Now we just need some version of this to escape into the wild like Stable Diffusion did so the technology doesn't end up suppressed by transgender communists.
This is in my area of interest. I may end up doing a project of my own once I become proficient in this area. If I do it the licensing will be a free for all so I'm sure it will end up screaming about niggers with enough prompting.
Getting training data is often more than half the battle. I'm studying the computer vision side of things myself, but there's likely a decent amount of crossover that's relevant to NLP.
Not anymore, they have been taking out sexually explicit stuff and ruined their 2.0 because of it. It's not going to stop people from making their own with what we have now though.
A group of users who don't like the censorship/licensing policies just started soliciting funding to develop an alternative model called "Unstable Diffusion".
Now we just need some version of this to escape into the wild like Stable Diffusion did so the technology doesn't end up suppressed by transgender communists.
This is in my area of interest. I may end up doing a project of my own once I become proficient in this area. If I do it the licensing will be a free for all so I'm sure it will end up screaming about niggers with enough prompting.
Give it the +NIGGER License.
The only license that guarantees a project will remain based.
The N world is the only indication that somebody is free to speak.
(yes - I am in a jurisdiction that would arrest me for saying it)
You are doing the Lord's work.
Getting training data is often more than half the battle. I'm studying the computer vision side of things myself, but there's likely a decent amount of crossover that's relevant to NLP.
Stable diffusion is based!
Not anymore, they have been taking out sexually explicit stuff and ruined their 2.0 because of it. It's not going to stop people from making their own with what we have now though.
Thank God for other models, SD is just one.
A group of users who don't like the censorship/licensing policies just started soliciting funding to develop an alternative model called "Unstable Diffusion".
Jup, heard of that one. Unsure how it'll turn out but I have my eye on it. Until then I will still use the other models that are out there.
Is there a list of still-based image AIs?