Release the full power of AMD GPU using Linux!
👉ⓢⓤⓑⓢⓒⓡⓘⓑⓔ Thank you for watching! please consider to subscribe. thank you!
👉🏽Update: Mar 23. Please join the discord server at [ Ссылка ] where you can ask questions and also try out my recently developed free chatgpt bot and stable diffusion bot!
If you would like to support me, here is my Kofi link: [ Ссылка ] and Patreon page: [ Ссылка ]

Affiliate links: buy on Amazon
4600G [ Ссылка ]
5600G: [ Ссылка ] (same iGPU, better CPU)
5700G: [ Ссылка ] (better iGPU, and better CPU)
ssd drive: [ Ссылка ]
DDR4 drive: [ Ссылка ]
AM4 motherboard: [ Ссылка ]
PSU (power supply unit): [ Ссылка ]
PC Case: [ Ссылка ]
if you are interested in discrete GPU: [ Ссылка ]

NOTES: consumer GPUs are in general not supported by ROCm, so YMMV.
The GPU model I use in tutorial: 6700XT. It can achieve faster than 6.3 it/s. It should also apply to other Radeon RX 6000 series such as 6500,6600,6750,6800,6900,6950, etc.
Linux (Ubuntu) installation tutorials: [ Ссылка ]
Linux vs Windows stable diffusion: [ Ссылка ]
0:6 My Kofi link. thank you!
0:22 Intro.
2:00 Install ROCm 5.4.2. AMD ROCm 5.4.2 version official installation guide: [ Ссылка ]
16:53 Install Anaconda, and create python virtual environment
20:12 Install Pytorch-ROCm
26:55 Install diffusers, transformers, accelerate
29:30 test run for stable diffusion
35:20 results

свернуть