Skip to content

Fix mixed_input_fmha_decode example #2986

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
anakinxc wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix mixed_input_fmha_decode example #2986

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3953e10
Fix mixed_input_fmha_decode example
anakinxc Jan 26, 2026
8557ce4
Update mixed_input_fmha_decode.py
anakinxc Jan 26, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion examples/python/CuTeDSL/blackwell/mixed_input_fmha/mixed_input_fmha_decode.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1163,7 +1163,7 @@ def decode(
tSrM = cute.make_rmem_tensor_like(tSsM)
tSrM_lane = Float32(0) # Avoid dynamic register indexing
for i in cutlass.range_constexpr(cute.size(tSrS)):
tSrM[i] = warp_fmax(tSrS[i])
tSrM[i] = self.warp_fmax(tSrS[i])
if i == lane_idx:
tSrM_lane = tSrM[i]

Expand Down Expand Up @@ -1453,6 +1453,7 @@ def warp_fmax(val : Float32):
)
return Float32(ret)

@staticmethod
@cute.jit
def smem_fmax(ptr : Pointer, val : Float32):
# https://stackoverflow.com/a/72461459
Expand Down