Merge develop

CMakeLists.txt

@@ -24,6 +24,8 @@ message(STATUS "CXX compiler: ${CMAKE_CXX_COMPILER}, version: "
         "${CMAKE_CXX_COMPILER_ID} ${CMAKE_CXX_COMPILER_VERSION}")
 message(STATUS "C compiler: ${CMAKE_C_COMPILER}, version: "
         "${CMAKE_C_COMPILER_ID} ${CMAKE_C_COMPILER_VERSION}")
+message(STATUS "AR tools: ${CMAKE_AR}")
+
 if(WIN32)
     set(CMAKE_SUPPRESS_REGENERATION ON)
     set(CMAKE_STATIC_LIBRARY_PREFIX lib)

cmake/operators.cmake

@@ -110,7 +110,7 @@ function(op_library TARGET)
     # Define operators that don't need pybind here.
     foreach(manual_pybind_op "compare_op" "logical_op" "nccl_op"
 "tensor_array_read_write_op" "tensorrt_engine_op" "conv_fusion_op"
-"fusion_transpose_flatten_concat_op" "fusion_conv_inception_op")
+"fusion_transpose_flatten_concat_op" "fusion_conv_inception_op" "sync_batch_norm_op")
         if ("${TARGET}" STREQUAL "${manual_pybind_op}")
             set(pybind_flag 1)
         endif()

paddle/fluid/API.spec

@@ -77,7 +77,7 @@ paddle.fluid.initializer.MSRAInitializer.__init__ (ArgSpec(args=['self', 'unifor
 paddle.fluid.initializer.force_init_on_cpu (ArgSpec(args=[], varargs=None, keywords=None, defaults=None), ('document', '6d0f3e22c90d9d500d36ff57daf056ee'))
 paddle.fluid.initializer.init_on_cpu (ArgSpec(args=[], varargs=None, keywords=None, defaults=None), ('document', 'a6d7011ca3d8c0d454dac3a56eae0c29'))
 paddle.fluid.initializer.NumpyArrayInitializer.__init__ (ArgSpec(args=['self', 'value'], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
-paddle.fluid.layers.fc (ArgSpec(args=['input', 'size', 'num_flatten_dims', 'param_attr', 'bias_attr', 'act', 'is_test', 'name'], varargs=None, keywords=None, defaults=(1, None, None, None, False, None)), ('document', '1929058262994f212620599c63aea6bd'))
+paddle.fluid.layers.fc (ArgSpec(args=['input', 'size', 'num_flatten_dims', 'param_attr', 'bias_attr', 'act', 'is_test', 'name'], varargs=None, keywords=None, defaults=(1, None, None, None, False, None)), ('document', '424e898365195e3ccbc2e7dc8b63605e'))
 paddle.fluid.layers.embedding (ArgSpec(args=['input', 'size', 'is_sparse', 'is_distributed', 'padding_idx', 'param_attr', 'dtype'], varargs=None, keywords=None, defaults=(False, False, None, None, 'float32')), ('document', '89c2c55a0b0656b106064048e068e77a'))
 paddle.fluid.layers.dynamic_lstm (ArgSpec(args=['input', 'size', 'h_0', 'c_0', 'param_attr', 'bias_attr', 'use_peepholes', 'is_reverse', 'gate_activation', 'cell_activation', 'candidate_activation', 'dtype', 'name'], varargs=None, keywords=None, defaults=(None, None, None, None, True, False, 'sigmoid', 'tanh', 'tanh', 'float32', None)), ('document', 'dfbb624f85015df29e994ca6999e8ff6'))
 paddle.fluid.layers.dynamic_lstmp (ArgSpec(args=['input', 'size', 'proj_size', 'param_attr', 'bias_attr', 'use_peepholes', 'is_reverse', 'gate_activation', 'cell_activation', 'candidate_activation', 'proj_activation', 'dtype', 'name', 'h_0', 'c_0', 'cell_clip', 'proj_clip'], varargs=None, keywords=None, defaults=(None, None, True, False, 'sigmoid', 'tanh', 'tanh', 'tanh', 'float32', None, None, None, None, None)), ('document', 'b4b608b986eb9617aa0525e1be21d32d'))
@@ -100,7 +100,7 @@ paddle.fluid.layers.pool2d (ArgSpec(args=['input', 'pool_size', 'pool_type', 'po
 paddle.fluid.layers.pool3d (ArgSpec(args=['input', 'pool_size', 'pool_type', 'pool_stride', 'pool_padding', 'global_pooling', 'use_cudnn', 'ceil_mode', 'name', 'exclusive'], varargs=None, keywords=None, defaults=(-1, 'max', 1, 0, False, True, False, None, True)), ('document', '043de7333b79ee0ac55053c14ed81625'))
 paddle.fluid.layers.adaptive_pool2d (ArgSpec(args=['input', 'pool_size', 'pool_type', 'require_index', 'name'], varargs=None, keywords=None, defaults=('max', False, None)), ('document', '859b887174d06f361658f69cb7c06d95'))
 paddle.fluid.layers.adaptive_pool3d (ArgSpec(args=['input', 'pool_size', 'pool_type', 'require_index', 'name'], varargs=None, keywords=None, defaults=('max', False, None)), ('document', '120f4323a3d7ed9c0916f15a59f0e497'))
-paddle.fluid.layers.batch_norm (ArgSpec(args=['input', 'act', 'is_test', 'momentum', 'epsilon', 'param_attr', 'bias_attr', 'data_layout', 'in_place', 'name', 'moving_mean_name', 'moving_variance_name', 'do_model_average_for_mean_and_var', 'fuse_with_relu', 'use_global_stats'], varargs=None, keywords=None, defaults=(None, False, 0.9, 1e-05, None, None, 'NCHW', False, None, None, None, False, False, False)), ('document', 'c527b71b8a4c60dca8df8a745c2b598d'))
+paddle.fluid.layers.batch_norm (ArgSpec(args=['input', 'act', 'is_test', 'momentum', 'epsilon', 'param_attr', 'bias_attr', 'data_layout', 'in_place', 'name', 'moving_mean_name', 'moving_variance_name', 'do_model_average_for_mean_and_var', 'fuse_with_relu', 'use_global_stats'], varargs=None, keywords=None, defaults=(None, False, 0.9, 1e-05, None, None, 'NCHW', False, None, None, None, False, False, False)), ('document', '320c6973b02ea179fa89fecc80796464'))
 paddle.fluid.layers.data_norm (ArgSpec(args=['input', 'act', 'epsilon', 'param_attr', 'data_layout', 'in_place', 'name', 'moving_mean_name', 'moving_variance_name', 'do_model_average_for_mean_and_var'], varargs=None, keywords=None, defaults=(None, 1e-05, None, 'NCHW', False, None, None, None, False)), ('document', 'e45e09e65a2658e07cad987222f0d9ab'))
 paddle.fluid.layers.beam_search_decode (ArgSpec(args=['ids', 'scores', 'beam_size', 'end_id', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', 'b0b8d53821716cd50c42e09b593f3feb'))
 paddle.fluid.layers.conv2d_transpose (ArgSpec(args=['input', 'num_filters', 'output_size', 'filter_size', 'padding', 'stride', 'dilation', 'groups', 'param_attr', 'bias_attr', 'use_cudnn', 'act', 'name'], varargs=None, keywords=None, defaults=(None, None, 0, 1, 1, None, None, None, True, None, None)), ('document', '03993955ab1e6d3044c44e6f17fc85e9'))
@@ -339,7 +339,8 @@ paddle.fluid.layers.generate_mask_labels (ArgSpec(args=['im_info', 'gt_classes',
 paddle.fluid.layers.iou_similarity (ArgSpec(args=['x', 'y', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '587845f60c5d97ffdf2dfd21da52eca1'))
 paddle.fluid.layers.box_coder (ArgSpec(args=['prior_box', 'prior_box_var', 'target_box', 'code_type', 'box_normalized', 'name', 'axis'], varargs=None, keywords=None, defaults=('encode_center_size', True, None, 0)), ('document', '032d0f4b7d8f6235ee5d91e473344f0e'))
 paddle.fluid.layers.polygon_box_transform (ArgSpec(args=['input', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '0e5ac2507723a0b5adec473f9556799b'))
-paddle.fluid.layers.yolov3_loss (ArgSpec(args=['x', 'gtbox', 'gtlabel', 'anchors', 'anchor_mask', 'class_num', 'ignore_thresh', 'downsample_ratio', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '991e934c3e09abf0edec7c9c978b4691'))
+paddle.fluid.layers.yolov3_loss (ArgSpec(args=['x', 'gtbox', 'gtlabel', 'anchors', 'anchor_mask', 'class_num', 'ignore_thresh', 'downsample_ratio', 'gtscore', 'use_label_smooth', 'name'], varargs=None, keywords=None, defaults=(None, True, None)), ('document', '57fa96922e42db8f064c3fb77f2255e8'))
+paddle.fluid.layers.yolo_box (ArgSpec(args=['x', 'img_size', 'anchors', 'class_num', 'conf_thresh', 'downsample_ratio', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '5566169a5ab993d177792c023c7fb340'))
 paddle.fluid.layers.box_clip (ArgSpec(args=['input', 'im_info', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '397e9e02b451d99c56e20f268fa03f2e'))
 paddle.fluid.layers.multiclass_nms (ArgSpec(args=['bboxes', 'scores', 'score_threshold', 'nms_top_k', 'keep_top_k', 'nms_threshold', 'normalized', 'nms_eta', 'background_label', 'name'], varargs=None, keywords=None, defaults=(0.3, True, 1.0, 0, None)), ('document', 'ca7d1107b6c5d2d6d8221039a220fde0'))
 paddle.fluid.layers.distribute_fpn_proposals (ArgSpec(args=['fpn_rois', 'min_level', 'max_level', 'refer_level', 'refer_scale', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '7bb011ec26bace2bc23235aa4a17647d'))
@@ -376,7 +377,7 @@ paddle.fluid.contrib.BeamSearchDecoder.read_array (ArgSpec(args=['self', 'init',
 paddle.fluid.contrib.BeamSearchDecoder.update_array (ArgSpec(args=['self', 'array', 'value'], varargs=None, keywords=None, defaults=None), ('document', '5754e9b3212b7c09497151516a0de5a7'))
 paddle.fluid.contrib.memory_usage (ArgSpec(args=['program', 'batch_size'], varargs=None, keywords=None, defaults=None), ('document', '8fcb2f93bb743693baa8d4860a5ccc47'))
 paddle.fluid.contrib.op_freq_statistic (ArgSpec(args=['program'], varargs=None, keywords=None, defaults=None), ('document', '4d43687113c4bf5b29d15aee2f4e4afa'))
-paddle.fluid.contrib.QuantizeTranspiler.__init__ (ArgSpec(args=['self', 'weight_bits', 'activation_bits', 'activation_quantize_type', 'weight_quantize_type', 'window_size'], varargs=None, keywords=None, defaults=(8, 8, 'abs_max', 'abs_max', 10000)), ('document', '14b39f1fcd5667ff556b1aad94357d1d'))
+paddle.fluid.contrib.QuantizeTranspiler.__init__ (ArgSpec(args=['self', 'weight_bits', 'activation_bits', 'activation_quantize_type', 'weight_quantize_type', 'window_size', 'moving_rate'], varargs=None, keywords=None, defaults=(8, 8, 'abs_max', 'abs_max', 10000, 0.9)), ('document', '14b39f1fcd5667ff556b1aad94357d1d'))
 paddle.fluid.contrib.QuantizeTranspiler.convert_to_int8 (ArgSpec(args=['self', 'program', 'place', 'scope'], varargs=None, keywords=None, defaults=(None,)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.contrib.QuantizeTranspiler.freeze_program (ArgSpec(args=['self', 'program', 'place', 'fuse_bn', 'scope'], varargs=None, keywords=None, defaults=(False, None)), ('document', '909675a1ab055c69b436a7893fcae4fd'))
 paddle.fluid.contrib.QuantizeTranspiler.training_transpile (ArgSpec(args=['self', 'program', 'startup_program'], varargs=None, keywords=None, defaults=(None, None)), ('document', '6dd9909f10b283ba2892a99058a72884'))
@@ -401,9 +402,9 @@ paddle.fluid.contrib.MagnitudePruner.__init__ (ArgSpec(args=['self', 'threshold'
 paddle.fluid.contrib.MagnitudePruner.prune (ArgSpec(args=['self', 'param', 'threshold'], varargs=None, keywords=None, defaults=(None,)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.contrib.RatioPruner.__init__ (ArgSpec(args=['self', 'ratios'], varargs=None, keywords=None, defaults=(None,)), ('document', 'e7a81a325b296a9ca502ee5adb4fc85d'))
 paddle.fluid.contrib.RatioPruner.prune (ArgSpec(args=['self', 'param', 'ratio'], varargs=None, keywords=None, defaults=(None,)), ('document', '358cbf2978c91028fb96a195a9884645'))
-paddle.fluid.contrib.load_persistables_for_increment (ArgSpec(args=['dirname', 'executor', 'program', 'lookup_table_var', 'lookup_table_var_path'], varargs=None, keywords=None, defaults=None), ('document', '11fbf7e8dd2289805de291b453a33ee7'))
-paddle.fluid.contrib.load_persistables_for_inference (ArgSpec(args=['dirname', 'executor', 'program', 'lookup_table_var_name'], varargs=None, keywords=None, defaults=None), ('document', '5b5577bb3d24070da819674255d16196'))
-paddle.fluid.contrib.convert_dist_to_sparse_program (ArgSpec(args=['program'], varargs=None, keywords=None, defaults=None), ('document', '4efbd93876832d4d35497cdbc7a1e6d8'))
+paddle.fluid.contrib.load_persistables_for_increment (ArgSpec(args=['dirname', 'executor', 'program', 'lookup_table_var', 'lookup_table_var_path'], varargs=None, keywords=None, defaults=None), ('document', '2ab36d4f7a564f5f65e455807ad06c67'))
+paddle.fluid.contrib.load_persistables_for_inference (ArgSpec(args=['dirname', 'executor', 'program', 'lookup_table_var_name'], varargs=None, keywords=None, defaults=None), ('document', '59066bac9db0ac6ce414d05780b7333f'))
+paddle.fluid.contrib.convert_dist_to_sparse_program (ArgSpec(args=['program'], varargs=None, keywords=None, defaults=None), ('document', '74c39c595dc70d6be2f16d8e462d282b'))
 paddle.fluid.contrib.HDFSClient.__init__ (ArgSpec(args=['self', 'hadoop_home', 'configs'], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.contrib.HDFSClient.delete (ArgSpec(args=['self', 'hdfs_path'], varargs=None, keywords=None, defaults=None), ('document', 'c3721aa2d4d9ef5a857dd47b2681c03e'))
 paddle.fluid.contrib.HDFSClient.download (ArgSpec(args=['self', 'hdfs_path', 'local_path', 'overwrite', 'unzip'], varargs=None, keywords=None, defaults=(False, False)), ('document', 'ca55bde92184d3fd0f9f5c963b25e634'))

paddle/fluid/framework/details/build_strategy.cc

@@ -16,6 +16,7 @@ limitations under the License. */
 
 #include <glog/logging.h>
 #include <memory>
+#include <utility>
 
 #include "paddle/fluid/framework/details/memory_optimize_helper.h"
 #include "paddle/fluid/framework/details/multi_devices_graph_pass.h"
@@ -49,6 +50,11 @@ class ParallelExecutorPassBuilder : public ir::PassBuilder {
       AppendPass("sequential_execution_pass");
     }
 
+    // Add op fusion.
+    if (strategy.sync_batch_norm_) {
+      AppendPass("sync_batch_norm_pass");
+    }
+
     // Add op fusion.
     if (strategy.fuse_relu_depthwise_conv_) {
       AppendPass("fuse_relu_depthwise_conv_pass");
@@ -227,6 +233,7 @@ std::unique_ptr<ir::Graph> BuildStrategy::Apply(
 }  // namespace framework
 }  // namespace paddle
 
+USE_PASS(sync_batch_norm_pass);
 USE_PASS(fuse_relu_depthwise_conv_pass);
 USE_PASS(fuse_elewise_add_act_pass);
 USE_PASS(graph_viz_pass);

paddle/fluid/framework/details/build_strategy.h

@@ -77,6 +77,8 @@ struct BuildStrategy {
 
   bool fuse_relu_depthwise_conv_{false};
 
+  bool sync_batch_norm_{false};
+
   bool memory_optimize_{true};
   // TODO(dzhwinter):
   // make enable_inplace, memory_optimize_

paddle/fluid/framework/details/memory_optimize_helper.cc

@@ -337,7 +337,6 @@ bool NodeCanReused(const VarDesc& node) {
   auto type = node.GetType();
   // only these types holds bulk of gpu memory
   if (!(type == proto::VarType::LOD_TENSOR ||
-        type == proto::VarType::SELECTED_ROWS ||
         type == proto::VarType::LOD_TENSOR_ARRAY)) {
     return false;
   }

paddle/fluid/framework/grad_op_desc_maker.h

@@ -14,7 +14,9 @@ limitations under the License. */
 
 #pragma once
 #include <algorithm>
+#include <memory>
 #include <string>
+#include <unordered_map>
 #include <unordered_set>
 #include <vector>
 #include "paddle/fluid/framework/op_desc.h"
@@ -55,11 +57,11 @@ class GradOpDescMakerBase {
                    std::back_inserter(ret_val),
                    [this](const std::string& fwd_var_name) -> std::string {
                      auto g_name = GradVarName(fwd_var_name);
-                     if (no_grad_set_.count(g_name)) {
-                       return kEmptyVarName;
-                     } else {
+                     if (no_grad_set_.empty() || !no_grad_set_.count(g_name)) {
                        (*this->grad_to_var_)[g_name] = fwd_var_name;
                        return g_name;
+                     } else {
+                       return kEmptyVarName;
                      }
                    });
     if (!drop_empty_grad) {

paddle/fluid/framework/ir/CMakeLists.txt

@@ -46,6 +46,8 @@ cc_library(fuse_pass_base SRCS fuse_pass_base.cc DEPS pass)
 pass_library(graph_to_program_pass base)
 pass_library(graph_viz_pass base)
 pass_library(lock_free_optimize_pass base)
+pass_library(cpu_quantize_pass inference)
+pass_library(cpu_quantize_squash_pass inference)
 pass_library(fc_fuse_pass inference)
 pass_library(attention_lstm_fuse_pass inference)
 pass_library(infer_clean_graph_pass inference)
@@ -66,6 +68,7 @@ pass_library(conv_elementwise_add_fuse_pass inference)
 pass_library(conv_affine_channel_fuse_pass inference)
 pass_library(transpose_flatten_concat_fuse_pass inference)
 pass_library(identity_scale_op_clean_pass base)
+pass_library(sync_batch_norm_pass base)
 
 # There may be many transpose-flatten structures in a model, and the output of
 # these structures will be used as inputs to the concat Op. This pattern will
@@ -100,6 +103,11 @@ cc_test(test_graph_pattern_detector SRCS graph_pattern_detector_tester.cc DEPS g
 cc_test(test_fc_fuse_pass SRCS fc_fuse_pass_tester.cc DEPS fc_fuse_pass framework_proto)
 cc_test(test_seqpool_concat_fuse_pass SRCS seqpool_concat_fuse_pass_tester.cc DEPS seqpool_concat_fuse_pass framework_proto)
 cc_test(test_is_test_pass SRCS is_test_pass_tester.cc DEPS is_test_pass)
+cc_test(test_cpu_quantize_pass SRCS cpu_quantize_pass_tester.cc DEPS cpu_quantize_pass naive_executor)
+cc_test(test_cpu_quantize_squash_pass SRCS cpu_quantize_squash_pass_tester.cc DEPS cpu_quantize_squash_pass naive_executor)
+if(NOT WIN32)
+    cc_test(test_sync_batch_norm_pass SRCS sync_batch_norm_pass_tester.cc DEPS sync_batch_norm_pass)
+endif()
 if (WITH_MKLDNN)
     cc_test(test_depthwise_conv_mkldnn_pass SRCS mkldnn/depthwise_conv_mkldnn_pass_tester.cc DEPS depthwise_conv_mkldnn_pass)
     cc_test(test_conv_bias_mkldnn_fuse_pass SRCS mkldnn/conv_bias_mkldnn_fuse_pass_tester.cc DEPS conv_bias_mkldnn_fuse_pass naive_executor)

