Kendra example that I shared allows for this -
retriever = AmazonKendraRetriever(index_id=kendra_index_id, region_name=region)
prompt_template = """
The following is a friendly conversation between a human and an AI.
The AI is talkative and provides lots of specific details from its context.
If the AI does not know the answer to a question, it truthfully says it
does not know.
{context}
Instruction: Based on the above documents, provide a detailed answer for, {question} Answer "don't know"
if not present in the document.
Solution:"""
PROMPT = PromptTemplate(
template=prompt_template, input_variables=["context", "question"]
)
condense_qa_template = """
Given the following conversation and a follow up question, rephrase the follow up question
to be a standalone question.
Chat History:
{chat_history}
Follow Up Input: {question}
Standalone question:"""
standalone_question_prompt = PromptTemplate.from_template(condense_qa_template)
qa = ConversationalRetrievalChain.from_llm(
llm=llm,
retriever=retriever,
condense_question_prompt=standalone_question_prompt,
return_source_documents=True,
combine_docs_chain_kwargs={"prompt":PROMPT})
return qa This is great because it allows for condensing the question, and also tune the final prompt to LLM.
