[XPU]Conv transpose fp16 && fix unittest (#53626)

* fix as review, add fp16 conv2d_transpose

* fix unittest of bn and reduce_mean

* fix bn unittest

* fix ci

* fix ci

paddle/fluid/framework/ir/xpu/conv2d_xpu_fuse_pass.cc

@@ -34,25 +34,6 @@ class Scope;
 }  // namespace framework
 }  // namespace paddle
 
-namespace {
-
-template <typename T1, typename T2>
-void ConvertTensorType(phi::DenseTensor* tensor) {
-  phi::DenseTensor tmp_tensor;
-  tmp_tensor.set_type(phi::CppTypeToDataType<T2>::Type());
-  tmp_tensor.Resize(tensor->dims());
-  auto* tmp_data = tmp_tensor.mutable_data<T2>(paddle::platform::CPUPlace());
-  auto* data = tensor->mutable_data<T1>(paddle::platform::CPUPlace());
-  for (int i = 0; i < tensor->numel(); i++) {
-    tmp_data[i] = static_cast<T2>(data[i]);
-  }
-  tensor->clear();
-  paddle::framework::TensorCopySync(
-      tmp_tensor, paddle::platform::CPUPlace(), tensor);
-}
-
-}  // namespace
-
 namespace paddle {
 namespace framework {
 namespace ir {
@@ -450,7 +431,7 @@ int Conv2dXPUFusePass::ApplyImpl(ir::Graph* graph,
     // conv_filter fp16 --> fp32
     auto tensor_type = filter_t->dtype();
     if (tensor_type == phi::DataType::FLOAT16) {
-      ConvertTensorType<float16, float>(filter_t);
+      CastToFp32(filter_t, nullptr);
     }
     auto filter_dims = filter_t->dims();
     bool has_bias = with_bn || with_conv_bias;
@@ -529,9 +510,6 @@ int Conv2dXPUFusePass::ApplyImpl(ir::Graph* graph,
         }
       }
     }
-    if (tensor_type == phi::DataType::FLOAT16) {
-      ConvertTensorType<float, float16>(filter_t);
-    }
     // filter max
     Node* filter_int16 = nullptr;
     Node* filter_max = nullptr;

paddle/phi/backends/xpu/xpu2_op_list.cc

@@ -164,7 +164,8 @@ XPUOpMap& get_kl2_ops() {
       {"conv3d",
        XPUKernelSet({phi::DataType::FLOAT32, phi::DataType::FLOAT16})},
       {"conv2d_transpose_grad", XPUKernelSet({phi::DataType::FLOAT32})},
-      {"conv2d_transpose", XPUKernelSet({phi::DataType::FLOAT32})},
+      {"conv2d_transpose",
+       XPUKernelSet({phi::DataType::FLOAT32, phi::DataType::FLOAT16})},
       {"cumsum",
        XPUKernelSet({phi::DataType::FLOAT32,
                      phi::DataType::FLOAT16,

paddle/phi/kernels/reduce_mean_kernel.cc

@@ -67,5 +67,6 @@ PD_REGISTER_KERNEL(
 #endif
 
 #if defined(PADDLE_WITH_XPU)
-PD_REGISTER_KERNEL(mean, XPU, ALL_LAYOUT, phi::MeanKernel, float) {}
+PD_REGISTER_KERNEL(
+    mean, XPU, ALL_LAYOUT, phi::MeanKernel, float, phi::dtype::float16) {}
 #endif

paddle/phi/kernels/xpu/activation_kernel.cc

@@ -195,13 +195,6 @@ void PowKernel(const Context& dev_ctx,
                const DenseTensor& x,
                const Scalar& factor,
                DenseTensor* out) {
-  // using XPUType = typename XPUTypeTrait<T>::Type;
-  // // dev_ctx.template Alloc<T>(out);
-  // auto pow_factor = factor.to<T>();
-  // const auto* x_data = reinterpret_cast<const XPUType*>(x.data<T>());
-  // auto* y_data = reinterpret_cast<XPUType*>(dev_ctx.template Alloc<T>(out));
-  // // const T* x_data = x.data<T>();
-  // // T* y_data = out->data<T>();
   dev_ctx.template Alloc<T>(out);
   float pow_factor = factor.to<float>();
   const T* x_data = x.data<T>();
@@ -578,16 +571,11 @@ PD_REGISTER_KERNEL(
 
