Kromoss is a full fine-tune of MOSS-Image-Base at 1024 px. It uses the same architecture, inference pipeline, text encoder (Qwen3.5-9B) and VAE (FLUX.2 small decoder) as MOSS-Image-SFT, and this repo has the same layout, so it is a drop-in replacement.

this should be loadable as is and run in 8 steps
the base and separate dmd lora are in the huggingface https://huggingface.co/lodestones/Kromoss