-
Notifications
You must be signed in to change notification settings - Fork 493
In Memory query engine for DocAgent #1137
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
@AgentGenie and @Eric-Shang, please review when you get a chance, thanks! Changes to protocol and naming of classes: Protocol simplified and aligned. Names were just what I could think of, you may have better/more relevant names. |
AgentGenie
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it is good to keep all 4 methods from the protocol in autogen/agentchat/contrib/rag/query_engine.py because common vector db query engine will need those 4 methods.
inmemory query engine is a special case.
You can add init_db and connect_db methods in inmemory query engine and explicitly raise NotImplementedError to remind the user.
I'll put back into protocol implement the NotImplementError exceptions... are you okay with naming changes? |
|
Why are these changes needed?
In the original plan for DocAgent, an in-memory option was to be included, allowing ingestion into the system prompt and then querying directly from that.
This PR creates an in memory query engine that stores ingested text into memory and runs queries from an internal agent's system prompt.
Related issue number
N/A
Checks