Re:Lawsuit challenging Stable Diffusion filed against Devian...
by synu on Saturday January 14, 2023 @05:14AM
(#34378996)
The important distinction then is using another program or device to analyze the bits but without copying them, that takes its own new impression? Like using a camera?
Re:Lawsuit challenging Stable Diffusion filed against Devian...
by astrange on Saturday January 14, 2023 @05:27AM
(#34379083)
Well, theoretically more like a vague memory of it or taking notes on it.
Re:Lawsuit challenging Stable Diffusion filed against Devian...
by derangedHorse on Saturday January 14, 2023 @12:45PM
(#34382135)
“any fractional bits would be in there if you want to look.”
What do you mean by this in the context of generating images via prompt? “Fractional bits” don’t make sense and it’s more misleading if anything. Regardless, a model violating criteria for being within fair use will always be judged by the outputs it generates rather than its composing bytes (which can be independent)
Since SD is trained by gradient updating against several different images at the same time, it of course never copies any image bits straight into it. Since it's a latent-diffusion model, any actual "image"ness is limited to the VAE, so any fractional bits would be in there.
The text encoder (LAION OpenCLIP) does have bits from elsewhere copied straight into it to build the tokens list.
https://huggingface.co/stabilityai/stable-diffusion-2-1/raw/...