111–120 of 199 posts
Why is "trending on artstation" a keyword that influences the ML model? seems weird
Earlier quoted context omitted.
how is this possible without a dedicated gPU? thought stable diffusion would require far more horsepower
They have integrated GPU that I believe Apple claimed was comparable to RTX 3090 (perhaps since debunked though or at the least maybe a misleading claim).
M1 ultra was compared to RTX 3090 which was a larger stretch.
The M1 max deliver about 10.5 tflops The M1 ultra about 21 tflops.
The desktop RTX 3080 delivers about 30 tflops and RTX 3090 about 40.
Apple’s comparison graph showed the speed of the M1s vs. RTXs at increasing power levels, with the M1s being more efficient at the same watt levels (which is probably true). However, since the graph stopped before the RTX GPUs reached full potential, the graph was somewhat misleading.
The M1 max and Ultra have extra video processing modules that make them faster than the RTX GPUs at some video tasks though.
Earlier quoted context omitted.
It works, well, incredibly.
how is this possible without a dedicated gPU? thought stable diffusion would require far more horsepower
Both of my M1 Macs only have 8G of RAM. Is it a waste of time trying to run this with 8G?
Yes, it's a waste of time. It works, kinda, sorta, but only after shutting down every other app, and even then occasionally it seems to fall over. The model is larger than 8GB in memory, so it's agonizing.
This really is a one-click installer, which is excellent. As far as I know it's the first and only one for M1. It's quite a polished UI, but there are a lot of features missing. The first ones that come to mind are: - No way to specify a seed. This is an important part of SD workflows, letting you redo an image with a slightly tweaked a prompt. - No way to specify a custom model. Alternative models (such as Waifu Dif…
Can it be built from source? (Am thinking if yes, then the seed feature should be easy to add).
For those of you following the lstein fork (especially the development branch), it has been making great progress. I'm not getting black images anymore and the speed has gone up significantly. Not one click install by a long shot, but the documentation is pretty clear to follow. Anyone with a bit of CLI experience can do it and if you don't have that, this is a great way to kind of stumble your way towards something…
As a slightly shameless plug, I've been hacking up a UI specifically for that fork with a focus on a more efficient workflow for image synthesis. A demonstration video can be found here: https://vimeo.com/748114237 If people have any sort of feedback I'd love to hear it, or if people have some specific features that are missing from the other UIs :)
This really is a one-click installer, which is excellent. As far as I know it's the first and only one for M1. It's quite a polished UI, but there are a lot of features missing. The first ones that come to mind are: - No way to specify a seed. This is an important part of SD workflows, letting you redo an image with a slightly tweaked a prompt. - No way to specify a custom model. Alternative models (such as Waifu Dif…
has anyone got a frok with img2img working for M1? I keep getting errors no matter what I try.
Earlier quoted context omitted.
I feel like I'm maybe missing some context here, because I've never had to install "conda Python" on any Mac ever. I had the system Python and a Homebrew-installed Python happily co-existing, and at various points have had "virtualized" Pythons installed via virtualenv or pipx. On Apple Silicon Macs there's no Python installed by default at all. Now, I've seen an awful lot of programs written in Python that decide to…
> On Apple Silicon Macs there's no Python installed by default at all. Not true. There's no python2 anymore, but there is a python3. You can find it at /usr/bin/python3 (most likely your homebrew install preceeds it in your PATH)