The model is fp8 and based on Flux.1D. It comes without VAE, Clip and textencoder to save disc space. So be sure to load those files in your workflow.

Suggest parameters for image generation:

Sampler: dpmpp_2s_ancestral/euler, Scheduler: beta/sgm_uniform, Steps: 20, Distilled CFG: 3.5

ComfyUI workflow for every image ... please find the prompts and the metadata there. I don't use Forge, sorry.