Skip to content
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

Jump to bottom

feat: Add converter for aten::log2 #1866

Merged
merged 1 commit into from
Apr 27, 2023
Merged

feat: Add converter for aten::log2 #1866

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
30 changes: 30 additions & 0 deletions core/conversion/converters/impl/unary.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,36 @@ auto reciprocal_registration TORCHTRT_UNUSED = RegisterNodeConversionPatterns().
return true;
}});

auto log2_registration TORCHTRT_UNUSED = RegisterNodeConversionPatterns().pattern(
{"aten::log2(Tensor self) -> Tensor", [](ConversionCtx* ctx, const torch::jit::Node* n, args& args) -> bool {
const static float ln2 = 0.693147180559945309; // same constant onnx uses
auto in = args[0].ITensorOrFreeze(ctx);
auto tensor_type = util::TRTDataTypeToScalarType(in->getType());
if (in->getType() == nvinfer1::DataType::kINT32) {
// pytorch implicitly casts to float for aten::log2(int)
in = castITensor(ctx, in, nvinfer1::DataType::kFLOAT);
tensor_type = at::kFloat;
}

auto log_layer = ctx->net->addUnary(*in, nvinfer1::UnaryOperation::kLOG);
TORCHTRT_CHECK(log_layer, "Unable to create log layer from node: " << *n);
log_layer->setName((util::node_info(n) + "_log").c_str());

std::vector<int64_t> ln2_dims(in->getDimensions().nbDims, 1);
auto ln2_tensor = at::full(ln2_dims, ln2, at::TensorOptions().dtype(tensor_type));
auto ln2_itensor = converters::tensor_to_const(ctx, ln2_tensor);

auto div_layer = add_elementwise(
ctx,
nvinfer1::ElementWiseOperation::kDIV,
log_layer->getOutput(0),
ln2_itensor,
(util::node_info(n) + "_div").c_str());
auto out_tensor = ctx->AssociateValueAndTensor(n->outputs()[0], div_layer->getOutput(0));
LOG_DEBUG("Output tensor shape: " << out_tensor->getDimensions());
return true;
}});

auto logical_not_registration TORCHTRT_UNUSED = RegisterNodeConversionPatterns().pattern(
{"aten::logical_not(Tensor self) -> Tensor", [](ConversionCtx* ctx, const torch::jit::Node* n, args& args) -> bool {
auto in = args[0].ITensorOrFreeze(ctx);
Expand Down
16 changes: 16 additions & 0 deletions tests/core/conversion/converters/test_unary.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,6 +47,21 @@ TEST(Converters, ATenReciprocalIntConvertsCorrectly) {
ASSERT_TRUE(torch_tensorrt::tests::util::exactlyEqual(jit_results[0], trt_results[0]));
}

TEST(Converters, ATenLog2IntConvertsCorrectly) {
const auto graph = gen_test_graph("log2");
auto g = std::make_shared<torch::jit::Graph>();
torch::jit::parseIR(graph, g.get());

auto in = at::tensor({1, 2, 7, 25, 50}, {at::kCUDA}).to(torch::kInt32);
auto params = torch_tensorrt::core::ir::get_static_params(g->inputs(), {});
auto jit_results = torch_tensorrt::tests::util::RunGraph(g, params, {in});

in = at::clone(in);
params = torch_tensorrt::core::ir::get_static_params(g->inputs(), {});
auto trt_results = torch_tensorrt::tests::util::RunGraphEngine(g, params, {in});
ASSERT_TRUE(torch_tensorrt::tests::util::almostEqual(jit_results[0], trt_results[0]));
}

TEST(Converters, ATenSignConvertsCorrectly) {
const auto graph = gen_test_graph("sign");
auto g = std::make_shared<torch::jit::Graph>();
Expand Down Expand Up @@ -129,6 +144,7 @@ test_unary(abs, Abs);
test_unary(floor, Floor);
test_unary(reciprocal, Reciprocal);
test_unary(log, Log);
test_unary(log2, Log2);
test_unary(ceil, Ceil);
test_unary(sqrt, Sqrt);
test_unary(exp, Exp);
Expand Down