I wan to combine power of GPT-3.5-turbo model and text embedding.
E.g I have large PHP code base, I want to analyze the code and find bugs. I can do it with direct GPT-3.5-trubo model directly with openAI. But it slow, I want convert PHP code to vector and insert into pincone db, then based on the vector data need to analyze the code and find bugs, but need to apply the GPT-3.5-turbo model on vector data and get the desire result.
Can you some one please suggest on this?