[Sparse] Rename and fix doc (#46853)

paddle/phi/api/yaml/sparse_backward.yaml

@@ -124,8 +124,8 @@
            cast_csr_grad {sparse_csr, sparse_csr -> sparse_csr}
     data_type : out_grad
 
-- backward_op : conv3d_coo_grad
-  forward : conv3d_coo (Tensor x, Tensor kernel, int[] paddings, int[] dilations, int[] strides, int groups, bool subm, str key) -> Tensor(out), Tensor(rulebook), Tensor(counter)
+- backward_op : conv3d_grad
+  forward : conv3d (Tensor x, Tensor kernel, int[] paddings, int[] dilations, int[] strides, int groups, bool subm, str key) -> Tensor(out), Tensor(rulebook), Tensor(counter)
   args : (Tensor x, Tensor kernel, Tensor out, Tensor rulebook, Tensor counter, Tensor out_grad, int[] paddings, int[] dilations, int[] strides, int groups, bool subm, str key)
   output : Tensor(x_grad), Tensor(kernel_grad)
   infer_meta :

paddle/phi/api/yaml/sparse_ops.yaml

@@ -119,7 +119,7 @@
     func : conv3d_coo{sparse_coo, dense -> sparse_coo, dense, dense}
     layout : x
   intermediate: rulebook, counter
-  backward : conv3d_coo_grad
+  backward : conv3d_grad
 
 - op : divide
   args : (Tensor x, Tensor y)
@@ -139,8 +139,8 @@
     func : UnchangedInferMeta
     param : [x]
   kernel :
-    func : divide_coo_scalar{sparse_coo -> sparse_coo},
-           divide_csr_scalar{sparse_csr -> sparse_csr}
+    func : divide_scalar_coo{sparse_coo -> sparse_coo},
+           divide_scalar_csr{sparse_csr -> sparse_csr}
   backward : divide_scalar_grad
 
 - op : expm1
@@ -393,7 +393,7 @@
   infer_meta :
     func : UnchangedInferMeta
   kernel :
-    func: coalesce{sparse_coo -> sparse_coo}
+    func: coalesce_coo{sparse_coo -> sparse_coo}
     layout : x
 
 - op: full_like
@@ -403,8 +403,8 @@
     func : CreateLikeInferMeta
     param : [x, dtype]
   kernel :
-    func : coo_full_like{sparse_coo -> sparse_coo},
-           csr_full_like{sparse_csr -> sparse_csr}
+    func : full_like_coo{sparse_coo -> sparse_coo},
+           full_like_csr{sparse_csr -> sparse_csr}
     layout : x
     data_type : dtype
 

paddle/phi/kernels/sparse/coalesce_kernel.h

@@ -22,14 +22,14 @@ namespace phi {
 namespace sparse {
 
 template <typename T, typename Context>
-void CoalesceKernel(const Context& dev_ctx,
-                    const SparseCooTensor& x,
-                    SparseCooTensor* out);
+void CoalesceCooKernel(const Context& dev_ctx,
+                       const SparseCooTensor& x,
+                       SparseCooTensor* out);
 
 template <typename T, typename Context>
-SparseCooTensor Coalesce(const Context& dev_ctx, const SparseCooTensor& x) {
+SparseCooTensor CoalesceCoo(const Context& dev_ctx, const SparseCooTensor& x) {
   SparseCooTensor coo;
-  CoalesceKernel<T, Context>(dev_ctx, x, &coo);
+  CoalesceCooKernel<T, Context>(dev_ctx, x, &coo);
   return coo;
 }
 

paddle/phi/kernels/sparse/cpu/coalesce_kernel.cc

@@ -22,9 +22,9 @@ namespace phi {
 namespace sparse {
 
 template <typename T, typename IntT>
-void CoalesceCPUKernel(const CPUContext& dev_ctx,
-                       const SparseCooTensor& x,
-                       SparseCooTensor* out) {
+void CoalesceCooCPUKernel(const CPUContext& dev_ctx,
+                          const SparseCooTensor& x,
+                          SparseCooTensor* out) {
   const DenseTensor& x_indices = x.indices();
   const DenseTensor& x_values = x.values();
   DenseTensor out_indices = phi::EmptyLike<IntT>(dev_ctx, x_indices);
@@ -95,21 +95,22 @@ void CoalesceCPUKernel(const CPUContext& dev_ctx,
 }
 
 template <typename T, typename Context>
-void CoalesceKernel(const Context& dev_ctx,
-                    const SparseCooTensor& x,
-                    SparseCooTensor* out) {
-  PD_VISIT_BASE_INTEGRAL_TYPES(x.indices().dtype(), "CoalesceCPUKernel", ([&] {
-                                 CoalesceCPUKernel<T, data_t>(dev_ctx, x, out);
-                               }));
+void CoalesceCooKernel(const Context& dev_ctx,
+                       const SparseCooTensor& x,
+                       SparseCooTensor* out) {
+  PD_VISIT_BASE_INTEGRAL_TYPES(
+      x.indices().dtype(), "CoalesceCooCPUKernel", ([&] {
+        CoalesceCooCPUKernel<T, data_t>(dev_ctx, x, out);
+      }));
 }
 
