OpenAI functions

The openai-functions Python project simplifies the usage of OpenAI's ChatGPT function calling feature. It abstracts away the complexity of parsing function signatures and docstrings by providing developers with a clean and intuitive interface.

Tests

Installation

You can install openai-functions from PyPI using pip:

pip install openai-functions

Usage

Import the necessary modules and provide your API key:

import enum
import openai
from openai_functions import Conversation

openai.api_key = "<YOUR_API_KEY>"

Create a Conversation instance:

conversation = Conversation()

Define your functions using the @conversation.add_function decorator:

class Unit(enum.Enum):
    FAHRENHEIT = "fahrenheit"
    CELSIUS = "celsius"

@conversation.add_function()
def get_current_weather(location: str, unit: Unit = Unit.FAHRENHEIT) -> dict:
    """Get the current weather in a given location.

    Args:
        location (str): The city and state, e.g., San Francisco, CA
        unit (Unit): The unit to use, e.g., fahrenheit or celsius
    """
    return {
        "location": location,
        "temperature": "72",
        "unit": unit.value,
        "forecast": ["sunny", "windy"],
    }

Ask the AI a question:

response = conversation.ask("What's the weather in San Francisco?")
# Should return something like:
# The current weather in San Francisco is 72 degrees Fahrenheit and it is sunny and windy.

You can read more about how to use Conversation here.

More barebones use - just schema generation and result parsing:

from openai_functions import FunctionWrapper

wrapper = FunctionWrapper(get_current_weather)
schema = wrapper.schema
result = wrapper({"location": "San Francisco, CA"})

Or you could use skills.

Another use case: data extraction

Import the necessary modules and provide your API key:

from dataclasses import dataclass
import openai
from openai_functions import nlp

openai.api_key = "<YOUR_API_KEY>"

Define your data container using the @nlp decorator:

@nlp
@dataclass
class Person:
    """Extract personal info"""

    name: str
    age: int

Ask the AI for the extracted data:

person = Person.from_natural_language("I'm Jack and I'm 20 years old.")

You can read more about @nlp here.

Note: mypy does not parse class decorators (#3135), so you might have trouble getting type checking when using it like this. Consider using something like nlp(Person).from_natural_language to get proper type support.

How it Works

openai-functions takes care of the following tasks:

Parsing the function signatures (with type annotations) and docstrings.
Sending the conversation and function descriptions to the OpenAI model.
Deciding whether to call a function based on the model's response.
Calling the appropriate function with the provided arguments.
Updating the conversation with the function response.
Repeating the process until the model generates a user-facing message.

This abstraction allows developers to focus on defining their functions and adding user messages without worrying about the details of function calling.

Note

Please note that openai-functions is an unofficial project not maintained by OpenAI. Use it at your discretion.

Metadata

Release files for openai-functions 1.0.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for openai-functions 1.0.2
File	Size	Uploaded
openai_functions-1.0.2.tar.gz	19.6 kB	Sep 29, 2023	Details

Built distribution (wheel)

Table of built distributions (wheels) for openai-functions 1.0.2
File	Interpreter	ABI	Platform	Reset
openai_functions-1.0.2-py3-none-any.whl 29.6 kB Sep 29, 2023	Python 3	none	any	Details

Total release size: 49.2 kB

Release files / openai_functions-1.0.2.tar.gz

Download URL	openai_functions-1.0.2.tar.gz
Size	19.6 kB
Tags	Source
SHA-256 checksum How to use checksums	`fc1ee6d578cd2f30d2ea5f9ec9c083cd4804cdae0aa3649b0358a9b6a6a553e8`
BLAKE2b-256 checksum How to use checksums	`f168155642bd0ed69746bacf1dc548a2b4492df7f0b3c4f190da898d4e614bde`
Upload date	Sep 29, 2023
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`poetry/1.4.2 CPython/3.11.5 Linux/6.4.12-zen1-1-zen`

Release files / openai_functions-1.0.2-py3-none-any.whl

Download URL	openai_functions-1.0.2-py3-none-any.whl
Size	29.6 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`693c68a5dbb3199334c520a8c18e55634e1980ecf26b2f49912e1cb996ece138`
BLAKE2b-256 checksum How to use checksums	`18766a4ca5760420e69707b2055d1387fb80948bb750f772b3881cb516bcb4e5`
Upload date	Sep 29, 2023
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`poetry/1.4.2 CPython/3.11.5 Linux/6.4.12-zen1-1-zen`