This is an aesthetic enhancement LoRA adapter for the Qwen-Image model, trained using differential LoRA techniques. It follows a similar workflow to <a rel="ugc" href="https://civarchive.com/models/1862167/artaug-lora-flux1-dev">ArtAug-lora for FLUX.1dev,</a> with code based on <a rel="ugc" href="https://github.com/modelscope/DiffSynth-Studio">DiffSynth-Studio</a>, to improve the visual quality, details, and overall artistic appeal of generated images.The model is mirrored here for convenience.

Files with SHA256: 263a826a754cfdece87ac901345cae14abc691f84266387d96994672d8e7ce69