Dear community,
On a text of mine (~ 16000 tokens), I would like to ask to an LLM around 50 different questions. Afterwards, extract the key words from the retrieved answers.
I would like to ask you about general advise for this Data Process Pipeline.
For the 1st stage, the ~50 questions, what should be better, ChatOpenAI class? Even if I do not want any chatbot? … Or better just LLMChain?
Then, for the 2nd stage KeyLLM?
And the ChatOpenAI or LLMChain nested to the KeyLLM with a SequentialChain?
Thank you very much,