afourney
There are a few issues possibly at play here. Since the original issue, graph group chat, initiate chats, and nested chats should each support this type of orchestration. The group...
I had a conversation with @sonichi about this PR today, and only now realized that the validation specifically excludes self loops. In many scenarios (e.g., with the WebSurferAgent) self-loops are...
Is this the Spider nl2sql benchmark? That would probably fit better under autogenbench
> I really like the PR. Can we also add a really simple Notebook to illustrate the use case? Yes, I can add a Notebook next week when I have...
@gagb I added a notebook and link.
A super significant element of #890 is that there is a hook that allows agents to consider more than just the last message when converting an inner-chat into a single...
Where does GroupChat fit in? Supposing we have a sophisticated orchestrator (and not just round-robin), how might I leverage that here? Does it automatically kick-in if Chat() has more than...
SocietyOfMindAgent was my attempt at this -- at least for GroupChat: https://github.com/microsoft/autogen/pull/890 I found it to be useful to be able to run the inner monologue, then do a final...
Using a function for termination like that can work in some scenarios. I don't think it would for the benchmark scenarios I've been chasing. In particular, in a fork of...
Generally our requirements are specified in setup.py or pyproject, including optional dependencies. I'm wondering how this would integrate with such systems.