 PD_REGISTER_ACTIVATION_KERNEL(exp, ExpKernel)  // no grad
 PD_REGISTER_ACTIVATION_KERNEL(floor, FloorKernel)
-// PD_REGISTER_ACTIVATION_KERNEL(leaky_relu, LeakyReluKernel)
-// PD_REGISTER_ACTIVATION_KERNEL(hard_sigmoid, HardSigmoidKernel)
 PD_REGISTER_ACTIVATION_KERNEL(hardswish, HardSwishKernel)
 PD_REGISTER_ACTIVATION_KERNEL(mish, MishKernel)
 PD_REGISTER_ACTIVATION_KERNEL(pow, PowKernel)
 PD_REGISTER_ACTIVATION_KERNEL(reciprocal, ReciprocalKernel)
 PD_REGISTER_ACTIVATION_KERNEL(relu6_raw, Relu6RawKernel)
-// PD_REGISTER_ACTIVATION_KERNEL(sigmoid, SigmoidKernel)
-// PD_REGISTER_ACTIVATION_KERNEL(sqrt, SqrtKernel)
-// PD_REGISTER_ACTIVATION_KERNEL(swish_raw, SwishRawKernel)
 PD_REGISTER_ACTIVATION_KERNEL(softplus, SoftplusKernel)
 PD_REGISTER_ACTIVATION_KERNEL(sin, SinKernel)
 PD_REGISTER_ACTIVATION_KERNEL(cos, CosKernel)

paddle/phi/kernels/xpu/conv_transpose_kernel.cc

@@ -123,11 +123,11 @@ void Conv2dTransposeKernel(const Context& ctx,
         true);
     PADDLE_ENFORCE_XDNN_SUCCESS(r, "conv2d_transpose_v2");
   } else {
-    int r = xpu::conv2d_transpose_v2<float, float, float, int16_t>(
+    int r = xpu::conv2d_transpose_v2<XPUT, XPUT, XPUT, int16_t>(
         ctx.x_context(),
-        x.data<float>(),
-        filter_.data<float>(),
-        out->data<float>(),
+        reinterpret_cast<const XPUT*>(x.data<T>()),
+        reinterpret_cast<const XPUT*>(filter.data<T>()),
+        reinterpret_cast<XPUT*>(out->data<T>()),
         batch_size,
         img_yc,
         img_xh,
@@ -148,5 +148,9 @@ void Conv2dTransposeKernel(const Context& ctx,
 
 }  // namespace phi
 
-PD_REGISTER_KERNEL(
-    conv2d_transpose, XPU, ALL_LAYOUT, phi::Conv2dTransposeKernel, float) {}
+PD_REGISTER_KERNEL(conv2d_transpose,
+                   XPU,
+                   ALL_LAYOUT,
+                   phi::Conv2dTransposeKernel,
+                   float,
+                   phi::dtype::float16) {}

test/xpu/test_batch_norm_op_xpu.py

@@ -154,7 +154,6 @@ class XPUTestBatchNormOp(XPUOpTestWrapper):
     class TestBatchNormOp(unittest.TestCase):
         def setUp(self):
             self.op_type = "batch_norm"
-            self.dtype = np.float32
             self.shape = [2, 3, 4, 5]
             self.data_layout = "NCHW"
             self.epsilon = 1e-05
@@ -162,6 +161,9 @@ class XPUTestBatchNormOp(XPUOpTestWrapper):
             self.init_dtype()
             self.set_xpu()
             self.set_attrs()
+            self.rtol = 1e-5
+            if self.dtype == np.float16:
+                self.rtol = 1e-2
 
