[XPU] add transpose bridge and unit test (#2630)

* [XPU] add transpose bridge and unit test

test=develop

lite/backends/xpu/device.cc

@@ -30,7 +30,12 @@ std::unique_ptr<xtcl::network::xRuntimeInstance> Device::Build(
 
   // The XPU compiler build the graph and fill all of the constant params, only
   // one output is supported now.
-  xtcl::xNetwork network = builder->FinalizeNetwork(*((*outputs)[0]));
+  xtcl::Array<xtcl::xExpr> all_outs;
+  for (size_t i = 0; i < outputs->size(); i++) {
+    all_outs.push_back(*outputs->at(i));
+  }
+  xtcl::xNetwork network =
+      builder->FinalizeNetwork(xtcl::relay::TupleNode::make(all_outs));
   auto target = xtcl::Target::Create(device_name_);
   auto compiler = xtcl::network::xTensorCompiler(network, target);
   compiler.SetParams(*params);  // Set the data of constant tensors

lite/core/arena/framework.cc

@@ -35,12 +35,12 @@ void TestCase::CreateInstruction() {
     op_desc_.reset(new cpp::OpDesc());
     op_desc_->SetType("subgraph");
     op_desc_->SetAttr<int32_t>("sub_block", sub_block_idx);
-    op_desc_->SetInput("Inputs", op_desc_->input_vars());
-    op_desc_->SetOutput("Outputs", op_desc_->output_vars());
-    op_desc_->SetAttr<std::vector<std::string>>(
-        "input_data_names", sub_block_op_desc->input_vars());
-    op_desc_->SetAttr<std::vector<std::string>>(
-        "output_data_names", sub_block_op_desc->output_vars());
+    auto in_names = sub_block_op_desc->input_vars();
+    auto out_names = sub_block_op_desc->output_vars();
+    op_desc_->SetInput("Inputs", in_names);
+    op_desc_->SetOutput("Outputs", out_names);
+    op_desc_->SetAttr<std::vector<std::string>>("input_data_names", in_names);
+    op_desc_->SetAttr<std::vector<std::string>>("output_data_names", out_names);
     op = LiteOpRegistry::Global().Create(op_desc().Type());
     static_cast<operators::SubgraphOp*>(op.get())->SetSubBlock(sub_block_desc);
   } else {

lite/core/arena/framework.h

@@ -188,13 +188,17 @@ class Arena {
     tester_->Prepare();
   }
 
-  bool TestPrecision() {
+  bool TestPrecision(const std::vector<std::string>& exclude_outs = {}) {
     tester_->RunBaseline(tester_->baseline_scope());
     tester_->RunInstruction();
 
     bool success = true;
     for (auto& out : tester_->op_desc().OutputArgumentNames()) {
       for (auto& var : tester_->op_desc().Output(out)) {
+        if (std::find(exclude_outs.begin(), exclude_outs.end(), var) !=
+            exclude_outs.end()) {
+          continue;
+        }
         success = success && CompareTensor(out, var);
       }
     }

lite/kernels/xpu/bridges/CMakeLists.txt

@@ -14,6 +14,7 @@ lite_cc_library(subgraph_bridge_pool_op_xpu SRCS pool_op.cc DEPS ${subgraph_brid
 lite_cc_library(subgraph_bridge_softmax_op_xpu SRCS softmax_op.cc DEPS ${subgraph_bridge_deps_xpu})
 lite_cc_library(subgraph_bridge_mul_op_xpu SRCS mul_op.cc DEPS ${xpu_subgraph_bridge_deps})
 lite_cc_library(subgraph_bridge_batch_norm_op_xpu SRCS batch_norm_op.cc DEPS ${xpu_subgraph_bridge_deps})
+lite_cc_library(subgraph_bridge_transpose_op_xpu SRCS transpose_op.cc DEPS ${xpu_subgraph_bridge_deps})
 
 set(xpu_subgraph_bridges
         subgraph_bridge_registry
@@ -26,6 +27,7 @@ set(xpu_subgraph_bridges
         subgraph_bridge_softmax_op_xpu
         subgraph_bridge_mul_op_xpu
         subgraph_bridge_batch_norm_op_xpu
+        subgraph_bridge_transpose_op_xpu
         CACHE INTERNAL "xpu_subgraph_bridges")
 
 message(STATUS "+++++ xpu_subgraph_bridges: ${xpu_subgraph_bridges}")