paddle/fluid/framework/ir/cpu_quantize_pass.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 "paddle/fluid/framework/ir/cpu_quantize_pass.h"
+#include <utility>
+#include <vector>
+#include "paddle/fluid/framework/eigen.h"
+#include "paddle/fluid/string/pretty_log.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+namespace {
+
+void UnlinkNodes(ir::Node* a, ir::Node* b) {
+  a->outputs.erase(std::remove(a->outputs.begin(), a->outputs.end(), b),
+                   a->outputs.end());
+  b->inputs.erase(std::remove(b->inputs.begin(), b->inputs.end(), a),
+                  b->inputs.end());
+}
+
+}  // namespace
+
+enum { U8_MAX = 255, S8_MAX = 127 };
+
+using EigenVectorArrayMap = Eigen::Map<Eigen::Array<double, Eigen::Dynamic, 1>>;
+using string::PrettyLogDetail;
+
+void CPUQuantizePass::QuantizeInput(Graph* g, Node* op, Node* input,
+                                    std::string input_name, double scale_to_one,
+                                    bool is_unsigned,
+                                    std::string scale_attr_name) const {
+  unsigned max = is_unsigned ? U8_MAX : S8_MAX;
+  float scale = scale_to_one * max;
+
+  // Create quantize output variable
+  VarDesc quantize_out_desc(patterns::PDNodeName("quantize", "out"));
+  auto* quantize_out_node = g->CreateVarNode(&quantize_out_desc);
+
+  // create a quantize op node
+  OpDesc q_desc;
+  q_desc.SetType("quantize");
+  q_desc.SetInput("Input", std::vector<std::string>({input->Name()}));
+  q_desc.SetOutput("Output",
+                   std::vector<std::string>({quantize_out_node->Name()}));
+  q_desc.SetAttr("Scale", scale);
+  q_desc.SetAttr("is_negative_input", !is_unsigned);
+  auto quantize_op = g->CreateOpNode(&q_desc);  // OpDesc will be copied.
+
+  // update op's input
+  op->Op()->SetInput(input_name,
+                     std::vector<std::string>({quantize_out_node->Name()}));
+
+  // link quantize op
+  UnlinkNodes(input, op);
+  IR_NODE_LINK_TO(input, quantize_op);
+  IR_NODE_LINK_TO(quantize_op, quantize_out_node);
+  IR_NODE_LINK_TO(quantize_out_node, op);
+
+  if (!scale_attr_name.empty()) op->Op()->SetAttr(scale_attr_name, scale);
+}
+
+void CPUQuantizePass::DequantizeOutput(Graph* g, Node* op, Node* output,
+                                       std::string output_name,
+                                       double scale_to_one, bool is_unsigned,
+                                       std::string scale_attr_name) const {
+  unsigned max = is_unsigned ? U8_MAX : S8_MAX;
+  float scale = scale_to_one * max;
+
+  // Create dequantize input variable
+  VarDesc dequantize_in_desc(patterns::PDNodeName("dequantize", "in"));
+  auto* dequantize_in_node = g->CreateVarNode(&dequantize_in_desc);
+
+  // create a dequantize op node for output.
+  OpDesc deq_desc;
+  deq_desc.SetType("dequantize");
+  deq_desc.SetInput("Input",
+                    std::vector<std::string>({dequantize_in_node->Name()}));
+  deq_desc.SetOutput("Output", std::vector<std::string>({output->Name()}));
+  deq_desc.SetAttr("Scale", scale);
+  auto dequantize_op = g->CreateOpNode(&deq_desc);  // OpDesc will be copied.
+
+  // update op's output
+  op->Op()->SetOutput(output_name,
+                      std::vector<std::string>({dequantize_in_node->Name()}));
+
+  // link dequantize op
+  UnlinkNodes(op, output);
+  IR_NODE_LINK_TO(op, dequantize_in_node);
+  IR_NODE_LINK_TO(dequantize_in_node, dequantize_op);
+  IR_NODE_LINK_TO(dequantize_op, output);
+
+  if (!scale_attr_name.empty()) op->Op()->SetAttr(scale_attr_name, scale);
+}
+
+void CPUQuantizePass::QuantizeConv(Graph* graph,
+                                   bool with_residual_data) const {
+  GraphPatternDetector gpd;
+  auto pattern = gpd.mutable_pattern();
+  patterns::ConvResidual conv_pattern{pattern, name_scope_};
+  conv_pattern(with_residual_data);
+
+  int quantize_conv_count = 0;
+  auto handler = [&](const GraphPatternDetector::subgraph_t& subgraph,
+                     Graph* g) {
+    VLOG(4) << "Quantize conv2d op";
+    GET_IR_NODE_FROM_SUBGRAPH(conv_op, conv_op, conv_pattern);
+    auto* conv_op_desc = conv_op->Op();
+
+    // skip if should not be quantized
+    if (!conv_op_desc->HasAttr("use_quantizer") ||
+        !boost::get<bool>(conv_op_desc->GetAttr("use_quantizer")))
+      return;
+
+    GET_IR_NODE_FROM_SUBGRAPH(conv_filter, conv_filter, conv_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(conv_input, conv_input, conv_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(conv_output, conv_output, conv_pattern);
+
+    // get scales calculated after warmup, they scale variables to MAX=1.0
+    auto scales = Get<VarQuantScale>("quant_var_scales");
+
+    auto input_scale = scales[conv_input->Name()].second.data<double>()[0];
+    bool is_input_unsigned = scales[conv_input->Name()].first;
+    QuantizeInput(g, conv_op, conv_input, "Input", input_scale,
+                  is_input_unsigned, "Scale_in");
+
+    auto filter_scale_tensor = scales[conv_filter->Name()].second;
+    EigenVectorArrayMap eigen_tensor{filter_scale_tensor.data<double>(),
+                                     filter_scale_tensor.numel(), 1};
+    eigen_tensor *= static_cast<double>(S8_MAX);
+    std::vector<float> filter_scale{
+        filter_scale_tensor.data<double>(),
+        filter_scale_tensor.data<double>() + filter_scale_tensor.numel()};
+
+    conv_op->Op()->SetAttr("Scale_weights", filter_scale);
+
+    if (with_residual_data) {
+      GET_IR_NODE_FROM_SUBGRAPH(conv_residual_data, conv_residual_data,
+                                conv_pattern);
+      auto residual_scale =
+          scales[conv_residual_data->Name()].second.data<double>()[0];
+      bool is_residual_unsigned = scales[conv_residual_data->Name()].first;
+
+      QuantizeInput(g, conv_op, conv_residual_data, "ResidualData",
+                    residual_scale, is_residual_unsigned, "Scale_in_eltwise");
+    }
+
+    auto output_scale = scales[conv_output->Name()].second.data<double>()[0];
+    bool is_output_unsigned = scales[conv_output->Name()].first;
+    DequantizeOutput(g, conv_op, conv_output, "Output", output_scale,
+                     is_output_unsigned, "Scale_out");
+
+    ++quantize_conv_count;
+  };
+
+  gpd(graph, handler);
+  AddStatis(quantize_conv_count);
+
+  std::stringstream msg_ss;
+  msg_ss << "---    quantized " << quantize_conv_count << " conv2d ops";
+  if (with_residual_data) msg_ss << " with residual connection";
+  PrettyLogDetail(msg_ss.str().c_str());
+}
+
+void CPUQuantizePass::QuantizePool(Graph* graph) const {
+  GraphPatternDetector gpd;
+  auto pattern = gpd.mutable_pattern();
+  patterns::Pool pool_pattern{pattern, name_scope_};
+  pool_pattern();
+
+  int quantize_pool_count = 0;
+  auto handler = [&](const GraphPatternDetector::subgraph_t& subgraph,
+                     Graph* g) {
+    VLOG(4) << "Quantize pool2d op";
+    GET_IR_NODE_FROM_SUBGRAPH(pool_op, pool_op, pool_pattern);
+    auto* pool_op_desc = pool_op->Op();
+
+    // skip if should not be quantized
+    if (!pool_op_desc->HasAttr("use_quantizer") ||
+        !boost::get<bool>(pool_op_desc->GetAttr("use_quantizer")))
+      return;
+
+    GET_IR_NODE_FROM_SUBGRAPH(pool_input, pool_input, pool_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(pool_output, pool_output, pool_pattern);
+
+    // get scales calculated after warmup, they scale variables to MAX=1.0
+    auto scales = Get<VarQuantScale>("quant_var_scales");
+
+    auto input_scale = scales[pool_input->Name()].second.data<double>()[0];
+    bool is_input_unsigned = scales[pool_input->Name()].first;
+    QuantizeInput(g, pool_op, pool_input, "X", input_scale, is_input_unsigned);
+
+    auto output_scale = scales[pool_output->Name()].second.data<double>()[0];
+    bool is_output_unsigned = scales[pool_output->Name()].first;
+    DequantizeOutput(g, pool_op, pool_output, "Out", output_scale,
+                     is_output_unsigned);
+
+    ++quantize_pool_count;
+  };
+
+  gpd(graph, handler);
+  AddStatis(quantize_pool_count);
+
+  PrettyLogDetail("---    quantized %d pool2d ops", quantize_pool_count);
+}
+
+std::unique_ptr<ir::Graph> CPUQuantizePass::ApplyImpl(
+    std::unique_ptr<ir::Graph> graph) const {
+  VLOG(3) << "Quantizing the graph.";
+  PADDLE_ENFORCE(graph.get());
+  FusePassBase::Init(name_scope_, graph.get());
+
+  PADDLE_ENFORCE(param_scope());
+
+  QuantizeConv(graph.get(), true /* with_residual_data */);
+  QuantizeConv(graph.get());
+  QuantizePool(graph.get());
+
+  return graph;
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+REGISTER_PASS(cpu_quantize_pass, paddle::framework::ir::CPUQuantizePass)
+    .RequirePassAttr("quant_var_scales");

paddle/fluid/framework/ir/cpu_quantize_pass.h

+// 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.
+
+#pragma once
+
+#include <memory>
+#include <string>
+#include <unordered_map>
+#include <utility>
+#include "paddle/fluid/framework/ir/fuse_pass_base.h"
+#include "paddle/fluid/framework/ir/graph.h"
+#include "paddle/fluid/framework/ir/graph_pattern_detector.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+/*
+ * Map variable name to tensor of scaling factors scaling it to MAX=1.0.
+ * bool denotes whether quantization of the variable should be done to unsigned
+ * type.
+ */
+using VarQuantScale =
+    std::unordered_map<std::string, std::pair<bool, LoDTensor>>;
+
+/*
+ * Quantize all supported operators.
+ */
+class CPUQuantizePass : public FusePassBase {
+ public:
+  virtual ~CPUQuantizePass() {}
+
+ protected:
+  std::unique_ptr<ir::Graph> ApplyImpl(
+      std::unique_ptr<ir::Graph> graph) const override;
+
+  void QuantizeConv(Graph* graph, bool with_residual_data = false) const;
+
+  void QuantizePool(Graph* graph) const;
+
+  void QuantizeInput(Graph* g, Node* op, Node* input, std::string input_name,
+                     double scale_to_one, bool is_unsigned,
+                     std::string scale_attr_name = "") const;
+
+  void DequantizeOutput(Graph* g, Node* op, Node* output,
+                        std::string output_name, double scale_to_one,
+                        bool is_unsigned,
+                        std::string scale_attr_name = "") const;
+
+  const std::string name_scope_{"quantize"};
+};
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle

paddle/fluid/framework/ir/cpu_quantize_pass_tester.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 "paddle/fluid/framework/ir/cpu_quantize_pass.h"
+#include <gtest/gtest.h>
+#include "paddle/fluid/framework/naive_executor.h"
+#include "paddle/fluid/platform/place.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+void SetOp(ProgramDesc* prog, const std::string& type, const std::string& name,
+           const std::vector<std::string>& inputs,
+           const std::vector<std::string>& outputs, bool use_mkldnn,
+           bool use_quantizer = false) {
+  auto* op = prog->MutableBlock(0)->AppendOp();
+  op->SetType(type);
+  op->SetAttr("use_mkldnn", use_mkldnn);
+  op->SetAttr("name", name);
+  if (type == "conv2d") {
+    op->SetInput("Input", {inputs[0]});
+    op->SetInput("Filter", {inputs[1]});
+    if (inputs.size() > 2)
+      op->SetInput("Bias", {inputs[2]});
+    else
+      op->SetInput("Bias", {});
+    if (inputs.size() > 3) {
+      op->SetInput("ResidualData", {inputs[3]});
+      op->SetAttr("fuse_residual_connection", true);
+    } else {
+      op->SetInput("ResidualData", {});
+      op->SetAttr("fuse_residual_connection", false);
+    }
+    op->SetOutput("Output", {outputs[0]});
+    op->SetAttr("use_quantizer", use_quantizer);
+    op->SetAttr("Scale_in", 1.0f);
+    op->SetAttr("Scale_out", 1.0f);
+    op->SetAttr("Scale_weights", std::vector<float>{1.0f});
+  } else if (type == "pool2d") {
+    op->SetInput("X", {inputs[0]});
+    op->SetOutput("Out", {outputs[0]});
+    op->SetAttr("use_quantizer", use_quantizer);
+  } else if (type == "dropout") {
+    op->SetInput("X", {inputs[0]});
+    op->SetOutput("Out", {outputs[0]});
+  } else if (type == "fc") {
+    op->SetInput("Input", {inputs[0]});
+    if (inputs.size() > 1) op->SetInput("W", {inputs[1]});
+    if (inputs.size() > 2) op->SetInput("Bias", {inputs[2]});
+    op->SetOutput("Out", {outputs[0]});
+  }
+}
+
+static const std::initializer_list<std::string> variable_names{
+    "a", "w1", "c",  "d", "w2", "e",  "f", "g",
+    "h", "w3", "b1", "i", "j",  "w4", "b2"};
+// (a,w1)->Conv1->c and c->Pool1->d
+//
+// (d,w2)->Conv2->e and e->Pool2->f
+//
+// d->Dropout1->g and g->Fc1->h and (h,w3,b1,i)->Conv3->j
+//
+// (d,w4, b2)->Conv4->i
+ProgramDesc BuildProgramDesc(bool use_mkldnn, bool use_quantizer) {
+  ProgramDesc prog;
+  for (auto& v : variable_names) {
+    auto* var = prog.MutableBlock(0)->Var(v);
+    if (v.find("w") == 0 || v.find("b") == 0) {
+      var->SetPersistable(true);
+    }
+  }
+
+  SetOp(&prog, "conv2d", "Conv1", {"a", "w1"}, {"c"}, use_mkldnn,
+        use_quantizer);
+  SetOp(&prog, "pool2d", "Pool1", {"c"}, {"d"}, use_mkldnn, use_quantizer);
+
+  SetOp(&prog, "conv2d", "Conv2", {"d", "w2"}, {"e"}, use_mkldnn,
+        use_quantizer);
+  SetOp(&prog, "pool2d", "Pool2", {"e"}, {"f"}, use_mkldnn, use_quantizer);
+
+  SetOp(&prog, "dropout", "Dropout1", {"d"}, {"g"}, use_mkldnn);
+  SetOp(&prog, "fc", "Fc1", {"g"}, {"h"}, use_mkldnn);
+  SetOp(&prog, "conv2d", "Conv3", {"h", "w3", "b1", "i"}, {"j"}, use_mkldnn,
+        use_quantizer);
+
+  SetOp(&prog, "conv2d", "Conv4", {"c", "w4", "b2"}, {"i"}, use_mkldnn,
+        use_quantizer);
+
+  return prog;
+}
+
+void InitTensorHolder(Scope* scope, const paddle::platform::Place& place,
+                      const char* var_name) {
+  auto x = scope->Var(var_name);
+  auto tensor = x->GetMutable<LoDTensor>();
+  tensor->mutable_data(place, proto::VarType::FP32,
+                       ::paddle::memory::Allocator::kDefault, 1);
+}
+
+void MainTest(const ProgramDesc& prog, int conv_count, int pool_count,
+              int quant_count, int dequant_count, int added_nodes_count,
+              float scale) {
+  std::unique_ptr<ir::Graph> graph(new ir::Graph(prog));
+
+  // Init scope, as it is used in pass
+  auto place = paddle::platform::CPUPlace();
+  NaiveExecutor exe{place};
+  Scope scope;
+  exe.CreateVariables(prog, 0, true, &scope);
+
+  auto* scales = new VarQuantScale();
+
+  for (auto& v : variable_names) {
+    InitTensorHolder(&scope, place, v.c_str());
+    LoDTensor tensor;
+    tensor.Resize({1});
+    auto* ptr = tensor.mutable_data<double>(place);
+    ptr[0] = 2.0;
+
+    (*scales)[v] = std::make_pair(false, std::move(tensor));
+  }
+
+  graph->Set(kParamScopeAttr, new framework::Scope*(&scope));
+
+  auto pass = PassRegistry::Instance().Get("cpu_quantize_pass");
+  pass->Set("quant_var_scales", scales);
+
+  int original_nodes_num = graph->Nodes().size();
+
+  graph = pass->Apply(std::move(graph));
+
+  int current_nodes_num = graph->Nodes().size();
+
+  int quantize_nodes_count = 0;
+  int dequantize_nodes_count = 0;
+  int conv2d_nodes_count = 0;
+  int pool2d_nodes_count = 0;
+  for (auto* node : graph->Nodes()) {
+    if (node->IsOp()) {
+      auto* op = node->Op();
+      if (op->Type() == "conv2d") {
+        conv2d_nodes_count++;
+        auto op_name = boost::get<std::string>(op->GetAttr("name"));
+        EXPECT_EQ(boost::get<float>(op->GetAttr("Scale_in")), scale)
+            << "Scale_in for node '" + op_name + "'.";
+        EXPECT_EQ(boost::get<float>(op->GetAttr("Scale_out")), scale)
+            << "Scale_out for node '" + op_name + "'.";
+        EXPECT_EQ(
+            boost::get<std::vector<float>>(op->GetAttr("Scale_weights"))[0],
+            scale)
+            << "Scale_weights for node '" + op_name + "'.";
+      } else if (op->Type() == "pool2d") {
+        pool2d_nodes_count++;
+      } else if (op->Type() == "quantize") {
+        quantize_nodes_count++;
+      } else if (op->Type() == "dequantize") {
+        dequantize_nodes_count++;
+      }
+    }
+  }
+  EXPECT_EQ(conv2d_nodes_count, conv_count);
+  EXPECT_EQ(pool2d_nodes_count, pool_count);
+  EXPECT_EQ(quantize_nodes_count, quant_count);
+  EXPECT_EQ(dequantize_nodes_count, dequant_count);
+  EXPECT_EQ(original_nodes_num + added_nodes_count, current_nodes_num);
+}
+
+TEST(CpuQuantizePass, quantize) {
+  bool use_mkldnn = true;
+  bool use_quantizer = true;
+  // (a->QUANT1->IN1,w1)->Conv1->OUT1->DEQUANT1->c and
+  // c->QUANT2->IN2->Pool1->OUT2->DEQUANT2->d
+  //
+  // (d->QUANT3->IN3,w2)->Conv2->OUT3->DEQUANT3->e and
+  // e->QUANT4->IN4->Pool2->OUT4->DEQUANT4->f
+  //
+  // d->Dropout1->g and g->Fc1->h and
+  // (h->QUANT5->IN5,w3,b1,i->QUANT6->IN6)->Conv3->OUT5->DEQUANT5->j
+  //
+  // (d->QUANT7->IN7,w4, b2)->Conv4->DEQUANT6->OUT6->i
+  // Insert nodes: 7 Quant + 7 IN + 6 OUT + 6 DEQUANT
+  int added_nodes = 7 + 7 + 6 + 6;
+  MainTest(BuildProgramDesc(use_mkldnn, use_quantizer), 4, 2, 7, 6, added_nodes,
+           2.0f * 127);
+}
+
+TEST(CpuQuantizePass, do_not_quantize) {
+  bool use_mkldnn = true;
+  bool use_quantizer = false;
+  int added_nodes = 0;
+  MainTest(BuildProgramDesc(use_mkldnn, use_quantizer), 4, 2, 0, 0, added_nodes,
+           1.0f);
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+USE_PASS(cpu_quantize_pass);

paddle/fluid/framework/ir/cpu_quantize_squash_pass.cc

+// Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file eint8_outcept 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 eint8_outpress or
+// implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#include "paddle/fluid/framework/ir/cpu_quantize_squash_pass.h"
+#include <string>
+#include <vector>
+#include "paddle/fluid/platform/enforce.h"
+#include "paddle/fluid/string/pretty_log.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+using string::PrettyLogDetail;
+
+void CPUQuantizeSquashPass::FindNodesToKeep(
+    Graph* graph,
+    std::unordered_map<const Node*, int>* nodes_keep_counter) const {
+  GraphPatternDetector gpd;
+  patterns::DequantAny deq_any_pattern{gpd.mutable_pattern(), "deqant_any"};
+  deq_any_pattern();
+
+  int found_count = 0;
+  auto handler = [&](const GraphPatternDetector::subgraph_t& subgraph,
+                     Graph* g) {
+    GET_IR_NODE_FROM_SUBGRAPH(dequant_out, dequant_out, deq_any_pattern);
+
+    if (nodes_keep_counter->find(dequant_out) == nodes_keep_counter->end())
+      (*nodes_keep_counter)[dequant_out] = 1;
+    else
+      (*nodes_keep_counter)[dequant_out] += 1;
+
+    found_count++;
+  };
+  gpd(graph, handler);
+  AddStatis(found_count);
+}
+
+void CPUQuantizeSquashPass::Squash(
+    Graph* graph,
+    std::unordered_map<const Node*, int>* nodes_keep_counter) const {
+  GraphPatternDetector gpd;
+  patterns::DequantQuantAny squash_pattern{gpd.mutable_pattern(), "squash"};
+  squash_pattern();
+
+  int found_squash_count = 0;
+  auto handler = [&](const GraphPatternDetector::subgraph_t& subgraph,
+                     Graph* g) {
+    VLOG(4) << "squash requantize-quantize ops pair";
+
+    GET_IR_NODE_FROM_SUBGRAPH(dequant_in, dequant_in, squash_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(dequant_op, dequant_op, squash_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(dequant_out, dequant_out, squash_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(quant_op, quant_op, squash_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(quant_out, quant_out, squash_pattern);
+    GET_IR_NODE_FROM_SUBGRAPH(next_op, next_op, squash_pattern);
+
+    auto* next_op_desc = next_op->Op();
+    float dequant_scale = boost::get<float>(dequant_op->Op()->GetAttr("Scale"));
+    float quant_scale = boost::get<float>(quant_op->Op()->GetAttr("Scale"));
+    PADDLE_ENFORCE(nodes_keep_counter->find(dequant_out) !=
+                   nodes_keep_counter->end());
+
+    // check if dequantize op should be kept or removed, decrease the counter
+    bool keep_dequant = (*nodes_keep_counter)[dequant_out]-- > 1;
+
+    if (dequant_scale == quant_scale) {
+      // squash dequantize-quantize to nothing
+      auto quant_out_var_name = quant_out->Name();
+      auto next_op_inputs = next_op_desc->InputNames();
+      for (const auto& name : next_op_inputs) {
+        auto var_name = next_op_desc->Input(name)[0];
+        if (var_name.compare(quant_out_var_name) == 0) {
+          next_op_desc->SetInput(
+              name, std::vector<std::string>({dequant_in->Name()}));
+          break;
+        }
+      }
+
+      if (keep_dequant)
+        GraphSafeRemoveNodes(graph, {quant_op, quant_out});
+      else
+        GraphSafeRemoveNodes(graph,
+                             {dequant_op, quant_op, dequant_out, quant_out});
+
+      IR_NODE_LINK_TO(dequant_in, next_op);
+
+      found_squash_count++;
+    } else {
+      // squash dequantize-quantize to requantize op
+      OpDesc desc;
+      desc.SetType("requantize");
+      desc.SetInput("Input", std::vector<std::string>({dequant_in->Name()}));
+      desc.SetOutput("Output", std::vector<std::string>({quant_out->Name()}));
+      desc.SetAttr("Scale_in", dequant_scale);
+      desc.SetAttr("Scale_out", quant_scale);
+
+      auto requant_op = g->CreateOpNode(&desc);
+
+      if (keep_dequant)
+        GraphSafeRemoveNodes(graph, {quant_op});
+      else
+        GraphSafeRemoveNodes(graph, {dequant_op, quant_op, dequant_out});
+
+      IR_NODE_LINK_TO(dequant_in, requant_op);
+      IR_NODE_LINK_TO(requant_op, quant_out);
+
+      found_squash_count++;
+    }
+  };
+  gpd(graph, handler);
+  AddStatis(found_squash_count);
+  PrettyLogDetail("---    squashed %d dequantize-quantize pairs",
+                  found_squash_count);
+}
+
+std::unique_ptr<ir::Graph> CPUQuantizeSquashPass::ApplyImpl(
+    std::unique_ptr<ir::Graph> graph) const {
+  PADDLE_ENFORCE(graph.get());
+  FusePassBase::Init("cpu_quantize_squash_pass", graph.get());
+
+  std::unordered_map<const Node*, int> nodes_keep_counter;
+  FindNodesToKeep(graph.get(), &nodes_keep_counter);
+  Squash(graph.get(), &nodes_keep_counter);
+
+  return graph;
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+REGISTER_PASS(cpu_quantize_squash_pass,
+              paddle::framework::ir::CPUQuantizeSquashPass);

paddle/fluid/framework/ir/cpu_quantize_squash_pass.h

+// 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.
+
+#pragma once
+
+#include <memory>
+#include <string>
+#include <unordered_map>
+#include "paddle/fluid/framework/ir/fuse_pass_base.h"
+#include "paddle/fluid/framework/ir/graph.h"
+#include "paddle/fluid/framework/ir/graph_pattern_detector.h"
+#include "paddle/fluid/framework/ir/pass.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+/*
+ * Squash dequantize->quantize pair pattern into requantize op
+ */
+class CPUQuantizeSquashPass : public FusePassBase {
+ public:
+  virtual ~CPUQuantizeSquashPass() {}
+
+ protected:
+  std::unique_ptr<ir::Graph> ApplyImpl(
+      std::unique_ptr<ir::Graph> graph) const override;
+
+  /*
+   * For each dequantize's output find the number of operators it is an input to
+   */
+  void FindNodesToKeep(
+      Graph* graph,
+      std::unordered_map<const Node*, int>* nodes_keep_counter) const;
+
+  /*
+   * Squash dequantize-quantize ops pairs into requantize or nothing
+   */
+  void Squash(Graph* graph,
+              std::unordered_map<const Node*, int>* nodes_keep_counter) const;
+
+  const std::string name_scope_{"squash"};
+};
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle

paddle/fluid/framework/ir/cpu_quantize_squash_pass_tester.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 "paddle/fluid/framework/ir/cpu_quantize_squash_pass.h"
+#include <gtest/gtest.h>
+#include "paddle/fluid/framework/naive_executor.h"
+#include "paddle/fluid/platform/place.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+void SetOp(ProgramDesc* prog, const std::string& type, const std::string& name,
+           const std::vector<std::string>& inputs,
+           const std::vector<std::string>& outputs, bool use_mkldnn,
+           float scale = 0) {
+  auto* op = prog->MutableBlock(0)->AppendOp();
+  op->SetType(type);
+  op->SetAttr("use_mkldnn", use_mkldnn);
+  op->SetAttr("name", name);
+  if (type == "conv2d") {
+    op->SetInput("Input", {inputs[0]});
+    if (inputs.size() > 1) op->SetInput("Filter", {inputs[1]});
+    if (inputs.size() > 2) op->SetInput("Bias", {inputs[2]});
+    op->SetOutput("Output", {outputs[0]});
+  } else if (type == "quantize") {
+    op->SetInput("Input", {inputs[0]});
+    op->SetOutput("Output", {outputs[0]});
+    op->SetAttr("Scale", scale);
+  } else if (type == "dequantize") {
+    op->SetInput("Input", {inputs[0]});
+    op->SetOutput("Output", {outputs[0]});
+    op->SetAttr("Scale", scale);
+  }
+}
+
+// (a,w1,b1)->Conv1->d
+// d->Dequant->e
+// e->Quant->f
+// (f,w2,b2)->Conv2->i
+ProgramDesc BuildProgramDesc(bool use_mkldnn, float scale1, float scale2) {
+  ProgramDesc prog;
+  for (auto& v : std::initializer_list<std::string>(
+           {"a", "w1", "b1", "d", "e", "f", "w2", "b2", "i"})) {
+    auto* var = prog.MutableBlock(0)->Var(v);
+    if (v.find("w") == 0 || v.find("b") == 0) {
+      var->SetPersistable(true);
+    }
+  }
+
+  SetOp(&prog, "conv2d", "Conv1", {"a", "w1", "b1"}, {"d"}, use_mkldnn);
+  SetOp(&prog, "dequantize", "Dequant", {"d"}, {"e"}, use_mkldnn, scale1);
+  SetOp(&prog, "quantize", "Quant", {"e"}, {"f"}, use_mkldnn, scale2);
+  SetOp(&prog, "conv2d", "Conv2", {"f", "w2", "b2"}, {"i"}, use_mkldnn);
+  return prog;
+}
+
+static const std::initializer_list<std::string> variable_names{
+    "a", "b", "c", "d", "e", "f", "g", "h"};
+// a->Conv1->b
+// b->Dequant->c
+//
+// c->Quant1->d and d->Conv2->e
+//
+// c->Conv3->f
+//
+// c->Quant2->g and g->Conv4->h
+//
+ProgramDesc BuildProgramDesc2(bool use_mkldnn, float scale1, float scale2,
+                              float scale3) {
+  ProgramDesc prog;
+  for (auto& v : variable_names) {
+    prog.MutableBlock(0)->Var(v);
+  }
+
+  SetOp(&prog, "conv2d", "Conv1", {"a"}, {"b"}, use_mkldnn);
+  SetOp(&prog, "dequantize", "Dequant", {"b"}, {"c"}, use_mkldnn, scale1);
+
+  SetOp(&prog, "quantize", "Quant1", {"c"}, {"d"}, use_mkldnn, scale2);
+  SetOp(&prog, "conv2d", "Conv2", {"d"}, {"e"}, use_mkldnn);
+
+  SetOp(&prog, "conv2d", "Conv3", {"c"}, {"f"}, use_mkldnn);
+
+  SetOp(&prog, "quantize", "Quant2", {"c"}, {"g"}, use_mkldnn, scale3);
+  SetOp(&prog, "conv2d", "Conv4", {"g"}, {"h"}, use_mkldnn);
+
+  return prog;
+}
+
+void InitTensorHolder(Scope* scope, const paddle::platform::Place& place,
+                      const char* var_name) {
+  auto x = scope->Var(var_name);
+  auto tensor = x->GetMutable<LoDTensor>();
+  tensor->mutable_data(place, proto::VarType::FP32,
+                       ::paddle::memory::Allocator::kDefault, 1);
+}
+
+void MainTest(const ProgramDesc& prog, int removed_nodes_num) {
+  std::unique_ptr<ir::Graph> graph(new ir::Graph(prog));
+
+  // Init scope, as it is used in pass
+  auto place = paddle::platform::CPUPlace();
+  NaiveExecutor exe{place};
+  Scope scope;
+  exe.CreateVariables(prog, 0, true, &scope);
+
+  for (auto& v : variable_names) {
+    InitTensorHolder(&scope, place, v.c_str());
+  }
+
+  graph->Set(kParamScopeAttr, new framework::Scope*(&scope));
+
+  auto pass = PassRegistry::Instance().Get("cpu_quantize_squash_pass");
+
+  int original_nodes_num = graph->Nodes().size();
+
+  graph = pass->Apply(std::move(graph));
+
+  int current_nodes_num = graph->Nodes().size();
+
+  EXPECT_EQ(original_nodes_num - removed_nodes_num, current_nodes_num);
+}
+
+TEST(CpuQuantizeSquashPass, equal_scales) {
+  auto scale = 1.2345f;
+  auto use_mkldnn = true;
+  // Remove 4 nodes: Dequant, Quant, e, f
+  auto remove_nodes = 4;
+  MainTest(BuildProgramDesc(use_mkldnn, scale, scale), remove_nodes);
+
+  use_mkldnn = !use_mkldnn;
+  MainTest(BuildProgramDesc(use_mkldnn, scale, scale), remove_nodes);
+}
+
+TEST(CpuQuantizeSquashPass, inequal_scales) {
+  auto scale1 = 1.2345f;
+  auto scale2 = 21.0f;
+  auto use_mkldnn = true;
+  // Remove 3 nodes: Dequant, Quant, e
+  // Insert 1 node: requantize
+  auto remove_nodes = 2;
+  MainTest(BuildProgramDesc(use_mkldnn, scale1, scale2), remove_nodes);
+
+  use_mkldnn = !use_mkldnn;
+  MainTest(BuildProgramDesc(use_mkldnn, scale1, scale2), remove_nodes);
+}
+
+TEST(CpuQuantizeSquashPass, branch_to_equal_inequal_and_fp32) {
+  // Delete both quantize ops,
+  // bypass dequantize in both branches,
+  // insert requantize on one branch
+  auto scale = 1.2345f;
+  auto scale2 = 21.0f;
+  auto use_mkldnn = true;
+  // Remove 3 nodes: Quant1, Quant2, g
+  // Insert 1 node: requantize
+  auto remove_nodes = 2;
+  MainTest(BuildProgramDesc2(use_mkldnn, scale, scale, scale2), remove_nodes);
+
+  use_mkldnn = !use_mkldnn;
+  MainTest(BuildProgramDesc2(use_mkldnn, scale, scale, scale2), remove_nodes);
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+USE_PASS(cpu_quantize_squash_pass);

paddle/fluid/framework/ir/graph_pattern_detector.cc

@@ -90,7 +90,8 @@ void GraphPatternDetector::operator()(Graph *graph,
   ValidateByNodeRole(&subgraphs);
 
   if (subgraphs.empty()) return;
-  PrettyLogEndl(Style::detail(), "---  detect %d subgraphs", subgraphs.size());
+  PrettyLogEndl(Style::detail(), "---  detected %d subgraphs",
+                subgraphs.size());
   int id = 0;
   for (auto &g : subgraphs) {
     VLOG(3) << "optimizing #" << id++ << " subgraph";
@@ -1074,9 +1075,53 @@ PDNode *patterns::Conv::operator()() {
                         ->AsOutput()
                         ->assert_is_op_output("conv2d", "Output");
 
-  conv_op->LinksFrom({input_var, filter_var});
-  conv_op->LinksTo({output_var});
+  conv_op->LinksFrom({input_var, filter_var}).LinksTo({output_var});
+  return output_var;
+}
+
+PDNode *patterns::ConvResidual::operator()(bool with_residual_data) {
+  auto conv_op = pattern->NewNode(conv_op_repr())->assert_is_op("conv2d");
+
+  if (!with_residual_data)
+    conv_op->assert_op_attr("fuse_residual_connection", false);
+
+  auto input_var = pattern->NewNode(conv_input_repr())
+                       ->AsInput()
+                       ->assert_is_op_input("conv2d", "Input");
+
+  auto filter_var = pattern->NewNode(conv_filter_repr())
+                        ->AsInput()
+                        ->assert_is_op_input("conv2d", "Filter");
+
+  auto output_var = pattern->NewNode(conv_output_repr())
+                        ->AsOutput()
+                        ->assert_is_op_output("conv2d", "Output");
+
+  std::vector<PDNode *> links_from{input_var, filter_var};
+
+  if (with_residual_data) {
+    auto res_conn_var = pattern->NewNode(conv_residual_data_repr())
+                            ->AsInput()
+                            ->assert_is_op_input("conv2d", "ResidualData");
+    links_from.push_back(res_conn_var);
+  }
+
+  conv_op->LinksFrom(links_from).LinksTo({output_var});
+  return output_var;
+}
+
+PDNode *patterns::Pool::operator()() {
+  auto pool_op = pattern->NewNode(pool_op_repr())->assert_is_op("pool2d");
+
+  auto input_var = pattern->NewNode(pool_input_repr())
+                       ->AsInput()
+                       ->assert_is_op_input("pool2d", "X");
+
+  auto output_var = pattern->NewNode(pool_output_repr())
+                        ->AsOutput()
+                        ->assert_is_op_output("pool2d", "Out");
 
+  pool_op->LinksFrom({input_var}).LinksTo({output_var});
   return output_var;
 }
 
@@ -1301,6 +1346,51 @@ PDNode *patterns::ConvAffineChannel::operator()(
   return ac_out_var;
 }
 
+PDNode *patterns::DequantQuantAny::operator()() {
+  auto *dequant_in = pattern->NewNode(dequant_in_repr())
+                         ->AsInput()
+                         ->assert_is_op_input("dequantize", "Input");
+
+  auto *dequant_op =
+      pattern->NewNode(dequant_op_repr())->assert_is_op("dequantize");
+
+  auto *dequant_out = pattern->NewNode(dequant_out_repr())
+                          ->AsOutput()
+                          ->assert_is_op_output("dequantize", "Output");
+
+  auto *quant_op = pattern->NewNode(quant_op_repr())
+                       ->assert_is_op("quantize")
+                       ->AsIntermediate();
+
+  auto *quant_out = pattern->NewNode(quant_out_repr())
+                        ->AsOutput()
+                        ->assert_is_op_output("quantize");
+
+  auto *next_op = pattern->NewNode(next_op_repr())->assert_is_op();
+
+  dequant_op->LinksFrom({dequant_in}).LinksTo({dequant_out});
+  quant_op->LinksFrom({dequant_out}).LinksTo({quant_out});
+  next_op->LinksFrom({quant_out});
+
+  return quant_out;
+}
+
+PDNode *patterns::DequantAny::operator()() {
+  auto *dequant_op =
+      pattern->NewNode(dequant_op_repr())->assert_is_op("dequantize");
+
+  auto *dequant_out = pattern->NewNode(dequant_out_repr())
+                          ->AsOutput()
+                          ->assert_is_op_output("dequantize", "Output");
+
+  auto *next_op = pattern->NewNode(next_op_repr())->assert_is_op();
+
+  dequant_op->LinksTo({dequant_out});
+  next_op->LinksFrom({dequant_out});
+
+  return dequant_out;
+}
+
 // a -> transpose_op(1) -> transpose_out_a -> flatten_op(1) -> flatten_out_a
 // b -> transpose_op(2) -> transpose_out_b -> flatten_op(2) -> flatten_out_b
 // ...

paddle/fluid/framework/ir/graph_pattern_detector.h

@@ -18,8 +18,11 @@
 #include <gtest/gtest_prod.h>
 #endif
 
+#include <memory>
 #include <numeric>
 #include <string>
+#include <unordered_map>
+#include <unordered_set>
 #include <utility>
 #include <vector>
 #include "paddle/fluid/framework/ir/graph.h"
@@ -656,6 +659,35 @@ struct Conv : public PatternBase {
   PATTERN_DECL_NODE(conv_output);
 };
 
+// Convolution op with residual data
+struct ConvResidual : public PatternBase {
+  ConvResidual(PDPattern* pattern, const std::string& name_scope)
+      : PatternBase(pattern, name_scope, "conv_residual") {}
+
+  PDNode* operator()(bool with_residual_data);
+
+  PATTERN_DECL_NODE(conv_op);
+  PATTERN_DECL_NODE(conv_input);
+  PATTERN_DECL_NODE(conv_filter);
+  PATTERN_DECL_NODE(conv_residual_data);
+  PATTERN_DECL_NODE(conv_output);
+};
+
+// Pool op
+// Forward pass for pooling.
+// pool_input is the input.
+// pool_output is a result of the operator.
+struct Pool : public PatternBase {
+  Pool(PDPattern* pattern, const std::string& name_scope)
+      : PatternBase(pattern, name_scope, "pooling") {}
+
+  PDNode* operator()();
+
+  PATTERN_DECL_NODE(pool_op);
+  PATTERN_DECL_NODE(pool_input);
+  PATTERN_DECL_NODE(pool_output);
+};
+
 // ElementwiseAdd used in residual connections.
 // y_var is used and convolution output.
 // The operator is removed, when residual
@@ -766,6 +798,34 @@ struct ConvAffineChannel : public PatternBase {
   PATTERN_DECL_NODE(ac_out);  // Out
 };
 
+// Dequantize + Quantize + anyOP
+// This pattern is used for squashing the dequantize-quantize pairs.
+struct DequantQuantAny : public PatternBase {
+  DequantQuantAny(PDPattern* pattern, const std::string& name_scope)
+      : PatternBase(pattern, name_scope, "dequant_quant_any") {}
+  PDNode* operator()();
+
+  PATTERN_DECL_NODE(dequant_in);
+  PATTERN_DECL_NODE(dequant_op);
+  PATTERN_DECL_NODE(dequant_out);
+  PATTERN_DECL_NODE(quant_op);
+  PATTERN_DECL_NODE(quant_out);
+  PATTERN_DECL_NODE(next_op);
+};
+
+// Dequantize + anyOP
+// This quantize is used for getting number of ops the Dequantize's
+// output is an input to.
+struct DequantAny : public PatternBase {
+  DequantAny(PDPattern* pattern, const std::string& name_scope)
+      : PatternBase(pattern, name_scope, "dequant_any") {}
+  PDNode* operator()();
+
+  PATTERN_DECL_NODE(dequant_op);
+  PATTERN_DECL_NODE(dequant_out);
+  PATTERN_DECL_NODE(next_op);
+};
+
 struct TransposeFlattenConcat : public PatternBase {
   TransposeFlattenConcat(PDPattern* pattern, const std::string& name_scope)
       : PatternBase(pattern, name_scope, "transpose_flatten_concat") {}

paddle/fluid/framework/ir/sync_batch_norm_pass.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 "paddle/fluid/framework/ir/sync_batch_norm_pass.h"
+#include <memory>
+#include <string>
+#include <utility>
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+std::unique_ptr<ir::Graph> SyncBatchNormPass::ApplyImpl(
+    std::unique_ptr<ir::Graph> graph) const {
+  VLOG(3) << "Use synchronous batch norm";
+  for (const Node* n : graph->Nodes()) {
+    if (n->IsOp()) {
+      auto* op = n->Op();
+      if (op->Type() == "batch_norm") {
+        op->SetType("sync_batch_norm");
+      }
+      if (op->Type() == "batch_norm_grad") {
+        op->SetType("sync_batch_norm_grad");
+      }
+    }
+  }
+  return graph;
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+REGISTER_PASS(sync_batch_norm_pass, paddle::framework::ir::SyncBatchNormPass);

paddle/fluid/framework/ir/sync_batch_norm_pass.h

+/* 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. */
+
+#pragma once
+
+#include <memory>
+#include "paddle/fluid/framework/ir/pass.h"
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+class SyncBatchNormPass : public Pass {
+ protected:
+  std::unique_ptr<ir::Graph> ApplyImpl(
+      std::unique_ptr<ir::Graph> graph) const override;
+};
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle

paddle/fluid/framework/ir/sync_batch_norm_pass_tester.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 "paddle/fluid/framework/ir/sync_batch_norm_pass.h"
+#include <gtest/gtest.h>
+
+namespace paddle {
+namespace framework {
+namespace ir {
+
+void SetOp(ProgramDesc* prog, const std::string& type, const std::string& name,
+           const std::vector<std::string>& inputs,
+           const std::vector<std::string>& outputs) {
+  auto* op = prog->MutableBlock(0)->AppendOp();
+  op->SetType(type);
+  op->SetAttr("name", name);
+  op->SetInput("X", inputs);
+  op->SetOutput("Out", outputs);
+}
+
+// (a, conv_w)->conv2d->b
+// (b, bn_scale, bn_bias, mean, var)->batch_norm
+//     ->(c, mean, var, save_mean, save_inv_var)
+ProgramDesc BuildProgramDesc() {
+  ProgramDesc prog;
+  for (auto& v : std::vector<std::string>({"a", "conv_w", "b", "bn_scale",
+                                           "bn_bias", "mean", "var", "c",
+                                           "save_mean", "save_inv_var"})) {
+    auto* var = prog.MutableBlock(0)->Var(v);
+    if (v == "conv_w" || v == "bn_scale" || v == "bn_bias" || v == "mean" ||
+        v == "var") {
+      var->SetPersistable(true);
+    }
+  }
+
+  SetOp(&prog, "conv2d", "conv", std::vector<std::string>({"a", "conv_w"}),
+        std::vector<std::string>({"b"}));
+  SetOp(&prog, "batch_norm", "bn",
+        std::vector<std::string>({"b", "bn_scale", "bn_bias", "mean", "var"}),
+        std::vector<std::string>(
+            {"c", "mean", "var", "save_mean", "save_inv_var"}));
+  return prog;
+}
+
+TEST(IsTestPass, basic) {
+  auto prog = BuildProgramDesc();
+
+  std::unique_ptr<ir::Graph> graph(new ir::Graph(prog));
+
+  auto pass = PassRegistry::Instance().Get("sync_batch_norm_pass");
+
+  graph = pass->Apply(std::move(graph));
+
+  for (auto* node : graph->Nodes()) {
+    if (node->IsOp()) {
+      auto* op = node->Op();
+      auto op_name = boost::get<std::string>(op->GetAttr("name"));
+      if (op_name == "bn") {
+        ASSERT_EQ(op->Type(), "sync_batch_norm");
+      }
+    }
+  }
+}
+
+}  // namespace ir
+}  // namespace framework
+}  // namespace paddle
+
+USE_PASS(sync_batch_norm_pass);

paddle/fluid/framework/parallel_executor.cc

@@ -14,8 +14,10 @@ limitations under the License. */
 
 #include "paddle/fluid/framework/parallel_executor.h"
 #include <algorithm>
+#include <memory>
 #include <string>
 #include <tuple>
+#include <utility>
 #include <vector>
 #include "paddle/fluid/framework/ir/graph_helper.h"
 
@@ -251,6 +253,20 @@ ParallelExecutor::ParallelExecutor(const std::vector<platform::Place> &places,
     member_->nccl_ctxs_.reset(new platform::NCCLContextMap(
         member_->places_, nccl_id, build_strategy.num_trainers_,
         build_strategy.trainer_id_));
+
+    std::unique_ptr<platform::NCCLContextMap> dev_nccl_ctxs;
+    dev_nccl_ctxs.reset(new platform::NCCLContextMap(member_->places_));
+    // Initialize device context's nccl comm
+    // Note, more than one ParallelExecutor with same place, the nccl comm will
+    // be rewrite and there will be some problem.
+    for (size_t dev_id = 0; dev_id < member_->places_.size(); ++dev_id) {
+      auto &nccl_ctx = dev_nccl_ctxs->at(dev_id);
+      platform::DeviceContextPool &pool =
+          platform::DeviceContextPool::Instance();
+      auto *dev_ctx = static_cast<platform::CUDADeviceContext *>(
+          pool.Get(member_->places_[dev_id]));
+      dev_ctx->set_nccl_comm(nccl_ctx.comm());
+    }
 #else
     PADDLE_THROW("Not compiled with CUDA");
 #endif

paddle/fluid/imperative/CMakeLists.txt

@@ -2,4 +2,5 @@ if(WITH_PYTHON)
 cc_library(layer SRCS layer.cc DEPS proto_desc operator device_context blas pybind)
 cc_library(tracer SRCS tracer.cc DEPS proto_desc device_context pybind)
 cc_library(engine SRCS engine.cc)
+cc_library(imperative_profiler SRCS profiler.cc)
 endif()

paddle/fluid/imperative/layer.cc

@@ -214,10 +214,8 @@ framework::LoDTensor& VarBase::GradValue() {
 }
 
 std::map<std::string, std::vector<VarBase*>> OpBase::ApplyGrad() {
-  if (grad_op_descs_.empty() && backward_id_ <= 0) {
-    VLOG(3) << "op with no grad: " << Type();
-    return {};
-  }
+  PADDLE_ENFORCE(!grad_op_descs_.empty() || backward_id_ > 0,
+                 "%s has no backward implementation", Type());
 
   VLOG(3) << "apply op grad: " << Type();
   std::vector<framework::VariableValueMap> tmp_grad_outputs;
@@ -239,7 +237,7 @@ std::map<std::string, std::vector<VarBase*>> OpBase::ApplyGrad() {
       VLOG(3) << "apply grad op " << grad_op_desc->Type();
 
       // Allocate tmp grad output variable
-      for (auto it : grad_output_variable_map) {
+      for (const auto& it : grad_output_variable_map) {
         auto& outputs = tmp_grad_outputs[k][it.first];
         outputs.reserve(it.second.size());
         for (size_t i = 0; i < it.second.size(); ++i) {
@@ -273,9 +271,9 @@ std::map<std::string, std::vector<VarBase*>> OpBase::ApplyGrad() {
 
   // Add tmp grad outputs to original grad vars
   for (size_t k = 0; k < grad_output_vars_.size(); ++k) {
-    for (auto it : grad_output_vars_[k]) {
+    for (const auto& it : grad_output_vars_[k]) {
       auto& outputs = tmp_grad_outputs[k][it.first];
-      auto& origin_outputs = it.second;
+      const auto& origin_outputs = it.second;
       PADDLE_ENFORCE_EQ(outputs.size(), origin_outputs.size());
 
       for (size_t i = 0; i < outputs.size(); ++i) {

paddle/fluid/imperative/layer.h

@@ -294,17 +294,23 @@ class PYBIND11_HIDDEN OpBase {
 
   void InvokeBackwardHooks();
 
