[Python] Request for Training and Fine-Tuning Workflow for Control-Color (Grayscale Image...

I am working on my thesis related to grayscale image colorization and am basing my work on the Control Color repository.

I want to retrain the model from scratch first, and then fine-tune and modify the architecture for further improvements. However, the repository documentation is incomplete regarding full training workflows. I need clear guidance or step-by-step instructions to achieve a working training and fine-tuning pipeline.

I have cloned and explored this project, but it is too complex for me to handle.

Expected: A working, end-to-end training guide.

Continue reading...