[WIP] update openai version to v1 & add image eval feature

[uni-zhuan]

[WIP]

• Test demos and verify stability

[DONE]

• Update openai version to v1.1.1, modify API functions
• Tested headline and animal demo, which are stable.
• Remove Litellm for now, waiting for an stable update
  -------------------------commit [6e93038]

[WIP]

• Waiting for litellm stable version

[DONE]

• modify MultimodalOutput image schema
• add openai_based_image_eval feature
• add a new demobuilding_design to test image eval
  --------------------------commit [3db4a85]

[motin]

A couple of suggestions:

• Split this PR into one that updates the OpenAI client and another one for adding an image eval feature
• Instead of removing litellm, convert existing demos etc to use an updated version of litellm so that there are no hard coded uses of the old OpenAI client. This makes it trivial to update the OpenAI version since it there will less outdated code to update.