// clang-format off // Generated file (from: conv_float.mod.py). Do not edit // clang-format off // Generated file (from: conv_float.mod.py). Do not edit // Generated from: conv_float.mod.py. namespace conv_float { // Generated conv_float test #include "-" // Generated model constructor #include "-" } // namespace conv_float // Create the model Model createTestModel_relu() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 0, 64, 0, 0, 64, 64, 0, 0, 128, 64, 0, 0, 160, 64, 0, 0, 192, 64, 0, 0, 224, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 224, 64, 0, 0, 192, 64, 0, 0, 160, 64, 0, 0, 128, 64, 0, 0, 64, 64, 0, 0, 0, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 120.0f, 94.0f, 104.0f, 70.0f, 164.0f, 23.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu) { generated_tests::Execute(device, conv_float::createTestModel_relu, conv_float::is_ignored_relu, conv_float::examples_relu); } // Create the model Model createTestModel_relu_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 0, 64, 0, 0, 64, 64, 0, 0, 128, 64, 0, 0, 160, 64, 0, 0, 192, 64, 0, 0, 224, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 224, 64, 0, 0, 192, 64, 0, 0, 160, 64, 0, 0, 128, 64, 0, 0, 64, 64, 0, 0, 0, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_relu_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 120.0f, 94.0f, 104.0f, 70.0f, 164.0f, 23.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_relu_relaxed, conv_float::is_ignored_relu_relaxed, conv_float::examples_relu_relaxed); } // Create the model Model createTestModel_relu_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 16}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 24, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 28, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 32, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 36, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 132, 136, 140, 144, 148, 152, 156, 160, 160, 156, 152, 148, 144, 140, 136, 132, 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 132, 134, 136, 138, 140, 142, 144}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {202, 160, 147, 152, 135, 182, 112, 156}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu_quant8) { generated_tests::Execute(device, conv_float::createTestModel_relu_quant8, conv_float::is_ignored_relu_quant8, conv_float::examples_relu_quant8); } // Create the model Model createTestModel_relu_weight_as_input() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu_weight_as_input(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu_weight_as_input = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}, {1, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f, 8.0f, 7.0f, 6.0f, 5.0f, 4.0f, 3.0f, 2.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 120.0f, 94.0f, 104.0f, 70.0f, 164.0f, 23.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu_weight_as_input) { generated_tests::Execute(device, conv_float::createTestModel_relu_weight_as_input, conv_float::is_ignored_relu_weight_as_input, conv_float::examples_relu_weight_as_input); } // Create the model Model createTestModel_relu_weight_as_input_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_relu_weight_as_input_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu_weight_as_input_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}, {1, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f, 8.0f, 7.0f, 6.0f, 5.0f, 4.0f, 3.0f, 2.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 120.0f, 94.0f, 104.0f, 70.0f, 164.0f, 23.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu_weight_as_input_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_relu_weight_as_input_relaxed, conv_float::is_ignored_relu_weight_as_input_relaxed, conv_float::examples_relu_weight_as_input_relaxed); } // Create the model Model createTestModel_relu_weight_as_input_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu_weight_as_input_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu_weight_as_input_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 132, 134, 136, 138, 140, 142, 144}}, {1, {132, 136, 140, 144, 148, 152, 156, 160, 160, 156, 152, 148, 144, 140, 136, 132}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {202, 160, 147, 152, 135, 182, 112, 156}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu_weight_as_input_quant8) { generated_tests::Execute(device, conv_float::createTestModel_relu_weight_as_input_quant8, conv_float::is_ignored_relu_weight_as_input_quant8, conv_float::examples_relu_weight_as_input_quant8); } // Create the model Model createTestModel_relu6() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 0, 64, 0, 0, 64, 64, 0, 0, 128, 64, 0, 0, 160, 64, 0, 0, 192, 64, 0, 0, 224, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 224, 64, 0, 0, 192, 64, 0, 0, 160, 64, 0, 0, 128, 64, 0, 0, 64, 64, 0, 0, 0, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu6(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6) { generated_tests::Execute(device, conv_float::createTestModel_relu6, conv_float::is_ignored_relu6, conv_float::examples_relu6); } // Create the model Model createTestModel_relu6_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 0, 64, 0, 0, 64, 64, 0, 0, 128, 64, 0, 0, 160, 64, 0, 0, 192, 64, 0, 0, 224, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 224, 64, 0, 0, 192, 64, 0, 0, 160, 64, 0, 0, 128, 64, 0, 0, 64, 64, 0, 0, 0, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_relu6_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_relu6_relaxed, conv_float::is_ignored_relu6_relaxed, conv_float::examples_relu6_relaxed); } // Create the model Model createTestModel_relu6_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 16}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 24, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 28, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 32, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 36, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 132, 136, 140, 144, 148, 152, 156, 160, 160, 156, 152, 148, 144, 140, 136, 132, 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu6_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 132, 134, 136, 138, 140, 142, 144}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {103, 103, 103, 103, 103, 103, 103, 103}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6_quant8) { generated_tests::Execute(device, conv_float::createTestModel_relu6_quant8, conv_float::is_ignored_relu6_quant8, conv_float::examples_relu6_quant8); } // Create the model Model createTestModel_relu6_weight_as_input() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu6_weight_as_input(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6_weight_as_input = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}, {1, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f, 8.0f, 7.0f, 6.0f, 5.0f, 4.0f, 3.0f, 2.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6_weight_as_input) { generated_tests::Execute(device, conv_float::createTestModel_relu6_weight_as_input, conv_float::is_ignored_relu6_weight_as_input, conv_float::examples_relu6_weight_as_input); } // Create the model Model createTestModel_relu6_weight_as_input_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_relu6_weight_as_input_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6_weight_as_input_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f}}, {1, {1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f, 8.0f, 7.0f, 6.0f, 5.0f, 4.0f, 3.0f, 2.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6_weight_as_input_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_relu6_weight_as_input_relaxed, conv_float::is_ignored_relu6_weight_as_input_relaxed, conv_float::examples_relu6_weight_as_input_relaxed); } // Create the model Model createTestModel_relu6_weight_as_input_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_relu6_weight_as_input_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_relu6_weight_as_input_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 132, 134, 136, 138, 140, 142, 144}}, {1, {132, 136, 140, 144, 148, 152, 156, 160, 160, 156, 152, 148, 144, 140, 136, 132}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {103, 103, 103, 103, 103, 103, 103, 103}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_relu6_weight_as_input_quant8) { generated_tests::Execute(device, conv_float::createTestModel_relu6_weight_as_input_quant8, conv_float::is_ignored_relu6_weight_as_input_quant8, conv_float::examples_relu6_weight_as_input_quant8); } // Create the model Model createTestModel_nchw_relu() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 64, 64, 0, 0, 160, 64, 0, 0, 224, 64, 0, 0, 0, 64, 0, 0, 128, 64, 0, 0, 192, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 192, 64, 0, 0, 128, 64, 0, 0, 0, 64, 0, 0, 224, 64, 0, 0, 160, 64, 0, 0, 64, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 94.0f, 70.0f, 23.0f, 120.0f, 104.0f, 164.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu, conv_float::is_ignored_nchw_relu, conv_float::examples_nchw_relu); } // Create the model Model createTestModel_nchw_relu_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 64, 64, 0, 0, 160, 64, 0, 0, 224, 64, 0, 0, 0, 64, 0, 0, 128, 64, 0, 0, 192, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 192, 64, 0, 0, 128, 64, 0, 0, 0, 64, 0, 0, 224, 64, 0, 0, 160, 64, 0, 0, 64, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_nchw_relu_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 94.0f, 70.0f, 23.0f, 120.0f, 104.0f, 164.