mindspore/tests/ut/cpp/pynative/pynative_execute_test.cc

155 lines
6.0 KiB
C++

/**
* Copyright 2022 Huawei Technologies Co., Ltd
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include <iostream>
#include <memory>
#include "common/common_test.h"
#include "pybind_api/ir/primitive_py.h"
#include "include/common/utils/python_adapter.h"
#include "include/common/utils/utils.h"
#include "include/common/utils/convert_utils_py.h"
#include "pipeline/jit/parse/data_converter.h"
#include "frontend/operator/ops.h"
#include "pipeline/pynative/pynative_execute.h"
#include "pipeline/pynative/forward/do_infer.h"
#include "pipeline/pynative/base.h"
#include "utils/ms_context.h"
namespace py = pybind11;
using pybind11::literals::operator"" _a;
using PrimitivePy = mindspore::PrimitivePy;
using Tensor = mindspore::tensor::Tensor;
using TensorPtr = mindspore::tensor::TensorPtr;
using BaseOpRunInfo = mindspore::pynative::BaseOpRunInfo;
using FrontendOpRunInfo = mindspore::pynative::FrontendOpRunInfo;
using InferOperation = mindspore::pynative::InferOperation;
namespace mindspore {
namespace pynative {
class TestPynativeExecute : public UT::Common {
public:
TestPynativeExecute() {}
};
inline ValuePtr PyAttrValue(const py::object &obj) {
ValuePtr converted_ret;
bool converted = parse::ConvertData(obj, &converted_ret);
if (!converted) {
MS_LOG(EXCEPTION) << "attribute convert error with type:" << std::string(py::str(obj));
}
return converted_ret;
}
FrontendOpRunInfoPtr ConstructOpExecInfo() {
py::str op_name = "Conv2D";
py::object tensor_py_module = py::module::import("mindspore.common.tensor").attr("Tensor");
py::object np_py_module = py::module::import("numpy");
py::object np_ones = np_py_module.attr("ones");
py::object np_float32 = np_py_module.attr("float32");
py::tuple weight_dim = py::make_tuple(64, 3, 3, 3);
py::object weight = tensor_py_module(np_float32(np_ones(weight_dim)));
py::tuple op_params = py::make_tuple(weight);
py::tuple inputs_dim = py::make_tuple(1, 3, 6, 6);
py::object input = tensor_py_module(np_float32(np_ones(inputs_dim)));
py::tuple op_inputs = py::make_tuple(input, weight);
py::tuple kernel_size = py::make_tuple(3, 3);
py::dict op_attrs = py::dict("out_channel"_a = 64, "kernel_size"_a = kernel_size, "mode"_a = 1, "pad_mode"_a = "same",
"stride"_a = 1, "dilation"_a = 1, "group"_a = 1, "data_format"_a = kOpFormat_NCHW);
auto conv_obj = prim::GetPythonOps("conv2d_prim", "gtest_input.pynative");
py::none py_none;
py::args args = py::make_tuple(conv_obj, op_name, op_inputs);
return PyNativeExecutor::GetInstance()->forward_executor()->GenerateOpRunInfo(args);
}
/// Feature: Test pynative create context
/// Description: Test pynative create context interface
/// Expectation: success
TEST_F(TestPynativeExecute, TestCreateContext) {
auto context = MsContext::GetInstance();
MS_EXCEPTION_IF_NULL(context);
context->set_param<std::string>(MS_CTX_DEVICE_TARGET, "CPU");
auto ctx3 = MsContext::GetInstance();
ASSERT_EQ(ctx3->backend_policy(), "vm");
ASSERT_EQ(ctx3->get_param<std::string>(MS_CTX_DEVICE_TARGET), "CPU");
ctx3->set_backend_policy("ge_only");
ctx3->set_param<std::string>(MS_CTX_DEVICE_TARGET, "GPU");
auto ctx4 = MsContext::GetInstance();
ASSERT_EQ(ctx3.get(), ctx4.get());
ASSERT_EQ(ctx4->backend_policy(), "ge_only");
ASSERT_EQ(ctx4->get_param<std::string>(MS_CTX_DEVICE_TARGET), "GPU");
}
/// Feature: Test pynative default context
/// Description: Test pynative default context interface
/// Expectation: success
TEST_F(TestPynativeExecute, TestDefaultContext) {
auto ctx = MsContext::GetInstance();
ASSERT_EQ(std::string(ctx->backend_policy()), "ge_only");
auto ctx2 = MsContext::GetInstance();
ASSERT_EQ(ctx.get(), ctx2.get());
}
/// Feature: Test pynative infer operation
/// Description: Test pynative infer interface by using `matmul` ops
/// Expectation: success
TEST_F(TestPynativeExecute, TestInferOperator) {
auto conv_obj = prim::GetPythonOps("matmul", "gtest_input.pynative");
auto t1 = prim::GetPythonOps("tensor1", "gtest_input.pynative");
auto t2 = prim::GetPythonOps("tensor2", "gtest_input.pynative");
// Get run op info.
auto op_run_info = std::make_shared<FrontendOpRunInfo>();
op_run_info->base_op_run_info.op_name = "MatMul";
op_run_info->op_prim = conv_obj->cast<PrimitivePyPtr>();
ASSERT_NE(op_run_info->op_prim, nullptr);
(void)op_run_info->input_value.emplace_back(t1);
(void)op_run_info->input_value.emplace_back(t2);
// call infer operator.
auto infer_operator = std::make_shared<InferOperation>();
ValuePtr infer_value = infer_operator->DoInfer(op_run_info);
// Check abstract.
ASSERT_NE(infer_value, nullptr);
ASSERT_EQ(infer_value->isa<AnyValue>(), true);
auto output_abs = op_run_info->base_op_run_info.abstract;
ASSERT_NE(output_abs, nullptr);
ASSERT_EQ(output_abs->isa<abstract::AbstractTensor>(), true);
auto abs_tensor = output_abs->cast<abstract::AbstractTensorPtr>();
ASSERT_NE(abs_tensor, nullptr);
// Check type.
auto base_type = abs_tensor->BuildType();
ASSERT_NE(base_type, nullptr);
auto tensor_type = base_type->cast<TensorTypePtr>();
ASSERT_NE(tensor_type, nullptr);
ASSERT_EQ(tensor_type->element()->type_id(), kNumberTypeFloat32);
// Check shape.
auto base_shape = abs_tensor->BuildShape();
ASSERT_NE(base_shape, nullptr);
auto shape = base_shape->cast<abstract::ShapePtr>();
ASSERT_NE(shape, nullptr);
auto shape_v = shape->shape();
ASSERT_EQ(shape_v.size(), 2);
ASSERT_EQ(shape_v[0], 1);
ASSERT_EQ(shape_v[1], 1);
}
} // namespace pynative
} // namespace mindspore