For exclusive models, follow me on
The "anima" version was a herculean task for me—and I might not do it again, given how exhausting it was to edit, upscale, and clean up over 2,000 images (especially those with terrible resolution)—but the result speaks for itself.
Read the instructions
-In this version, the images were carefully scaled to preserve the style.
-It was manually tagged with female Disney characters supported on Gelbooru.
-It was trained on 2,400 images using a low learning rate over approximately 10,000 steps.
cartoonvalley-style-(anima)
Adetailer is recommended (For full-body generations, it is essential)
upscaler: R-ESRGAN 4x+ Anime6B (10-15 steps)
recommended weight: 1
Sampling: Euler a- Normal
Steps: 20-30
CFG Scale: 5
Recommended
Keyword:
- cartoonvalley
- cartoonvalley,shiny skin
Negative (optional)
-worst quality, low quality, score_1, score_2, score_3, artist name,blurry, jpeg artifacts, lowres,bad anatomy,bad hands
cartoonvalley-style-(illustrious)
Adetailer is recommended (For full-body generations, it is essential)
upscaler: R-ESRGAN 4x+ Anime6B (10-15 steps)
recommended weight: 1
Sampling: Euler a-Karras
Steps: 20-30
CFG Scale: 5-7
Recommended
Keyword:
- cartoonvalley
- cartoonvalley,shiny skin
Negative (optional)
-worst quality, low quality, artist name,blurry, jpeg artifacts, lowres,bad anatomy,bad hands
If you like my work, you can buy me a coffee.
I also accept commissions