 }  // namespace sparse
 }  // namespace phi
 
-PD_REGISTER_KERNEL(coalesce,
+PD_REGISTER_KERNEL(coalesce_coo,
                    CPU,
                    ALL_LAYOUT,
-                   phi::sparse::CoalesceKernel,
+                   phi::sparse::CoalesceCooKernel,
                    float,
                    double,
                    phi::dtype::float16,

paddle/phi/kernels/sparse/cpu/full_kernel.cc

@@ -31,7 +31,7 @@ void FullValue(const Context& dev_ctx, DenseTensor* tensor, T val) {
 }
 
 template <typename T, typename Context>
-void CooFullLikeKernel(const Context& dev_ctx,
+void FullLikeCooKernel(const Context& dev_ctx,
                        const SparseCooTensor& x,
                        const Scalar& val,
                        DataType dtype,
@@ -51,7 +51,7 @@ void CooFullLikeKernel(const Context& dev_ctx,
 }
 
 template <typename T, typename Context>
-void CsrFullLikeKernel(const Context& dev_ctx,
+void FullLikeCsrKernel(const Context& dev_ctx,
                        const SparseCsrTensor& x,
                        const Scalar& val,
                        DataType dtype,
@@ -78,10 +78,10 @@ void CsrFullLikeKernel(const Context& dev_ctx,
 
 }  // namespace phi
 
-PD_REGISTER_KERNEL(coo_full_like,
+PD_REGISTER_KERNEL(full_like_coo,
                    CPU,
                    ALL_LAYOUT,
-                   phi::CooFullLikeKernel,
+                   phi::FullLikeCooKernel,
                    float,
                    double,
                    uint8_t,
@@ -96,10 +96,10 @@ PD_REGISTER_KERNEL(coo_full_like,
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_COO);
 }
 
-PD_REGISTER_KERNEL(csr_full_like,
+PD_REGISTER_KERNEL(full_like_csr,
                    CPU,
                    ALL_LAYOUT,
-                   phi::CsrFullLikeKernel,
+                   phi::FullLikeCsrKernel,
                    float,
                    double,
                    uint8_t,

paddle/phi/kernels/sparse/cpu/unary_kernel.cc

@@ -25,7 +25,7 @@ namespace phi {
 namespace sparse {
 
 template <typename T, typename Context>
-void DivCooScalarKernel(const Context& dev_ctx,
+void DivScalarCooKernel(const Context& dev_ctx,
                         const SparseCooTensor& x,
                         float scalar,
                         SparseCooTensor* out) {
@@ -41,7 +41,7 @@ void DivCooScalarKernel(const Context& dev_ctx,
 }
 
 template <typename T, typename Context>
-void DivCsrScalarKernel(const Context& dev_ctx,
+void DivScalarCsrKernel(const Context& dev_ctx,
                         const SparseCsrTensor& x,
                         float scalar,
                         SparseCsrTensor* out) {
@@ -97,19 +97,19 @@ PD_REGISTER_SPARSE_UNARY_CPU_KERNEL(expm1, Expm1)
 PD_REGISTER_SPARSE_UNARY_CPU_KERNEL(relu6, Relu6)
 PD_REGISTER_SPARSE_UNARY_CPU_KERNEL(leaky_relu, LeakyRelu)
 
-PD_REGISTER_KERNEL(divide_coo_scalar,
+PD_REGISTER_KERNEL(divide_scalar_coo,
                    CPU,
                    ALL_LAYOUT,
-                   phi::sparse::DivCooScalarKernel,
+                   phi::sparse::DivScalarCooKernel,
                    float,
                    double) {
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_COO);
 }
 
