Fama
Well-known member
A new model called Flux has been released this month. Great image fidelity and prompt following, not yet totally sold on how cool and artistic pictures it can generate compared to SDXL, but it follows prompts much better, does well with hands and can actually handle text quite well.
It's heavier than SDXL I think, but I can run it locally with a RTX 3070 using https://github.com/lllyasviel/stable-diffusion-webui-forge and https://huggingface.co/lllyasviel/flux1-dev-bnb-nf4/tree/main (I use v1 since apparently it's a little lighter than v2).
Some examples, not sure if I made the mistake in the text or if the AI model did:
One of these will probably be used for the Stoner Bee (in case you didn't know, the AR-15 was designed by Eugene Stoner).
Maybe something like this for the Mofeta (why yes, I do like cool looking castles, who doesn't?).
I did have a nice generation session with Fooocus too, not sure if these are useful for pedals but they definitely look cool:
This one (or a variation of it) could be used for a project I've been cooking up:
It's heavier than SDXL I think, but I can run it locally with a RTX 3070 using https://github.com/lllyasviel/stable-diffusion-webui-forge and https://huggingface.co/lllyasviel/flux1-dev-bnb-nf4/tree/main (I use v1 since apparently it's a little lighter than v2).
Some examples, not sure if I made the mistake in the text or if the AI model did:
One of these will probably be used for the Stoner Bee (in case you didn't know, the AR-15 was designed by Eugene Stoner).
Maybe something like this for the Mofeta (why yes, I do like cool looking castles, who doesn't?).
I did have a nice generation session with Fooocus too, not sure if these are useful for pedals but they definitely look cool:
This one (or a variation of it) could be used for a project I've been cooking up:
