Skip to content

Fix edge checking in attention mask #3002

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
jagrit06 wants to merge 2 commits into
base: main
Choose a base branch
from
+14 −8
Open

Fix edge checking in attention mask #3002

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
097abca
Fix edge checking in attention mask
jagrit06 Jan 15, 2026
12f766e
Fix edge checking in attention mask
jagrit06 Jan 15, 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
22 changes: 14 additions & 8 deletions mlx/backend/metal/kernels/steel/attn/kernels/steel_attention_nax.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -349,14 +349,20 @@ template <
const short col_pos = base_col + ik * UK + sn;

MSubTile mfrag;
mfrag.load_safe(
mask,
int64_t(mask_params->M_strides[2]),
Int<1>{},
params->qL,
params->kL,
row_pos,
col_pos);
if ((!align_Q && is_last_q) || (!align_K && is_last_k)) {
mfrag.load_safe(
mask,
int64_t(mask_params->M_strides[2]),
Int<1>{},
params->qL,
params->kL,
row_pos,
col_pos);
} else {
const int M_str = int(mask_params->M_strides[2]);
Copy link
Member

@awni awni Jan 16, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That changed to int from int64_t. Is that reasonable?

const int M_load_off = row_pos * M_str + col_pos;
mfrag.load(mask + M_load_off, M_str, Int<1>{});
}

thread auto& fg = Ptile.subtile_at(iq, ik).frag_at(0, 0);

Expand Down
Loading