import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("nadanainone/istolemyownart", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]Trained Dreambooth model using all my own art currently uploaded to my Tumblr (https://nadanainone.tumblr.com/, link only for reference, imagine plugging your own work for clout l m a o) and drawn over the course of around 10 years. The results are an inconsistent mess of half animu, half cartoony and a complete artistic downgrade over using literally any other model. Can be massively improved by adding real artists to your prompts.
I love this model and the tech btw. Enjoy.
1076 images (including flipped copies), 10k steps, 1e-6 learning rate
Raw examples:
Examples using real artists in the prompts:
Token (not really needed I think): ludditesbtfo
Pair it with the included hypernetwork for even more awful results.
Does this give me and everyone else the rights over your entire art and dataset and all your future artworks both made manually and with AI?
lol no
Can I use anything generated with this model for whatever I want though?
Can't stop fair use and derivative works
Can I sell your model?
No, all models should always be free and open source
Can I train another model using your art?
Sure, it'll probably be better than mine
Do I own that model if I make it?
lol no, see above
Can I use this to impersonate you and make bad faith arguments to make MY stunning and brave anti AI argument win over your disgusting pro AI argument, chud?
lole
Well your art sucks!
Yes
- Downloads last month
- 10