Test interactive functionality

[jyfeng1]

Dear Authors,

Thank you for releasing this model as open-source! It's a valuable resource for the community.

I tested the inference_examples_RGB.ipynb code you provided, and it runs smoothly. However, I noticed that in the demo provided by your team, the model is used in a conversational format for input and output, but I didn't find this functionality in the code you shared. So, I would like to ask how this is implemented. I look forward to your reply. Thank you!

[theodore-zhao]

Thanks for your interest! The demo is a wrapper around the core inference code with GPT-4 assistance. The model itself isn't conversational. We are thinking of adding the demo code to the repo while requiring the users to spin up there own GPT-4 endpoints. Let me know if that's something of interest to you.