Update OptiGuide, solutions, and examples to use autogen and pyomo instead of FLAML and gurobipy

[alfredoihernandez]

Do the following for pyomo without disrupting currently implementation with gurobipy.

### Tasks
- [ ] Enable OptiGuide to be installed with pip using splinter packages for either pyomo or gurobipy
- [x] Create new coffee.py and optiguide.py with pyomo #12 
- [ ] Update optiguide example notebook showing how to use either gurobipy or pyomo
- [ ] Update the README
- [x] Update to use pyautogen instead of FLAML in package installation and setup.py #10 

[pphuc25]

Are you still doing this? shall we coauthored for contribute these tasks?
However @alfredoihernandez , why would you like to change from flaml to autogen? because with my knowledge, flaml already has autogen in the library, so we just call it? correct me if I wrong.

[BeibinLi]

@pphuc25 For updating to AutoGen: it is because some new features from autogen are only included in the autogen library but not the original flaml library.

I guess @alfredoihernandez is on vacation right now. Let's wait for a week; otherwise, I can also make the updates.

[adam-behrendt]

Interested in this as well. Currently having issues with FLAML resulting in: InvalidRequestError: Invalid URL (POST /v1/openai/deployments/gpt-35-turbo/chat/completions)
See my latest comment in the related issue microsoft/autogen#573

[BeibinLi]

@adam-behrendt Thanks for your comments.
@pphuc25 and @alfredoihernandez I just updated the main branch with Pyautogen 0.2.2. Please take a look.

There are more issues regarding test cases, setup pipy, etc. All of which will be resolved within this week.

[pphuc25]

At now the open sources seem work really well, the need of use openai api maybe less needed. Should we custom to use autogen with open source model? @BeibinLi
I have researched and I seen that at now we can already do it,

link tutorial: https://www.youtube.com/watch?v=10FCv-gCKug&fbclid=IwAR0NmQhizie6RTtE4DKDmQsnMFCgPUngK5QDWLvRmurfa5ZRkCEQd2TWr0U

[adam-behrendt]

Thank you @BeibinLi, looking forward to this Pyomo update! It will be very useful.

[BeibinLi]

About Open Source LLM

@pphuc25 Yes, AutoGen supports open source models and also Gemini (soon, I am working on it recently). The only thing need to be changed is the OAI_CONFIG_LIST. Feel free to raise another issue here about OSS or create a PR (with notebook).

About Pyomo

@adam-behrendt @alfredoihernandez

Please review the PR #12. Thanks to @pphuc25 !

[alfredoihernandez]

@pphuc25 Left a review on #12

Also tagging #10 to this issue and updated issue description to better split up work