#output trimmed in terminal

9 messages · Page 1 of 1 (latest)

grave meadow
#

Hi, am new to autogen and been playing around with some of the samples, am doing so in a linux terminal running .py scripts directly, eg. samples/simple_chat.py . No matter, which sample/code I have been playing with the output from the agent will be trimmed to one line - also I keep getting those IOStream warning. Will this only work properly in a notebook setting or is there some argument I need to provide, that I missed, to enable more output? (I am using azure AI API)

"python3 samples/simple_chat.py
Provide feedback to agent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: Give me investment suggestions in 3 bullet points
No default IOStream has been set, defaulting to IOConsole.
user (to agent):

Give me investment suggestion in 3 bullet points


No default IOStream has been set, defaulting to IOConsole.

USING AUTO REPLY...
No default IOStream has been set, defaulting to IOConsole.
No default IOStream has been set, defaulting to IOConsole.
agent (to user):

  • Research and invest in index funds or ETFs that track the overall market for

No default IOStream has been set, defaulting to IOConsole.
No default IOStream has been set, defaulting to IOConsole."

obtuse mountain
#

What version of Autogen are you on?

grave meadow
#

pip show pyautogen
Name: pyautogen
Version: 0.2.21

obtuse mountain
#

Try updating to .22 - should at a minimum solve the "no default console" warnings.

Are you by any chance trying to stream the responses?

grave meadow
#

thanks, will give it a go, am simply running "python3 samples/simple_chat.py" out of the box in a terminal

#

and you are right, it did get rid of the console warnings, but am still not able to receive/output a multiline output from the agent

obtuse mountain
#

Does your OAI_CONFIG_LIST have anything in it other than model and api_key?

grave meadow
#

yes, because it is azure, I have
"base_url": "https://XXXXX.openai.azure.com/",
"api_type": "azure",
"api_version": "2023-07-01-preview"
as well, I can also test with openai, but think I already did last week, but will give it another go just in case

#

same thing, only model and openai api_key and get this

python3 samples/simple_chat.py
agent (to user):

How can I help you today?


Provide feedback to agent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: Give me investment suggestion in 3 bullet points
user (to agent):

Give me investment suggestion in 3 bullet points


USING AUTO REPLY...
agent (to user):

  • Research and consider investing in a diversified portfolio of low-cost index funds or exchange

Provide feedback to agent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: