Hello everyone. I am Dr. Furkan Gözükara. PhD Computer Engineer. SECourses is a dedicated YouTube channel for the following topics : Tech, AI, News, Science, Robotics, Singularity, ComfyUI, SwarmUI, ML, Artificial Intelligence, Humanoid Robots, Wan 2.2, FLUX, Krea, Qwen Image, VLMs, Stable Diffusion
Just a quick question on Supir- it's amazing. Running locally on a RTX3080, and it takes forever. Can you use multi-gpu with Supir, so I can throw batches up on Massed compute? Is the speed more a function of gpu compute or vram?
I'm quite happy with it, not to mention it looks fancy, got a G.Skill Trident Z Royal kit cuz I figured if I'm gonna put out the money for 96GB of RAM might as well get the fancy looking one with the good timings lol Question though, how much RAM usage do you standardly see while running Flux?
It is bright as hell though, and needs active cooling for what settings in pushing it to already. Haven't stability tested it yet but I'll do that on my days off. So far it seems rock sold. I'm only doing CL30 at 6200MHz 1.35v, the kit is rated for CL32 at 6400MHz 1.35v but I've yet to get 6400 to run stable because of AMD's fun memory controller stuff.
I am using this tutorial in runpod, when i try to load "Rank_1_28700MB_Slow.json" i get connection error and i can't proceed any further. No matter how many times i try.