Below are 78 of the 400+ works used in instruction.
Original size 3871x2635
Learning Outcomes
The model was trained over 800 steps with a generation step size of 25. SDXL LoRA is capable of producing vibrant, high-quality reproductions and mimics original brushstrokes quite well. SDXL handles details less proficiently and sometimes merges detailed objects into a jumble.
Link to Colab
The model was trained on stable-diffusion-xl-base-1.0
https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0