Skip to content

Add CastLike grad #26185

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

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Add CastLike grad #26185

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
230dc92
Add CastLike grad implementation
youzow3 Sep 27, 2025
1ba0e04
Improve CastLikeGrad test
youzow3 Sep 28, 2025
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: 12 additions & 0 deletions orttraining/orttraining/core/graph/gradient_builder.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,18 @@
{MakeAttribute("to", int64_t(IElemType(0)))})};
}

IMPLEMENT_GRADIENT_BUILDER(GetCastLikeGradient) {
std::vector<NodeDef> result;
result.push_back(
NodeDef("Cast",
{GO(0)}, {GI(0)},
{MakeAttribute("to", int64_t(IElemType(0)))}));

Check warning on line 84 in orttraining/orttraining/core/graph/gradient_builder.cc

View workflow job for this annotation

GitHub Actions / Optional Lint C++ 

[cpplint] reported by reviewdog 🐶
 Using deprecated casting style.  Use static_cast<int64_t>(...) instead  [readability/casting] [4]

Raw Output:
orttraining/orttraining/core/graph/gradient_builder.cc:84:  Using deprecated casting style.  Use static_cast<int64_t>(...) instead  [readability/casting] [4]
result.push_back(ConstantScalarNode(0, "0", IElemType(1)));
result.push_back(NodeDef("Shape", {I(1)}, {ArgDef("gi_shape")}));
result.push_back(NodeDef("Expand", {ArgDef("0"), ArgDef("gi_shape")}, {GI(1)}));
return result;
}

IMPLEMENT_GRADIENT_BUILDER(GetSinGradient) {
std::vector<NodeDef> result;
result.push_back(NodeDef("Cos", {I(0)}, {IA("Cos_O0")}));
Expand Down
1 change: 1 addition & 0 deletions orttraining/orttraining/core/graph/gradient_builder.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@ namespace training {
};

DECLARE_GRADIENT_BUILDER(GetCastGradient)
DECLARE_GRADIENT_BUILDER(GetCastLikeGradient)
DECLARE_GRADIENT_BUILDER(GetSinGradient)
DECLARE_GRADIENT_BUILDER(GetCosGradient)
DECLARE_GRADIENT_BUILDER(GetLogGradient)
Expand Down
1 change: 1 addition & 0 deletions orttraining/orttraining/core/graph/gradient_builder_registry.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@ GradientDef GetGradientForOp(const GradientGraphConfiguration& gradient_graph_co
void GradientBuilderRegistry::RegisterGradientBuilders() {
// Register gradient builders here.
REGISTER_GRADIENT_BUILDER("Cast", GetCastGradient);
REGISTER_GRADIENT_BUILDER("CastLike", GetCastLikeGradient);
REGISTER_GRADIENT_BUILDER("Sin", GetSinGradient);
REGISTER_GRADIENT_BUILDER("Cos", GetCosGradient);
REGISTER_GRADIENT_BUILDER("Log", GetLogGradient);
Expand Down
62 changes: 62 additions & 0 deletions orttraining/orttraining/test/gradient/gradient_ops_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -700,6 +700,68 @@ TEST(GradientCheckerTest, CastGrad) {
}
}

TEST(GradientCheckerTest, CastLikeGrad) {
OpDef op_def{"CastLike", kOnnxDomain, 15};
float error_tolerance = 1e-3f;
// dummy test like CastGrad
{
TensorShape shape({2, 3, 4});
float max_error;
GradientChecker<float, float, float> gradient_checker;

ASSERT_STATUS_OK(gradient_checker.ComputeGradientError(op_def, {shape, shape}, {shape}, &max_error));
EXPECT_IS_TINIER_THAN(max_error, error_tolerance);
}

// float <-> double
{
TensorInfo info_f({2, 3, 4}, true, nullptr, DataTypeImpl::GetTensorType<float>());
TensorInfo info_d({2, 3, 4}, true, nullptr, DataTypeImpl::GetTensorType<double>());
// float -> double
{
float max_error;
GradientChecker<float, float, float> gradient_checker;

ASSERT_STATUS_OK(gradient_checker.ComputeGradientError(op_def, {info_f, info_d}, {info_d}, &max_error));
EXPECT_IS_TINIER_THAN(max_error, error_tolerance);
}
// double -> float
{
float max_error;
GradientChecker<float, float, float> gradient_checker;

ASSERT_STATUS_OK(gradient_checker.ComputeGradientError(op_def, {info_d, info_f}, {info_f}, &max_error));
EXPECT_IS_TINIER_THAN(max_error, error_tolerance);
}
}

// float <-> int32_t
{
TensorInfo info_i({2, 3, 4}, false, nullptr, DataTypeImpl::GetTensorType<int32_t>());
TensorInfo info_f({2, 3, 4}, true, nullptr, DataTypeImpl::GetTensorType<float>());
/*
// float -> int32_t
// This part causes programming error
{
float max_error;
GradientChecker<float, float, float> gradient_checker;
TensorInfo info_i_g({2, 3, 4}, true, nullptr, DataTypeImpl::GetTensorType<int32_t>());

ASSERT_STATUS_OK(gradient_checker.ComputeGradientError(op_def, {info_f, info_i}, {info_i_g}, &max_error));
EXPECT_IS_TINIER_THAN(max_error, error_tolerance);
}
*/
// int32_t -> float
{
float max_error;
GradientChecker<float, float, float> gradient_checker;

ASSERT_STATUS_OK(gradient_checker.ComputeGradientError(op_def, {info_i, info_f}, {info_f}, &max_error));
EXPECT_IS_TINIER_THAN(max_error, error_tolerance);
}
}
}

TEST(GradientCheckerTest, SplitGrad) {
TensorShape shape({9, 5});
float max_error;
Expand Down
Loading