             if self.data_layout == "NHWC":
                 channel_size = self.shape[3]
@@ -175,15 +177,15 @@ class XPUTestBatchNormOp(XPUOpTestWrapper):
             np.random.seed(1024)
             self.x_np = np.random.random_sample(self.shape).astype(self.dtype)
             self.scale_np = np.random.random_sample([channel_size]).astype(
-                self.dtype
+                np.float32
             )
             self.bias_np = np.random.random_sample([channel_size]).astype(
-                self.dtype
+                np.float32
             )
-            self.mean_np = np.zeros([channel_size]).astype(self.dtype)
-            self.variance_np = np.ones([channel_size]).astype(self.dtype)
-            self.saved_mean_np = np.zeros([channel_size]).astype(self.dtype)
-            self.saved_variance_np = np.ones([channel_size]).astype(self.dtype)
+            self.mean_np = np.zeros([channel_size]).astype(np.float32)
+            self.variance_np = np.ones([channel_size]).astype(np.float32)
+            self.saved_mean_np = np.zeros([channel_size]).astype(np.float32)
+            self.saved_variance_np = np.ones([channel_size]).astype(np.float32)
 
         def set_attrs(self):
             pass
@@ -244,7 +246,110 @@ class XPUTestBatchNormOp(XPUOpTestWrapper):
                 self.epsilon,
                 self.data_layout,
             )
-            np.testing.assert_allclose(y_np_ref, y_np, rtol=1e-05)
+            np.testing.assert_allclose(y_np_ref, y_np, rtol=self.rtol)
+
+    class TestBatchNormOpUseGlobalStats(unittest.TestCase):
+        def setUp(self):
+            self.places = [paddle.XPUPlace(0)]
+            self.init_test()
+
+        # train mode
+        def init_test(self):
+            self.use_global_stats = True
+            self.trainable_statistics = False
+
+        def test_global_stats(self):
+            for p in self.places:
+                with fluid.dygraph.guard(p):
+                    x = paddle.randn([2, 6, 6, 4])
+                    net1 = paddle.nn.BatchNorm(
+                        6,
+                        param_attr=fluid.ParamAttr(
+                            initializer=paddle.nn.initializer.Constant(1.0)
+                        ),
+                        use_global_stats=self.use_global_stats,
+                        trainable_statistics=self.trainable_statistics,
+                    )
+                    net2 = paddle.nn.BatchNorm2D(
+                        6, use_global_stats=self.use_global_stats
+                    )
+                    net2.weight = net1.weight
+                    net2.bias = net1.bias
+                    if self.trainable_statistics:
+                        net1.training = False
+                        net2.training = False
+                    y1 = net1(x)
+                    y2 = net2(x)
+                    np.testing.assert_allclose(
+                        y1.numpy(), y2.numpy(), rtol=1e-5
+                    )
+
+    class TestBatchNormOpUseGlobalStats1(TestBatchNormOpUseGlobalStats):
+        # test mode
+        def init_test(self):
+            self.use_global_stats = True
+            self.trainable_statistics = True
+
+    class TestBatchNormUseGlobalStats2(TestBatchNormOpUseGlobalStats):
+        # train mode
+        def init_test(self):
+            self.use_global_stats = True
+            self.trainable_statistics = False
+
+
+support_types = get_xpu_op_support_types('batch_norm')
+for stype in support_types:
+    create_test_class(globals(), XPUTestBatchNormOp, stype)
+
+
+class XPUTestBatchNormGradOp(XPUOpTestWrapper):
+    def __init__(self):
+        self.op_name = 'batch_norm'
+        self.use_dynamic_create_class = False
+
+    class TestBatchNormGradOp(unittest.TestCase):
+        def setUp(self):
+            self.op_type = "batch_norm"
+            self.shape = [2, 3, 4, 5]
+            self.data_layout = "NCHW"
+            self.epsilon = 1e-05
+            self.momentum = 0.9
+            self.init_dtype()
+            self.set_xpu()
+            self.set_attrs()
+
+            if self.data_layout == "NHWC":
+                channel_size = self.shape[3]
+            elif self.data_layout == "NCHW":
+                channel_size = self.shape[1]
+            else:
+                raise ValueError(
+                    "Unsupported data layout! Only NCHW and NHWC is supported, but received "
+                    + self.data_layout
+                )
+            np.random.seed(1024)
+            self.x_np = np.random.random_sample(self.shape).astype(self.dtype)
+            self.scale_np = np.random.random_sample([channel_size]).astype(
+                np.float32
+            )
+            self.bias_np = np.random.random_sample([channel_size]).astype(
+                np.float32
+            )
+            self.mean_np = np.zeros([channel_size]).astype(np.float32)
+            self.variance_np = np.ones([channel_size]).astype(np.float32)
+            self.saved_mean_np = np.zeros([channel_size]).astype(np.float32)
+            self.saved_variance_np = np.ones([channel_size]).astype(np.float32)
+
+        def set_attrs(self):
+            pass
+
+        def init_dtype(self):
+            self.dtype = self.in_type
+
+        def set_xpu(self):
+            self.__class__.use_xpu = True
+            self.__class__.op_type = self.in_type
+            self.place = paddle.XPUPlace(0)
 
