From 6c54e0e86ec93367b2e2fedd0ad58177f16a17f4 Mon Sep 17 00:00:00 2001
From: Zhang Jun <ewalker@live.cn>
Date: Wed, 16 Nov 2022 14:45:04 +0800
Subject: [PATCH] [inference][trt] update trt hardswish plugin to layer
 (#47745)

---
 .../tensorrt/convert/hard_swish_op.cc         | 43 +++++++++++--------
 1 file changed, 26 insertions(+), 17 deletions(-)

diff --git a/paddle/fluid/inference/tensorrt/convert/hard_swish_op.cc b/paddle/fluid/inference/tensorrt/convert/hard_swish_op.cc
index add9e5638f6..5ce386c8a71 100644
--- a/paddle/fluid/inference/tensorrt/convert/hard_swish_op.cc
+++ b/paddle/fluid/inference/tensorrt/convert/hard_swish_op.cc
@@ -13,7 +13,6 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #include "paddle/fluid/inference/tensorrt/convert/op_converter.h"
-#include "paddle/fluid/inference/tensorrt/plugin/hard_swish_op_plugin.h"
 
 namespace paddle {
 namespace framework {
@@ -41,7 +40,6 @@ class HardSwishOpConverter : public OpConverter {
 
     framework::OpDesc op_desc(op, nullptr);
     // Declare inputs
-    int input_num = op_desc.Input("X").size();
     auto* input = engine_->GetITensor(op_desc.Input("X")[0]);
 
     const float threshold =
@@ -69,21 +67,32 @@ class HardSwishOpConverter : public OpConverter {
                                nvinfer1::ElementWiseOperation::kPROD);
       layer = eltwise_layer;
     } else {
-      if (engine_->with_dynamic_shape()) {
-#if IS_TRT_VERSION_GE(6000)
-        plugin::HardSwishPluginDynamic* plugin =
-            new plugin::HardSwishPluginDynamic(threshold, scale, offset);
-        layer = engine_->AddDynamicPlugin(&input, input_num, plugin);
-#else
-        PADDLE_THROW(platform::errors::Fatal(
-            "You are running the TRT Dynamic Shape mode, need to confirm that "
-            "your TRT version is no less than 6.0"));
-#endif
-      } else {
-        plugin::HardSwishPlugin* plugin =
-            new plugin::HardSwishPlugin(threshold, scale, offset);
-        layer = engine_->AddPlugin(&input, input_num, plugin);
-      }
+      int32_t rank = input->getDimensions().nbDims;
+      nvinfer1::Dims constant_shape;
+      constant_shape.nbDims = rank;
+      std::fill(constant_shape.d, constant_shape.d + rank, 1);
+      std::vector<float> weight_threshold_data{threshold};
+      std::vector<float> weight_scale_data{scale};
+      std::vector<float> weight_offset_data{offset};
+      std::vector<float> weight_zero_data{0.f};
+      auto* threshold_data =
+          AddConstantLayer(weight_threshold_data.data(), constant_shape);
+      auto* scale_data =
+          AddConstantLayer(weight_scale_data.data(), constant_shape);
+      auto* offset_data =
+          AddConstantLayer(weight_offset_data.data(), constant_shape);
+      auto* zero_data =
+          AddConstantLayer(weight_zero_data.data(), constant_shape);
+
+      auto* input_sum_with_offset = Sum(input, offset_data);
+      auto* pre_max_with_zero = Max(input_sum_with_offset, zero_data);
+      auto* pre_min_with_threshold = Min(pre_max_with_zero, threshold_data);
+      auto* pre_prod_with_input = Prod(pre_min_with_threshold, input);
+      layer = TRT_ENGINE_ADD_LAYER(engine_,
+                                   ElementWise,
+                                   *pre_prod_with_input,
+                                   *scale_data,
+                                   nvinfer1::ElementWiseOperation::kDIV);
     }
     auto output_name = op_desc.Output("Out")[0];
     RreplenishLayerAndOutput(layer, "hard_swish", {output_name}, test_mode);
-- 
GitLab