I’m not sure what AI OP is using – he didn’t indicate – but I think that he might be using Stable Diffusion, which one can run locally.
If you look at OP’s images from about five months back, they were 512x512. Now they’re 1024x1024.
That roughly coincides with Stable Diffusion’s model switching from a base model trained on 512x512 images to SDXL, which is trained on 1024x1024.
Stable Diffusion doesn’t have restrictions as such. The base model isn’t trained on nudes (though I suspect that it’d have no trouble creating something like OP’s image), but people have trained derived models on all sorts of things.
howd u get it to generate this? for me stuff like this trips the content safety filter
I’m not sure what AI OP is using – he didn’t indicate – but I think that he might be using Stable Diffusion, which one can run locally.
If you look at OP’s images from about five months back, they were 512x512. Now they’re 1024x1024.
That roughly coincides with Stable Diffusion’s model switching from a base model trained on 512x512 images to SDXL, which is trained on 1024x1024.
Stable Diffusion doesn’t have restrictions as such. The base model isn’t trained on nudes (though I suspect that it’d have no trouble creating something like OP’s image), but people have trained derived models on all sorts of things.
Post description says “DallE 3”. Thats why I’m asking.
My older posts were Stable Diffusion but I am branching out into other models at this time.