I haven’t tried it, but the outputs are pretty oversaturated, but don’t look that bad. It also should be more easily trainable than Flux, since they released models that weren’t distilled. You can see what people have uploaded to Civitai so far if you filter by SD 3.5.

It feels like all SAI models are poisoned in some way. I wonder what’s wrong with this one?

I’m about to install it too. I just switched to SwarmUI.

I was using this for a while on A1111. It seems to have a way to use it as a standalone executable.

Same. My A1111 started having problems I couldn’t deal with anymore, so I gave it a shot. I just miss my plugins.

I have a feeling this experiment would sooner get the axe than have ads injected. There was initially a waiting list, but just a few days in it was completely open to the public.

Sometimes they just do research. Like when their employees made transformers and nothing came of them until Open AI capitalized on it.

There are ads in them now? I didn’t encounter any when using it a few days ago.

I just generated one and didn’t hear any ads.

It’s only decipherable to people who have kept up with the last two years of IMG Gen jargon.

I found this guide on how to make an inpainting model out of any model. Though it’s pretty out of date.

Yeah, this seems like the last confirmation we didn’t really need.

What are his feelings on open source? That’s my question.

If you’re using the same UI and metadata, you should be able to reproduce images with only slight differences and then upscale them with hires fix or something else.

Doesn’t seem like it.

That’s kind of unbelievable given what they say it can do.

They said they would be open sourcing it.

That was really cool.

Those might just be LoRA merged models, not full fine-tuning. From what I heard, fine-tuning doesn’t work because the models are distilled. You’d have to find a way to undistill them to train them.

Last I heard, LoRAs cause catastrophic forgetting in the model, and full fine-tuning doesn’t really work.

I don’t think so. They’re going to have to do a lot better than a tutorial to win people back. That said, the two Flux models being distilled making them close to impossible to fine-tune sucks too.