Skip to content

[L2Norm] Avoid recompilation for variable-length inputs #669

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

Merged
yzhangcs merged 4 commits into from
Dec 1, 2025
Merged

[L2Norm] Avoid recompilation for variable-length inputs #669

Changes from 3 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
8c63c19
remove unnecessary tl.constexpr signature for l2_norm kernel
retonym Nov 28, 2025
1165df2
restore NB as autotune key
retonym Dec 1, 2025
0b94179
Add do_not_specialize option to Triton kernels
yzhangcs Dec 1, 2025
3d8af38
Merge branch 'main' into yunfei/l2norm_recompile_varlen
yzhangcs Dec 1, 2025
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
8 changes: 4 additions & 4 deletions fla/modules/l2norm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,13 +83,13 @@ def l2norm_bwd_kernel1(
key=['D', 'NB'],
**autotune_cache_kwargs,
)
@triton.jit
@triton.jit(do_not_specialize=['T'])
def l2norm_fwd_kernel(
x,
y,
rstd,
eps,
T: tl.constexpr,
T,
D: tl.constexpr,
BD: tl.constexpr,
NB: tl.constexpr,
Expand Down Expand Up @@ -117,14 +117,14 @@ def l2norm_fwd_kernel(
key=['D', 'NB'],
**autotune_cache_kwargs,
)
@triton.jit
@triton.jit(do_not_specialize=['T'])
def l2norm_bwd_kernel(
y,
rstd,
dy,
dx,
eps,
T: tl.constexpr,
T,
D: tl.constexpr,
BD: tl.constexpr,
NB: tl.constexpr,
Expand Down
Loading