[Issue]: Warning from the def _num_token_from_messages is verbose and hard to silence

[colaso96]

Describe the issue

Here is an example output:
[I 2024-04-27 00:02:01,357.357 autogen.token_count_utils] gpt-4 may update over time. Returning num tokens assuming gpt-4-0613.

However, this is very verbose and frustrating to see printed. Can we either dynamically retrieve the number of tokens from the model count or toggle this output? I can work around it but it seems like a fixable bit of code.

Steps to reproduce

Initialize a compressible agent or series of compressible agents

Screenshots and logs

Screenshots in comment

Additional Information

No response

[colaso96]

[WaelKarkoub]

Hi @colaso96! We are deprecating compressible agents in favor of TransformMessages https://microsoft.github.io/autogen/docs/topics/long_contexts/. I'm implementing llmlingua as a new transform in #2225 to compress text.