Upwork is hiring a Looking for ML consultant: Tuning LLMs/text-image models to produce new images based on image set

Hello,

I am exploring a few ideas leveraging LLMs. I'm looking for somebody with a strong background in machine learning to help me figure out how to build these out.

I am a software engineer myself, but I want to accelerate the learning process. You will play the role of architect/consultant to aid my ideation/system design process.

The first project I want to tackle is:

Given a set of ~100,000 images with text labels, how do I train an LLM/text-to-image model to generate images that match those in the set based on a prompt?

E.g. imagine I have a set of images showing kitchens my firm has designed. The following prompt should show a kitchen that matches the style of kitchens from that image set: "a kitchen with granite counters and aluminum stove".

Please only reach out if you have a clear idea of how to implement this. You will be paid hourly for at least 2-3 hours to as much as 15-20.