Trained on 32 recent images from renowned "AI thief" (his words, not mine) Henry Daubrez, with no permission asked. This LoRA produces excellent painterly images that trend toward surreal and abstract with beautiful textures and expressive swirls. Images were captioned via GPTV and edited for best practices. Training was done using the prodigy optimizer for 40 epochs with a batch size of 4 and a gradient accumulation of 4. Seems to work well with a variety of models and schedulers. Make sure to follow @henrydaubrez on X to see more of his excellent original work.
Trigger words
You should use painterly style to trigger the image generation.
You should use surreal to trigger the image generation.
You should use abstract to trigger the image generation.
Download model
Weights for this model are available in Safetensors format.