kernl
kernl copied to clipboard
feat: add bw pass on layernorm/rmsnorm
I notice this PR has been open for a while; what's its status?
@samhavens hi, no real issue, just focusing on inference right now... Is it something you are interested into?
Not having to deal with CUDA for an RMSnorm kernel is appealing, yeah 😄 it's not high priority currently, but I wanted to make sure to keep tabs on this PR 👀