GitHub - zby/LLMEasyTools: OpenAI tools and functions with no fuss
Apr 06, 2024 - github.com
LLMEasyTools is a minimal Python package designed to simplify interaction with the OpenAI Python API library. The package focuses on 'tools' and 'function calls', offering a non-intrusive design that allows developers to debug their code and optimize communication with the LLM. It integrates Pydantic for robust data validation and schema generation. Features include schema generation, structured data from LLM, function name mapping, dispatching function calls, stateful tools, and no patching.
The installation and usage of LLMEasyTools are straightforward. Developers can clone the repository and install it using pip. The package allows developers to create a toolbox instance, register their tool, and directly invoke functions based on LLM response structures. The package is open-sourced under the Apache License.
Key takeaways:
LLMEasyTools is a Python package designed for seamless interaction with the OpenAI Python API library, focusing on 'tools' and 'function calls' without being intrusive.
It offers features such as schema generation, structured data from LLM, function name mapping, dispatching function calls, stateful tools, and no patching.
It provides examples of how to use the package, including getting structured data from LLM and dispatching to a function.
LLMEasyTools is open-sourced under the Apache License.
