Trained on edited cute furry & kemono illustrations + upscaled illustrious generations + edited aesthetically appealing anime screenshots, all with e621, danbooru, and natural language captions. Trigger words "high quality", "best quality", and "masterpiece" used to control aesthetic & quality.
Prompt with E621 & booru tags and natural language wherever you want to be more specific, in any order desired.
Use a CFG of 3.5 to 7.0 with 5.0 being a good middle ground.
I personally generate at SD1.5 resolutions (Chroma is trained at both SD1.5 and SDXL sizes), upscale to the actual resolution I want, and then resample again for 11-14 steps.
Description
V1 is trained on 83 images, some of these are repeats of the most aesthetic images with captions manually shuffled to keep natural language order intact