Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
gaoqiong
composable_kernel
Commits
9904929b
Commit
9904929b
authored
Aug 11, 2022
by
Jing Zhang
Browse files
fixed splitK
parent
e731e8c0
Changes
4
Show whitespace changes
Inline
Side-by-side
Showing
4 changed files
with
3 additions
and
8 deletions
+3
-8
example/15_grouped_gemm/grouped_gemm_xdl_bfp16.cpp
example/15_grouped_gemm/grouped_gemm_xdl_bfp16.cpp
+1
-1
example/24_batched_gemm/batched_gemm_xdl_bfp16.cpp
example/24_batched_gemm/batched_gemm_xdl_bfp16.cpp
+1
-1
include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
...operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
+1
-1
include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp
...tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp
+0
-5
No files found.
example/15_grouped_gemm/grouped_gemm_xdl_bfp16.cpp
View file @
9904929b
example/24_batched_gemm/batched_gemm_xdl_bfp16.cpp
View file @
9904929b
include/ck/tensor_operation/gpu/device/device_gemm_xdl_splitk_c_shuffle.hpp
View file @
9904929b
...
@@ -95,7 +95,7 @@ struct DeviceGemmXdlSplitKCShuffle : public DeviceGemmSplitK<ALayout,
...
@@ -95,7 +95,7 @@ struct DeviceGemmXdlSplitKCShuffle : public DeviceGemmSplitK<ALayout,
const
auto
a_grid_desc_m_kpad
=
transform_tensor_descriptor
(
const
auto
a_grid_desc_m_kpad
=
transform_tensor_descriptor
(
a_grid_desc_m_k
,
a_grid_desc_m_k
,
make_tuple
(
make_
right_pad_transform
(
K
,
KPad
-
K
),
make_pass_through_transform
(
M
)),
make_tuple
(
make_
pass_through_transform
(
M
),
make_right_pad_transform
(
K
,
KPad
-
K
)),
make_tuple
(
Sequence
<
0
>
{},
Sequence
<
1
>
{}),
make_tuple
(
Sequence
<
0
>
{},
Sequence
<
1
>
{}),
make_tuple
(
Sequence
<
0
>
{},
Sequence
<
1
>
{}));
make_tuple
(
Sequence
<
0
>
{},
Sequence
<
1
>
{}));
...
...
include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp
View file @
9904929b
...
@@ -547,11 +547,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
...
@@ -547,11 +547,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
static_cast
<
FloatC
*>
(
p_shared_block
),
static_cast
<
FloatC
*>
(
p_shared_block
),
c_block_desc_mblock_mperblock_nblock_nperblock
.
GetElementSpaceSize
());
c_block_desc_mblock_mperblock_nblock_nperblock
.
GetElementSpaceSize
());
static_assert
(
M1
==
MWave
,
""
);
static_assert
(
N1
==
NWave
,
""
);
static_assert
(
M2
*
M3
*
M4
==
MPerXDL
,
""
);
static_assert
(
N2
==
NPerXDL
,
""
);
constexpr
auto
c_block_desc_m0_n0_m1_n1_m2_m3_m4_n2
=
transform_tensor_descriptor
(
constexpr
auto
c_block_desc_m0_n0_m1_n1_m2_m3_m4_n2
=
transform_tensor_descriptor
(
c_block_desc_mblock_mperblock_nblock_nperblock
,
c_block_desc_mblock_mperblock_nblock_nperblock
,
make_tuple
(
make_tuple
(
...
...
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