-  void TrackPreOp(const VarBase* inp_var, const std::string& inp_name) {
-    if (inp_var->PreOp() && !inp_var->IsStopGradient()) {
-      VLOG(3) << "add pre op " << inp_var->PreOp()->Type() << " in slot "
-              << inp_name;
-      pre_ops_[inp_name].push_back(inp_var->PreOp());
-      pre_ops_out_idx_[inp_name].push_back(inp_var->PreOpOutIdx());
-    } else {
-      VLOG(3) << "no pre op in slot " << inp_name
-              << " input var stop_gradient: " << inp_var->IsStopGradient();
-      pre_ops_[inp_name].push_back(nullptr);
-      // pre_ops_out_idx_[inp_name].push_back(-1);
+  void TrackPreOp(const std::string& inp_name,
+                  const std::vector<VarBase*>& inputs) {
+    auto& pre_ops_list = pre_ops_[inp_name];
+    pre_ops_list.reserve(inputs.size());
+    auto& pre_ops_out_idx_list = pre_ops_out_idx_[inp_name];
+    for (VarBase* inp_var : inputs) {
+      if (inp_var->PreOp() && !inp_var->IsStopGradient()) {
+        VLOG(3) << "add pre op " << inp_var->PreOp()->Type() << " in slot "
+                << inp_name;
+        pre_ops_list.emplace_back(inp_var->PreOp());
+        pre_ops_out_idx_list.push_back(inp_var->PreOpOutIdx());
+      } else {
+        VLOG(3) << "no pre op in slot " << inp_name
+                << " input var stop_gradient: " << inp_var->IsStopGradient();
+        pre_ops_list.emplace_back(nullptr);
+        // pre_ops_out_idx_list.push_back(-1);
+      }
     }
   }
 

paddle/fluid/imperative/profiler.cc

+// Copyright (c) 2018 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 "paddle/fluid/imperative/profiler.h"
+
+#ifdef WITH_GPERFTOOLS
+#include "gperftools/profiler.h"
+#endif
+#include <gflags/gflags.h>
+#include <glog/logging.h>
+#include <mutex>   // NOLINT
+#include <thread>  // NOLINT
+
+DEFINE_string(
+    tracer_profile_fname, "xxgperf",
+    "Profiler filename for imperative tracer, which generated by gperftools."
+    "Only valid when compiled `WITH_PROFILER=ON`. Empty if disable.");
+
+namespace paddle {
+namespace imperative {
+
+static std::once_flag gTracerProfileOnce;
+#ifdef WITH_GPERFTOOLS
+static bool gTracerProfilerStarted = false;
+#endif
+
+void StartProfile() {
+  if (!FLAGS_tracer_profile_fname.empty()) {
+    std::call_once(gTracerProfileOnce, [] {
+#ifdef WITH_GPERFTOOLS
+      ProfilerStart(FLAGS_tracer_profile_fname.c_str());
+      gTracerProfilerStarted = true;
+#else
+      LOG(WARNING) << "Paddle is not compiled with gperftools. "
+                      "FLAGS_tracer_profile_fname will be ignored";
+#endif
+    });
+  }
+}
+
+void StopProfile() {
+#ifdef WITH_GPERFTOOLS
+  ProfilerFlush();
+#else
+  LOG(WARNING) << "Paddle is not compiled with gperftools. "
+                  "FLAGS_tracer_profile_fname will be ignored";
+#endif
+}
+
+}  // namespace imperative
+}  // namespace paddle

paddle/fluid/imperative/profiler.h

+// Copyright (c) 2018 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.
+
+#pragma once
+
+namespace paddle {
+namespace imperative {
+
+extern void StartProfile();
+
+extern void StopProfile();
+
+}  // namespace imperative
+}  // namespace paddle

paddle/fluid/imperative/tracer.cc

@@ -23,34 +23,21 @@
 #include "paddle/fluid/platform/device_context.h"
 #include "paddle/fluid/platform/enforce.h"
 
-#ifdef WITH_GPERFTOOLS
-#include "gperftools/profiler.h"
-#endif
-
-DEFINE_string(
-    tracer_profile_fname, "",
-    "Profiler filename for imperative tracer, which generated by gperftools."
-    "Only valid when compiled `WITH_PROFILER=ON`. Empty if disable.");
-
 namespace paddle {
 namespace imperative {
 
-static std::once_flag gTracerProfileOnce;
-#ifdef WITH_GPERFTOOLS
-static bool gTracerProfilerStarted = false;
-#endif
-
 void CreateGradOp(const framework::OpDesc& op_desc,
                   const std::unordered_set<std::string>& no_grad_set,
                   const std::vector<framework::BlockDesc*>& grad_sub_block,
                   std::vector<framework::OpDesc*>* grad_op_descs,
                   std::unordered_map<std::string, std::string>* grad_to_var) {
   PADDLE_ENFORCE(grad_op_descs->empty());
-  std::vector<std::unique_ptr<framework::OpDesc>> descs =
-      framework::OpInfoMap::Instance()
-          .Get(op_desc.Type())
-          .GradOpMaker()(op_desc, no_grad_set, grad_to_var, grad_sub_block);
+  const framework::OpInfo& op_info =
+      framework::OpInfoMap::Instance().Get(op_desc.Type());
+  if (!op_info.grad_op_maker_) return;
 
+  std::vector<std::unique_ptr<framework::OpDesc>> descs =
+      op_info.GradOpMaker()(op_desc, no_grad_set, grad_to_var, grad_sub_block);
   for (auto& desc : descs) {
     grad_op_descs->emplace_back(desc.release());
   }
@@ -145,31 +132,13 @@ framework::VariableNameMap CreateOutputVarNameMap(
   return result;
 }
 
-Tracer::Tracer(framework::BlockDesc* root_block) : root_block_(root_block) {
-  if (!FLAGS_tracer_profile_fname.empty()) {
-    std::call_once(gTracerProfileOnce, [] {
-#ifdef WITH_GPERFTOOLS
-      ProfilerStart(FLAGS_tracer_profile_fname.c_str());
-      gTracerProfilerStarted = true;
-#else
-      LOG(WARNING) << "Paddle is not compiled with gperftools. "
-                      "FLAGS_tracer_profile_fname will be ignored";
-#endif
-    });
-  }
-}
+Tracer::Tracer(framework::BlockDesc* root_block) : root_block_(root_block) {}
 
 std::set<std::string> Tracer::Trace(OpBase* op, const VarBasePtrMap& inputs,
                                     const VarBasePtrMap& outputs,
                                     framework::AttributeMap attrs_map,
                                     const platform::Place expected_place,
                                     const bool stop_gradient) {
-#ifdef WITH_GPERFTOOLS
-  if (gTracerProfilerStarted) {
-    ProfilerFlush();
-  }
-#endif
-
   framework::VariableValueMap invars_map;
   framework::VariableValueMap outvars_map;
 
@@ -184,7 +153,6 @@ std::set<std::string> Tracer::Trace(OpBase* op, const VarBasePtrMap& inputs,
                               inp->Name());
 
       invars.emplace_back(inp->var_);
-      op->TrackPreOp(inp, it.first);
       if (!stop_gradient) {
         current_vars_map[inp->Name()] = inp;
       }
@@ -192,6 +160,7 @@ std::set<std::string> Tracer::Trace(OpBase* op, const VarBasePtrMap& inputs,
               << " inited: " << inp->var_->IsInitialized()
               << " stop_grad: " << inp->IsStopGradient();
     }
+    op->TrackPreOp(it.first, it.second);
   }
 
   op->output_vars_ = outputs;
@@ -319,9 +288,7 @@ std::vector<VarBase*> Tracer::PyTrace(OpBase* op,
   std::vector<framework::Variable*> ret_vars =
       PyLayer::Apply(op->forward_id_, inputs);
 
-  for (VarBase* inp : inputs) {
-    op->TrackPreOp(inp, PyLayer::kFwdInp);
-  }
+  op->TrackPreOp(PyLayer::kFwdInp, inputs);
 
   std::vector<VarBase*>& outputs = op->output_vars_[PyLayer::kFwdOut];
   outputs.reserve(ret_vars.size());

paddle/fluid/inference/CMakeLists.txt

@@ -91,5 +91,5 @@ if(WITH_TESTING)
   add_subdirectory(tests/book)
   if(WITH_INFERENCE_API_TEST)
     add_subdirectory(tests/api)
-  endif()  
+  endif()
 endif()

paddle/fluid/inference/analysis/argument.h

@@ -27,6 +27,7 @@
 #include <string>
 #include <unordered_map>
 #include <unordered_set>
+#include <utility>
 #include <vector>
 
 #include "paddle/fluid/framework/ir/graph.h"
@@ -38,7 +39,10 @@
 namespace paddle {
 namespace inference {
 namespace analysis {
+
 using framework::ir::Graph;
+using VarQuantScale =
+    std::unordered_map<std::string, std::pair<bool, framework::LoDTensor>>;
 
 /*
  * The argument definition of both Pass and PassManagers.
@@ -127,6 +131,8 @@ struct Argument {
   // Pass a set of op types to enable its mkldnn kernel
   DECL_ARGUMENT_FIELD(mkldnn_enabled_op_types, MKLDNNEnabledOpTypes,
                       std::unordered_set<std::string>);
+  // Scales for variables to be quantized
+  DECL_ARGUMENT_FIELD(quant_var_scales, QuantVarScales, VarQuantScale);
 
   // Passed from config.
   DECL_ARGUMENT_FIELD(use_gpu, UseGPU, bool);

paddle/fluid/inference/analysis/ir_pass_manager.cc

@@ -14,6 +14,7 @@
 
 #include "paddle/fluid/inference/analysis/ir_pass_manager.h"
 #include <string>
+#include <unordered_map>
 #include <vector>
 #include "paddle/fluid/framework/ir/fuse_pass_base.h"
 #include "paddle/fluid/framework/ir/graph.h"
@@ -55,14 +56,14 @@ void IRPassManager::CreatePasses(Argument *argument,
                                   ".dot";
       pass->Set("graph_viz_path", new std::string(std::move(dot_file_path)));
       pass_num++;
-    }
-    if (pass_name == "mkldnn_placement_pass") {
+    } else if (pass_name == "mkldnn_placement_pass") {
       pass->Set("mkldnn_enabled_op_types",
                 new std::unordered_set<std::string>(
                     argument->mkldnn_enabled_op_types()));
-    }
-
-    if (pass_name == "tensorrt_subgraph_pass") {
+    } else if (pass_name == "cpu_quantize_pass") {
+      pass->Set("quant_var_scales",
+                new VarQuantScale(argument->quant_var_scales()));
+    } else if (pass_name == "tensorrt_subgraph_pass") {
       pass->Set("workspace_size", new int(argument->tensorrt_workspace_size()));
       pass->Set("max_batch_size", new int(argument->tensorrt_max_batch_size()));
       pass->Set("min_subgraph_size",

paddle/fluid/inference/api/analysis_config.cc

@@ -219,7 +219,14 @@ void AnalysisConfig::Update() {
   }
 
   if (enable_memory_optim_) {
-    pass_builder()->AppendAnalysisPass("memory_optimize_pass");
+    auto analysis_passes = pass_builder()->AnalysisPasses();
+    auto memory_opti_pass_name = "memory_optimize_pass";
+    bool already_exists =
+        std::find(analysis_passes.begin(), analysis_passes.end(),
+                  memory_opti_pass_name) != analysis_passes.end();
+    if (!already_exists) {
+      pass_builder()->AppendAnalysisPass(memory_opti_pass_name);
+    }
   }
 
   if (ir_debug_) {

paddle/fluid/operators/CMakeLists.txt

@@ -44,10 +44,10 @@ if (WITH_DISTRIBUTE)
     SET(OP_PREFETCH_DEPS ${OP_PREFETCH_DEPS} parameter_prefetch)
 endif()
 
-register_operators(EXCLUDES py_func_op warpctc_op conv_fusion_op DEPS ${OP_HEADER_DEPS} ${OP_PREFETCH_DEPS})
+register_operators(EXCLUDES py_func_op warpctc_op conv_fusion_op sync_batch_norm_op DEPS ${OP_HEADER_DEPS} ${OP_PREFETCH_DEPS})
 
-# warpctc_op needs cudnn 7 above
 if (WITH_GPU)
+    # warpctc_op needs cudnn 7 above
     if (${CUDNN_MAJOR_VERSION} VERSION_LESS 7)
         op_library(warpctc_op DEPS dynload_warpctc sequence_padding sequence_scale SRCS warpctc_op.cc warpctc_op.cu.cc)
     else()
@@ -58,6 +58,10 @@ if (WITH_GPU)
         op_library(conv_fusion_op)
         file(APPEND ${pybind_file} "USE_CUDA_ONLY_OP(conv2d_fusion);\n")
     endif()
+    if (NOT WIN32)
+        op_library(sync_batch_norm_op)
+        file(APPEND ${pybind_file} "USE_CUDA_ONLY_OP(sync_batch_norm);\n")
+    endif()
 else()
     op_library(warpctc_op DEPS dynload_warpctc sequence_padding sequence_scale)
 endif()

paddle/fluid/operators/batch_norm_op.cu

@@ -33,26 +33,6 @@ using CudnnDataType = platform::CudnnDataType<T>;
 template <typename T>
 using BatchNormParamType = typename CudnnDataType<T>::BatchNormParamType;
 
-void ExtractNCWHD(const framework::DDim &dims, const DataLayout &data_layout,
-                  int *N, int *C, int *H, int *W, int *D) {
-  *N = dims[0];
-  if (dims.size() == 2) {
-    *C = dims[1];
-    *H = 1;
-    *W = 1;
-    *D = 1;
-  } else {
-    *C = data_layout == DataLayout::kNCHW ? dims[1] : dims[dims.size() - 1];
-    *H = data_layout == DataLayout::kNCHW ? dims[2] : dims[1];
-    *W = dims.size() > 3
-             ? (data_layout == DataLayout::kNCHW ? dims[3] : dims[2])
-             : 1;
-    *D = dims.size() > 4
-             ? (data_layout == DataLayout::kNCHW ? dims[4] : dims[3])
-             : 1;
-  }
-}
-
 template <typename T>
 class BatchNormKernel<platform::CUDADeviceContext, T>
     : public framework::OpKernel<T> {
@@ -196,22 +176,6 @@ class BatchNormKernel<platform::CUDADeviceContext, T>
   }
 };
 
-template <typename T, framework::DataLayout layout>
-static __global__ void KeBNBackwardData(const T *dy,
-                                        const BatchNormParamType<T> *scale,
-                                        const BatchNormParamType<T> *variance,
-                                        const double epsilon, const int C,
-                                        const int HxW, const int num, T *dx) {
-  int gid = blockIdx.x * blockDim.x + threadIdx.x;
-  int stride = blockDim.x * gridDim.x;
-  for (int i = gid; i < num; i += stride) {
-    const int c = layout == framework::DataLayout::kNCHW ? i / HxW % C : i % C;
-    BatchNormParamType<T> inv_var = 1.0 / sqrt(variance[c] + epsilon);
-    dx[i] = static_cast<T>(static_cast<BatchNormParamType<T>>(dy[i]) *
-                           scale[c] * inv_var);
-  }
-}
-
 template <typename T, int BlockDim, framework::DataLayout layout>
 static __global__ void KeBNBackwardScaleBias(
     const T *dy, const T *x, const BatchNormParamType<T> *mean,
@@ -248,6 +212,22 @@ static __global__ void KeBNBackwardScaleBias(
   }
 }
 
+template <typename T, framework::DataLayout layout>
+static __global__ void KeBNBackwardData(const T *dy,
+                                        const BatchNormParamType<T> *scale,
+                                        const BatchNormParamType<T> *variance,
+                                        const double epsilon, const int C,
+                                        const int HxW, const int num, T *dx) {
+  int gid = blockIdx.x * blockDim.x + threadIdx.x;
+  int stride = blockDim.x * gridDim.x;
+  for (int i = gid; i < num; i += stride) {
+    const int c = layout == framework::DataLayout::kNCHW ? i / HxW % C : i % C;
+    BatchNormParamType<T> inv_var = 1.0 / sqrt(variance[c] + epsilon);
+    dx[i] = static_cast<T>(static_cast<BatchNormParamType<T>>(dy[i]) *
+                           scale[c] * inv_var);
+  }
+}
+
 template <typename T>
 class BatchNormGradKernel<platform::CUDADeviceContext, T>
     : public framework::OpKernel<T> {
@@ -383,7 +363,7 @@ class BatchNormGradKernel<platform::CUDADeviceContext, T>
           KeBNBackwardScaleBias<T, block, framework::DataLayout::kNCHW><<<
               grid2, block, 0, dev_ctx.stream()>>>(
               d_y->data<T>(), x->data<T>(), running_mean_data, running_var_data,
-              epsilon, C, H * W, num, d_scale->data<BatchNormParamType<T>>(),
+              epsilon, N, C, H * W * D, d_scale->data<BatchNormParamType<T>>(),
               d_bias->data<BatchNormParamType<T>>());
         }
       } else {
@@ -394,10 +374,10 @@ class BatchNormGradKernel<platform::CUDADeviceContext, T>
               running_var_data, epsilon, C, H * W, num, d_x->data<T>());
         }
         if (d_scale && d_bias) {
-          KeBNBackwardScaleBias<T, block, framework::DataLayout::kNCHW><<<
+          KeBNBackwardScaleBias<T, block, framework::DataLayout::kNHWC><<<
               grid2, block, 0, dev_ctx.stream()>>>(
               d_y->data<T>(), x->data<T>(), running_mean_data, running_var_data,
-              epsilon, C, H * W, num, d_scale->data<BatchNormParamType<T>>(),
+              epsilon, N, C, H * W * D, d_scale->data<BatchNormParamType<T>>(),
               d_bias->data<BatchNormParamType<T>>());
         }
       }

paddle/fluid/operators/batch_norm_op.h

@@ -13,6 +13,9 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #pragma once
+#include <memory>
+#include <string>
+#include <unordered_map>
 #include "paddle/fluid/framework/eigen.h"
 #include "paddle/fluid/framework/op_registry.h"
 
@@ -35,17 +38,84 @@ template <typename T>
 using ConstEigenVectorArrayMap =
     Eigen::Map<const Eigen::Array<T, Eigen::Dynamic, 1>>;
 
+class BatchNormOp : public framework::OperatorWithKernel {
+ public:
+  using framework::OperatorWithKernel::OperatorWithKernel;
+  void InferShape(framework::InferShapeContext *ctx) const override;
+
+ protected:
+  framework::OpKernelType GetExpectedKernelType(
+      const framework::ExecutionContext &ctx) const override;
+};
+
+class BatchNormGradOp : public framework::OperatorWithKernel {
+ public:
+  using framework::OperatorWithKernel::OperatorWithKernel;
+  void InferShape(framework::InferShapeContext *ctx) const override;
+
+ protected:
+  framework::OpKernelType GetExpectedKernelType(
+      const framework::ExecutionContext &ctx) const override;
+};
+
+class BatchNormOpMaker : public framework::OpProtoAndCheckerMaker {
+ public:
+  void Make() override;
+};
+
+class BatchNormGradMaker : public framework::SingleGradOpDescMaker {
+ public:
+  using framework::SingleGradOpDescMaker::SingleGradOpDescMaker;
+
+ protected:
+  std::unique_ptr<framework::OpDesc> Apply() const override;
+
+  virtual std::string GradOpType() const {
+    return this->ForwardOpType() + "_grad";
+  }
+};
+
+class BatchNormOpInferVarType
+    : public framework::PassInDtypeAndVarTypeToOutput {
+ protected:
+  std::unordered_map<std::string, std::string> GetInputOutputWithSameType()
+      const override {
+    return std::unordered_map<std::string, std::string>{{"X", /*->*/ "Y"}};
+  }
+};
+
 template <typename DeviceContext, typename T>
 class BatchNormKernel : public framework::OpKernel<T> {
  public:
-  void Compute(const framework::ExecutionContext& ctx) const override;
+  void Compute(const framework::ExecutionContext &ctx) const override;
 };
 
 template <typename DeviceContext, typename T>
 class BatchNormGradKernel : public framework::OpKernel<T> {
  public:
-  void Compute(const framework::ExecutionContext& ctx) const override;
+  void Compute(const framework::ExecutionContext &ctx) const override;
 };
 
+inline void ExtractNCWHD(const framework::DDim &dims,
+                         const DataLayout &data_layout, int *N, int *C, int *H,
+                         int *W, int *D) {
+  *N = dims[0];
+  if (dims.size() == 2) {
+    *C = dims[1];
+    *H = 1;
+    *W = 1;
+    *D = 1;
+  } else {
+    *C = data_layout == DataLayout::kNCHW ? dims[1] : dims[dims.size() - 1];
+    *H = data_layout == DataLayout::kNCHW ? dims[2] : dims[1];
+    *W = dims.size() > 3
+             ? (data_layout == DataLayout::kNCHW ? dims[3] : dims[2])
+             : 1;
+    *D = dims.size() > 4
+             ? (data_layout == DataLayout::kNCHW ? dims[4] : dims[3])
+             : 1;
+  }
+}
+
 }  // namespace operators
 }  // namespace paddle

paddle/fluid/operators/conv_op.cc

@@ -14,6 +14,7 @@ limitations under the License. */
 
 #include "paddle/fluid/operators/conv_op.h"
 
+#include <memory>
 #include <string>
 #include <vector>
 
@@ -194,6 +195,12 @@ void Conv2DOpMaker::Make() {
   AddAttr<bool>("use_mkldnn",
                 "(bool, default false) Only used in mkldnn kernel")
       .SetDefault(false);
+  AddAttr<bool>("use_quantizer",
+                "(bool, default false) "
+                "Set to true for operators that should be quantized and use "
+                "int8 kernel. "
+                "Only used on CPU.")
+      .SetDefault(false);
   AddAttr<bool>("fuse_relu", "(bool, default false) Only used in mkldnn kernel")
       .SetDefault(false);
   AddAttr<bool>("fuse_residual_connection",

paddle/fluid/operators/cross_entropy_op.cc

@@ -13,18 +13,21 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #include "paddle/fluid/operators/cross_entropy_op.h"
+#include <memory>
 #include <string>
+#include <unordered_map>
 
 namespace paddle {
 namespace operators {
 
-class CrossEntropyOp : public framework::OperatorWithKernel {
+class CrossEntropyOpBase : public framework::OperatorWithKernel {
  public:
   using framework::OperatorWithKernel::OperatorWithKernel;
 
   void InferShape(framework::InferShapeContext* ctx) const override {
     PADDLE_ENFORCE(ctx->HasInput("X"), "Input(X) should be not null.");
     PADDLE_ENFORCE(ctx->HasInput("Label"), "Input(Label) should be not null.");
+
     PADDLE_ENFORCE(ctx->HasOutput("Y"), "Output(Y) should be not null.");
 
     auto x_dims = ctx->GetInputDim("X");
@@ -43,7 +46,8 @@ class CrossEntropyOp : public framework::OperatorWithKernel {
                         "Input(X) and Input(Label) shall have the same shape "
                         "except the last dimension.");
     }
-    if (ctx->Attrs().Get<bool>("soft_label")) {
+
+    if (IsSoftLabel(ctx)) {
       if (check) {
         PADDLE_ENFORCE_EQ(x_dims[rank - 1], label_dims[rank - 1],
                           "If Attr(soft_label) == true, the last dimension of "
@@ -69,21 +73,24 @@ class CrossEntropyOp : public framework::OperatorWithKernel {
     return framework::OpKernelType(ctx.Input<Tensor>("X")->type(),
                                    ctx.device_context());
   }
+
+  virtual bool IsSoftLabel(framework::InferShapeContext* ctx) const {
+    return ctx->Attrs().Get<bool>("soft_label");
+  }
 };
 
-class CrossEntropyGradientOp : public framework::OperatorWithKernel {
+class CrossEntropyGradientOpBase : public framework::OperatorWithKernel {
  public:
   using framework::OperatorWithKernel::OperatorWithKernel;
 
