The schnell live demo is impressive. My decade-old rig takes minutes per step on Flux, so this is actually a useful way for me to check for known concepts in almost real-time. It even had time to show me a quick Phở when I was typing in photo.
comfy
joined 3 years ago
Not communist obviously
I find it's useful to select more descriptive terms than use the literal dozens of varying definitions of 'socialism' and 'communism'. The terms by themselves can be so vague that I can truthfully state this - "communism is the goal of communism!" A communist society, for example, is different from a communist party or a communist state (aka. Marxist–Leninist state), which are only parts of the communist movement and the communist school of thought. Obviously no-one looks at the PRC and sees a stateless, classless society, but that's an understandable (albeit condescending) interpretation of when people say "China is communist".
(Pinging @xnx@slrpnk.net as I'm also replying to their comment)
I didn't even think of dual cards, because I have an old & budget motherboard with one slot. But 2 x 16GB GPUs and a new motherboard (and if necessary, new CPU) and PSU and it might even still be cheaper than a 24GB NVIDIA for me. Of course I'd have to explore the trade-offs in detail because I've never looked into how dual cards work.
(but truth be told, I just as easily could settle for a 1x16 GB if I'm confident it would be able to train, even if slowly, AuraFlow or FLEX LoRas for the upcoming Pony v7 model. It's just a hobby.)
I am using a lot of Pony* models at the moment and Pony v7 looks like it will switch to AuraFlow or FLUX [1] so it's useful to hear your experience with it on a 3080Ti.
Good to know about CUDA/Direct ML.
I found a couple of 2022 posts recommending 3090s, especially since cryptocoin miners were selling lots of them cheap. Thanks for the heads up about the 5000 release, I suspect it will be above my budget but it will net me better deals on a 4090 :P
Changed my mind lol
I haven't determined that. I only have one device set up to run SD and haven't organized any test with someone else.
I mean, if the concern is that tiling is a factor relative to non-tiling, okay, but if someone else is tiling, l’d think that they’d get the same output.
That's true, I'll check to see if the metadata mentions the tiling was used.
Good call-out. My (naïve) understanding is that tools like tiling VAE to handle low VRAM, and lowing steps in the more stable of the samplers, are going to have a generally negative impact on the result, and a very similar image with better detail could be remade using similar variables on better hardware. Maybe that's a bit idealistic. Like you said, the seed mode usually changes images with size. (You said 'usually', is there a way to minimize this?)
edit: I'm aware 'better' and 'higher quality' are vague and even subjective terms. But I'm trying to convey something beyond merely higher resolution.
Sure, I agree, but at the end of the day it's useful to be able to search and watch YouTube videos so long as it's a popular platform because it still has by far the bulk of topics covered.
No contradiction. Law is a dumb basis for deciding what you like and don't like.
IMO, the worst thing about "Minetest" is that it sounded like it was just a test creation, a prototype or experiment. It's certainly well beyond that now. The announcement introduction mentions people associate it with being a Minecraft clone or alpha release, but even further, to me the name initially gave me the impression it was [still] someone's small hobby project. 'Luanti' is much better.