The ability of LLMs to execute commands through plain language (e.g. English) has enabled agentic systems that can complete a user query by orchestrating the right set of tools (e.g. ToolFormer, Gorilla). This, along with the recent multi-modal efforts such as the GPT-4o or Gemini-1.5 model, has expanded the realm of possibilities with AI agents. While this is quite exciting, the large model size
![The Berkeley Artificial Intelligence Research Blog](https://cdn-ak-scissors.b.st-hatena.com/image/square/6af74858763fb861d40e3d5b19b14270aa256a22/height=288;version=1;width=512/http%3A%2F%2Fbair.berkeley.edu%2Fblog%2Fassets%2FBAIR_Logo.png)