turnwind/BOPRO

LLM for Bayesian Optimization

BOPRO

BoPRO is a simple attempt at LLM for Bayesian Optimization. By generating meta-prompts to enhance the model's understanding of tasks, knowledge transfer and meta-learning are achieved.

How to use

As you can see, it's very simple.

⚠️ Note: Due to issues with the formatting of LLM output, some functions may experience abnormalities. This will be addressed through the new feature (response_format) provided by OpenAI.