-  void InferShape(framework::InferShapeContext* ctx) const override {
-    PADDLE_ENFORCE(ctx->HasInput("X"), "Input(X) should be not null.");
+  void InferShape(framework::InferShapeContext* ctx) const {
     PADDLE_ENFORCE(ctx->HasInput("Label"), "Input(Label) should be not null.");
     PADDLE_ENFORCE(ctx->HasInput(framework::GradVarName("Y")),
                    "Input(Y@GRAD) shoudl be not null.");
     PADDLE_ENFORCE(ctx->HasOutput(framework::GradVarName("X")),
                    "Output(X@GRAD) should be not null.");
 
-    auto x_dims = ctx->GetInputDim("X");
+    auto x_dims = GetXDim(ctx);
     auto label_dims = ctx->GetInputDim("Label");
     auto dy_dims = ctx->GetInputDim(framework::GradVarName("Y"));
     int rank = x_dims.size();
@@ -108,9 +115,7 @@ class CrossEntropyGradientOp : public framework::OperatorWithKernel {
                         "The Input(X) and Input(Y@Grad) should have the same "
                         "shape except the last dimension.");
     }
-    PADDLE_ENFORCE_EQ(dy_dims[rank - 1], 1,
-                      "The last dimension of Input(Y@Grad) should be 1.");
-    if (ctx->Attrs().Get<bool>("soft_label")) {
+    if (IsSoftLabel(ctx)) {
       if (check) {
         PADDLE_ENFORCE_EQ(
             x_dims[rank - 1], label_dims[rank - 1],
@@ -123,7 +128,10 @@ class CrossEntropyGradientOp : public framework::OperatorWithKernel {
                         "Input(Label) should be 1.");
     }
     ctx->SetOutputDim(framework::GradVarName("X"), x_dims);
-    ctx->ShareLoD("X", framework::GradVarName("X"));
+    PADDLE_ENFORCE_EQ(dy_dims[rank - 1], 1,
+                      "The last dimension of Input(Y@Grad) should be 1.");
+    ctx->SetOutputDim(framework::GradVarName("X"), x_dims);
+    ctx->ShareLoD(VarNameWithXLoD(), framework::GradVarName("X"));
   }
 
  protected:
@@ -131,8 +139,28 @@ class CrossEntropyGradientOp : public framework::OperatorWithKernel {
   // is determined by its input "X".
   framework::OpKernelType GetExpectedKernelType(
       const framework::ExecutionContext& ctx) const override {
-    return framework::OpKernelType(ctx.Input<Tensor>("X")->type(),
-                                   ctx.device_context());
+    return framework::OpKernelType(
+        ctx.Input<Tensor>(framework::GradVarName("Y"))->type(),
+        ctx.device_context());
+  }
+
+  virtual framework::DDim GetXDim(framework::InferShapeContext* ctx) const {
+    return ctx->GetInputDim("X");
+  }
+
+  virtual const char* VarNameWithXLoD() const { return "X"; }
+
+  virtual bool IsSoftLabel(framework::InferShapeContext* ctx) const {
+    return ctx->Attrs().Get<bool>("soft_label");
+  }
+};
+
+class CrossEntropyOpInferVarType
+    : public framework::PassInDtypeAndVarTypeToOutput {
+ protected:
+  std::unordered_map<std::string, std::string> GetInputOutputWithSameType()
+      const override {
+    return std::unordered_map<std::string, std::string>{{"X", /*->*/ "Y"}};
   }
 };
 
@@ -200,22 +228,132 @@ or not. But the output only shares the LoD information with input X.
   }
 };
 
-class CrossEntropyOpInferVarType
-    : public framework::PassInDtypeAndVarTypeToOutput {
+class CrossEntropyGradientOp : public CrossEntropyGradientOpBase {
+ public:
+  using CrossEntropyGradientOpBase::CrossEntropyGradientOpBase;
+
+  void InferShape(framework::InferShapeContext* ctx) const override {
+    PADDLE_ENFORCE(ctx->HasInput("X"), "Input(X) should be not null.");
+    CrossEntropyGradientOpBase::InferShape(ctx);
+  }
+};
+
+class CrossEntropyOp2 : public CrossEntropyOpBase {
+ public:
+  using CrossEntropyOpBase::CrossEntropyOpBase;
+
+  void InferShape(framework::InferShapeContext* ctx) const override {
+    CrossEntropyOpBase::InferShape(ctx);
+
+    PADDLE_ENFORCE(ctx->HasOutput("XShape"),
+                   "Output(XShape) should be not null.");
+
+    PADDLE_ENFORCE(ctx->HasOutput("MatchX"),
+                   "Output(MatchX) should be not null.");
+    auto x_dims = ctx->GetInputDim("X");
+    auto x_dims_vec = framework::vectorize(x_dims);
+    x_dims_vec.push_back(0);
+    ctx->SetOutputDim("XShape", framework::make_ddim(x_dims_vec));
+    x_dims[x_dims.size() - 1] = 1;
+    ctx->SetOutputDim("MatchX", x_dims);
+    ctx->ShareLoD("X", /*->*/ "XShape");
+  }
+
  protected:
-  std::unordered_map<std::string, std::string> GetInputOutputWithSameType()
-      const override {
-    return std::unordered_map<std::string, std::string>{{"X", /*->*/ "Y"}};
+  bool IsSoftLabel(framework::InferShapeContext* ctx) const override {
+    return false;
+  }
+};
+
+class CrossEntropyGradientOp2 : public CrossEntropyGradientOpBase {
+ public:
+  using CrossEntropyGradientOpBase::CrossEntropyGradientOpBase;
+  void InferShape(framework::InferShapeContext* ctx) const override {
+    PADDLE_ENFORCE(ctx->HasInput("MatchX"), "Input(MatchX) must exist");
+    CrossEntropyGradientOpBase::InferShape(ctx);
+  }
+
+ protected:
+  virtual framework::DDim GetXDim(framework::InferShapeContext* ctx) const {
+    auto x_shape = ctx->GetInputDim("XShape");
+    return framework::DDim(x_shape.Get(), x_shape.size() - 1);
+  }
+
+  virtual const char* VarNameWithXLoD() const { return "XShape"; }
+
+  virtual bool IsSoftLabel(framework::InferShapeContext* ctx) const {
+    return false;
   }
 };
+
+class CrossEntropyOpMaker2 : public framework::OpProtoAndCheckerMaker {
+ public:
+  void Make() override {
+    AddInput("X",
+             "(Tensor, default Tensor<float>), a tensor whose last dimension "
+             "size is equal to the number of classes. This input is a "
+             "probability computed by the previous operator, which is almost "
+             "always the result of a softmax operator.");
+    AddInput(
+        "Label",
+        "(Tensor), the tensor which represents the ground truth. It has the "
+        "same shape with 'X' except the last dimension. One hot Tensor.");
+    AddOutput("Y",
+              "(Tensor, default Tensor<float>), a tensor whose shape is same "
+              "with 'X' except that the last dimension size is 1. It "
+              "represents the cross entropy loss.");
+    AddOutput("XShape", "Temporaily variable to save shape and LoD of X.");
+    AddOutput("MatchX",
+              "X value that matches label, used for gradient computation.");
+    AddAttr<int>("ignore_index",
+                 "(int, default -100), Specifies a target value that is"
+                 "ignored and does not contribute to the input gradient."
+                 "Only valid if soft_label is set to False")
+        .SetDefault(-100);
+    AddComment(R"DOC(
+Hard-label CrossEntropy Operator.
+
+The input 'X' and 'Label' will first be logically flattened to 2-D matrixs. 
+The matrix's second dimension(row length) is as same as the original last 
+dimension, and the first dimension(column length) is the product of all other 
+original dimensions. Then the softmax computation will take palce on each raw 
+of flattened matrixs.
+
+Only support hard label.
+
+Both the input X and Label can carry the LoD (Level of Details) information,
+or not. But the output only shares the LoD information with input X.
+
+)DOC");
+  }
+};
+
+class CrossEntropyGradOpDescMaker2 : public framework::SingleGradOpDescMaker {
+ public:
+  using framework::SingleGradOpDescMaker::SingleGradOpDescMaker;
+
+ protected:
+  std::unique_ptr<framework::OpDesc> Apply() const override {
+    std::unique_ptr<framework::OpDesc> op(new framework::OpDesc());
+    op->SetType("cross_entropy_grad2");
+    op->SetInput("Label", Input("Label"));
+    op->SetInput("MatchX", Output("MatchX"));
+    op->SetInput("XShape", Output("XShape"));
+    op->SetInput(framework::GradVarName("Y"), OutputGrad("Y"));
+    op->SetOutput(framework::GradVarName("X"), InputGrad("X"));
+    op->SetAttrMap(Attrs());
+    return op;
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle
 
 namespace ops = paddle::operators;
 using CPUCtx = paddle::platform::CPUDeviceContext;
 
-REGISTER_OPERATOR(cross_entropy, ops::CrossEntropyOp, ops::CrossEntropyOpMaker,
-                  ops::CrossEntropyOpInferVarType,
+REGISTER_OPERATOR(cross_entropy, ops::CrossEntropyOpBase,
+                  ops::CrossEntropyOpMaker, ops::CrossEntropyOpInferVarType,
                   paddle::framework::DefaultGradOpDescMaker<true>);
 REGISTER_OPERATOR(cross_entropy_grad, ops::CrossEntropyGradientOp);
 REGISTER_OP_CPU_KERNEL(cross_entropy, ops::CrossEntropyOpKernel<CPUCtx, float>,
@@ -223,3 +361,14 @@ REGISTER_OP_CPU_KERNEL(cross_entropy, ops::CrossEntropyOpKernel<CPUCtx, float>,
 REGISTER_OP_CPU_KERNEL(cross_entropy_grad,
                        ops::CrossEntropyGradientOpKernel<CPUCtx, float>,
                        ops::CrossEntropyGradientOpKernel<CPUCtx, double>);
+
+REGISTER_OPERATOR(cross_entropy2, ops::CrossEntropyOp2,
+                  ops::CrossEntropyOpMaker2, ops::CrossEntropyOpInferVarType,
+                  ops::CrossEntropyGradOpDescMaker2);
+REGISTER_OPERATOR(cross_entropy_grad2, ops::CrossEntropyGradientOp2);
+REGISTER_OP_CPU_KERNEL(cross_entropy2,
+                       ops::CrossEntropyOpKernel2<CPUCtx, float>,
+                       ops::CrossEntropyOpKernel2<CPUCtx, double>);
+REGISTER_OP_CPU_KERNEL(cross_entropy_grad2,
+                       ops::CrossEntropyGradientOpKernel2<CPUCtx, float>,
+                       ops::CrossEntropyGradientOpKernel2<CPUCtx, double>);

paddle/fluid/operators/cross_entropy_op.cu

@@ -27,3 +27,13 @@ REGISTER_OP_CUDA_KERNEL(
     cross_entropy_grad, ops::CrossEntropyGradientOpKernel<CUDACtx, float>,
     ops::CrossEntropyGradientOpKernel<CUDACtx, double>,
     ops::CrossEntropyGradientOpKernel<CUDACtx, plat::float16>);
+
+REGISTER_OP_CUDA_KERNEL(cross_entropy2,
+                        ops::CrossEntropyOpKernel2<CUDACtx, float>,
+                        ops::CrossEntropyOpKernel2<CUDACtx, double>,
+                        ops::CrossEntropyOpKernel2<CUDACtx, plat::float16>);
+
+REGISTER_OP_CUDA_KERNEL(
+    cross_entropy_grad2, ops::CrossEntropyGradientOpKernel2<CUDACtx, float>,
+    ops::CrossEntropyGradientOpKernel2<CUDACtx, double>,
+    ops::CrossEntropyGradientOpKernel2<CUDACtx, plat::float16>);

paddle/fluid/operators/cross_entropy_op.h

@@ -15,6 +15,7 @@ limitations under the License. */
 #pragma once
 #include "paddle/fluid/framework/eigen.h"
 #include "paddle/fluid/framework/op_registry.h"
+#include "paddle/fluid/operators/math.h"
 #include "paddle/fluid/operators/math/cross_entropy.h"
 #include "paddle/fluid/operators/math/math_function.h"
 #include "paddle/fluid/platform/for_range.h"
@@ -137,5 +138,124 @@ class CrossEntropyGradientOpKernel : public framework::OpKernel<T> {
   }
 };
 
+template <typename T>
+struct HardLabelCrossEntropyForwardFunctor {
+  HardLabelCrossEntropyForwardFunctor(const T* x, T* y, T* match_x,
+                                      const int64_t* label,
+                                      int64_t ignore_index,
+                                      int64_t feature_size)
+      : x_(x),
+        y_(y),
+        match_x_(match_x),
+        label_(label),
+        ignore_index_(ignore_index),
+        feature_size_(feature_size) {}
+
+  HOSTDEVICE void operator()(int64_t idx) const {
+    auto label = label_[idx];
+    if (label != ignore_index_) {
+      auto match_x = x_[idx * feature_size_ + label];
+      y_[idx] = -math::TolerableValue<T>()(real_log(match_x));
+      match_x_[idx] = match_x;
+    } else {
+      y_[idx] = 0;
+      match_x_[idx] = 0;  // any value is ok
+    }
+  }
+
+  const T* x_;
+  T* y_;
+  T* match_x_;
+  const int64_t* label_;
+  int64_t ignore_index_;
+  int64_t feature_size_;
+};
+
+template <typename T>
+struct HardLabelCrossEntropyBackwardFunctor {
+  HardLabelCrossEntropyBackwardFunctor(T* dx, const T* dy, const T* match_x,
+                                       const int64_t* label,
+                                       int64_t ignore_index,
+                                       int64_t feature_size)
+      : dx_(dx),
+        dy_(dy),
+        match_x_(match_x),
+        label_(label),
+        ignore_index_(ignore_index),
+        feature_size_(feature_size) {}
+
+  HOSTDEVICE void operator()(int64_t idx) const {
+    auto row_idx = idx / feature_size_;
+    auto col_idx = idx % feature_size_;
+    auto label = label_[row_idx];
+    if (label == col_idx && label != ignore_index_) {
+      dx_[idx] = -dy_[row_idx] / match_x_[row_idx];
+    } else {
+      dx_[idx] = 0;
+    }
+  }
+
+  T* dx_;
+  const T* dy_;
+  const T* match_x_;
+  const int64_t* label_;
+  int64_t ignore_index_;
+  int64_t feature_size_;
+};
+
+template <typename DeviceContext, typename T>
+class CrossEntropyOpKernel2 : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* x = ctx.Input<Tensor>("X");
+    auto* label = ctx.Input<Tensor>("Label");
+    auto* y = ctx.Output<Tensor>("Y");
+    auto* match_x = ctx.Output<Tensor>("MatchX");
+
+    auto& x_dims = x->dims();
+    auto feature_size = x_dims[x_dims.size() - 1];
+    auto batch_size = framework::product(x->dims()) / feature_size;
+
+    auto* p_x = x->data<T>();
+    auto* p_label = label->data<int64_t>();
+    auto* p_y = y->mutable_data<T>(ctx.GetPlace());
+    auto* p_match_x = match_x->mutable_data<T>(ctx.GetPlace());
+
+    auto ignore_index = ctx.Attr<int>("ignore_index");
+
+    platform::ForRange<DeviceContext> for_range(
+        ctx.template device_context<DeviceContext>(), batch_size);
+    for_range(HardLabelCrossEntropyForwardFunctor<T>(
+        p_x, p_y, p_match_x, p_label, ignore_index, feature_size));
+  }
+};
+
+template <typename DeviceContext, typename T>
+class CrossEntropyGradientOpKernel2 : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* dx = ctx.Output<Tensor>(framework::GradVarName("X"));
+    auto* dy = ctx.Input<Tensor>(framework::GradVarName("Y"));
+    auto* match_x = ctx.Input<Tensor>("MatchX");
+    auto* label = ctx.Input<Tensor>("Label");
+
+    auto* p_dx = dx->mutable_data<T>(ctx.GetPlace());
+    auto* p_dy = dy->data<T>();
+    auto* p_match_x = match_x->data<T>();
+    auto* p_label = label->data<int64_t>();
+
+    int64_t ignore_index = ctx.Attr<int>("ignore_index");
+    int rank = dx->dims().size();
+    int64_t feature_size = dx->dims()[rank - 1];
+    int64_t batch_size = framework::product(dx->dims()) / feature_size;
+
+    platform::ForRange<DeviceContext> for_range(
+        ctx.template device_context<DeviceContext>(),
+        batch_size * feature_size);
+    for_range(HardLabelCrossEntropyBackwardFunctor<T>(
+        p_dx, p_dy, p_match_x, p_label, ignore_index, feature_size));
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle

paddle/fluid/operators/detection/CMakeLists.txt

@@ -33,6 +33,7 @@ detection_library(rpn_target_assign_op SRCS rpn_target_assign_op.cc)
 detection_library(generate_proposal_labels_op SRCS generate_proposal_labels_op.cc)
 detection_library(box_clip_op SRCS box_clip_op.cc box_clip_op.cu)
 detection_library(yolov3_loss_op SRCS yolov3_loss_op.cc)
+detection_library(yolo_box_op SRCS yolo_box_op.cc yolo_box_op.cu)
 detection_library(box_decoder_and_assign_op SRCS box_decoder_and_assign_op.cc box_decoder_and_assign_op.cu)
 
 if(WITH_GPU)

paddle/fluid/operators/detection/box_coder_op.cc

@@ -60,14 +60,15 @@ class BoxCoderOp : public framework::OperatorWithKernel {
     } else if (code_type == BoxCodeType::kDecodeCenterSize) {
       PADDLE_ENFORCE_EQ(target_box_dims.size(), 3,
                         "The rank of Input TargetBox must be 3");
-      if (axis == 0) {
-        PADDLE_ENFORCE_EQ(target_box_dims[1], prior_box_dims[0]);
-      } else if (axis == 1) {
-        PADDLE_ENFORCE_EQ(target_box_dims[0], prior_box_dims[0]);
-      } else {
-        PADDLE_THROW("axis must be 0 or 1.");
+      PADDLE_ENFORCE(axis == 0 || axis == 1, "axis must be 0 or 1");
+      if (ctx->IsRuntime()) {
+        if (axis == 0) {
+          PADDLE_ENFORCE_EQ(target_box_dims[1], prior_box_dims[0]);
+        } else if (axis == 1) {
+          PADDLE_ENFORCE_EQ(target_box_dims[0], prior_box_dims[0]);
+        }
+        PADDLE_ENFORCE_EQ(target_box_dims[2], prior_box_dims[1]);
       }
-      PADDLE_ENFORCE_EQ(target_box_dims[2], prior_box_dims[1]);
       ctx->ShareDim("TargetBox", /*->*/ "OutputBox");
     }
 

paddle/fluid/operators/detection/yolo_box_op.cc

+/* Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserve.
+   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 "paddle/fluid/operators/detection/yolo_box_op.h"
+#include "paddle/fluid/framework/op_registry.h"
+
+namespace paddle {
+namespace operators {
+
+using framework::Tensor;
+
+class YoloBoxOp : public framework::OperatorWithKernel {
+ public:
+  using framework::OperatorWithKernel::OperatorWithKernel;
+  void InferShape(framework::InferShapeContext* ctx) const override {
+    PADDLE_ENFORCE(ctx->HasInput("X"),
+                   "Input(X) of YoloBoxOp should not be null.");
+    PADDLE_ENFORCE(ctx->HasInput("ImgSize"),
+                   "Input(ImgSize) of YoloBoxOp should not be null.");
+    PADDLE_ENFORCE(ctx->HasOutput("Boxes"),
+                   "Output(Boxes) of YoloBoxOp should not be null.");
+    PADDLE_ENFORCE(ctx->HasOutput("Scores"),
+                   "Output(Scores) of YoloBoxOp should not be null.");
+
+    auto dim_x = ctx->GetInputDim("X");
+    auto dim_imgsize = ctx->GetInputDim("ImgSize");
+    auto anchors = ctx->Attrs().Get<std::vector<int>>("anchors");
+    int anchor_num = anchors.size() / 2;
+    auto class_num = ctx->Attrs().Get<int>("class_num");
+
+    PADDLE_ENFORCE_EQ(dim_x.size(), 4, "Input(X) should be a 4-D tensor.");
+    PADDLE_ENFORCE_EQ(
+        dim_x[1], anchor_num * (5 + class_num),
+        "Input(X) dim[1] should be equal to (anchor_mask_number * (5 "
+        "+ class_num)).");
+    PADDLE_ENFORCE_EQ(dim_imgsize.size(), 2,
+                      "Input(ImgSize) should be a 2-D tensor.");
+    PADDLE_ENFORCE_EQ(
+        dim_imgsize[0], dim_x[0],
+        "Input(ImgSize) dim[0] and Input(X) dim[0] should be same.");
+    PADDLE_ENFORCE_EQ(dim_imgsize[1], 2, "Input(ImgSize) dim[1] should be 2.");
+    PADDLE_ENFORCE_GT(anchors.size(), 0,
+                      "Attr(anchors) length should be greater than 0.");
+    PADDLE_ENFORCE_EQ(anchors.size() % 2, 0,
+                      "Attr(anchors) length should be even integer.");
+    PADDLE_ENFORCE_GT(class_num, 0,
+                      "Attr(class_num) should be an integer greater than 0.");
+
+    int box_num = dim_x[2] * dim_x[3] * anchor_num;
+    std::vector<int64_t> dim_boxes({dim_x[0], box_num, 4});
+    ctx->SetOutputDim("Boxes", framework::make_ddim(dim_boxes));
+
+    std::vector<int64_t> dim_scores({dim_x[0], box_num, class_num});
+    ctx->SetOutputDim("Scores", framework::make_ddim(dim_scores));
+  }
+
+ protected:
+  framework::OpKernelType GetExpectedKernelType(
+      const framework::ExecutionContext& ctx) const override {
+    return framework::OpKernelType(ctx.Input<Tensor>("X")->type(),
+                                   ctx.GetPlace());
+  }
+};
+
+class YoloBoxOpMaker : public framework::OpProtoAndCheckerMaker {
+ public:
+  void Make() override {
+    AddInput("X",
+             "The input tensor of YoloBox operator is a 4-D tensor with "
+             "shape of [N, C, H, W]. The second dimension(C) stores "
+             "box locations, confidence score and classification one-hot "
+             "keys of each anchor box. Generally, X should be the output "
+             "of YOLOv3 network.");
+    AddInput("ImgSize",
+             "The image size tensor of YoloBox operator, "
+             "This is a 2-D tensor with shape of [N, 2]. This tensor holds "
+             "height and width of each input image used for resizing output "
+             "box in input image scale.");
+    AddOutput("Boxes",
+              "The output tensor of detection boxes of YoloBox operator, "
+              "This is a 3-D tensor with shape of [N, M, 4], N is the "
+              "batch num, M is output box number, and the 3rd dimension "
+              "stores [xmin, ymin, xmax, ymax] coordinates of boxes.");
+    AddOutput("Scores",
+              "The output tensor of detection boxes scores of YoloBox "
+              "operator, This is a 3-D tensor with shape of "
+              "[N, M, :attr:`class_num`], N is the batch num, M is "
+              "output box number.");
+
+    AddAttr<int>("class_num", "The number of classes to predict.");
+    AddAttr<std::vector<int>>("anchors",
+                              "The anchor width and height, "
+                              "it will be parsed pair by pair.")
+        .SetDefault(std::vector<int>{});
+    AddAttr<int>("downsample_ratio",
+                 "The downsample ratio from network input to YoloBox operator "
+                 "input, so 32, 16, 8 should be set for the first, second, "
+                 "and thrid YoloBox operators.")
+        .SetDefault(32);
+    AddAttr<float>("conf_thresh",
+                   "The confidence scores threshold of detection boxes. "
+                   "Boxes with confidence scores under threshold should "
+                   "be ignored.")
+        .SetDefault(0.01);
+    AddComment(R"DOC(
+         This operator generates YOLO detection boxes from output of YOLOv3 network.
+         
+         The output of previous network is in shape [N, C, H, W], while H and W
+         should be the same, H and W specify the grid size, each grid point predict 
+         given number boxes, this given number, which following will be represented as S,
+         is specified by the number of anchors. In the second dimension(the channel
+         dimension), C should be equal to S * (5 + class_num), class_num is the object 
+         category number of source dataset(such as 80 in coco dataset), so the 
+         second(channel) dimension, apart from 4 box location coordinates x, y, w, h, 
+         also includes confidence score of the box and class one-hot key of each anchor 
+         box.
+
+         Assume the 4 location coordinates are :math:`t_x, t_y, t_w, t_h`, the box 
+         predictions should be as follows:
+
+         $$
+         b_x = \\sigma(t_x) + c_x
+         $$
+         $$
+         b_y = \\sigma(t_y) + c_y
+         $$
+         $$
+         b_w = p_w e^{t_w}
+         $$
+         $$
+         b_h = p_h e^{t_h}
+         $$
+
+         in the equation above, :math:`c_x, c_y` is the left top corner of current grid
+         and :math:`p_w, p_h` is specified by anchors.
+
+         The logistic regression value of the 5th channel of each anchor prediction boxes
+         represents the confidence score of each prediction box, and the logistic
+         regression value of the last :attr:`class_num` channels of each anchor prediction 
+         boxes represents the classifcation scores. Boxes with confidence scores less than
+         :attr:`conf_thresh` should be ignored, and box final scores is the product of 
+         confidence scores and classification scores.
+
+         $$
+         score_{pred} = score_{conf} * score_{class}
+         $$
+
+         )DOC");
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle
+
+namespace ops = paddle::operators;
+REGISTER_OPERATOR(yolo_box, ops::YoloBoxOp, ops::YoloBoxOpMaker,
+                  paddle::framework::EmptyGradOpMaker);
+REGISTER_OP_CPU_KERNEL(yolo_box, ops::YoloBoxKernel<float>,
+                       ops::YoloBoxKernel<double>);

