Testing (and therefore using) Google's Gemini model with a valid Gemini API key is not working in versions later than autogen==0.5.0

[toygark]

What happened?

Testing (and therefore using) Google's Gemini model with a valid Gemini API key is not working in versions later than autogen==0.5.0.

The process fails with the following error message: 'Error occurred while testing model: Please install google-generativeai and 'vertexai' to use Google's API.'.

In versions before (and including) autogen==0.5.0, testing (and therefore using) Google's Gemini model with a valid Gemini API key works fine.

What did you expect to happen?

Testing (and therefore using) Google's Gemini model with a valid Gemini API key should work fine in versions later than autogen==0.5.0, including the latest 0.6.0 -- which is not the case right now.

How can we reproduce it (as minimally and precisely as possible)?

Just enter a valid Gemini API key in Google's Gemini model and press the 'Test Model' button.

AutoGen version

0.5.1 upto latest

Which package was this bug in

Core

Model used

gemini-1.5-pro-latest

Python version

No response

Operating system

No response

Any additional info you think would be helpful for fixing this bug

No response