lite/kernels/xpu/bridges/paddle_use_bridges.h

@@ -22,3 +22,5 @@ USE_SUBGRAPH_BRIDGE(XPU, pool2d);
 USE_SUBGRAPH_BRIDGE(XPU, softmax);
 USE_SUBGRAPH_BRIDGE(XPU, mul);
 USE_SUBGRAPH_BRIDGE(XPU, batch_norm);
+USE_SUBGRAPH_BRIDGE(XPU, transpose);
+USE_SUBGRAPH_BRIDGE(XPU, transpose2);

lite/kernels/xpu/bridges/transpose_op.cc

+// Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
+//
+// 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 "lite/kernels/npu/bridges/registry.h"
+#include "lite/kernels/xpu/bridges/graph.h"
+#include "lite/kernels/xpu/bridges/utility.h"
+
+namespace paddle {
+namespace lite {
+namespace subgraph {
+namespace xpu {
+
+int TransposeConverter(void* ctx, OpLite* op) {
+  CHECK(ctx != nullptr);
+  CHECK(op != nullptr);
+  auto graph = static_cast<Graph*>(ctx);
+  auto op_info = op->op_info();
+  auto op_type = op_info->Type();
+  VLOG(3) << "[XPU] Converting " + op_type + "...";
+
+  // Create node and set params from op
+  auto x_var_name = op_info->Input("X").front();
+  auto out_var_name = op_info->Output("Out").front();
+
+  auto axis = op_info->GetAttr<std::vector<int>>("axis");
+
+  CHECK(graph->HasNode(x_var_name));
+  graph->AddNode(
+      out_var_name,
+      graph->builder_.CreateTranspose(
+          *graph->GetNode(x_var_name),
+          Cvt2ArrayInt(std::vector<int64_t>(axis.begin(), axis.end()))));
+
+  return SUCCESS;
+}
+
+}  // namespace xpu
+}  // namespace subgraph
+}  // namespace lite
+}  // namespace paddle
+
+REGISTER_SUBGRAPH_BRIDGE(XPU,
+                         transpose,
+                         paddle::lite::subgraph::xpu::TransposeConverter);
+REGISTER_SUBGRAPH_BRIDGE(XPU,
+                         transpose2,
+                         paddle::lite::subgraph::xpu::TransposeConverter);

lite/kernels/xpu/bridges/utility.cc

@@ -125,6 +125,18 @@ std::shared_ptr<xtcl::xNDArray> CvtTensor(const Tensor& in_tensor,
   return out_tensor;
 }
 
+xtcl::Array<xtcl::Integer> Cvt2ArrayInt(const std::vector<int64_t>& input) {
+  xtcl::Array<xtcl::Integer> output;
+  for (auto i : input) {
+    output.push_back(i);
+  }
+  return output;
+}
+
+xtcl::Array<xtcl::Integer> Cvt2ArrayInt(const DDim& input) {
+  return Cvt2ArrayInt(input.Vectorize());
+}
+
 }  // namespace xpu
 }  // namespace subgraph
 }  // namespace lite

lite/kernels/xpu/bridges/utility.h

@@ -47,6 +47,9 @@ std::shared_ptr<xtcl::xNDArray> CvtTensor(
     PrecisionType in_ptype = PRECISION(kFloat),
     DataLayoutType in_ltype = DATALAYOUT(kNCHW));
 
+xtcl::Array<xtcl::Integer> Cvt2ArrayInt(const std::vector<int64_t>& input);
+xtcl::Array<xtcl::Integer> Cvt2ArrayInt(const DDim& input);
+
 }  // namespace xpu
 }  // namespace subgraph
 }  // namespace lite

