4 ms·
Bananas. Thanks so much... to everyone involved. It works. 14 seconds to generate an image on an M1 Max with the given instructions (`--n_samples 1 --n_iter 1`
by gregsadetsky 4y ago
Bananas. Thanks so much... to everyone involved. It works.
14 seconds to generate an image on an M1 Max with the given instructions (`--n_samples 1 --n_iter 1`)
Also, interesting/curious small note: images generated with this script are "invisibly watermarked" i.e. steganographied!
See https://github.com/bfirsh/stable-diffusion/blob/main/scripts/txt2img.py#L253 https://github.com/bfirsh/stable-diffusion/blob/main/scripts...
- grishka 4y ago> Also, interesting/curious small note: images generated with this script are "invisibly watermarked" i.e. steganographied! Why?
- Retr0id 4y agoSo that future iterations of StableDiffusion (or similar models) don't end up getting trained on their own outputs.
- hifikuno 4y agoOh wow, I didn't even think of that. I am pretty sure a few of the repo's have turned off the invisible watermark, I wonder if that will have consequences down the line for training data.
- gregsadetsky 4y ago... so this means that watermarking an image you own is probably the only way to avoid it being used for training further models? :-)