IVGCVSW-3649 Add TfLite parser support for Prelu layer
Signed-off-by: Narumol Prangnawarat <narumol.prangnawarat@arm.com>
Change-Id: I3dedcc86efe1a67c709d9da636953e2fc400107b
diff --git a/src/armnnTfLiteParser/test/Prelu.cpp b/src/armnnTfLiteParser/test/Prelu.cpp
new file mode 100644
index 0000000..b4aa8d7
--- /dev/null
+++ b/src/armnnTfLiteParser/test/Prelu.cpp
@@ -0,0 +1,159 @@
+//
+// Copyright © 2021 Arm Ltd. All rights reserved.
+// SPDX-License-Identifier: MIT
+//
+
+#include <boost/test/unit_test.hpp>
+#include "ParserFlatbuffersFixture.hpp"
+#include "../TfLiteParser.hpp"
+
+#include <string>
+
+BOOST_AUTO_TEST_SUITE(TensorflowLiteParser)
+
+struct PreluFixture : public ParserFlatbuffersFixture
+{
+ explicit PreluFixture(const std::string& inputShape,
+ const std::string& alphaShape,
+ const std::string& outputShape,
+ const std::string& inputIndex,
+ const std::string& alphaData)
+ {
+ m_JsonString = R"(
+ {
+ "version": 3,
+ "operator_codes": [
+ {
+ "builtin_code": "PRELU",
+ "version": 1
+ }
+ ],
+ "subgraphs": [
+ {
+ "tensors": [
+ {
+ "shape": )" + inputShape + R"(,
+ "type": "FLOAT32",
+ "buffer": 1,
+ "name": "input0",
+ "quantization": {
+ "details_type": "NONE",
+ "quantized_dimension": 0
+ },
+ "is_variable": false
+ },
+ {
+ "shape": )" + alphaShape + R"(,
+ "type": "FLOAT32",
+ "buffer": 2,
+ "name": "input1",
+ "quantization": {
+ "details_type": "NONE",
+ "quantized_dimension": 0
+ },
+ "is_variable": false
+ },
+ {
+ "shape": )" + outputShape + R"(,
+ "type": "FLOAT32",
+ "buffer": 3,
+ "name": "output",
+ "quantization": {
+ "details_type": "NONE",
+ "quantized_dimension": 0
+ },
+ "is_variable": false
+ }
+ ],
+ "inputs": )" + inputIndex + R"(,
+ "outputs": [
+ 2
+ ],
+ "operators": [
+ {
+ "opcode_index": 0,
+ "inputs": [
+ 0,
+ 1
+ ],
+ "outputs": [
+ 2
+ ],
+ "builtin_options_type": "NONE",
+ "custom_options_format": "FLEXBUFFERS"
+ }
+ ],
+ "name": "main"
+ }
+ ],
+ "description": "MLIR Converted.",
+ "buffers": [
+ {
+ },
+ {
+ },
+ { )" + alphaData + R"(
+ },
+ {
+ }
+ ]
+ }
+ )";
+ Setup();
+ }
+};
+
+struct SimplePreluFixture : PreluFixture
+{
+ SimplePreluFixture() : PreluFixture("[ 2, 3 ]",
+ "[ 1, 1 ]",
+ "[ 2, 3 ]",
+ "[ 0, 1 ]",
+ "") {}
+};
+
+struct PreluConstAlphaFixture : PreluFixture
+{
+ PreluConstAlphaFixture() : PreluFixture(
+ "[ 2, 3 ]",
+ "[ 2, 3 ]",
+ "[ 2, 3 ]",
+ "[ 0 ]",
+ "\"data\": [ 0, 0, 128, 62, 0, 0, 128, 62, 0, 0, 128, 62, 0, 0, 128, 62, 0, 0, 128, 62, 0, 0, 128, 62 ]"){}
+};
+
+struct PreluDynamicTensorFixture : PreluFixture
+{
+ PreluDynamicTensorFixture() : PreluFixture("[ 2, 3 ]",
+ "[ 1, 1 ]",
+ "[]",
+ "[ 0 ]",
+ "\"data\": [ 0, 0, 128, 62 ]") {}
+};
+
+BOOST_FIXTURE_TEST_CASE(SimplePrelu, SimplePreluFixture)
+{
+ RunTest<2, armnn::DataType::Float32>(
+ 0,
+ {{"input0", { -14.f, 2.f, 0.f, 1.f, -5.f, 14.f }},{"input1", { 0.25f }}},
+ {{"output", { -3.5f, 2.f, 0.f, 1.f, -1.25f, 14.f }}});
+}
+
+BOOST_FIXTURE_TEST_CASE(PreluConstAlpha, PreluConstAlphaFixture)
+{
+ RunTest<2, armnn::DataType::Float32>(
+ 0,
+ {{"input0", { -14.f, 2.f, 0.f, 1.f, -5.f, 14.f }}},
+ {{"output", { -3.5f, 2.f, 0.f, 1.f, -1.25f, 14.f }}});
+}
+
+BOOST_FIXTURE_TEST_CASE(PreluDynamicTensor, PreluDynamicTensorFixture)
+{
+ RunTest<2, armnn::DataType::Float32, armnn::DataType::Float32>(
+ 0,
+ {{"input0", { -14.f, 2.f, 0.f, 1.f, -5.f, 14.f }}},
+ {{"output", { -3.5f, 2.f, 0.f, 1.f, -1.25f, 14.f }}},
+ true);
+}
+
+BOOST_AUTO_TEST_SUITE_END()