Training a V2 Right Now, With double the dataset, will upload once done training. hopefully before 01/08/2025 I created this LoRA to help generate heavily stylized shots inspired by Wong Kar-wai, especially ones that capture a sense of motion blur. Flux tends to struggle with that look or maybe I just haven’t found the right prompt for it. So I trained this LoRA on bunch of film photography Dataset. At higher weights, it leans into a retro aesthetic, thanks to the film-based dataset. If you don't want that please use it at a lower strength (around 0.7) Recommended Tags : Motion_blur, Heavily Stylized blur, Shaky Cam,

Files with SHA256: d087098d534a2fe2bb60364a32e56c30d791ac6df6265a516fae1c4f4f740c3d