paddle/fluid/operators/detection/yolo_box_op.cu

+/* 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 "paddle/fluid/operators/detection/yolo_box_op.h"
+#include "paddle/fluid/operators/math/math_function.h"
+
+namespace paddle {
+namespace operators {
+
+using Tensor = framework::Tensor;
+
+template <typename T>
+__global__ void KeYoloBoxFw(const T* input, const int* imgsize, T* boxes,
+                            T* scores, const float conf_thresh,
+                            const int* anchors, const int n, const int h,
+                            const int w, const int an_num, const int class_num,
+                            const int box_num, int input_size) {
+  int tid = blockIdx.x * blockDim.x + threadIdx.x;
+  int stride = blockDim.x * gridDim.x;
+  T box[4];
+  for (; tid < n * box_num; tid += stride) {
+    int grid_num = h * w;
+    int i = tid / box_num;
+    int j = (tid % box_num) / grid_num;
+    int k = (tid % grid_num) / w;
+    int l = tid % w;
+
+    int an_stride = (5 + class_num) * grid_num;
+    int img_height = imgsize[2 * i];
+    int img_width = imgsize[2 * i + 1];
+
+    int obj_idx =
+        GetEntryIndex(i, j, k * w + l, an_num, an_stride, grid_num, 4);
+    T conf = sigmoid<T>(input[obj_idx]);
+    if (conf < conf_thresh) {
+      continue;
+    }
+
+    int box_idx =
+        GetEntryIndex(i, j, k * w + l, an_num, an_stride, grid_num, 0);
+    GetYoloBox<T>(box, input, anchors, l, k, j, h, input_size, box_idx,
+                  grid_num, img_height, img_width);
+    box_idx = (i * box_num + j * grid_num + k * w + l) * 4;
+    CalcDetectionBox<T>(boxes, box, box_idx, img_height, img_width);
+
+    int label_idx =
+        GetEntryIndex(i, j, k * w + l, an_num, an_stride, grid_num, 5);
+    int score_idx = (i * box_num + j * grid_num + k * w + l) * class_num;
+    CalcLabelScore<T>(scores, input, label_idx, score_idx, class_num, conf,
+                      grid_num);
+  }
+}
+
+template <typename T>
+class YoloBoxOpCUDAKernel : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* input = ctx.Input<Tensor>("X");
+    auto* img_size = ctx.Input<Tensor>("ImgSize");
+    auto* boxes = ctx.Output<Tensor>("Boxes");
+    auto* scores = ctx.Output<Tensor>("Scores");
+
+    auto anchors = ctx.Attr<std::vector<int>>("anchors");
+    int class_num = ctx.Attr<int>("class_num");
+    float conf_thresh = ctx.Attr<float>("conf_thresh");
+    int downsample_ratio = ctx.Attr<int>("downsample_ratio");
+
+    const int n = input->dims()[0];
+    const int h = input->dims()[2];
+    const int w = input->dims()[3];
+    const int box_num = boxes->dims()[1];
+    const int an_num = anchors.size() / 2;
+    int input_size = downsample_ratio * h;
+
+    auto& dev_ctx = ctx.cuda_device_context();
+    auto& allocator =
+        platform::DeviceTemporaryAllocator::Instance().Get(dev_ctx);
+    int bytes = sizeof(int) * anchors.size();
+    auto anchors_ptr = allocator.Allocate(sizeof(int) * anchors.size());
+    int* anchors_data = reinterpret_cast<int*>(anchors_ptr->ptr());
+    const auto gplace = boost::get<platform::CUDAPlace>(ctx.GetPlace());
+    const auto cplace = platform::CPUPlace();
+    memory::Copy(gplace, anchors_data, cplace, anchors.data(), bytes,
+                 dev_ctx.stream());
+
+    const T* input_data = input->data<T>();
+    const int* imgsize_data = img_size->data<int>();
+    T* boxes_data = boxes->mutable_data<T>({n, box_num, 4}, ctx.GetPlace());
+    T* scores_data =
+        scores->mutable_data<T>({n, box_num, class_num}, ctx.GetPlace());
+    math::SetConstant<platform::CUDADeviceContext, T> set_zero;
+    set_zero(dev_ctx, boxes, static_cast<T>(0));
+    set_zero(dev_ctx, scores, static_cast<T>(0));
+
+    int grid_dim = (n * box_num + 512 - 1) / 512;
+    grid_dim = grid_dim > 8 ? 8 : grid_dim;
+
+    KeYoloBoxFw<T><<<grid_dim, 512, 0, ctx.cuda_device_context().stream()>>>(
+        input_data, imgsize_data, boxes_data, scores_data, conf_thresh,
+        anchors_data, n, h, w, an_num, class_num, box_num, input_size);
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle
+
+namespace ops = paddle::operators;
+REGISTER_OP_CUDA_KERNEL(yolo_box, ops::YoloBoxOpCUDAKernel<float>,
+                        ops::YoloBoxOpCUDAKernel<double>);

paddle/fluid/operators/detection/yolo_box_op.h

+/* Copyright (c) 2019 PaddlePaddle Authors. All Rights Reserve.
+   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. */
+
+#pragma once
+#include <algorithm>
+#include <vector>
+#include "paddle/fluid/framework/op_registry.h"
+#include "paddle/fluid/platform/hostdevice.h"
+
+namespace paddle {
+namespace operators {
+
+using Tensor = framework::Tensor;
+
+template <typename T>
+HOSTDEVICE inline T sigmoid(T x) {
+  return 1.0 / (1.0 + std::exp(-x));
+}
+
+template <typename T>
+HOSTDEVICE inline void GetYoloBox(T* box, const T* x, const int* anchors, int i,
+                                  int j, int an_idx, int grid_size,
+                                  int input_size, int index, int stride,
+                                  int img_height, int img_width) {
+  box[0] = (i + sigmoid<T>(x[index])) * img_width / grid_size;
+  box[1] = (j + sigmoid<T>(x[index + stride])) * img_height / grid_size;
+  box[2] = std::exp(x[index + 2 * stride]) * anchors[2 * an_idx] * img_width /
+           input_size;
+  box[3] = std::exp(x[index + 3 * stride]) * anchors[2 * an_idx + 1] *
+           img_height / input_size;
+}
+
+HOSTDEVICE inline int GetEntryIndex(int batch, int an_idx, int hw_idx,
+                                    int an_num, int an_stride, int stride,
+                                    int entry) {
+  return (batch * an_num + an_idx) * an_stride + entry * stride + hw_idx;
+}
+
+template <typename T>
+HOSTDEVICE inline void CalcDetectionBox(T* boxes, T* box, const int box_idx,
+                                        const int img_height,
+                                        const int img_width) {
+  boxes[box_idx] = box[0] - box[2] / 2;
+  boxes[box_idx + 1] = box[1] - box[3] / 2;
+  boxes[box_idx + 2] = box[0] + box[2] / 2;
+  boxes[box_idx + 3] = box[1] + box[3] / 2;
+
+  boxes[box_idx] = boxes[box_idx] > 0 ? boxes[box_idx] : static_cast<T>(0);
+  boxes[box_idx + 1] =
+      boxes[box_idx + 1] > 0 ? boxes[box_idx + 1] : static_cast<T>(0);
+  boxes[box_idx + 2] = boxes[box_idx + 2] < img_width - 1
+                           ? boxes[box_idx + 2]
+                           : static_cast<T>(img_width - 1);
+  boxes[box_idx + 3] = boxes[box_idx + 3] < img_height - 1
+                           ? boxes[box_idx + 3]
+                           : static_cast<T>(img_height - 1);
+}
+
+template <typename T>
+HOSTDEVICE inline void CalcLabelScore(T* scores, const T* input,
+                                      const int label_idx, const int score_idx,
+                                      const int class_num, const T conf,
+                                      const int stride) {
+  for (int i = 0; i < class_num; i++) {
+    scores[score_idx + i] = conf * sigmoid<T>(input[label_idx + i * stride]);
+  }
+}
+
+template <typename T>
+class YoloBoxKernel : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* input = ctx.Input<Tensor>("X");
+    auto* imgsize = ctx.Input<Tensor>("ImgSize");
+    auto* boxes = ctx.Output<Tensor>("Boxes");
+    auto* scores = ctx.Output<Tensor>("Scores");
+    auto anchors = ctx.Attr<std::vector<int>>("anchors");
+    int class_num = ctx.Attr<int>("class_num");
+    float conf_thresh = ctx.Attr<float>("conf_thresh");
+    int downsample_ratio = ctx.Attr<int>("downsample_ratio");
+
+    const int n = input->dims()[0];
+    const int h = input->dims()[2];
+    const int w = input->dims()[3];
+    const int box_num = boxes->dims()[1];
+    const int an_num = anchors.size() / 2;
+    int input_size = downsample_ratio * h;
+
+    const int stride = h * w;
+    const int an_stride = (class_num + 5) * stride;
+
+    Tensor anchors_;
+    auto anchors_data =
+        anchors_.mutable_data<int>({an_num * 2}, ctx.GetPlace());
+    std::copy(anchors.begin(), anchors.end(), anchors_data);
+
+    const T* input_data = input->data<T>();
+    const int* imgsize_data = imgsize->data<int>();
+    T* boxes_data = boxes->mutable_data<T>({n, box_num, 4}, ctx.GetPlace());
+    memset(boxes_data, 0, boxes->numel() * sizeof(T));
+    T* scores_data =
+        scores->mutable_data<T>({n, box_num, class_num}, ctx.GetPlace());
+    memset(scores_data, 0, scores->numel() * sizeof(T));
+
+    T box[4];
+    for (int i = 0; i < n; i++) {
+      int img_height = imgsize_data[2 * i];
+      int img_width = imgsize_data[2 * i + 1];
+
+      for (int j = 0; j < an_num; j++) {
+        for (int k = 0; k < h; k++) {
+          for (int l = 0; l < w; l++) {
+            int obj_idx =
+                GetEntryIndex(i, j, k * w + l, an_num, an_stride, stride, 4);
+            T conf = sigmoid<T>(input_data[obj_idx]);
+            if (conf < conf_thresh) {
+              continue;
+            }
+
+            int box_idx =
+                GetEntryIndex(i, j, k * w + l, an_num, an_stride, stride, 0);
+            GetYoloBox<T>(box, input_data, anchors_data, l, k, j, h, input_size,
+                          box_idx, stride, img_height, img_width);
+            box_idx = (i * box_num + j * stride + k * w + l) * 4;
+            CalcDetectionBox<T>(boxes_data, box, box_idx, img_height,
+                                img_width);
+
+            int label_idx =
+                GetEntryIndex(i, j, k * w + l, an_num, an_stride, stride, 5);
+            int score_idx = (i * box_num + j * stride + k * w + l) * class_num;
+            CalcLabelScore<T>(scores_data, input_data, label_idx, score_idx,
+                              class_num, conf, stride);
+          }
+        }
+      }
+    }
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle

paddle/fluid/operators/detection/yolov3_loss_op.cc

@@ -10,6 +10,7 @@
    limitations under the License. */
 
 #include "paddle/fluid/operators/detection/yolov3_loss_op.h"
+#include <memory>
 #include "paddle/fluid/framework/op_registry.h"
 
 namespace paddle {
@@ -72,6 +73,18 @@ class Yolov3LossOp : public framework::OperatorWithKernel {
     PADDLE_ENFORCE_GT(class_num, 0,
                       "Attr(class_num) should be an integer greater then 0.");
 
+    if (ctx->HasInput("GTScore")) {
+      auto dim_gtscore = ctx->GetInputDim("GTScore");
+      PADDLE_ENFORCE_EQ(dim_gtscore.size(), 2,
+                        "Input(GTScore) should be a 2-D tensor");
+      PADDLE_ENFORCE_EQ(
+          dim_gtscore[0], dim_gtbox[0],
+          "Input(GTBox) and Input(GTScore) dim[0] should be same");
+      PADDLE_ENFORCE_EQ(
+          dim_gtscore[1], dim_gtbox[1],
+          "Input(GTBox) and Input(GTScore) dim[1] should be same");
+    }
+
     std::vector<int64_t> dim_out({dim_x[0]});
     ctx->SetOutputDim("Loss", framework::make_ddim(dim_out));
 
@@ -112,6 +125,12 @@ class Yolov3LossOpMaker : public framework::OpProtoAndCheckerMaker {
              "This is a 2-D tensor with shape of [N, max_box_num], "
              "and each element should be an integer to indicate the "
              "box class id.");
+    AddInput("GTScore",
+             "The score of GTLabel, This is a 2-D tensor in same shape "
+             "GTLabel, and score values should in range (0, 1). This "
+             "input is for GTLabel score can be not 1.0 in image mixup "
+             "augmentation.")
+        .AsDispensable();
     AddOutput("Loss",
               "The output yolov3 loss tensor, "
               "This is a 1-D tensor with shape of [N]");
@@ -143,6 +162,9 @@ class Yolov3LossOpMaker : public framework::OpProtoAndCheckerMaker {
     AddAttr<float>("ignore_thresh",
                    "The ignore threshold to ignore confidence loss.")
         .SetDefault(0.7);
+    AddAttr<bool>("use_label_smooth",
+                  "Whether to use label smooth. Default True.")
+        .SetDefault(true);
     AddComment(R"DOC(
          This operator generates yolov3 loss based on given predict result and ground
          truth boxes.
@@ -204,6 +226,15 @@ class Yolov3LossOpMaker : public framework::OpProtoAndCheckerMaker {
          loss = (loss_{xy} + loss_{wh}) * weight_{box}
               + loss_{conf} + loss_{class}
          $$
+
+         While :attr:`use_label_smooth` is set to be :attr:`True`, the classification
+         target will be smoothed when calculating classification loss, target of 
+         positive samples will be smoothed to :math:`1.0 - 1.0 / class\_num` and target of
+         negetive samples will be smoothed to :math:`1.0 / class\_num`.
+
+         While :attr:`GTScore` is given, which means the mixup score of ground truth 
+         boxes, all losses incured by a ground truth box will be multiplied by its 
+         mixup score.
          )DOC");
   }
 };
@@ -240,6 +271,7 @@ class Yolov3LossGradMaker : public framework::SingleGradOpDescMaker {
     op->SetInput("X", Input("X"));
     op->SetInput("GTBox", Input("GTBox"));
     op->SetInput("GTLabel", Input("GTLabel"));
+    op->SetInput("GTScore", Input("GTScore"));
     op->SetInput(framework::GradVarName("Loss"), OutputGrad("Loss"));
     op->SetInput("ObjectnessMask", Output("ObjectnessMask"));
     op->SetInput("GTMatchMask", Output("GTMatchMask"));
@@ -249,6 +281,7 @@ class Yolov3LossGradMaker : public framework::SingleGradOpDescMaker {
     op->SetOutput(framework::GradVarName("X"), InputGrad("X"));
     op->SetOutput(framework::GradVarName("GTBox"), {});
     op->SetOutput(framework::GradVarName("GTLabel"), {});
+    op->SetOutput(framework::GradVarName("GTScore"), {});
     return std::unique_ptr<framework::OpDesc>(op);
   }
 };

paddle/fluid/operators/detection/yolov3_loss_op.h

@@ -37,8 +37,8 @@ static T SigmoidCrossEntropy(T x, T label) {
 }
 
 template <typename T>
-static T L2Loss(T x, T y) {
-  return 0.5 * (y - x) * (y - x);
+static T L1Loss(T x, T y) {
+  return std::abs(y - x);
 }
 
 template <typename T>
@@ -47,8 +47,8 @@ static T SigmoidCrossEntropyGrad(T x, T label) {
 }
 
 template <typename T>
-static T L2LossGrad(T x, T y) {
-  return x - y;
+static T L1LossGrad(T x, T y) {
+  return x > y ? 1.0 : -1.0;
 }
 
 static int GetMaskIndex(std::vector<int> mask, int val) {
@@ -121,47 +121,49 @@ template <typename T>
 static void CalcBoxLocationLoss(T* loss, const T* input, Box<T> gt,
                                 std::vector<int> anchors, int an_idx,
                                 int box_idx, int gi, int gj, int grid_size,
-                                int input_size, int stride) {
+                                int input_size, int stride, T score) {
   T tx = gt.x * grid_size - gi;
   T ty = gt.y * grid_size - gj;
   T tw = std::log(gt.w * input_size / anchors[2 * an_idx]);
   T th = std::log(gt.h * input_size / anchors[2 * an_idx + 1]);
 
-  T scale = (2.0 - gt.w * gt.h);
+  T scale = (2.0 - gt.w * gt.h) * score;
   loss[0] += SigmoidCrossEntropy<T>(input[box_idx], tx) * scale;
   loss[0] += SigmoidCrossEntropy<T>(input[box_idx + stride], ty) * scale;
-  loss[0] += L2Loss<T>(input[box_idx + 2 * stride], tw) * scale;
-  loss[0] += L2Loss<T>(input[box_idx + 3 * stride], th) * scale;
+  loss[0] += L1Loss<T>(input[box_idx + 2 * stride], tw) * scale;
+  loss[0] += L1Loss<T>(input[box_idx + 3 * stride], th) * scale;
 }
 
 template <typename T>
 static void CalcBoxLocationLossGrad(T* input_grad, const T loss, const T* input,
                                     Box<T> gt, std::vector<int> anchors,
                                     int an_idx, int box_idx, int gi, int gj,
-                                    int grid_size, int input_size, int stride) {
+                                    int grid_size, int input_size, int stride,
+                                    T score) {
   T tx = gt.x * grid_size - gi;
   T ty = gt.y * grid_size - gj;
   T tw = std::log(gt.w * input_size / anchors[2 * an_idx]);
   T th = std::log(gt.h * input_size / anchors[2 * an_idx + 1]);
 
-  T scale = (2.0 - gt.w * gt.h);
+  T scale = (2.0 - gt.w * gt.h) * score;
   input_grad[box_idx] =
       SigmoidCrossEntropyGrad<T>(input[box_idx], tx) * scale * loss;
   input_grad[box_idx + stride] =
       SigmoidCrossEntropyGrad<T>(input[box_idx + stride], ty) * scale * loss;
   input_grad[box_idx + 2 * stride] =
-      L2LossGrad<T>(input[box_idx + 2 * stride], tw) * scale * loss;
+      L1LossGrad<T>(input[box_idx + 2 * stride], tw) * scale * loss;
   input_grad[box_idx + 3 * stride] =
-      L2LossGrad<T>(input[box_idx + 3 * stride], th) * scale * loss;
+      L1LossGrad<T>(input[box_idx + 3 * stride], th) * scale * loss;
 }
 
 template <typename T>
 static inline void CalcLabelLoss(T* loss, const T* input, const int index,
                                  const int label, const int class_num,
-                                 const int stride) {
+                                 const int stride, const T pos, const T neg,
+                                 T score) {
   for (int i = 0; i < class_num; i++) {
     T pred = input[index + i * stride];
-    loss[0] += SigmoidCrossEntropy<T>(pred, (i == label) ? 1.0 : 0.0);
+    loss[0] += SigmoidCrossEntropy<T>(pred, (i == label) ? pos : neg) * score;
   }
 }
 
@@ -169,11 +171,13 @@ template <typename T>
 static inline void CalcLabelLossGrad(T* input_grad, const T loss,
                                      const T* input, const int index,
                                      const int label, const int class_num,
-                                     const int stride) {
+                                     const int stride, const T pos, const T neg,
+                                     T score) {
   for (int i = 0; i < class_num; i++) {
     T pred = input[index + i * stride];
     input_grad[index + i * stride] =
-        SigmoidCrossEntropyGrad<T>(pred, (i == label) ? 1.0 : 0.0) * loss;
+        SigmoidCrossEntropyGrad<T>(pred, (i == label) ? pos : neg) * score *
+        loss;
   }
 }
 
@@ -188,8 +192,8 @@ static inline void CalcObjnessLoss(T* loss, const T* input, const T* objness,
         for (int l = 0; l < w; l++) {
           T obj = objness[k * w + l];
           if (obj > 1e-5) {
-            // positive sample: obj = 1
-            loss[i] += SigmoidCrossEntropy<T>(input[k * w + l], 1.0);
+            // positive sample: obj = mixup score
+            loss[i] += SigmoidCrossEntropy<T>(input[k * w + l], 1.0) * obj;
           } else if (obj > -0.5) {
             // negetive sample: obj = 0
             loss[i] += SigmoidCrossEntropy<T>(input[k * w + l], 0.0);
@@ -215,7 +219,8 @@ static inline void CalcObjnessLossGrad(T* input_grad, const T* loss,
           T obj = objness[k * w + l];
           if (obj > 1e-5) {
             input_grad[k * w + l] =
-                SigmoidCrossEntropyGrad<T>(input[k * w + l], 1.0) * loss[i];
+                SigmoidCrossEntropyGrad<T>(input[k * w + l], 1.0) * obj *
+                loss[i];
           } else if (obj > -0.5) {
             input_grad[k * w + l] =
                 SigmoidCrossEntropyGrad<T>(input[k * w + l], 0.0) * loss[i];
@@ -252,6 +257,7 @@ class Yolov3LossKernel : public framework::OpKernel<T> {
     auto* input = ctx.Input<Tensor>("X");
     auto* gt_box = ctx.Input<Tensor>("GTBox");
     auto* gt_label = ctx.Input<Tensor>("GTLabel");
+    auto* gt_score = ctx.Input<Tensor>("GTScore");
     auto* loss = ctx.Output<Tensor>("Loss");
     auto* objness_mask = ctx.Output<Tensor>("ObjectnessMask");
     auto* gt_match_mask = ctx.Output<Tensor>("GTMatchMask");
@@ -260,6 +266,7 @@ class Yolov3LossKernel : public framework::OpKernel<T> {
     int class_num = ctx.Attr<int>("class_num");
     float ignore_thresh = ctx.Attr<float>("ignore_thresh");
     int downsample_ratio = ctx.Attr<int>("downsample_ratio");
+    bool use_label_smooth = ctx.Attr<bool>("use_label_smooth");
 
     const int n = input->dims()[0];
     const int h = input->dims()[2];
@@ -272,6 +279,13 @@ class Yolov3LossKernel : public framework::OpKernel<T> {
     const int stride = h * w;
     const int an_stride = (class_num + 5) * stride;
 