         def test_train(self):
             y_grad_np = np.random.random_sample(self.shape).astype(self.dtype)
@@ -349,58 +454,10 @@ class XPUTestBatchNormOp(XPUOpTestWrapper):
                         outputs[name], outs[id], rtol=1e-05, atol=1e-4
                     )
 
-    class TestBatchNormOpUseGlobalStats(unittest.TestCase):
-        def setUp(self):
-            self.places = [paddle.XPUPlace(0)]
-            self.init_test()
-
-        # train mode
-        def init_test(self):
-            self.use_global_stats = True
-            self.trainable_statistics = False
-
-        def test_global_stats(self):
-            for p in self.places:
-                with fluid.dygraph.guard(p):
-                    x = paddle.randn([2, 6, 6, 4])
-                    net1 = paddle.nn.BatchNorm(
-                        6,
-                        param_attr=fluid.ParamAttr(
-                            initializer=paddle.nn.initializer.Constant(1.0)
-                        ),
-                        use_global_stats=self.use_global_stats,
-                        trainable_statistics=self.trainable_statistics,
-                    )
-                    net2 = paddle.nn.BatchNorm2D(
-                        6, use_global_stats=self.use_global_stats
-                    )
-                    net2.weight = net1.weight
-                    net2.bias = net1.bias
-                    if self.trainable_statistics:
-                        net1.training = False
-                        net2.training = False
-                    y1 = net1(x)
-                    y2 = net2(x)
-                    np.testing.assert_allclose(
-                        y1.numpy(), y2.numpy(), rtol=1e-05
-                    )
-
-    class TestBatchNormOpUseGlobalStats1(TestBatchNormOpUseGlobalStats):
-        # test mode
-        def init_test(self):
-            self.use_global_stats = True
-            self.trainable_statistics = True
-
-    class TestBatchNormUseGlobalStats2(TestBatchNormOpUseGlobalStats):
-        # train mode
-        def init_test(self):
-            self.use_global_stats = True
-            self.trainable_statistics = False
 
-
-support_types = get_xpu_op_support_types('batch_norm')
-for stype in support_types:
-    create_test_class(globals(), XPUTestBatchNormOp, stype)
+support_types_grad = get_xpu_op_support_types('batch_norm_grad')
+for stype_grad in support_types_grad:
+    create_test_class(globals(), XPUTestBatchNormGradOp, stype_grad)
 
 if __name__ == '__main__':
     unittest.main()