Introduction The biggest problem with LLMs for the longest time was not being able to access real-time data and get hallucinated responses. To fix this, OpenAI added a new feature called function calling to their chat completion API. This feature mad...

Markdown supported · ⌘↵ to post
No responses yet.