Flux updated.
This LoRA is trained on MidJourney images that incorporate my Personalized Style blended with Cyberpunk aesthetics. For this Flux version, I completely reworked both the dataset and the captions, resulting in a more refined and cohesive model.
The dataset has been expanded to ~240 images, and the glitch art effects present in earlier SDXL/Pony versions have been significantly reduced.
The signature "hazy industrial fog" and some minor glitch effects are still present, but much more subtle and controlled in this version.
The trigger words are ArsMJStyle, CyberPunk
Have a good effect in the 0.4 - 1.5 range.