMessages: A boolean value that indicates whether to return the message as a list of messages.

Examples

get_message(test_response)
# Output: 'The 2020 World Series was played in Arlington, Texas at the Globe Life Field, which was the new home stadium for the Texas Rangers.'

get_message(test_response, {'usage': True, 'role': True, 'isMessages': True})
# Output: {'usage': {'prompt_tokens': 56, 'completion_tokens': 31, 'total_tokens': 87},
#          'roles': ['assistant'],
#          'messages': ['The 2020 World Series was played in Arlington, Texas at the Globe Life Field, which was the new home stadium for the Texas Rangers.']}

get_message(test_response, {'usage': True, 'role': True})
# Output: {'usage': {'prompt_tokens': 56, 'completion_tokens': 31, 'total_tokens': 87},
#          'role': 'assistant',
#          'message': 'The 2020 World Series was played in Arlington, Texas at the Globe Life Field, which was the new home stadium for the Texas Rangers.'}

get_message(test_response, {'usage': True, 'isMessages': True})
# Output: {'usage': {'prompt_tokens': 56, 'completion_tokens': 31, 'total_tokens': 87},
#          'messages': ['The 2020 World Series was played in Arlington, Texas at the Globe Life Field, which was the new home stadium for the Texas Rangers.']}

About the Developer

This repository was written by Bram Adams, a writer and programmer based out of NYC.

Bram publishes a Zettelkasten, with a twice/weekly newsletter (which you can subscribe to here), is a community developer ambassador for OpenAI and does freeleance contracts (for hire!) related to AI/web dev/AR+VR.

Bram is also the creator of Stenography, a API and VSC Extension that automatically documents code on save.

You can learn more about him and his work on his website.