Works as expected

Works very well, I hooked up it up between the load diffusion model, the ModelSamplingFlux and clip text encode in the comfyUI with the full flux1_dev model.

I generated 7 images and these 3 yielded the best results. All had a striking resemblance to Elizabeth Olsen, it is just that my prompt could've done a better job to describe the scene.