-PD_REGISTER_KERNEL(divide_csr_scalar,
+PD_REGISTER_KERNEL(divide_scalar_csr,
                    CPU,
                    ALL_LAYOUT,
-                   phi::sparse::DivCsrScalarKernel,
+                   phi::sparse::DivScalarCsrKernel,
                    float,
                    double) {
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_CSR);

paddle/phi/kernels/sparse/full_kernel.h

@@ -22,14 +22,14 @@
 namespace phi {
 
 template <typename T, typename Context>
-void CooFullLikeKernel(const Context& dev_ctx,
+void FullLikeCooKernel(const Context& dev_ctx,
                        const SparseCooTensor& x,
                        const Scalar& val,
                        DataType dtype,
                        SparseCooTensor* out);
 
 template <typename T, typename Context>
-void CsrFullLikeKernel(const Context& dev_ctx,
+void FullLikeCsrKernel(const Context& dev_ctx,
                        const SparseCsrTensor& x,
                        const Scalar& val,
                        DataType dtype,

paddle/phi/kernels/sparse/gpu/coalesce_kernel.cu

@@ -27,9 +27,9 @@ namespace phi {
 namespace sparse {
 
 template <typename T, typename IntT>
-void CoalesceGPUKernel(const GPUContext& dev_ctx,
-                       const SparseCooTensor& x,
-                       SparseCooTensor* out) {
+void CoalesceCooGPUKernel(const GPUContext& dev_ctx,
+                          const SparseCooTensor& x,
+                          SparseCooTensor* out) {
   const DenseTensor& x_indices = x.indices();
   const DenseTensor& x_values = x.values();
   DenseTensor out_indices = phi::EmptyLike<IntT>(dev_ctx, x_indices);
@@ -172,20 +172,21 @@ void CoalesceGPUKernel(const GPUContext& dev_ctx,
 }
 
 template <typename T, typename Context>
-void CoalesceKernel(const Context& dev_ctx,
-                    const SparseCooTensor& x,
-                    SparseCooTensor* out) {
-  PD_VISIT_BASE_INTEGRAL_TYPES(x.indices().dtype(), "CoalesceGPUKernel", ([&] {
-                                 CoalesceGPUKernel<T, data_t>(dev_ctx, x, out);
-                               }));
+void CoalesceCooKernel(const Context& dev_ctx,
+                       const SparseCooTensor& x,
+                       SparseCooTensor* out) {
+  PD_VISIT_BASE_INTEGRAL_TYPES(
+      x.indices().dtype(), "CoalesceCooGPUKernel", ([&] {
+        CoalesceCooGPUKernel<T, data_t>(dev_ctx, x, out);
+      }));
 }
 }  // namespace sparse
 }  // namespace phi
 
-PD_REGISTER_KERNEL(coalesce,
+PD_REGISTER_KERNEL(coalesce_coo,
                    GPU,
                    ALL_LAYOUT,
-                   phi::sparse::CoalesceKernel,
+                   phi::sparse::CoalesceCooKernel,
                    float,
                    double,
                    phi::dtype::float16,

paddle/phi/kernels/sparse/gpu/full_kernel.cu

@@ -37,7 +37,7 @@ struct FullFunctor {
 };
 
 template <typename T, typename Context>
-void CooFullLikeKernel(const Context& dev_ctx,
+void FullLikeCooKernel(const Context& dev_ctx,
                        const SparseCooTensor& x,
                        const Scalar& val,
                        DataType dtype,
@@ -60,7 +60,7 @@ void CooFullLikeKernel(const Context& dev_ctx,
 }
 
 template <typename T, typename Context>
-void CsrFullLikeKernel(const Context& dev_ctx,
+void FullLikeCsrKernel(const Context& dev_ctx,
                        const SparseCsrTensor& x,
                        const Scalar& val,
                        DataType dtype,
@@ -87,10 +87,10 @@ void CsrFullLikeKernel(const Context& dev_ctx,
 
 }  // namespace phi
 
-PD_REGISTER_KERNEL(coo_full_like,
+PD_REGISTER_KERNEL(full_like_coo,
                    GPU,
                    ALL_LAYOUT,
-                   phi::CooFullLikeKernel,
+                   phi::FullLikeCooKernel,
                    float,
                    double,
                    uint8_t,
@@ -105,10 +105,10 @@ PD_REGISTER_KERNEL(coo_full_like,
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_COO);
 }
 
-PD_REGISTER_KERNEL(csr_full_like,
+PD_REGISTER_KERNEL(full_like_csr,
                    GPU,
                    ALL_LAYOUT,
-                   phi::CsrFullLikeKernel,
+                   phi::FullLikeCsrKernel,
                    float,
                    double,
                    uint8_t,

paddle/phi/kernels/sparse/gpu/unary_kernel.cu

@@ -34,7 +34,7 @@ struct DivScalarFunctor {
 };
 
 template <typename T, typename Context>
-void DivCooScalarKernel(const Context& dev_ctx,
+void DivScalarCooKernel(const Context& dev_ctx,
                         const SparseCooTensor& x,
                         float scalar,
                         SparseCooTensor* out) {
@@ -47,7 +47,7 @@ void DivCooScalarKernel(const Context& dev_ctx,
 }
 
 template <typename T, typename Context>
-void DivCsrScalarKernel(const Context& dev_ctx,
+void DivScalarCsrKernel(const Context& dev_ctx,
                         const SparseCsrTensor& x,
                         float scalar,
                         SparseCsrTensor* out) {
@@ -102,19 +102,19 @@ PD_REGISTER_SPARSE_UNARY_GPU_KERNEL(expm1, Expm1)
 PD_REGISTER_SPARSE_UNARY_GPU_KERNEL(relu6, Relu6)
 PD_REGISTER_SPARSE_UNARY_GPU_KERNEL(leaky_relu, LeakyRelu)
 
