Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
FlashMLA
Commits
84f83034
Commit
84f83034
authored
Mar 13, 2026
by
zhanghj2
Browse files
fix h_q < 16 bug
parent
40f4bf39
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
3 additions
and
2 deletions
+3
-2
csrc/api/sparse_decode.h
csrc/api/sparse_decode.h
+3
-2
No files found.
csrc/api/sparse_decode.h
View file @
84f83034
#pragma once
#pragma once
#include "common.h"
#include "common.h"
#include <cutlass/fast_math.h>
#include "params.h"
#include "params.h"
...
@@ -60,13 +61,13 @@ public:
...
@@ -60,13 +61,13 @@ public:
Arch
arch
=
Arch
();
Arch
arch
=
Arch
();
if
(
h_q
<=
16
)
{
if
(
h_q
<=
16
)
{
return
{
return
{
std
::
max
(
arch
.
num_sms
*
2
/
s_q
/
(
h_q
/
16
),
1
),
std
::
max
(
arch
.
num_sms
*
2
/
s_q
/
cutlass
::
ceil_div
(
h_q
,
16
),
1
),
5
,
5
,
64
64
};
};
}
}
return
{
return
{
std
::
max
(
arch
.
num_sms
/
s_q
/
(
h_q
/
64
),
1
),
std
::
max
(
arch
.
num_sms
/
s_q
/
cutlass
::
ceil_div
(
h_q
,
64
),
1
),
5
,
5
,
64
64
};
};
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment