Gorilla is an exciting new AI system that can understand natural language commands and interact with over 1,600 different APIs. Developed by UC Berkeley PhD student Shishir Patil, the Gorilla LLM aims to make API usage and integration easier than ever before.
Table of Contents
How Gorilla Works
Gorilla is built on top of two powerful language models – Falcon and MPT. These models have been fine-tuned specifically for parsing API syntax and calls. This gives Gorilla a few key advantages:
- Semantically correct API invocations: Gorilla can understand the meaning behind API commands and make the proper API calls. This reduces hallucination issues common in large language models like GPT-4.
- Syntactically correct code: Gorilla can parse Abstract Syntax Trees (ASTs) when generating code. This results in code that is syntactically sound and can be executed properly.
- Compatibility with 1,600+ APIs: Gorilla has access to a vast library of APIs spanning multiple domains. This gives developers access to a wide range of capabilities through natural language.
Use Gorilla LLM for your CLI and Spotlight Search
Gorilla powered CLI
Get started with pip install gorilla-cli
Use Cases
With its advanced natural language understanding and broad API coverage, Gorilla can enable many valuable use cases:
- Voice assistants & chatbots: Creating conversational interfaces that can interact with back-end APIs and systems.
- API exploration: Quickly trying out new APIs without reading documentation or writing code.
- Code generation: Generating code snippets, examples, and documentation using natural language prompts.
- API testing: Testing APIs by invoking them through natural language queries.
- Innovative applications: Building highly functional apps by leveraging Gorilla’s API capabilities.
Availability
Gorilla is open-source and free for commercial use. It can be used through Google Colab, CLI, and Python imports.
The project also provides APIBench – a curated collection of APIs for easy training. API developers can contribute their own APIs to expand Gorilla’s capabilities.
The Bottom Line
Gorilla represents an exciting innovation in leveraging large language models for practical API usage. By providing correct API interactions through natural language, Gorilla can save developers time, reduce errors, and open new possibilities. It’s a project worth keeping an eye on!
There are many new and different types of Large language models that have been created since the release of ChatGPT and they only keep getting better with being only 8 months in. Imagine where we will be next year!
Learn more about Gorilla LLM on the official Berkeley .edu website.
Join Gorilla’s Discord group.