Skip to content
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

[Feature Request]: Custom output for streaming #2637

Open
tyler-suard-parker opened this issue May 9, 2024 · 1 comment
Open

[Feature Request]: Custom output for streaming #2637

tyler-suard-parker opened this issue May 9, 2024 · 1 comment
Labels
enhancement New feature or request streaming Streaming output to websocket

Comments

@tyler-suard-parker
Copy link
Contributor

Is your feature request related to a problem? Please describe.

The current streaming output has lots of stuff that I do not want, for example

user_proxy (to assistant):

>>>>> USING AUTO_REPLY...

It would be nice to have a way to edit how the stream comes back. For example, adding an intro, like
Collecting information to answer your query...
And then only streaming back the last message from a particular agent.

Why this would be helpful: let's say I want one agent to do the searching and another agent to write the answer. I don't want the user to see all the data that gets returned by the search, I only want the user to see the written answer, streamed back from the writing agent.

Describe the solution you'd like

Just more flexible streaming in general. Right now it only allows streaming exactly what Autogen outputs, and it would be nice to have the ability to modify the format of what gets returned.

Additional context

No response

@tyler-suard-parker tyler-suard-parker added the enhancement New feature or request label May 9, 2024
@WaelKarkoub
Copy link
Collaborator

If I'm not mistaken, all outputs follow this format, whether they're streamed or not. However, you raise a valid point, the messages you are seeing are hardcoded in the ConversableAgent class and not easy to change (if it's even possible at all).

Maybe you can restructure your code where you can silence some conversations that you don't need the output of. Take a look at the send method in ConversableAgent.

@ekzhu ekzhu added the streaming Streaming output to websocket label May 10, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request streaming Streaming output to websocket
Projects
None yet
Development

No branches or pull requests

3 participants