Kani: A Lightweight and Highly Hackable Framework for Building Language Model Applications

TL;DR

Kani is an open-source framework designed to simplify and customize the development of complex language model applications, supporting features like chat management and function calling with high flexibility.

Contribution

It introduces a lightweight, model-agnostic framework that is highly customizable and easy to extend for building advanced language model applications.

Findings

Supports diverse language model features including tool usage and retrieval augmentation

Provides overridable core functions for customization

Facilitates rapid development for researchers and industry professionals

Abstract

Language model applications are becoming increasingly popular and complex, often including features like tool usage and retrieval augmentation. However, existing frameworks for such applications are often opinionated, deciding for developers how their prompts ought to be formatted and imposing limitations on customizability and reproducibility. To solve this we present Kani: a lightweight, flexible, and model-agnostic open-source framework for building language model applications. Kani helps developers implement a variety of complex features by supporting the core building blocks of chat interaction: model interfacing, chat management, and robust function calling. All Kani core functions are easily overridable and well documented to empower developers to customize functionality for their own needs. Kani thus serves as a useful tool for researchers, hobbyists, and industry professionals alike to accelerate their development while retaining interoperability and fine-grained control.