Gorilla LLM is an open-source project that connects LLMs to APIs, accurately returning API calls for a specific task provided by the user in natural language.
It has been developed by researchers from UC Berkeley and Microsoft Research.
Its primary objective is to improve the ability of LLMs to effectively utilise tools through API calls, an area that has yet to be fully realised despite recent advancements in LLMs. Gorilla excels in writing API calls and adapting to changes in test-time documents, allowing for flexible user updates or version changes, thereby outperforming GPT-4.
Gorilla is designed to connect LLMs with a variety of tools, services, and applications that are accessible through APIs. It can write precise API calls for various platforms including Kubernetes, GCP, AWS, Azure, OpenAPI, and more. Gorilla has shown to perform better than GPT-4, Chat-GPT, and Claude, significantly reducing hallucination errors.
In essence, Gorilla is a large language model that is connected with a multitude of APIs, aiming to enhance the potential of LLMs to effectively use tools via API calls. It outperforms GPT-4 in writing API calls and demonstrates a strong capability to adapt to changes in test-time documents, allowing for flexible user updates or version changes. Gorilla’s use of self-instruct fine-tuning and retrieval allows it to understand and reason about constraints, making it a robust and reliable tool for API calls.