Lora model inspired by Cocotama Kanna, but as a Human!
Model Development:
I used a Lora to generate images using Control Net in some of the bases and trained over it
The gold shape was challenging to get right, but it was a fun project to work on
Example Prompt:
(score_9, score_8_up, score_7_up) (Cocotama Kanna:1.2) (1girl:1.1) (fluffy blue hair:1.3) (pink eyes:1.2) (human:1.1) (fluffy white dress:1.3) (gold accents:1.2) (city:1.4) (busy streets:1.4) (daytime:1.1) (cars:1.2) (street lights:1.2) (buildings:1.2) (sidewalks:1.1) (people in background:0.8)
Generation Tips:
Images generated on Realistic Pony Diffusion - XL
May redo with a larger dataset and different training configs in the future
Try it Out:
Head over to Tensor.art to experiment with this model: https://tensor.art/models/743625702982609018
Support My Work:
If you enjoy using this model or would like to commission a custom model, please consider supporting me on Ko-fi: https://ko-fi.com/oliversouls. Your contributions will help me continue to create and improve AI models. Thank you for your support!