andybug@lemmy.worldtoTechnology@lemmy.ml•New ROCm™ 5.6 Release Brings Enhancements and Optimizations for AI and HPC WorkloadsEnglish
2·
1 year agoI’ve used ROCm for pytorch and stable diffusion. Using the pre-built docker images is definitely the way to go. Once setup, it’s worked great for me.
You just have to understand that most projects are going to assume Nvidia (CUDA), so you’ll have to jump through more hoops and still may not be able to do everything.
I use
chezmoi
for tracking dotfiles and used to use a fancy Ansible setup. Now, I just occasionally backup a list of explicitly installed packages and track the major changes made to the system in a simple Markdown file.