That task is called Named Entity Recognition and you can do it without training data using our library (you can use any LLM that exposes an OpenAI compatible API endpoint: https://github.com/plncmm/llmner

Can we take a look at the fine tuning dataset?

We are developing a library for this exact use case. https://github.com/plncmm/llmner

Your task is called Named Entity Recognition and llmNER is a library that uses the llm of your choice to extract entities from texts given a natural language description.

You can give the model a list of strings where to extract the entities and the modem will return a list of annotated strings for you to extract the information you need.

Be aware that our library is under active development right now.