lite/kernels/xpu/subgraph_compute.cc

@@ -60,9 +60,14 @@ int SubgraphEngine::BuildDeviceProgram() {
   // Obtain the output nodes of the XPU IR graph and build the graph to XPU
   // runtime
   std::vector<xtcl::xExpr*> output_nodes;
+  std::vector<std::string> valid_output_names;
   for (auto& output_name : output_names_) {
-    output_nodes.push_back(graph.GetNode(output_name).get());
+    if (graph.HasNode(output_name)) {
+      output_nodes.push_back(graph.GetNode(output_name).get());
+      valid_output_names.push_back(output_name);
+    }
   }
+  CHECK(!valid_output_names.empty()) << "[XPU] no valid output names";
   device_program_ = lite::xpu::Device::Global().Build(
       &graph.builder_, &graph.params_, &output_nodes);
   if (device_program_ == nullptr) {
@@ -73,16 +78,16 @@ int SubgraphEngine::BuildDeviceProgram() {
   // Query and check the dimensions of input and output tensors
   origin_idims_.resize(input_names_.size());
   origin_itensors_.resize(input_names_.size());
-  origin_odims_.resize(output_names_.size());
-  origin_otensors_.resize(output_names_.size());
+  origin_odims_.resize(valid_output_names.size());
+  origin_otensors_.resize(valid_output_names.size());
   for (int i = 0; i < input_names_.size(); i++) {
     origin_itensors_[i] = scope_->FindMutableTensor(input_names_[i]);
     CHECK(origin_itensors_[i]);
     origin_idims_[i] = origin_itensors_[i]->dims();
     VLOG(3) << "[XPU] Input dims[" << i << "]: " << origin_idims_[i];
   }
-  for (int i = 0; i < output_names_.size(); i++) {
-    origin_otensors_[i] = scope_->FindMutableTensor(output_names_[i]);
+  for (int i = 0; i < valid_output_names.size(); i++) {
+    origin_otensors_[i] = scope_->FindMutableTensor(valid_output_names[i]);
     CHECK(origin_otensors_[i]);
     origin_odims_[i] = origin_otensors_[i]->dims();
     VLOG(3) << "[XPU] Output dims[" << i << "]: " << origin_odims_[i];
@@ -113,7 +118,7 @@ int SubgraphEngine::LaunchDeviceProgram() {
   device_program_->Run();
   VLOG(3) << "[XPU] Process cost " << GetCurrentUS() - start_time << " us";
   // Copy the data of output XPU tensor to the buffer of origin output tensors
-  for (size_t i = 0; i < output_names_.size(); i++) {
+  for (size_t i = 0; i < origin_otensors_.size(); i++) {
     auto output_ndarray = device_program_->GetOutput(i);
     std::memcpy(origin_otensors_[i]->mutable_data<float>(),
                 static_cast<float*>(output_ndarray.ToDLPack()->dl_tensor.data),

lite/operators/transpose_op.cc

@@ -135,6 +135,15 @@ bool Transpose2Op::InferShape() const {
     out_dims[i] = x_dims[axis[i]];
   }
   param_.output->Resize(out_dims);
+
+  std::vector<DDim::value_type> xshape_dims(x_dims.size() + 1, 0);
+  for (size_t i = 0; i < x_dims.size(); i++) {
+    xshape_dims[i + 1] = x_dims[i];
+  }
+  param_.xshape->Resize(xshape_dims);
+  auto xshape_lod = param_.xshape->mutable_lod();
+  *xshape_lod = param_.x->lod();
+
   return true;
 }
 

lite/tests/kernels/CMakeLists.txt

-if((NOT LITE_WITH_OPENCL AND NOT LITE_WITH_FPGA AND NOT LITE_WITH_XPU) AND (LITE_WITH_X86 OR LITE_WITH_ARM))
+if((NOT LITE_WITH_OPENCL AND NOT LITE_WITH_FPGA) AND (LITE_WITH_X86 OR LITE_WITH_ARM))
     lite_cc_test(test_kernel_scale_compute SRCS scale_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
     lite_cc_test(test_kernel_power_compute SRCS power_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
     lite_cc_test(test_kernel_shuffle_channel_compute SRCS shuffle_channel_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
@@ -24,6 +24,7 @@ if((NOT LITE_WITH_OPENCL AND NOT LITE_WITH_FPGA AND NOT LITE_WITH_XPU) AND (LITE
     #lite_cc_test(test_kernel_write_to_array_compute SRCS write_to_array_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
     #lite_cc_test(test_kernel_read_from_array_compute SRCS read_from_array_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
     lite_cc_test(test_concat_compute SRCS concat_compute_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
+    lite_cc_test(test_kernel_transpose_compute SRCS transpose_compute_test.cc DEPS arena_framework ${xpu_kernels} ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})
 
 if(LITE_BUILD_EXTRA)
     lite_cc_test(test_gru_unit SRCS gru_unit_test.cc DEPS arena_framework ${x86_kernels} ${arm_kernels} ${lite_ops} ${host_kernels})

