The Gorilla repository includes evaluation APIs, evaluation code, and inference code for running Gorilla locally. The team behind Gorilla is planning to release the training code as soon as they can finalize the pipeline. The project roadmap includes the release of dataset and evaluation code, opening up APIZoo for community contributions, hosting Gorilla LLM chat for HF model APIs, and training a state-of-the-art Gorilla LLM with expanded APIBench and APIZoo.
Key takeaways:
- Gorilla is a large language model that enables the use of tools by invoking APIs. It can generate semantically and syntactically correct API calls based on a natural language query.
- The project has released APIBench, the largest collection of APIs, which are curated and easy to train on. They aim to expand this API store and teach large language models how to write them.
- Gorilla is commercially usable under the Apache 2.0 license, and can be installed for use in your CLI with the command 'pip install gorilla-cli'.
- The project is open to contributions and aims to build an open-source, one-stop-shop for all APIs that large language models can interact with.