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu_relaxed, conv_float::is_ignored_nchw_relu_relaxed, conv_float::examples_nchw_relu_relaxed); } // Create the model Model createTestModel_nchw_relu_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 16}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 24, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 28, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 32, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 36, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 132, 140, 148, 156, 136, 144, 152, 160, 160, 152, 144, 136, 156, 148, 140, 132, 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 134, 138, 142, 132, 136, 140, 144}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {202, 147, 135, 112, 160, 152, 182, 156}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu_quant8) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu_quant8, conv_float::is_ignored_nchw_relu_quant8, conv_float::examples_nchw_relu_quant8); } // Create the model Model createTestModel_nchw_relu_weight_as_input() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu_weight_as_input(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu_weight_as_input = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}, {1, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f, 8.0f, 6.0f, 4.0f, 2.0f, 7.0f, 5.0f, 3.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 94.0f, 70.0f, 23.0f, 120.0f, 104.0f, 164.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu_weight_as_input) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu_weight_as_input, conv_float::is_ignored_nchw_relu_weight_as_input, conv_float::examples_nchw_relu_weight_as_input); } // Create the model Model createTestModel_nchw_relu_weight_as_input_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_nchw_relu_weight_as_input_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu_weight_as_input_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}, {1, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f, 8.0f, 6.0f, 4.0f, 2.0f, 7.0f, 5.0f, 3.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {204.0f, 94.0f, 70.0f, 23.0f, 120.0f, 104.0f, 164.0f, 112.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu_weight_as_input_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu_weight_as_input_relaxed, conv_float::is_ignored_nchw_relu_weight_as_input_relaxed, conv_float::examples_nchw_relu_weight_as_input_relaxed); } // Create the model Model createTestModel_nchw_relu_weight_as_input_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu_weight_as_input_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu_weight_as_input_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 134, 138, 142, 132, 136, 140, 144}}, {1, {132, 140, 148, 156, 136, 144, 152, 160, 160, 152, 144, 136, 156, 148, 140, 132}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {202, 147, 135, 112, 160, 152, 182, 156}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu_weight_as_input_quant8) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu_weight_as_input_quant8, conv_float::is_ignored_nchw_relu_weight_as_input_quant8, conv_float::examples_nchw_relu_weight_as_input_quant8); } // Create the model Model createTestModel_nchw_relu6() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 64, 64, 0, 0, 160, 64, 0, 0, 224, 64, 0, 0, 0, 64, 0, 0, 128, 64, 0, 0, 192, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 192, 64, 0, 0, 128, 64, 0, 0, 0, 64, 0, 0, 224, 64, 0, 0, 160, 64, 0, 0, 64, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu6(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6, conv_float::is_ignored_nchw_relu6, conv_float::examples_nchw_relu6); } // Create the model Model createTestModel_nchw_relu6_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 64}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 64, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 68, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 72, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 76, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 80, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 84, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 128, 63, 0, 0, 64, 64, 0, 0, 160, 64, 0, 0, 224, 64, 0, 0, 0, 64, 0, 0, 128, 64, 0, 0, 192, 64, 0, 0, 0, 65, 0, 0, 0, 65, 0, 0, 192, 64, 0, 0, 128, 64, 0, 0, 0, 64, 0, 0, 224, 64, 0, 0, 160, 64, 0, 0, 64, 64, 0, 0, 128, 63, 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_nchw_relu6_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6_relaxed, conv_float::is_ignored_nchw_relu6_relaxed, conv_float::examples_nchw_relu6_relaxed); } // Create the model Model createTestModel_nchw_relu6_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 16}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 24, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 28, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 32, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 36, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0}; const std::vector outputIndexes = {8}; std::vector