+    T label_pos = 1.0;
+    T label_neg = 0.0;
+    if (use_label_smooth) {
+      label_pos = 1.0 - 1.0 / static_cast<T>(class_num);
+      label_neg = 1.0 / static_cast<T>(class_num);
+    }
+
     const T* input_data = input->data<T>();
     const T* gt_box_data = gt_box->data<T>();
     const int* gt_label_data = gt_label->data<int>();
@@ -283,6 +297,19 @@ class Yolov3LossKernel : public framework::OpKernel<T> {
     int* gt_match_mask_data =
         gt_match_mask->mutable_data<int>({n, b}, ctx.GetPlace());
 
+    const T* gt_score_data;
+    if (!gt_score) {
+      Tensor gtscore;
+      gtscore.mutable_data<T>({n, b}, ctx.GetPlace());
+      math::SetConstant<platform::CPUDeviceContext, T>()(
+          ctx.template device_context<platform::CPUDeviceContext>(), &gtscore,
+          static_cast<T>(1.0));
+      gt_score = &gtscore;
+      gt_score_data = gtscore.data<T>();
+    } else {
+      gt_score_data = gt_score->data<T>();
+    }
+
     // calc valid gt box mask, avoid calc duplicately in following code
     Tensor gt_valid_mask;
     bool* gt_valid_mask_data =
@@ -355,19 +382,20 @@ class Yolov3LossKernel : public framework::OpKernel<T> {
         int mask_idx = GetMaskIndex(anchor_mask, best_n);
         gt_match_mask_data[i * b + t] = mask_idx;
         if (mask_idx >= 0) {
+          T score = gt_score_data[i * b + t];
           int box_idx = GetEntryIndex(i, mask_idx, gj * w + gi, mask_num,
                                       an_stride, stride, 0);
           CalcBoxLocationLoss<T>(loss_data + i, input_data, gt, anchors, best_n,
-                                 box_idx, gi, gj, h, input_size, stride);
+                                 box_idx, gi, gj, h, input_size, stride, score);
 
           int obj_idx = (i * mask_num + mask_idx) * stride + gj * w + gi;
-          obj_mask_data[obj_idx] = 1.0;
+          obj_mask_data[obj_idx] = score;
 
           int label = gt_label_data[i * b + t];
           int label_idx = GetEntryIndex(i, mask_idx, gj * w + gi, mask_num,
                                         an_stride, stride, 5);
           CalcLabelLoss<T>(loss_data + i, input_data, label_idx, label,
-                           class_num, stride);
+                           class_num, stride, label_pos, label_neg, score);
         }
       }
     }
@@ -384,6 +412,7 @@ class Yolov3LossGradKernel : public framework::OpKernel<T> {
     auto* input = ctx.Input<Tensor>("X");
     auto* gt_box = ctx.Input<Tensor>("GTBox");
     auto* gt_label = ctx.Input<Tensor>("GTLabel");
+    auto* gt_score = ctx.Input<Tensor>("GTScore");
     auto* input_grad = ctx.Output<Tensor>(framework::GradVarName("X"));
     auto* loss_grad = ctx.Input<Tensor>(framework::GradVarName("Loss"));
     auto* objness_mask = ctx.Input<Tensor>("ObjectnessMask");
@@ -392,6 +421,7 @@ class Yolov3LossGradKernel : public framework::OpKernel<T> {
     auto anchor_mask = ctx.Attr<std::vector<int>>("anchor_mask");
     int class_num = ctx.Attr<int>("class_num");
     int downsample_ratio = ctx.Attr<int>("downsample_ratio");
+    bool use_label_smooth = ctx.Attr<bool>("use_label_smooth");
 
     const int n = input_grad->dims()[0];
     const int c = input_grad->dims()[1];
@@ -404,6 +434,13 @@ class Yolov3LossGradKernel : public framework::OpKernel<T> {
     const int stride = h * w;
     const int an_stride = (class_num + 5) * stride;
 
+    T label_pos = 1.0;
+    T label_neg = 0.0;
+    if (use_label_smooth) {
+      label_pos = 1.0 - 1.0 / static_cast<T>(class_num);
+      label_neg = 1.0 / static_cast<T>(class_num);
+    }
+
     const T* input_data = input->data<T>();
     const T* gt_box_data = gt_box->data<T>();
     const int* gt_label_data = gt_label->data<int>();
@@ -414,25 +451,41 @@ class Yolov3LossGradKernel : public framework::OpKernel<T> {
         input_grad->mutable_data<T>({n, c, h, w}, ctx.GetPlace());
     memset(input_grad_data, 0, input_grad->numel() * sizeof(T));
 
+    const T* gt_score_data;
+    if (!gt_score) {
+      Tensor gtscore;
+      gtscore.mutable_data<T>({n, b}, ctx.GetPlace());
+      math::SetConstant<platform::CPUDeviceContext, T>()(
+          ctx.template device_context<platform::CPUDeviceContext>(), &gtscore,
+          static_cast<T>(1.0));
+      gt_score = &gtscore;
+      gt_score_data = gtscore.data<T>();
+    } else {
+      gt_score_data = gt_score->data<T>();
+    }
+
     for (int i = 0; i < n; i++) {
       for (int t = 0; t < b; t++) {
         int mask_idx = gt_match_mask_data[i * b + t];
         if (mask_idx >= 0) {
+          T score = gt_score_data[i * b + t];
           Box<T> gt = GetGtBox(gt_box_data, i, b, t);
           int gi = static_cast<int>(gt.x * w);
           int gj = static_cast<int>(gt.y * h);
 
           int box_idx = GetEntryIndex(i, mask_idx, gj * w + gi, mask_num,
                                       an_stride, stride, 0);
-          CalcBoxLocationLossGrad<T>(
-              input_grad_data, loss_grad_data[i], input_data, gt, anchors,
-              anchor_mask[mask_idx], box_idx, gi, gj, h, input_size, stride);
+          CalcBoxLocationLossGrad<T>(input_grad_data, loss_grad_data[i],
+                                     input_data, gt, anchors,
+                                     anchor_mask[mask_idx], box_idx, gi, gj, h,
+                                     input_size, stride, score);
 
           int label = gt_label_data[i * b + t];
           int label_idx = GetEntryIndex(i, mask_idx, gj * w + gi, mask_num,
                                         an_stride, stride, 5);
           CalcLabelLossGrad<T>(input_grad_data, loss_grad_data[i], input_data,
-                               label_idx, label, class_num, stride);
+                               label_idx, label, class_num, stride, label_pos,
+                               label_neg, score);
         }
       }
     }

paddle/fluid/operators/expand_op.cc

@@ -13,6 +13,7 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #include "paddle/fluid/operators/expand_op.h"
+#include <memory>
 #include <vector>
 
 namespace paddle {
@@ -138,12 +139,28 @@ class ExpandGradOp : public framework::OperatorWithKernel {
   }
 };
 
+class ExpandGradOpDescMaker : public framework::SingleGradOpDescMaker {
+ public:
+  using framework::SingleGradOpDescMaker::SingleGradOpDescMaker;
+
+ protected:
+  std::unique_ptr<framework::OpDesc> Apply() const override {
+    std::unique_ptr<framework::OpDesc> op(new framework::OpDesc());
+    op->SetType("expand_grad");
+    op->SetInput("X", Input("X"));
+    op->SetInput(framework::GradVarName("Out"), OutputGrad("Out"));
+    op->SetOutput(framework::GradVarName("X"), InputGrad("X"));
+    op->SetAttrMap(Attrs());
+    return op;
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle
 
 namespace ops = paddle::operators;
 REGISTER_OPERATOR(expand, ops::ExpandOp, ops::ExpandOpMaker,
-                  paddle::framework::DefaultGradOpDescMaker<true>);
+                  ops::ExpandGradOpDescMaker);
 REGISTER_OPERATOR(expand_grad, ops::ExpandGradOp);
 REGISTER_OP_CPU_KERNEL(
     expand, ops::ExpandKernel<paddle::platform::CPUDeviceContext, float>,

paddle/fluid/operators/fake_quantize_op.cc

@@ -81,6 +81,30 @@ struct FindRangeAbsMaxFunctor<platform::CPUDeviceContext, T> {
 
 template struct FindRangeAbsMaxFunctor<platform::CPUDeviceContext, float>;
 
+template <typename T>
+struct FindMovingAverageAbsMaxFunctor<platform::CPUDeviceContext, T> {
+  void operator()(const platform::CPUDeviceContext& ctx,
+                  const framework::Tensor& in_accum,
+                  const framework::Tensor& in_state, const T* cur_scale,
+                  const float rate, framework::Tensor* out_state,
+                  framework::Tensor* out_accum, framework::Tensor* out_scale) {
+    T accum = in_accum.data<T>()[0];
+    T state = in_state.data<T>()[0];
+    T scale = cur_scale[0];
+
+    state = rate * state + 1;
+    accum = rate * accum + scale;
+    scale = accum / state;
+
+    out_state->mutable_data<T>(ctx.GetPlace())[0] = state;
+    out_accum->mutable_data<T>(ctx.GetPlace())[0] = accum;
+    out_scale->mutable_data<T>(ctx.GetPlace())[0] = scale;
+  }
+};
+
+template struct FindMovingAverageAbsMaxFunctor<platform::CPUDeviceContext,
+                                               float>;
+
 class FakeQuantizeAbsMaxOp : public framework::OperatorWithKernel {
  public:
   FakeQuantizeAbsMaxOp(const std::string& type,
@@ -255,6 +279,78 @@ $$Out = round(X/scale * range)$$
   }
 };
 
+class FakeQuantizeMovingAverageAbsMaxOp : public framework::OperatorWithKernel {
+ public:
+  FakeQuantizeMovingAverageAbsMaxOp(const std::string& type,
+                                    const framework::VariableNameMap& inputs,
+                                    const framework::VariableNameMap& outputs,
+                                    const framework::AttributeMap& attrs)
+      : OperatorWithKernel(type, inputs, outputs, attrs) {}
+
+  void InferShape(framework::InferShapeContext* ctx) const override {
+    PADDLE_ENFORCE(
+        ctx->HasInput("X"),
+        "Input(X) of FakeQuantizeMovingAverageAbsMaxOp should not be null.");
+    PADDLE_ENFORCE(
+        ctx->HasOutput("Out"),
+        "Output(Out) of FakeQuantizeMovingAverageAbsMaxOp should not be null.");
+    PADDLE_ENFORCE(ctx->HasOutput("OutScale"),
+                   "Output(OutScale) of FakeQuantizeMovingAverageAbsMaxOp "
+                   "should not be null");
+    if (ctx->HasOutput("OutState")) {
+      ctx->SetOutputDim("OutState", {1});
+    }
+    if (ctx->HasOutput("OutAccum")) {
+      ctx->SetOutputDim("OutAccum", {1});
+    }
+    ctx->SetOutputDim("Out", ctx->GetInputDim("X"));
+    ctx->SetOutputDim("OutScale", {1});
+    ctx->ShareLoD("X", /*->*/ "Out");
+  }
+
+ protected:
+  framework::OpKernelType GetExpectedKernelType(
+      const framework::ExecutionContext& ctx) const override {
+    return framework::OpKernelType(ctx.Input<framework::LoDTensor>("X")->type(),
+                                   ctx.device_context());
+  }
+};
+
+class FakeQuantizeMovingAverageAbsMaxOpMaker
+    : public framework::OpProtoAndCheckerMaker {
+ public:
+  void Make() override {
+    AddInput("X", "(Tensor) Input is float data type.");
+    AddInput("InScale", "Last scale.");
+    AddInput("InAccum", "Last accum.").AsDispensable();
+    AddInput("InState", "Last state.").AsDispensable();
+    AddOutput("Out", "(Tensor) Output of quantized low level tensor.");
+    AddOutput("OutScale", " Current scale");
+    AddOutput("OutState", "(Tensor) state buffer.").AsDispensable();
+    AddOutput("OutAccum", "(Tensor) accum buffer.").AsDispensable();
+    AddAttr<float>("moving_rate", "(float, default 0.9) moving rate.")
+        .SetDefault(0.9);
+    AddAttr<int>("bit_length", "(int, default 8), quantization bit number.")
+        .SetDefault(8)
+        .AddCustomChecker([](const int& bit_length) {
+          PADDLE_ENFORCE(bit_length >= 1 && bit_length <= 16,
+                         "'bit_length' should be between 1 and 16.");
+        });
+    AddAttr<bool>("is_test",
+                  "(bool, default false) Set to true for inference only, false "
+                  "for training. Some layers may run faster when this is true.")
+        .SetDefault(false);
+    AddComment(R"DOC(
+FakeQuantize operator is used in static quantization.
+
+$$scale = (0.9*max(abs(x))+accum)/(0.9*state+1)$$
+$$range = 2^{bit_length - 1} - 1$$
+$$Out = round(X/scale * range)$$
+
+)DOC");
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle
 
@@ -273,6 +369,12 @@ REGISTER_OPERATOR(fake_quantize_range_abs_max, ops::FakeQuantizeRangeAbsMaxOp,
 REGISTER_OP_CPU_KERNEL(fake_quantize_range_abs_max,
                        ops::FakeQuantizeRangeAbsMaxKernel<CPU, float>);
 
+REGISTER_OPERATOR(fake_quantize_moving_average_abs_max,
+                  ops::FakeQuantizeMovingAverageAbsMaxOp,
+                  ops::FakeQuantizeMovingAverageAbsMaxOpMaker,
+                  paddle::framework::EmptyGradOpMaker);
+REGISTER_OP_CPU_KERNEL(fake_quantize_moving_average_abs_max,
+                       ops::FakeQuantizeMovingAverageAbsMaxKernel<CPU, float>);
 REGISTER_OPERATOR(fake_channel_wise_quantize_abs_max,
                   ops::FakeChannelWiseQuantizeAbsMaxOp,
                   ops::FakeChannelWiseQuantizeAbsMaxOpMaker,

paddle/fluid/operators/fake_quantize_op.cu

@@ -147,6 +147,41 @@ struct FindRangeAbsMaxFunctor<platform::CUDADeviceContext, T> {
 
 template struct FindRangeAbsMaxFunctor<platform::CUDADeviceContext, float>;
 
+template <typename T>
+struct FindMovingAverageAbsMaxFunctor<platform::CUDADeviceContext, T> {
+  void operator()(const platform::CUDADeviceContext& ctx,
+                  const framework::Tensor& in_accum,
+                  const framework::Tensor& in_state, const T* cur_scale,
+                  const float rate, framework::Tensor* out_state,
+                  framework::Tensor* out_accum, framework::Tensor* out_scale) {
+    const auto gpu_place = boost::get<platform::CUDAPlace>(ctx.GetPlace());
+
+    T accum;
+    memory::Copy(platform::CPUPlace(), &accum, gpu_place, in_accum.data<T>(),
+                 sizeof(T), 0);
+    T state;
+    memory::Copy(platform::CPUPlace(), &state, gpu_place, in_state.data<T>(),
+                 sizeof(T), 0);
+    T scale;
+    memory::Copy(platform::CPUPlace(), &scale, gpu_place, cur_scale, sizeof(T),
+                 0);
+
+    state = rate * state + 1;
+    accum = rate * accum + scale;
+    scale = accum / state;
+
+    memory::Copy(gpu_place, out_accum->mutable_data<T>(gpu_place),
+                 platform::CPUPlace(), &accum, sizeof(T), 0);
+    memory::Copy(gpu_place, out_state->mutable_data<T>(gpu_place),
+                 platform::CPUPlace(), &state, sizeof(T), 0);
+    memory::Copy(gpu_place, out_scale->mutable_data<T>(gpu_place),
+                 platform::CPUPlace(), &scale, sizeof(T), 0);
+  }
+};
+
+template struct FindMovingAverageAbsMaxFunctor<platform::CUDADeviceContext,
+                                               float>;
+
 template <typename T>
 struct ClipAndFakeQuantFunctor<platform::CUDADeviceContext, T> {
   void operator()(const platform::CUDADeviceContext& ctx,
@@ -178,3 +213,6 @@ REGISTER_OP_CUDA_KERNEL(fake_channel_wise_quantize_abs_max,
                         ops::FakeChannelWiseQuantizeAbsMaxKernel<CUDA, float>);
 REGISTER_OP_CUDA_KERNEL(fake_quantize_range_abs_max,
                         ops::FakeQuantizeRangeAbsMaxKernel<CUDA, float>);
+REGISTER_OP_CUDA_KERNEL(
+    fake_quantize_moving_average_abs_max,
+    ops::FakeQuantizeMovingAverageAbsMaxKernel<CUDA, float>);

paddle/fluid/operators/fake_quantize_op.h

@@ -42,12 +42,20 @@ struct FindRangeAbsMaxFunctor {
                   framework::Tensor* scales_arr, framework::Tensor* out_scale);
 };
 
+template <typename DeviceContext, typename T>
+struct FindMovingAverageAbsMaxFunctor {
+  void operator()(const DeviceContext& ctx, const framework::Tensor& in_accum,
+                  const framework::Tensor& in_state,
+                  const framework::Tensor& cur_scale,
+                  framework::Tensor* out_state, framework::Tensor* out_accum,
+                  framework::Tensor* out_scale);
+};
+
 template <typename DeviceContext, typename T>
 class FakeQuantizeAbsMaxKernel : public framework::OpKernel<T> {
  public:
   void Compute(const framework::ExecutionContext& context) const override {
     auto* in = context.Input<framework::Tensor>("X");
-
     auto* out = context.Output<framework::Tensor>("Out");
     auto* out_scale = context.Output<framework::Tensor>("OutScale");
     T* out_s = out_scale->mutable_data<T>(context.GetPlace());
@@ -138,5 +146,54 @@ class FakeQuantizeRangeAbsMaxKernel : public framework::OpKernel<T> {
   }
 };
 
+template <typename DeviceContext, typename T>
+class FakeQuantizeMovingAverageAbsMaxKernel : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& context) const override {
+    auto* in = context.Input<framework::Tensor>("X");
+    auto* in_scale = context.Input<framework::Tensor>("InScale");
+    auto* out = context.Output<framework::Tensor>("Out");
+    out->mutable_data<T>(context.GetPlace());
+
+    bool is_test = context.Attr<bool>("is_test");
+    int bit_length = context.Attr<int>("bit_length");
+    int bin_cnt = std::pow(2, bit_length - 1) - 1;
+    auto& dev_ctx = context.template device_context<DeviceContext>();
+
+    // testing
+    if (is_test) {
+      ClipAndFakeQuantFunctor<DeviceContext, T>()(dev_ctx, *in, *in_scale,
+                                                  bin_cnt, out);
+      return;
+    }
+
+    // training
+    auto* in_accum = context.Input<framework::Tensor>("InAccum");
+    auto* in_state = context.Input<framework::Tensor>("InState");
+    auto& allocator =
+        platform::DeviceTemporaryAllocator::Instance().Get(dev_ctx);
+    auto cur_scale = allocator.Allocate(1 * sizeof(T));
+    T* cur_scale_data = static_cast<T*>(cur_scale->ptr());
+
+    FindAbsMaxFunctor<DeviceContext, T>()(dev_ctx, in->data<T>(), in->numel(),
+                                          cur_scale_data);
+
+    auto* out_state = context.Output<framework::Tensor>("OutState");
+    auto* out_accum = context.Output<framework::Tensor>("OutAccum");
+    auto* out_scale = context.Output<framework::Tensor>("OutScale");
+    out_state->mutable_data<T>(context.GetPlace());
+    out_accum->mutable_data<T>(context.GetPlace());
+    out_scale->mutable_data<T>(context.GetPlace());
+    float moving_rate = context.Attr<float>("moving_rate");
+
+    FindMovingAverageAbsMaxFunctor<DeviceContext, T>()(
+        dev_ctx, *in_accum, *in_state, cur_scale_data, moving_rate, out_state,
+        out_accum, out_scale);
+
+    ClipAndFakeQuantFunctor<DeviceContext, T>()(dev_ctx, *in, *out_scale,
+                                                bin_cnt, out);
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle

paddle/fluid/operators/fc_op.cc

@@ -55,17 +55,8 @@ void FCOp::InferShape(framework::InferShapeContext* ctx) const {
       "The input tensor Input's rank of FCOp should be larger than "
       "in_num_col_dims.");
 
-  auto in_mat_dims = framework::flatten_to_2d(in_dims, in_num_col_dims);
-  PADDLE_ENFORCE_EQ(
-      in_mat_dims[1], w_dims[0],
-      "Fully Connected input and weigth size do not match. %s, %s");
-
   std::vector<int64_t> output_dims;
-  output_dims.reserve(static_cast<size_t>(in_num_col_dims + 1));
-  for (int i = 0; i < in_num_col_dims; ++i) {
-    output_dims.push_back(in_dims[i]);
-  }
-  output_dims.push_back(w_dims[1]);
+  FCOutputSize(in_dims, w_dims, output_dims, in_num_col_dims);
 
   ctx->SetOutputDim("Out", framework::make_ddim(output_dims));
   ctx->ShareLoD("Input", "Out");
@@ -128,6 +119,9 @@ void FCOpMaker::Make() {
   AddAttr<bool>("use_mkldnn",
                 "(bool, default false) Only used in mkldnn kernel")
       .SetDefault(false);
+  AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape,
+                "Skip calling InferShape() function in the runtime.")
+      .SetDefault(true);
   AddComment(R"DOC(
   Fully Connected Operator.
 
