I am trying to improve my approach to GPUs for Local AI and could use some practical advice.
I am particularly interested in limitations and customisation. I would like to compare approaches that people have actually used rather than just theoretical suggestions.
What has worked best for you...
