Flux Recommended strength is 1. This version was trained (49 images) on Flux.1 Dev. Picked the last model at 4000 steps. Trained on an A40 on runpod - took 3 hours. This model seems to have unintentionally trained on the faces of the characters as well. If a generic model (earlier step) is preferred, will release that as an alternate version. Let me know in the comments. LoRA trained on 153 screencaps from the movie Tron Legacy. The general style is captured very well, the specifics like outfits, exact shapes are not entirely there. My first try at a style model, looks nice but can be improved (will keep at it for v2). All feedback is appreciated. Any tips on improving the model? Refer to the sample images for specific keywords to re-create images.

Files with SHA256: 065b92a2d85953bec41a8c3ab9d0cf3b82ab6c828976c39e0971fa60dc849168