xingyaoww / code-act

Official Repo for ICML 2024 paper "Executable Code Actions Elicit Better LLM Agents" by Xingyao Wang, Yangyi Chen, Lifan Yuan, Yizhe Zhang, Yunzhu Li, Hao Peng, Heng Ji.
MIT License
468 stars 38 forks source link

how to test with GPT model #9

Open seshubonam opened 4 months ago

seshubonam commented 4 months ago

Is it possible to replace the llm from the 2 provided open source options to other models from openai or clause to compare performance.

Found the method to serve for an openai compartiable api but did not understand how to replace the base llm used for codeact that plans and executes the python code.

thanks for the amazing work ⚡