SATchives
Well-known member
- Joined
- Jul 16, 2019
- Messages
- 1,376
I'm not very proficient in coding, I only know enough to put things together sometimes.
I was experimenting with A.I text to image. in this case specifically Stable Diffusion and dreambooth, recently someone shared the ability to train your images into the model using jewgles dreambooth methods, and I noticed this. per_img_token_list = [ hebrew dead powerless filth ] on the link below
https://github.com/JoePenna/Dreambooth-Stable-Diffusion/blob/main/ldm/data/personalized.py
it seems to me they are training the models and mixing in their shit jewish witchcraft into the a.i training process
if anyone more familiar with A.I and python than me wants to take a look and see if the tokens can be changed. i can post more links, and what I currently know.
I couldn't find any jewbrew filth on the stable diffusion yet, only this code responsible for training a model. which makes me think that the original latent space model itself is trained this way.
I only posted this because I think we all should be aware of the possibility of any A.I program being trained this way.
I was exploring this technology as it could be very useful in creating content, and now I'm disgusted.
I was experimenting with A.I text to image. in this case specifically Stable Diffusion and dreambooth, recently someone shared the ability to train your images into the model using jewgles dreambooth methods, and I noticed this. per_img_token_list = [ hebrew dead powerless filth ] on the link below
https://github.com/JoePenna/Dreambooth-Stable-Diffusion/blob/main/ldm/data/personalized.py
it seems to me they are training the models and mixing in their shit jewish witchcraft into the a.i training process
if anyone more familiar with A.I and python than me wants to take a look and see if the tokens can be changed. i can post more links, and what I currently know.
I couldn't find any jewbrew filth on the stable diffusion yet, only this code responsible for training a model. which makes me think that the original latent space model itself is trained this way.
I only posted this because I think we all should be aware of the possibility of any A.I program being trained this way.
I was exploring this technology as it could be very useful in creating content, and now I'm disgusted.