CUDA: fuse adds, fuse add with rms norm #15631
CUDA: fused add with rms_norm_mul
fbbd94c9
Non-broadcast fuse works
2dcc02d1
Add fused adds
69bcd48c
format
4d105783
am17an
force pushed
to
4d105783
293 days ago
Remove n_fuse from template params
5adf50ed
am17an
force pushed
293 days ago
Address review comments
b64ba1cc
am17an
force pushed
to
b64ba1cc
293 days ago
Move template inside binbcast
f4488188
am17an
merged
009b709d
into master 293 days ago
am17an
deleted the rms_norm_fused_add branch 293 days ago
CISC
commented
on 2025-08-29
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub