Skip to content

[Bugfix] Conv3Dtranspose default kernel layout should be IODHW #14340

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

Merged
merged 10 commits into from
Mar 30, 2023
Merged

[Bugfix] Conv3Dtranspose default kernel layout should be IODHW #14340

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
321ea37
fix conv3Dtranspose kernel layout
rebel-jangys Mar 20, 2023
87fd31d
fix typo
rebel-jangys Mar 20, 2023
ad2712f
Fix conv3dtranspose frontend relay converter for pytorch, keras
rebel-jangys Mar 21, 2023
b883be5
Add conv3dtranspose convert layout pass
rebel-jangys Mar 22, 2023
1731303
Fix conv3dtranspose frontend relay converter keras
rebel-jangys Mar 22, 2023
eec9409
lint
rebel-jangys Mar 22, 2023
c16a59d
fix dnnl test for conv3d transpose
rebel-jangys Mar 22, 2023
3b1c52a
fix tensorflow relay converter for conv3d_transpose op
rebel-jangys Mar 23, 2023
391e344
fix dnnl runtime
rebel-jangys Mar 30, 2023
4cafaa9
fix topi conv3d_transpose legalize
rebel-jangys Mar 30, 2023
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
12 changes: 6 additions & 6 deletions include/tvm/relay/attrs/nn.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -430,10 +430,10 @@ struct Conv3DTransposeAttrs : public tvm::AttrsNode<Conv3DTransposeAttrs> {
"dimensions respectively. Convolution is applied on the 'D', 'H' and"
"'W' dimensions.");
TVM_ATTR_FIELD(kernel_layout)
.set_default("OIDHW")
.set_default("IODHW")
.describe(
"Dimension ordering of data and weight. Can be 'OIDHW', 'OIDHW16o16i', etc."
"'O', 'I', 'D', 'H', 'W' stands for num_filter, input_channel, depth, height, and width"
"Dimension ordering of data and weight. Can be 'IODHW', 'IODHW16i16o', etc."
"'I', 'O', 'D', 'H', 'W' stands for input_channel, num_filter, depth, height, and width"
"dimensions respectively.");
TVM_ATTR_FIELD(out_layout)
.set_default("")
Expand Down Expand Up @@ -588,10 +588,10 @@ struct Conv2DTransposeAttrs : public tvm::AttrsNode<Conv2DTransposeAttrs> {
"dimensions respectively. Convolution is applied on the 'H' and"
"'W' dimensions.");
TVM_ATTR_FIELD(kernel_layout)
.set_default("OIHW")
.set_default("IOHW")
.describe(
"Dimension ordering of data and weight. Can be 'OIHW', 'OIHW16o16i', etc."
"'O', 'I', 'H', 'W' stands for num_filter, input_channel, height, and width"
"Dimension ordering of data and weight. Can be 'IOHW', 'OIHW16o16i', etc."
"'I', 'O', 'H', 'W' stands for input_channel, num_filter, height, and width"
"dimensions respectively.");
TVM_ATTR_FIELD(out_layout)
.set_default("")
Expand Down
2 changes: 1 addition & 1 deletion python/tvm/relay/op/nn/nn.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -436,7 +436,7 @@ def conv3d_transpose(
channels=None,
kernel_size=None,
data_layout="NCDHW",
kernel_layout="OIDHW",
kernel_layout="IODHW",
out_layout="",
output_padding=(0, 0, 0),
out_dtype="",
Expand Down
18 changes: 10 additions & 8 deletions src/relay/op/nn/convolution.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -594,7 +594,7 @@ bool Conv3DTransposeRel(const Array<Type>& types, int num_inputs, const Attrs& a
if (data == nullptr) return false;

static const Layout kNCDHW("NCDHW");
static const Layout kOIDHW("OIDHW");
static const Layout kIODHW("IODHW");

const Conv3DTransposeAttrs* param = attrs.as<Conv3DTransposeAttrs>();
ICHECK(param != nullptr);
Expand All @@ -606,9 +606,9 @@ bool Conv3DTransposeRel(const Array<Type>& types, int num_inputs, const Attrs& a
<< "Conv3d_transpose only support input layouts that are convertible from NCDHW."
<< " But got " << in_layout;

const auto trans_kernel_layout = tir::BijectiveLayout(kernel_layout, kOIDHW);
const auto trans_kernel_layout = tir::BijectiveLayout(kernel_layout, kIODHW);
ICHECK(trans_kernel_layout.defined())
<< "Conv3d_transpose only support kernel layouts that are convertible from OIDHW."
<< "Conv3d_transpose only support kernel layouts that are convertible from IODHW."
<< " But got " << kernel_layout;

Layout out_layout(param->out_layout == "" ? param->data_layout : param->out_layout);
Expand Down Expand Up @@ -651,16 +651,18 @@ bool Conv3DTransposeRel(const Array<Type>& types, int num_inputs, const Attrs& a
ICHECK(reporter->AssertEQ(param->kernel_size[0], wshape[2]) &&
reporter->AssertEQ(param->kernel_size[1], wshape[3]) &&
reporter->AssertEQ(param->kernel_size[2], wshape[4]))
<< "Conv3D: shape of weight is inconsistent with kernel_size, "
<< "Conv3DTransposed: shape of weight is inconsistent with kernel_size, "
<< " kernel_size=" << param->kernel_size << " wshape=" << Array<IndexExpr>(wshape);
}
if (param->channels.defined()) {
ICHECK(reporter->AssertEQ(param->channels, wshape[1]))
<< "Conv3D: shape of weight is inconsistent with channels, "
<< " channels=" << param->channels << " wshape=" << Array<IndexExpr>(wshape);
ICHECK(reporter->AssertEQ(indexdiv(param->channels, param->groups), wshape[1]))
<< "Conv3DTransposed: shape of weight is inconsistent out_channels, "
<< " out_channels // groups != weight.shape[1] "
<< " out_channels=" << param->channels << " groups=" << param->groups
<< " wshape=" << Array<IndexExpr>(wshape);
}
if (!dshape_ncdhw[1].as<tir::AnyNode>() && !wshape[0].as<tir::AnyNode>()) {
ICHECK(reporter->AssertEQ(indexdiv(dshape_ncdhw[1], param->groups), wshape[0]));
ICHECK(reporter->AssertEQ(dshape_ncdhw[1], wshape[0]));
}
channels = wshape[1];
dilated_ksize_d = 1 + (wshape[2] - 1) * param->dilation[0];
Expand Down