LLM agent for controlling real-world apps via RESTful APIs
Top 29.9% on sourcepulse
RestGPT enables large language models (LLMs) to autonomously control real-world applications by interacting with RESTful APIs. It targets developers and researchers building LLM-powered agents that require integration with external services, offering a framework for planning, API execution, and response parsing.
How It Works
RestGPT employs an iterative coarse-to-fine online planning framework. It comprises a Planner for generating sub-tasks, an API Selector to map these to specific API calls, an Executor to run the API calls, a Caller to manage parameters and documentation, and a Parser to process API responses based on provided schemas. This modular approach allows for structured interaction with diverse APIs.
Quick Start & Requirements
pip install langchain colorama tiktoken spotipy openai
config.yaml
file with API keys. Run python init_spotify.py
to initialize the Spotify environment (note: this will remove existing Spotify data).python run.py
to interact with the agent, or use python run_tmdb.py
and python run_spotify.py
for benchmark execution.Highlighted Details
Maintenance & Community
The project code was released in August 2023, with paper updates following. Links to community channels or roadmaps are not provided in the README.
Licensing & Compatibility
The README does not explicitly state a license. Compatibility for commercial use or closed-source linking is not specified.
Limitations & Caveats
The demo is still under construction. For Spotify benchmark execution, users must manually modify query_idx
in run_spotify.py
. The init_spotify.py
script will remove all existing data from the user's Spotify account.
1 year ago
1 day