Kernels
activation / tests

Commit History

feat: support sequence parallel with fused_add_rms_norm
151bb5a

wyldecat commited on

refactor(activation): change fused_add_rms_norm and fused_add_rms_norm_backward to out-place operations
7e4334d

wyldecat commited on

feat: support sequence parallel with rms_norm
06d6367

wyldecat commited on

feat: make rms_norm as out-place
9d0a235

wyldecat commited on

Fix fused add rms norm (#4)
a1e5ca8
unverified

TaehyunKim TaehyunKimMotif commited on

feat(rms-norm): Impl fused RMSNorm
f3b99fb

iamwyldecat commited on

feat(poly-norm): add perf test
d14fd4d

iamwyldecat commited on

fix(poly-norm): calc param grad explicitly
704692b

iamwyldecat commited on

chore(poly-norm): remove unnecessary file
552d415

iamwyldecat commited on

feat(poly-norm): Add PolyNorm
44e9845

iamwyldecat commited on