GitLab

If we try to load the CUDA library on an old GPU, it panics and crashes
the server.  This checks the compute capability before we load the
library so we can gracefully fall back to CPU mode.

* gpu: read memory info from all cuda devices

* add `LOOKUP_SIZE` constant

* better constant name

* address comments

This refines the gpu package error handling and fixes a bug with the
system memory lookup on windows.

If someone checks out the ollama repo and doesn't install the CUDA
library, this will ensure they can build a CPU only version