-PD_REGISTER_KERNEL(divide_coo_scalar,
+PD_REGISTER_KERNEL(divide_scalar_coo,
                    GPU,
                    ALL_LAYOUT,
-                   phi::sparse::DivCooScalarKernel,
+                   phi::sparse::DivScalarCooKernel,
                    float,
                    double) {
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_COO);
 }
 
-PD_REGISTER_KERNEL(divide_csr_scalar,
+PD_REGISTER_KERNEL(divide_scalar_csr,
                    GPU,
                    ALL_LAYOUT,
-                   phi::sparse::DivCsrScalarKernel,
+                   phi::sparse::DivScalarCsrKernel,
                    float,
                    double) {
   kernel->InputAt(0).SetDataLayout(phi::DataLayout::SPARSE_CSR);

paddle/phi/kernels/sparse/unary_kernel.h

@@ -74,13 +74,13 @@ void ScaleCsrKernel(const Context& dev_ctx,
                     SparseCsrTensor* out);
 
 template <typename T, typename Context>
-void DivCooScalarKernel(const Context& dev_ctx,
+void DivScalarCooKernel(const Context& dev_ctx,
                         const SparseCooTensor& x,
                         float scalar,
                         SparseCooTensor* out);
 
 template <typename T, typename Context>
-void DivCsrScalarKernel(const Context& dev_ctx,
+void DivScalarCsrKernel(const Context& dev_ctx,
                         const SparseCsrTensor& x,
                         float scalar,
                         SparseCsrTensor* out);

paddle/phi/tests/kernels/test_sparse_conv3d_dev_api.cc

@@ -212,7 +212,7 @@ void TestConv3dBase(const std::vector<IntT>& indices,
                                                "Conv3d",
                                                &d_rulebook,
                                                &d_counter);
-  SparseCooTensor tmp_d_out = sparse::Coalesce<T>(dev_ctx_gpu, d_out);
+  SparseCooTensor tmp_d_out = sparse::CoalesceCoo<T>(dev_ctx_gpu, d_out);
 
   ASSERT_EQ(correct_out_dims.size(), d_out.dims().size());
   ASSERT_EQ((int64_t)correct_out_features.size() / out_channels, d_out.nnz());

paddle/phi/tests/kernels/test_sparse_pool_dev_api.cc

@@ -161,7 +161,7 @@ void TestMaxPoolBase(const std::vector<IntT>& indices,
                                                 &d_rulebook,
                                                 &d_counter);
 
-  SparseCooTensor tmp_d_out = sparse::Coalesce<T>(dev_ctx_gpu, d_out);
+  SparseCooTensor tmp_d_out = sparse::CoalesceCoo<T>(dev_ctx_gpu, d_out);
 
   ASSERT_EQ(correct_out_dims.size(), d_out.dims().size());
   ASSERT_EQ((int64_t)correct_out_features.size() / out_channels, d_out.nnz());

python/paddle/incubate/sparse/nn/layer/conv.py

@@ -144,7 +144,7 @@ class Conv3D(_Conv3D):
     Parameters:
         in_channels(int): The number of input channels in the input image.
         out_channels(int): The number of output channels produced by the convolution.
-        kernel_size(int|list|tuple, optional): The size of the convolving kernel.
+        kernel_size(int|list|tuple): The size of the convolving kernel.
         stride(int|list|tuple, optional): The stride size. If stride is a list/tuple, it must
             contain three integers, (stride_D, stride_H, stride_W). Otherwise, the
             stride_D = stride_H = stride_W = stride. The default value is 1.
@@ -277,7 +277,7 @@ class SubmConv3D(_Conv3D):
     Parameters:
         in_channels(int): The number of input channels in the input image.
         out_channels(int): The number of output channels produced by the convolution.
-        kernel_size(int|list|tuple, optional): The size of the convolving kernel.
+        kernel_size(int|list|tuple): The size of the convolving kernel.
         stride(int|list|tuple, optional): The stride size. If stride is a list/tuple, it must
             contain three integers, (stride_D, stride_H, stride_W). Otherwise, the
             stride_D = stride_H = stride_W = stride. The default value is 1.