Comment by outlog

amazing work ♥!

any idea on M1 gpu support? (know it's mostly an upstream issue)

excited about "real"/intensive ML workflows locally..