Catgirl version 1.0

Trained 600+ sfw catgirl cosplay & catgirl lingerie model images on base model SD v1.5

Preprocess. BLIP did 100% the captions on each training image. I didn't edit not one caption, BLIP wrote whatever BLIP wanted on 600+ training images.

This is my first textual Inversion ever, so it's not perfect. More like an experiment to see if I could even do it.

Catgirl v1.0 turned out better then I imagined. Try it.

See what you can create yourself with Catgirl v1.

Comment & rate below if you like it.

I'm working on Catgirl v1.1 now and using another custom model and not using the base model SD v1.5 to train the textual embedding this time.

Catgirl v1.2. I might edit the 600+ training image captions with accurate details.

Catgirl v1.3. I might hand pick the best 50 training images with detailed captions.

Instructions on how to use Catgirl v1.0

1. Don’t use base models to generate your images. Look at the example images to see what Model I used. Custom models work well with Catgirl v1. Not all models work well, you got to find the right checkpoint model and try it with Catgirl.

2. You can install this textual embedding and then just use "catgirlv1" in your positive prompt box and you will get your Catgirl in her room with no other positive prompts required.

3. Want more? Catgirl v1 loves detailed & long prompts though. Catgirl standing by a Ferrari, won't be that great. Prompt with more details for the best results.