lite/tests/kernels/transpose_compute_test.cc

+// Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserved.
+//
+// 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 <gtest/gtest.h>
+#include "lite/api/paddle_use_kernels.h"
+#include "lite/api/paddle_use_ops.h"
+#include "lite/core/arena/framework.h"
+
+namespace paddle {
+namespace lite {
+
+int data_index(std::vector<int> pos, DDimLite dims) {
+  int d1 = dims[1];
+  int d2 = dims[2];
+  int d3 = dims[3];
+  return pos[3] + pos[2] * d3 + pos[1] * d3 * d2 + pos[0] * d3 * d2 * d1;
+}
+
+std::vector<int> pos_trans(std::vector<int> in_pos, std::vector<int> axis) {
+  std::vector<int> out_pos(in_pos.size());
+  for (int i = 0; i < axis.size(); i++) {
+    out_pos[axis[i]] = in_pos[i];
+  }
+  return out_pos;
+}
+
+class TransposeComputeTester : public arena::TestCase {
+ protected:
+  // common attributes for this op.
+  std::string op_type_ = "transpose2";
+  std::string input_ = "x";
+  std::string output_ = "out";
+  std::string xshape_ = "xshape";
+  DDim x_dims_;
+  std::vector<int> axis_;
+
+ public:
+  TransposeComputeTester(const Place& place,
+                         const std::string& alias,
+                         DDim x_dims,
+                         std::vector<int> axis)
+      : TestCase(place, alias), x_dims_(x_dims), axis_(axis) {}
+
+  void RunBaseline(Scope* scope) override {
+    auto* out = scope->NewTensor(output_);
+    CHECK(out);
+
+    auto* x = scope->FindTensor(input_);
+    auto x_dims = x->dims();
+
+    std::vector<int64_t> out_shape(x_dims.size(), 0);
+    for (size_t i = 0; i < x_dims.size(); i++) {
+      out_shape[i] = x_dims[axis_[i]];
+    }
+    out->Resize(out_shape);
+
+    auto y_dims = out->dims();
+
+    int input_n = x_dims[0];
+    int input_c = x_dims[1];
+    int input_h = x_dims[2];
+    int input_w = x_dims[3];
+
+    auto input_data = x->data<float>();
+    auto output_data = out->mutable_data<float>();
+
+    for (int n = 0; n < input_n; ++n) {
+      for (int c = 0; c < input_c; ++c) {
+        for (int h = 0; h < input_h; ++h) {
+          for (int w = 0; w < input_w; ++w) {
+            std::vector<int> in_pos{n, c, h, w};
+            std::vector<int> out_pos = pos_trans(in_pos, axis_);
+            int in_index = data_index(in_pos, x_dims);
+            int out_index = data_index(out_pos, y_dims);
+            output_data[out_index] = input_data[in_index];
+          }
+        }
+      }
+    }
+
+    if (op_type_ == "transpose2") {
+      auto* xshape = scope->NewTensor(xshape_);
+      auto xshape_dims = x_dims.Vectorize();
+      xshape_dims.insert(xshape_dims.begin(), 0);
+      xshape->Resize(xshape_dims);
+    }
+  }
+
+  void PrepareOpDesc(cpp::OpDesc* op_desc) {
+    op_desc->SetType(op_type_);
+    op_desc->SetInput("X", {input_});
+    op_desc->SetOutput("Out", {output_});
+    if (op_type_ == "transpose2") {
+      op_desc->SetOutput("XShape", {xshape_});
+    }
+    op_desc->SetAttr("axis", axis_);
+  }
+
+  void PrepareData() override {
+    std::vector<float> data(x_dims_.production());
+    for (int i = 0; i < x_dims_.production(); i++) {
+      data[i] = i * 1.1;
+    }
+    SetCommonTensor(input_, x_dims_, data.data());
+  }
+};
+
+TEST(Transpose, precision) {
+  LOG(INFO) << "test Transpose op";
+  float abs_error = 2e-5;
+  Place place;
+#ifdef LITE_WITH_XPU
+  place = TARGET(kXPU);
+#else
+  return;
+#endif
+
+  DDim x_dims{{2, 3, 4, 5}};
+  // [XPU]: {3, 1, 0, 2} is unsupported
+  std::vector<std::vector<int>> axes{
+      {0, 1, 2, 3}, {0, 1, 3, 2}, {0, 2, 1, 3}, {3, 1, 2, 0}};
+  for (auto axis : axes) {
+    std::unique_ptr<arena::TestCase> tester(
+        new TransposeComputeTester(place, "def", x_dims, axis));
+    arena::Arena arena(std::move(tester), place, abs_error);
+    arena.TestPrecision({"xshape"});
+  }
+}
+
+}  // namespace lite
+}  // namespace paddle