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_ROCM
Commits
e493ab00
Commit
e493ab00
authored
Dec 23, 2024
by
coderfeli
Browse files
add one
parent
50769720
Changes
2
Expand all
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
50 additions
and
48 deletions
+50
-48
example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8.cpp
...gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8.cpp
+3
-1
library/src/tensor_operation_instance/gpu/gemm_multiply_multiply/device_gemm_multiply_multiply_xdl_f8_f8_bf16/device_gemm_multiply_multiply_xdl_f8_f8_bf16_mk_nk_mn.hpp
...device_gemm_multiply_multiply_xdl_f8_f8_bf16_mk_nk_mn.hpp
+47
-47
No files found.
example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8.cpp
View file @
e493ab00
...
@@ -83,7 +83,9 @@ using DeviceOpInstance = ck::tensor_operation::device::DeviceGemmMultiD_Xdl_CShu
...
@@ -83,7 +83,9 @@ using DeviceOpInstance = ck::tensor_operation::device::DeviceGemmMultiD_Xdl_CShu
// kernel 1: 256->32x128x128
// kernel 1: 256->32x128x128
// < Row, Col, DsLayout, ELayout, A0DataType, B0DataType, DsDataType, EDataType, AccDataType, CShuffleDataType, AElementOp, BElementOp, CDEElementOp, GemmSpec, 256, 32, 128, 128, 16, 16, 32, 32, 1, 1, S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, 1, 1, S<1, 32, 1, 8>, S<8, 8, 1>, ck::BlockGemmPipelineScheduler::Interwave, ck::BlockGemmPipelineVersion::v1, FP8>;
// < Row, Col, DsLayout, ELayout, A0DataType, B0DataType, DsDataType, EDataType, AccDataType, CShuffleDataType, AElementOp, BElementOp, CDEElementOp, GemmSpec, 256, 32, 128, 128, 16, 16, 32, 32, 1, 1, S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, 1, 1, S<1, 32, 1, 8>, S<8, 8, 1>, ck::BlockGemmPipelineScheduler::Interwave, ck::BlockGemmPipelineVersion::v1, FP8>;
// kernel 2: 128->32x128x128
// kernel 2: 128->32x128x128
<
Row
,
Col
,
DsLayout
,
ELayout
,
A0DataType
,
B0DataType
,
DsDataType
,
EDataType
,
AccDataType
,
CShuffleDataType
,
AElementOp
,
BElementOp
,
CDEElementOp
,
GemmSpec
,
128
,
32
,
128
,
128
,
16
,
16
,
32
,
32
,
1
,
2
,
S
<
8
,
16
,
1
>
,
S
<
1
,
0
,
2
>
,
S
<
1
,
0
,
2
>
,
2
,
16
,
16
,
0
,
S
<
8
,
16
,
1
>
,
S
<
1
,
0
,
2
>
,
S
<
1
,
0
,
2
>
,
2
,
16
,
16
,
0
,
1
,
1
,
S
<
1
,
16
,
1
,
8
>
,
S
<
2
,
2
,
1
>
,
ck
::
BlockGemmPipelineScheduler
::
Interwave
,
ck
::
BlockGemmPipelineVersion
::
v1
,
FP8
>
;
// < Row, Col, DsLayout, ELayout, A0DataType, B0DataType, DsDataType, EDataType, AccDataType, CShuffleDataType, AElementOp, BElementOp, CDEElementOp, GemmSpec, 128, 32, 128, 128, 16, 16, 32, 32, 1, 2, S<8, 16, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, S<8, 16, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0, 1, 1, S<1, 16, 1, 8>, S<8, 8, 1>, ck::BlockGemmPipelineScheduler::Interwave, ck::BlockGemmPipelineVersion::v1, FP8>;
<
Row
,
Col
,
DsLayout
,
ELayout
,
A0DataType
,
B0DataType
,
DsDataType
,
EDataType
,
AccDataType
,
CShuffleDataType
,
AElementOp
,
BElementOp
,
CDEElementOp
,
GemmSpec
,
256
,
32
,
128
,
128
,
16
,
16
,
32
,
32
,
1
,
1
,
S
<
8
,
32
,
1
>
,
S
<
1
,
0
,
2
>
,
S
<
1
,
0
,
2
>
,
2
,
16
,
16
,
0
,
S
<
8
,
32
,
1
>
,
S
<
1
,
0
,
2
>
,
S
<
1
,
0
,
2
>
,
2
,
16
,
16
,
0
,
1
,
1
,
S
<
1
,
32
,
1
,
8
>
,
S
<
8
,
8
,
1
>
,
ck
::
BlockGemmPipelineScheduler
::
Intrawave
,
ck
::
BlockGemmPipelineVersion
::
v3
,
FP8
>
;
// clang-format on
// clang-format on
...
...
library/src/tensor_operation_instance/gpu/gemm_multiply_multiply/device_gemm_multiply_multiply_xdl_f8_f8_bf16/device_gemm_multiply_multiply_xdl_f8_f8_bf16_mk_nk_mn.hpp
View file @
e493ab00
This diff is collapsed.
Click to expand it.
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