This model imitates the style of Pixar cartoons.
Works only with people. No animals, objects or backgrounds.
It is advisable to use additional prompts and negative prompts.
Trained on 70 images.
Description
FAQ
Comments (31)
def gonna get this when the safetensor drops
Safetensors loaded!
I'm sorry. This is my first model
Good model, but I'd like to see a version more colourful.
Thanks for the feedback! You can use special prompts to get more colorful results. For example, add "vivid" to positive prompts and "muted colors" to negative prompts. I also advise you to avoid the "cinematic look" prompt, because it also dims the colors.
cool！
Would you make a LoRA instead? :)
Sorry, I don't have the opportunity to make a LoRA
Cross-eyed are more serious
What's the prompt word to make the character watching to the camera(viewer)
Try any or all of these: Eye Contact, Looking At Viewer, and Looking at Camera/Photographer, and messing with the weights () and positioning in your prompt.
5/5
a brilliant checkpoint, just little prompts but the production is awesome!
Thanks for the author!
Thank you for your work.
Was it trained on vanilla SD 1.5, or some mix?
It's important in the case of Add Difference merging with it.
You're welcome! I used "delibrate" for training this model
@KittyKnight
Deliberate 2?
@alexds9 Deliberate 1 if I'm not mistaken
@KittyKnight
Thank you!
@KittyKnight Does it mean you finetune based on "deliberate"?
@StablePika Um... yes
@KittyKnight @KittyKnight In the description, it says the base model is SD1.5, so I am kinda confused if you finetuned on "deliberate" or you finetuned on SD1.5 and merge it with "deliberate"
@StablePika
The options are only SD main versions like: 1.4, 1.5 and 2.1.
So he can't choose custom checkpoint.
Most checkpoints are mixes of 1.5 and 1.4 anyway, so I don't think this field is helpful at all.
@alexds9 Thanks, that explains.
What sampler did you use to gen these?
DPM++ 2M Karras
The SD scaling effect of the lora I trained is not good. Is there any way to solve it? To make the picture clearer.
Why do all the sample images look disappointed. Like I just farted and they heard me.
I read this and then got to the Batman image and actually teared up laughing hahaha
Good for text2image. Seemingly unworkable for image2image. But still excellent work!
Add onsite generation please 🫥
can I use this for my NSFW games / comics?
Do u have a py example of how this works? that is my first time working with this kind of feature. can u helpme shownwing how to use?
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.