operandValues = { 132, 140, 148, 156, 136, 144, 152, 160, 160, 152, 144, 136, 156, 148, 140, 132, 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu6_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 134, 138, 142, 132, 136, 140, 144}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {103, 103, 103, 103, 103, 103, 103, 103}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6_quant8) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6_quant8, conv_float::is_ignored_nchw_relu6_quant8, conv_float::examples_nchw_relu6_quant8); } // Create the model Model createTestModel_nchw_relu6_weight_as_input() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu6_weight_as_input(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6_weight_as_input = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}, {1, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f, 8.0f, 6.0f, 4.0f, 2.0f, 7.0f, 5.0f, 3.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6_weight_as_input) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6_weight_as_input, conv_float::is_ignored_nchw_relu6_weight_as_input, conv_float::examples_nchw_relu6_weight_as_input); } // Create the model Model createTestModel_nchw_relu6_weight_as_input_relaxed() { const std::vector operands = { { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_FLOAT32, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 0, 0, 72, 195, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, .relaxComputationFloat32toFloat16 = true, }; } bool is_ignored_nchw_relu6_weight_as_input_relaxed(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6_weight_as_input_relaxed = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f}}, {1, {1.0f, 3.0f, 5.0f, 7.0f, 2.0f, 4.0f, 6.0f, 8.0f, 8.0f, 6.0f, 4.0f, 2.0f, 7.0f, 5.0f, 3.0f, 1.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {{0, {6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f, 6.0f}}}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6_weight_as_input_relaxed) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6_weight_as_input_relaxed, conv_float::is_ignored_nchw_relu6_weight_as_input_relaxed, conv_float::examples_nchw_relu6_weight_as_input_relaxed); } // Create the model Model createTestModel_nchw_relu6_weight_as_input_quant8() { const std::vector operands = { { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 1, .scale = 0.5f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {2, 2, 2, 2}, .numberOfConsumers = 0, .scale = 0.25f, .zeroPoint = 128, .lifetime = OperandLifeTime::MODEL_INPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, }, { .type = OperandType::TENSOR_INT32, .dimensions = {1}, .numberOfConsumers = 1, .scale = 0.125f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 0, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 4, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 8, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 12, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 16, .length = 4}, }, { .type = OperandType::INT32, .dimensions = {}, .numberOfConsumers = 1, .scale = 0.0f, .zeroPoint = 0, .lifetime = OperandLifeTime::CONSTANT_COPY, .location = {.poolIndex = 0, .offset = 20, .length = 4}, }, { .type = OperandType::TENSOR_QUANT8_ASYMM, .dimensions = {1, 2, 2, 2}, .numberOfConsumers = 0, .scale = 2.0f, .zeroPoint = 100, .lifetime = OperandLifeTime::MODEL_OUTPUT, .location = {.poolIndex = 0, .offset = 0, .length = 0}, } }; const std::vector operations = { { .type = OperationType::CONV_2D, .inputs = {0, 1, 2, 3, 4, 5, 6, 7}, .outputs = {8}, } }; const std::vector inputIndexes = {0, 1}; const std::vector outputIndexes = {8}; std::vector operandValues = { 192, 249, 255, 255, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 3, 0, 0, 0, 1, 0, 0, 0 }; const std::vector pools = {}; return { .operands = operands, .operations = operations, .inputIndexes = inputIndexes, .outputIndexes = outputIndexes, .operandValues = operandValues, .pools = pools, }; } bool is_ignored_nchw_relu6_weight_as_input_quant8(int i) { static std::set ignore = {}; return ignore.find(i) != ignore.end(); } std::vector examples_nchw_relu6_weight_as_input_quant8 = { // Begin of an example { //Input(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {130, 134, 138, 142, 132, 136, 140, 144}}, {1, {132, 140, 148, 156, 136, 144, 152, 160, 160, 152, 144, 136, 156, 148, 140, 132}}} }, //Output(s) { // See tools/test_generator/include/TestHarness.h:MixedTyped // int -> FLOAT32 map {}, // int -> INT32 map {}, // int -> QUANT8_ASYMM map {{0, {103, 103, 103, 103, 103, 103, 103, 103}}} } }, // End of an example }; TEST_F(NeuralnetworksHidlTest, conv_float_nchw_relu6_weight_as_input_quant8) { generated_tests::Execute(device, conv_float::createTestModel_nchw_relu6_weight_as_input_quant8, conv_float::is_ignored_nchw_relu6_weight_as_input_quant8, conv_float::examples_nchw_relu6_weight_as_input_quant8); }