This Lora is an envy of mine to make a Coco. Even if there is very good loras about her (and I use them), it's not the same as to make it myself.

SD Pony : Yes, I know Pony already knows bandicoots. Your point is ?

Remember : With Pony, it's better to start your prompt with :

score_9, score_8_up, score_7_up, score_6_up, source_furry BREAK

and put your "real" prompt just after (as explained by Pony's author)

SDXL : SDXL version. Apparently, if Pony already has a Coco, that's not the case of most SDXL models. Then, here she is.

Same prompts than SD 1.5,

anthro, furry can force some non-furry SDXL checkpoints to do furries. Depends of checkpoints.

SD 1.5 Retrained : The all-new version, retrained with my all-new knowledge about training (I'll probably retrain Cleo with that, too).

Changed the points :

• Added one image (If memory serve)

• corrected some images in Krita/Gimp/'toshop

• Re-did the captioning

• Patched the numerous captioning errors, too (discovered Grep is a great help, here. grep "rectangle" *.txt and such works well

• Read a lot and re-did the settings.

• Choose Anylora as training model, in place of SD - Anime

• Trained a few times (My buzz ! I loved you so much !)

Tips :

• Using a weight of 0.75 is largely enough. Coco is a character without a lot of variations. and using 1 will make even her hair color unchangeable. It's better to weight more the triggers if she doesn't show well (like in my example in the snow.

• Generation with DPM++ 2M / SDE and Automatic / Karras works well

• 20 - 25 sampling steps should be largely enough.

• Upscaling upgrade the image a lot. I use R-ESRGAN 4x+ Anime6B, hires steps 0, Denoising strength 0.35, upscale by 2, and no refiner.

Triggers :

• cocobandicoot, but can works without. You try.

• COLOR eyes. Do goods eyes, but a "perfect COLOR eyes" is even better.

• COLOR hair. At LORA weight 0.75, works well. At 1, her hair are generally blonde, sometimes purple.

• bangs (the hair on the face), sidelocks (the hairs on the side), ponytail (the ponytail) are all captioned, as they are not present on all training images. Meaning they would be able to not appear if not prompted, or used in the negative prompt. Hats have a tendancy to mess with the bangs (as real hats do).

• bangs, sidelocks and ponytails can also be used with COLOR, but the result is random, for some reason.

• COLOR body fur, COLOR belly. Coco is so HEAVILY orange with a beige belly that is nearly impossible to change it, but I've still captioned it (a few training images with other colors)

• overalls. understand "naked overalls", too. Captioned "COLOR", but almost never works (Coco always wears a blue, after all)

• COLOR shirt. this one works pretty well (just put the LORA weight at 0.75. At 1, the one she wears is generaly white).

• hair ornament is often her flower. COLOR hair flower can be added (by default, its colors is like blue, pink or red).

• All images have been tagged with anthro and furry. The lora should works well with non-furry anime checkpoints, but you tell me.

• goggles on head

SD 1.5 - V1 : My first version.

You can choose it if you prefer, but it has a lots of errors in the captioning, and defects in the training images.