Skip to content

Support aten::sum with bool tensor input #1512

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 2 commits into from
Dec 2, 2022
Merged

Support aten::sum with bool tensor input #1512

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
e7bd9ac
Support aten::sum with bool tensor input
mfeliz-cruise Dec 1, 2022
20a0716
lint
mfeliz-cruise Dec 1, 2022
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
7 changes: 7 additions & 0 deletions core/conversion/converters/impl/reduce.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,13 @@ auto reduce_registrations TORCHTRT_UNUSED =
auto in_dims = util::toVec(in_tensor->getDimensions());
LOG_WARNING("Sum Converter disregards dtype");

if (in_tensor->getType() == nvinfer1::DataType::kBOOL) {
LOG_DEBUG(
"Found type " << in_tensor->getType() << " in aten::sum, casting to "
<< nvinfer1::DataType::kINT32 << " for compatibility.");
in_tensor = castITensor(ctx, in_tensor, nvinfer1::DataType::kINT32);
}

uint32_t axis_mask = (uint32_t)(((uint64_t)1 << in_dims.size()) - 1);

auto sum_layer = ctx->net->addReduce(*in_tensor, nvinfer1::ReduceOperation::kSUM, axis_mask, false);
Expand Down
10 changes: 10 additions & 0 deletions tests/core/conversion/converters/test_reduce.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -137,6 +137,16 @@ converts_keepdims_correctly(mean, Mean);

#undef converts_keepdims_correctly

TEST(Converters, ATenSumBoolConvertsCorrectly) {
const auto graph = R"IR(
graph(%0 : Tensor):
%4 : None = prim::Constant()
%5 : Tensor = aten::sum(%0, %4)
return (%5))IR";
auto in = at::randint(-1, 2, {4, 4, 4}, at::kCUDA).to(at::kBool);
test_body(graph, in);
}

TEST(Converters, ATenSumDimNegOneIndexConvertsCorrectly) {
const auto graph = R"IR(
graph(%0 : Tensor):
Expand Down