Skip to content
Projects
Groups
Snippets
Help
This project
Loading...
Sign in / Register
Toggle navigation
T
tic
Overview
Overview
Details
Activity
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Board
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Charts
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
wenyuanbo
tic
Commits
464ebb13
Commit
464ebb13
authored
Nov 20, 2019
by
Animesh Jain
Committed by
Zhi
Nov 20, 2019
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
[QNN] Lowering for Depthwise Convolution. (#4351)
parent
2672aad4
Expand all
Hide whitespace changes
Inline
Side-by-side
Showing
4 changed files
with
139 additions
and
9 deletions
+139
-9
src/relay/pass/pattern_util.h
+2
-0
src/relay/qnn/op/convolution.cc
+0
-0
tests/python/relay/test_op_qnn_conv2d.py
+132
-9
topi/python/topi/x86/conv2d_alter_op.py
+5
-0
No files found.
src/relay/pass/pattern_util.h
View file @
464ebb13
...
...
@@ -503,6 +503,8 @@ static inline Expr Tile(Expr data, Array<Integer> reps) {
Expr
MakeConcatenate
(
Expr
data
,
int
axis
);
Expr
MakeRepeat
(
Expr
data
,
int
repeats
,
int
axis
);
Expr
MakeStridedSlice
(
Expr
data
,
Array
<
Integer
>
begin
,
Array
<
Integer
>
end
,
Array
<
Integer
>
strides
);
Expr
MakeStack
(
Expr
data
,
int
axis
);
...
...
src/relay/qnn/op/convolution.cc
View file @
464ebb13
This diff is collapsed.
Click to expand it.
tests/python/relay/test_op_qnn_conv2d.py
View file @
464ebb13
This diff is collapsed.
Click to expand it.
topi/python/topi/x86/conv2d_alter_op.py
View file @
464ebb13
...
...
@@ -197,6 +197,11 @@ def _conv2d_legalize(attrs, inputs, arg_types):
if
not
(
dilation
[
0
]
==
1
and
dilation
[
1
]
==
1
):
return
None
# No legalization for depthwise convolutions yet.
groups
=
attrs
.
get_int
(
"groups"
)
if
groups
!=
1
:
return
None
# Collect the input tensors.
data_tensor
,
kernel_tensor
=
arg_types
[
0
],
arg_types
[
1
]
data_dtype
=
data_tensor
.
dtype
...
...
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