Find answers from the community

Updated 7 months ago

Hi @Logan M

Hi @Logan M

i have seen there is a caht engine in LlamaIndex. however, in all the examples it is used with indexes. Is there any solution to save the chat history of the results of NLSQL query engine or Pandasquery engine?
L
A
7 comments
You can plug any query engine into an agent

Create the QueryEngineTool, pass it to the agent, and off you go
Does agentic support Open-source LLMs?
I usually get a JSON decoder error when touching agents with Open-source models😆
I mean, it supports, but open-source is pretty bad at agentic behaviour lol

Optionally you can provide a retriever and use a context chat engine. But you'd have to wrap your query engine into a sort of custom retriever class for that to work
The thing is I have both PandasQueryEngine and SQLQueryEngine which cannot be used with retriever, yes?
In this case, is there any option than Agents?
You can wrap the query engine into a custom retriever if you wanted
I.e. the retrieve is just calling the query engine
Oh, ok.
I got your point I will give it a try. Thanks a lot
Add a reply
Sign up and join the conversation on Discord