Olivier Simard-Hanley

PSA. I haven't updated the dataset, but I've managed to integrate additional fields, including text descriptions. See here: https://github.com/malcolmosh/goodbooks-10k-extended

There should be a way to stop a chain or a sequence based on certain criteria too, such as a similarity threshold or the result of a moderation chain.

A chat_history object consisting of (user, human) string tuples passed to the ConversationalRetrievalChain.from_llm method will automatically be formatted through the [_get_chat_history function](https://github.com/hwchase17/langchain/blob/master/langchain/chains/conversational_retrieval/base.py#L22). In a chatbot, you can simply keep appending...