fix for conv2D training error (#38938)

PaddlePaddle · Jan 17, 2022 · 944ea43 · 944ea43
1 parent 05c98ec
commit 944ea43
Show file tree

Hide file tree

Showing 3 changed files with 42 additions and 2 deletions.
diff --git a/paddle/fluid/operators/mkldnn/conv_mkldnn_op.cc b/paddle/fluid/operators/mkldnn/conv_mkldnn_op.cc
@@ -613,7 +613,7 @@ class ConvMKLDNNHandlerT
     auto weights_mem_p = this->AcquireMemory("@weights_mem_p_target");
     if (is_test && weights_mem_p) {
       return weights_mem_p;
-    } else {
+    } else if (is_test) {
       const K* filter_data = filter->data<K>();
       auto weights_tz = framework::vectorize(filter->dims());
       platform::GetGroupConvWeightsTz(weights_tz, groups);
@@ -626,6 +626,19 @@ class ConvMKLDNNHandlerT
           user_src_md, this->fwd_pd_->weights_desc(),
           platform::to_void_cast<K>(filter_data), "@weights_mem_p", is_test, {},
           scale_data, mask);
+    } else {
+      const T* filter_data = filter->data<T>();
+      auto weights_tz = framework::vectorize(filter->dims());
+      platform::GetGroupConvWeightsTz(weights_tz, groups);
+
+      auto user_src_md = platform::MKLDNNMemDesc(
+          weights_tz, platform::MKLDNNGetDataType<T>(),
+          GetWeightsFormat(filter->format(), groups, is_conv3d));
+
+      return this->AcquireMemoryWithReorder(
+          user_src_md, this->fwd_pd_->weights_desc(),
+          platform::to_void_cast<T>(filter_data), "@weights_mem_p", is_test, {},
+          scale_data, mask);
     }
   }
 
@@ -1027,7 +1040,8 @@ REGISTER_OP_KERNEL_WITH_CUSTOM_TYPE(conv2d_grad, MKLDNN,
 REGISTER_OP_KERNEL_WITH_CUSTOM_TYPE(
     conv2d_grad, MKLDNN, ::paddle::platform::CPUPlace, BF16,
     ops::kConvMKLDNNFP32,
-    ops::ConvMKLDNNGradOpKernel<paddle::platform::bfloat16, float>);
+    ops::ConvMKLDNNGradOpKernel<paddle::platform::bfloat16,
+                                paddle::platform::bfloat16>);
 
 REGISTER_OP_KERNEL_WITH_CUSTOM_TYPE(depthwise_conv2d, MKLDNN,
                                     ::paddle::platform::CPUPlace, FP32,

diff --git a/paddle/fluid/platform/mkldnn_reuse.h b/paddle/fluid/platform/mkldnn_reuse.h
@@ -377,6 +377,14 @@ class MKLDNNHandlerT {
     if (bwd_pd_ == nullptr) {
       return false;
     } else {
+      if (std::is_same<TBackward_params, mkldnn_dummy_primitive>::value ==
+          false) {
+        const std::string key_bw_w_pd = key_ + "@bwd_w_pd";
+        bwd_w_pd_ =
+            std::static_pointer_cast<typename TBackward_params::primitive_desc>(
+                dev_ctx_.GetBlob(key_bw_w_pd));
+      }
+
       // When BWD is cached then still we need to Get FWD PD
       const std::string key_fpd = key_ + "@fwd_pd";
       fwd_pd_ = std::static_pointer_cast<typename TForward::primitive_desc>(

diff --git a/python/paddle/fluid/tests/unittests/mkldnn/test_conv2d_bf16_mkldnn_op.py b/python/paddle/fluid/tests/unittests/mkldnn/test_conv2d_bf16_mkldnn_op.py
@@ -50,6 +50,7 @@ def setUp(self):
         self.init_fuse_residual()
         self.init_data_type()
         self.init_force_fp32_output()
+        self.init_infer_or_train()
 
         self.conv2d_param = {
             'stride': self.stride,
@@ -83,6 +84,9 @@ def setUp(self):
         if self.input_type is not np.float32:
             self.input = convert_float_to_uint16(self.input)
 
+        if self.weight_type is not np.float32:
+            self.filter = convert_float_to_uint16(self.filter)
+
         self.inputs = {
             'Input': self.input,
             'Filter': OpTest.np_dtype_to_fluid_dtype(
@@ -105,6 +109,8 @@ def setUp(self):
             'fuse_residual_connection': self.fuse_residual
         }
 
+        self.init_additional_attrs()
+
     def test_check_output(self):
         self.check_output_with_place(core.CPUPlace())
 
@@ -141,6 +147,12 @@ def init_fuse_relu(self):
     def init_fuse_residual(self):
         self.fuse_residual = True
 
+    def init_infer_or_train(self):
+        self.weight_type = np.float32
+
+    def init_additional_attrs(self):
+        self.attrs['is_test'] = True
+
 
 @OpTestTool.skip_if_not_cpu_bf16()
 class TestConv2DWithGradBF16Op(TestConv2DBF16Op):
@@ -150,6 +162,12 @@ def init_fuse_relu(self):
     def init_fuse_residual(self):
         self.fuse_residual = None
 
+    def init_additional_attrs(self):
+        self.attrs['is_test'] = False
+
+    def init_infer_or_train(self):
+        self.weight_type = np.uint16
+
     def test_check_grad(self):
         dout = self.conv_output_float
         x = self.inputs_fp32['Input']