@@ -142,13 +136,20 @@ class FCOpKernel : public framework::OpKernel<T> {
   void Compute(const paddle::framework::ExecutionContext& ctx) const override {
     PADDLE_ENFORCE(platform::is_cpu_place(ctx.GetPlace()),
                    "It must use CPUPlace.");
-    auto input = ctx.Input<Tensor>("Input");
+    auto input = ctx.Input<framework::LoDTensor>("Input");
     auto w = ctx.Input<Tensor>("W");
     auto bias = ctx.Input<Tensor>("Bias");
-    auto output = ctx.Output<Tensor>("Out");
+    auto output = ctx.Output<framework::LoDTensor>("Out");
+    int in_num_col_dims = ctx.Attr<int>("in_num_col_dims");
     auto w_dims = w->dims();
+
+    std::vector<int64_t> output_dims;
+    FCOutputSize(input->dims(), w_dims, output_dims, in_num_col_dims);
+    output->Resize(framework::make_ddim(output_dims));
+    output->set_lod(input->lod());
+
     auto out_dims = output->dims();
-    int M = framework::product(out_dims) / out_dims[out_dims.size() - 1];
+    int M = framework::product(out_dims) / w_dims[1];
 
     const T* input_data = input->data<T>();
     const T* w_data = w->data<T>();

paddle/fluid/operators/fc_op.h

@@ -48,5 +48,21 @@ class FCOpMaker : public framework::OpProtoAndCheckerMaker {
   void Make() override;
 };
 
+inline void FCOutputSize(const framework::DDim& in_dims,
+                         const framework::DDim& w_dims,
+                         std::vector<int64_t>& out_dims,  // NOLINT
+                         int in_num_col_dims) {
+  auto in_mat_dims = framework::flatten_to_2d(in_dims, in_num_col_dims);
+  PADDLE_ENFORCE_EQ(
+      in_mat_dims[1], w_dims[0],
+      "Fully Connected input and weigth size do not match. %s, %s");
+
+  out_dims.reserve(static_cast<size_t>(in_num_col_dims + 1));
+  for (int i = 0; i < in_num_col_dims; ++i) {
+    out_dims.push_back(in_dims[i]);
+  }
+  out_dims.push_back(w_dims[1]);
+}
+
 }  // namespace operators
 }  // namespace paddle

paddle/fluid/operators/fused/fused_embedding_seq_pool_op.cc

@@ -88,7 +88,8 @@ class FusedEmbeddingSeqPoolOpMaker : public framework::OpProtoAndCheckerMaker {
                   "(boolean, default false) "
                   "Sparse update.")
         .SetDefault(false);
-    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape, "")
+    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape,
+                  "Skip calling InferShape() function in the runtime.")
         .SetDefault(true);
     AddComment(R"DOC(
 FusedEmbeddingSeqPool Operator.

paddle/fluid/operators/hash_op.cc

@@ -54,7 +54,8 @@ $$Out = scale * X$$
 )DOC");
     AddAttr<int>("num_hash", "").SetDefault(1);
     AddAttr<int>("mod_by", "").SetDefault(100000);
-    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape, "")
+    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape,
+                  "Skip calling InferShape() function in the runtime.")
         .SetDefault(true);
   }
 };

paddle/fluid/operators/math.h

+// 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.
+
+#pragma once
+
+#include "paddle/fluid/platform/float16.h"
+#include "paddle/fluid/platform/hostdevice.h"
+
+#include "math.h"  // NOLINT
+
+namespace paddle {
+namespace operators {
+
+inline HOSTDEVICE platform::float16 real_exp(platform::float16 x) {
+  return static_cast<platform::float16>(::expf(static_cast<float>(x)));
+}
+
+inline HOSTDEVICE float real_exp(float x) { return ::expf(x); }
+
+inline HOSTDEVICE double real_exp(double x) { return ::exp(x); }
+
+inline HOSTDEVICE platform::float16 real_log(platform::float16 x) {
+  return static_cast<platform::float16>(::logf(static_cast<float>(x)));
+}
+
+inline HOSTDEVICE float real_log(float x) { return ::logf(x); }
+
+inline HOSTDEVICE double real_log(double x) { return ::log(x); }
+
+}  // namespace operators
+}  // namespace paddle

paddle/fluid/operators/math/cross_entropy.cu

@@ -12,6 +12,7 @@ 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 "paddle/fluid/operators/math.h"
 #include "paddle/fluid/operators/math/cross_entropy.h"
 #include "paddle/fluid/platform/cuda_device_function.h"
 #include "paddle/fluid/platform/cuda_primitives.h"
@@ -20,17 +21,6 @@ namespace paddle {
 namespace operators {
 namespace math {
 
-namespace {
-
-__device__ __forceinline__ float real_log(float x) { return logf(x); }
-
-__device__ __forceinline__ double real_log(double x) { return log(x); }
-
-__device__ __forceinline__ platform::float16 real_log(
-    const platform::float16& val) {
-  return static_cast<platform::float16>(logf(static_cast<float>(val)));
-}
-
 template <typename T>
 __global__ void CrossEntropyKernel(T* Y, const T* X, const int64_t* label,
                                    const int N, const int D,
@@ -61,7 +51,6 @@ __global__ void SoftCrossEntropyKernel(T* Y, const T* X, const T* label,
     Y[blockIdx.x] = -val;
   }
 }
-}  // namespace
 
 template <typename T>
 class CrossEntropyFunctor<platform::CUDADeviceContext, T> {

paddle/fluid/operators/mkldnn/conv_mkldnn_op.cc

@@ -592,6 +592,7 @@ class ConvMKLDNNOpKernel : public paddle::framework::OpKernel<T> {
           platform::SetDstMemoryHandler<uint8_t>(ctx, output, handler,
                                                  &dst_memory_p);
         } else {
+          need_s8_to_u8 = fuse_relu;
           platform::SetDstMemoryHandler<int8_t>(ctx, output, handler,
                                                 &dst_memory_p);
         }

paddle/fluid/operators/mkldnn/fc_mkldnn_op.cc

@@ -123,7 +123,7 @@ class FCMKLDNNOpKernel : public paddle::framework::OpKernel<T> {
     auto& dev_ctx = ctx.template device_context<MKLDNNDeviceContext>();
     const auto& mkldnn_engine = dev_ctx.GetEngine();
 
-    auto input = ctx.Input<Tensor>("Input");
+    auto input = ctx.Input<framework::LoDTensor>("Input");
     auto w = ctx.Input<Tensor>("W");
     auto bias = ctx.Input<Tensor>("Bias");
 
@@ -151,7 +151,13 @@ class FCMKLDNNOpKernel : public paddle::framework::OpKernel<T> {
     const T* input_data = input->data<T>();
     const T* w_data = w->data<T>();
 
-    auto output = ctx.Output<Tensor>("Out");
+    auto output = ctx.Output<framework::LoDTensor>("Out");
+    int in_num_col_dims = ctx.Attr<int>("in_num_col_dims");
+    std::vector<int64_t> output_dims;
+    FCOutputSize(input->dims(), w->dims(), output_dims, in_num_col_dims);
+    output->Resize(framework::make_ddim(output_dims));
+    output->set_lod(input->lod());
+
     T* output_data = output->mutable_data<T>(ctx.GetPlace());
 
     auto dst_memory = mem.dst(output_data);
@@ -204,19 +210,21 @@ class FCMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
     Tensor* input_grad = ctx.Output<Tensor>(framework::GradVarName("Input"));
     Tensor* w_grad = ctx.Output<Tensor>(framework::GradVarName("W"));
 
+    const Tensor* input = ctx.Input<Tensor>("Input");
+    const T* input_data = input->data<T>();
+
+    const Tensor* w = ctx.Input<Tensor>("W");
+    const T* w_data = w->data<T>();
+
     if (input_grad) {
+      input_grad->Resize(input->dims());
       input_grad_data = input_grad->mutable_data<T>(ctx.GetPlace());
     }
     if (w_grad) {
+      w_grad->Resize(w->dims());
       w_grad_data = w_grad->mutable_data<T>(ctx.GetPlace());
     }
 
-    const Tensor* input = ctx.Input<Tensor>("Input");
-    const T* input_data = input->data<T>();
-
-    const Tensor* w = ctx.Input<Tensor>("W");
-    const T* w_data = w->data<T>();
-
     const Tensor* out_grad = ctx.Input<Tensor>(framework::GradVarName("Out"));
     const T* out_grad_data = out_grad->data<T>();
 

paddle/fluid/operators/mkldnn/transpose_mkldnn_op.cc

@@ -73,6 +73,29 @@ class TransposeMKLDNNOpKernel : public paddle::framework::OpKernel<T> {
   }
 };
 
+template <typename T>
+class TransposeINT8MKLDNNOpKernel : public paddle::framework::OpKernel<T> {
+ public:
+  void Compute(const paddle::framework::ExecutionContext& ctx) const override {
+    std::vector<int> axis = ctx.Attr<std::vector<int>>("axis");
+    std::vector<int> axis_int8 = {0, 2, 3, 1};
+    if (axis.size() != 1) {
+      PADDLE_ENFORCE_EQ(axis.size(), axis_int8.size());
+      for (size_t i = 0; i < axis.size(); i++) {
+        PADDLE_ENFORCE_EQ(axis[i], axis_int8[i],
+                          "Current INT8 MKLDNN Transpose kernel only surpport "
+                          "axis with [0, 2, 3, 1] due to MKL-DNN kernel "
+                          "implementation.");
+      }
+    }
+    auto* input = ctx.Input<Tensor>("X");
+    auto* output = ctx.Output<Tensor>("Out");
+    output->ShareDataWith(*input);
+    output->set_layout(DataLayout::kMKLDNN);
+    output->set_format(input->format());
+  }
+};
+
 template <typename T>
 class TransposeMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
  public:
@@ -140,7 +163,10 @@ class TransposeMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
 namespace ops = paddle::operators;
 
 REGISTER_OP_KERNEL(transpose2, MKLDNN, ::paddle::platform::CPUPlace,
-                   ops::TransposeMKLDNNOpKernel<float>);
+                   ops::TransposeMKLDNNOpKernel<float>,
+                   ops::TransposeINT8MKLDNNOpKernel<uint8_t>,
+                   ops::TransposeINT8MKLDNNOpKernel<int8_t>);
+
 REGISTER_OP_KERNEL(transpose, MKLDNN, ::paddle::platform::CPUPlace,
                    ops::TransposeMKLDNNOpKernel<float>);
 

paddle/fluid/operators/optimizers/adam_op.h

@@ -15,6 +15,7 @@ limitations under the License. */
 #pragma once
 #include <math.h>  // for sqrt in CPU and CUDA
 #include <Eigen/Dense>
+#include <unordered_map>
 #include <vector>
 #include "paddle/fluid/framework/op_registry.h"
 #include "paddle/fluid/framework/threadpool.h"
@@ -311,17 +312,17 @@ struct SparseAdamFunctor<T, CPUAdam> {
     T beta1_pow = *beta1_pow_;
     T beta2_pow = *beta2_pow_;
     lr *= sqrt(1 - beta2_pow) / (1 - beta1_pow);
-    size_t row_count = numel / row_numel_;
+    int64_t row_count = static_cast<int64_t>(numel / row_numel_);
 
-    for (size_t i = 0U, j = 0U; i != row_count; ++i) {
+    for (int64_t i = 0, j = 0; i != row_count; ++i) {
       if (i == *(rows_ + j)) {
-        for (size_t k = 0U; k != row_numel_; ++k) {
+        for (int64_t k = 0; k != row_numel_; ++k) {
           T g = grad_[j * row_numel_ + k];
           adam_update(i * row_numel_ + k, g);
         }
         ++j;
       } else {
-        for (size_t k = 0U; k != row_numel_; ++k) {
+        for (int64_t k = 0; k != row_numel_; ++k) {
           T mom1 = moment1_[i * row_numel_ + k];
           T mom2 = moment2_[i * row_numel_ + k];
           T p = param_[i * row_numel_ + k];
@@ -427,43 +428,23 @@ class AdamOpKernel : public framework::OpKernel<T> {
         }
       }
 
-      framework::SelectedRows cpu_grad_merge;
+      framework::SelectedRows tmp_grad_merge;
       const framework::SelectedRows* grad_merge_ptr;
       if (is_strict_sorted) {
         grad_merge_ptr = &grad;
       } else {
         // merge duplicated rows if any.
         // The rows of grad_merge have been sorted inside MergeAdd functor
-        framework::SelectedRows* grad_merge_var;
         scatter::MergeAdd<DeviceContext, T> merge_func;
-        if (platform::is_cpu_place(ctx.GetPlace())) {
-          grad_merge_var = &cpu_grad_merge;
-        } else {
-          // FIXME(qiao): GPU also need to fix this
-          grad_merge_var = const_cast<framework::Scope&>(ctx.scope())
-                               .Var()
-                               ->GetMutable<framework::SelectedRows>();
-        }
         merge_func(ctx.template device_context<DeviceContext>(), grad,
-                   grad_merge_var, true);
-        grad_merge_ptr = grad_merge_var;
+                   &tmp_grad_merge, true);
+        grad_merge_ptr = &tmp_grad_merge;
       }
 
       auto& grad_merge = *grad_merge_ptr;
       auto& grad_tensor = grad_merge.value();
       const T* grad_data = grad_tensor.template data<T>();
-      const int64_t* rows = nullptr;
-// When compiled without CUDA, the CUDAData() interface should not be
-// provided.
-#if defined(PADDLE_WITH_CUDA)
-      if (platform::is_gpu_place(ctx.GetPlace())) {
-        rows = grad_merge.rows().CUDAData(ctx.GetPlace());
-      } else {
-#endif
-        rows = grad_merge.rows().data();
-#if defined(PADDLE_WITH_CUDA)
-      }
-#endif
+      const int64_t* rows = grad_merge.rows().Data(ctx.GetPlace());
       auto row_numel = grad_tensor.numel() / grad_merge.rows().size();
 
       if (platform::is_cpu_place(ctx.GetPlace())) {
@@ -488,7 +469,7 @@ class AdamOpKernel : public framework::OpKernel<T> {
           }
         }
 #ifndef _WIN32
-        else if (FLAGS_inner_op_parallelism > 1 &&
+        else if (FLAGS_inner_op_parallelism > 1 &&  // NOLINT
                  min_row_size_to_use_multithread > 0 &&
                  param.dims()[0] > min_row_size_to_use_multithread) {
           VLOG(3) << "use multi thread, inner_op_parallelism="
@@ -516,11 +497,11 @@ class AdamOpKernel : public framework::OpKernel<T> {
           for (int i = 0; i < FLAGS_inner_op_parallelism; ++i) {
             int64_t start = i * line_in_each_thread;
             int64_t end = (i + 1) * line_in_each_thread;
-            if (start >= param_row_count) {
+            if (start >= static_cast<int64_t>(param_row_count)) {
               break;
             }
-            if (end > param_row_count) {
-              end = param_row_count;
+            if (end > static_cast<int64_t>(param_row_count)) {
+              end = static_cast<int64_t>(param_row_count);
             }
             fs.push_back(
                 framework::Async([&functor, &row_id_to_grad_row_offset,
@@ -545,8 +526,8 @@ class AdamOpKernel : public framework::OpKernel<T> {
           }
           for (size_t i = 0; i < fs.size(); ++i) fs[i].wait();
         }
-#endif  // !_WIN32
-        else {
+#endif          // !_WIN32
+        else {  // NOLINT
           functor(param.numel());
         }
       } else if (platform::is_gpu_place(ctx.GetPlace())) {

paddle/fluid/operators/optimizers/momentum_op.h

@@ -13,6 +13,7 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #pragma once
+#include <memory>
 #include <string>
 #include "paddle/fluid/framework/eigen.h"
 #include "paddle/fluid/framework/op_registry.h"
@@ -69,6 +70,7 @@ class MomentumOp : public framework::OperatorWithKernel {
     ctx->SetOutputDim("ParamOut", param_dim);
     ctx->SetOutputDim("VelocityOut", param_dim);
   }
+
   framework::OpKernelType GetExpectedKernelType(
       const framework::ExecutionContext& ctx) const override {
     auto input_data_type = framework::GetDataTypeOfVar(ctx.InputVar("Param"));
@@ -351,23 +353,14 @@ class MomentumOpKernel : public framework::OpKernel<T> {
         VLOG(3) << "Grad SelectedRows contains no data!";
         return;
       }
-      auto* merged_grad = const_cast<framework::Scope&>(ctx.scope())
-                              .Var()
-                              ->GetMutable<framework::SelectedRows>();
+
+      framework::SelectedRows tmp_merged_grad;
+      framework::SelectedRows* merged_grad = &tmp_merged_grad;
       math::scatter::MergeAdd<DeviceContext, T> merge_func;
       merge_func(ctx.template device_context<DeviceContext>(), *grad,
                  merged_grad);
 
-      const int64_t* rows = nullptr;
-#ifdef PADDLE_WITH_CUDA
-      if (platform::is_gpu_place(ctx.GetPlace())) {
-        rows = merged_grad->rows().CUDAData(ctx.GetPlace());
-      } else {
-#endif
-        rows = merged_grad->rows().data();
-#ifdef PADDLE_WITH_CUDA
-      }
-#endif
+      const int64_t* rows = merged_grad->rows().Data(ctx.GetPlace());
       int64_t row_numel =
           merged_grad->value().numel() / merged_grad->rows().size();
       platform::ForRange<DeviceContext> for_range(

paddle/fluid/operators/optimizers/rmsprop_op.h

@@ -216,24 +216,14 @@ class RmspropOpKernel : public framework::OpKernel<T> {
       }
     } else if (grad_var->IsType<framework::SelectedRows>()) {
       auto &grad = grad_var->Get<framework::SelectedRows>();
-      auto *merged_grad = const_cast<framework::Scope &>(ctx.scope())
-                              .Var()
-                              ->GetMutable<framework::SelectedRows>();
-
+      framework::SelectedRows tmp_merged_grad;
+      framework::SelectedRows *merged_grad = &tmp_merged_grad;
       math::scatter::MergeAdd<DeviceContext, T> merge_func;
       merge_func(dev_ctx, grad, merged_grad);
 
       platform::ForRange<DeviceContext> for_range(dev_ctx, limit);
-      const int64_t *rows;
-#ifdef PADDLE_WITH_CUDA
-      if (platform::is_gpu_place(ctx.GetPlace())) {
-        rows = merged_grad->rows().CUDAData(ctx.GetPlace());
-      } else {
-#endif
-        rows = merged_grad->rows().data();
-#ifdef PADDLE_WITH_CUDA
-      }
-#endif
+      const int64_t *rows = merged_grad->rows().Data(ctx.GetPlace());
+
       auto &merged_tensor = merged_grad->value();
       int64_t row_count = merged_grad->rows().size();
       int64_t row_numel = merged_tensor.numel() / row_count;

paddle/fluid/operators/pool_op.cc

@@ -13,6 +13,7 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 
 #include "paddle/fluid/operators/pool_op.h"
+#include <unordered_map>
 #ifdef PADDLE_WITH_CUDA
 #include "paddle/fluid/platform/cudnn_helper.h"
 #endif
@@ -212,6 +213,12 @@ void Pool2dOpMaker::Make() {
   AddAttr<bool>("use_mkldnn",
                 "(bool, default false) Only used in mkldnn kernel")
       .SetDefault(false);
+  AddAttr<bool>("use_quantizer",
+                "(bool, default false) "
+                "Set to true for operators that should be quantized and use "
+                "int8 kernel. "
+                "Only used on CPU.")
+      .SetDefault(false);
   AddAttr<std::string>(
       "data_format",
       "(string, default NCHW) Only used in "

paddle/fluid/operators/reshape_op.cc

@@ -219,14 +219,6 @@ class ReshapeKernel {
           std::vector<int>(shape_data, shape_data + shape_tensor->numel());
       out_dims = ReshapeOp::ValidateShape(shape, in->dims());
     }
-    if (!in->lod().empty()) {
-      PADDLE_ENFORCE_EQ(
-          out_dims[0], in->dims()[0],
-          "Reshape operator cannot reshape an input sequence batch "
-          "into an output sequence batch that has a different "
-          "number of time steps. Please consider using "
-          "sequence_reshape op.");
-    }
 
     out->mutable_data(ctx.GetPlace(), in->type());
     framework::TensorCopy(

paddle/fluid/operators/selu_op.h

@@ -15,13 +15,12 @@ limitations under the License. */
 #pragma once
 #include <string>
 #include "paddle/fluid/framework/op_registry.h"
+#include "paddle/fluid/operators/math.h"
 #include "paddle/fluid/platform/for_range.h"
+
 namespace paddle {
 namespace operators {
 
-static HOSTDEVICE float real_exp(float x) { return expf(x); }
-static HOSTDEVICE float real_exp(double x) { return exp(x); }
-
 template <typename T>
 struct SeluFunctor {
   SeluFunctor(const T* x_data_ptr, float alpha, float scale, T* y_data_ptr)

paddle/fluid/operators/sequence_ops/sequence_enumerate_op.cc

@@ -59,7 +59,8 @@ class SequenceEnumerateOpMaker : public framework::OpProtoAndCheckerMaker {
         });
     AddAttr<int>("pad_value", "(int) The enumerate sequence padding value.")
         .SetDefault(0);
-    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape, "")
+    AddAttr<bool>(framework::kAllKernelsMustComputeRuntimeShape,
+                  "Skip calling InferShape() function in the runtime.")
         .SetDefault(true);
     AddComment(R"DOC(
 Sequence Enumerate Operator.

paddle/fluid/operators/sequence_ops/sequence_softmax_op.cu

@@ -14,6 +14,7 @@ limitations under the License. */
 
 #include <algorithm>
 #include <cub/cub.cuh>  // NOLINT
+#include "paddle/fluid/operators/math.h"
 #include "paddle/fluid/operators/sequence_ops/sequence_softmax_op.h"
 
 namespace paddle {
@@ -21,9 +22,6 @@ namespace operators {
 
 using LoDTensor = framework::LoDTensor;
 
-__device__ __forceinline__ float real_exp(float x) { return expf(x); }
-__device__ __forceinline__ double real_exp(double x) { return exp(x); }
-
 template <typename T, int BlockDim>
 using BlockReduce = cub::BlockReduce<T, BlockDim>;
 

paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op.cu

@@ -12,6 +12,7 @@ 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 "cub/cub.cuh"
+#include "paddle/fluid/operators/math.h"
 #include "paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op.h"
 #include "paddle/fluid/platform/cuda_primitives.h"
 #include "paddle/fluid/platform/hostdevice.h"
@@ -21,11 +22,6 @@ namespace operators {
 
 using Tensor = framework::Tensor;
 
-static HOSTDEVICE float real_exp(float x) { return expf(x); }
-static HOSTDEVICE float real_exp(double x) { return exp(x); }
-static HOSTDEVICE float real_log(float x) { return logf(x); }
-static HOSTDEVICE float real_log(double x) { return log(x); }
-
 static constexpr int kNumCUDAThreads = 512;
 static constexpr int kNumMaxinumNumBlocks = 4096;
 

paddle/fluid/operators/slice_op.cu

@@ -12,18 +12,138 @@ 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 <thrust/device_vector.h>
+#include "paddle/fluid/operators/math/math_function.h"
 #include "paddle/fluid/operators/slice_op.h"
+#include "paddle/fluid/platform/cuda_device_function.h"
+#include "paddle/fluid/platform/cuda_primitives.h"
+#include "paddle/fluid/platform/float16.h"
+
+namespace paddle {
+namespace operators {
+
+using platform::PADDLE_CUDA_NUM_THREADS;
+
+template <size_t D>
+__global__ void Padding(const paddle::platform::float16* d_out,
+                        const int* out_dims, const int* in_dims,
+                        const int* offsets, int64_t n,
+                        paddle::platform::float16* d_in) {
+  int64_t out_idx = threadIdx.x + blockDim.x * blockIdx.x;
+  if (out_idx < n) {
+    int coords[D] = {0};
+    for (int i = D - 1; i >= 0; --i) {
+      coords[i] = out_idx % out_dims[i];
+      out_idx /= out_dims[i];
+      coords[i] += offsets[i];
+    }
+
+    int64_t in_idx = 0;
+    for (int i = 0; i < D - 1; ++i) {
+      in_idx += coords[i] * in_dims[i + 1];
+    }
+    in_idx += coords[D - 1];
+
+    d_in[in_idx] = d_out[out_idx];
+  }
+}
+
+template <>
+class SliceGradKernel<paddle::platform::CUDADeviceContext,
+                      paddle::platform::float16>
+    : public framework::OpKernel<paddle::platform::float16> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* d_out = ctx.Input<framework::Tensor>(framework::GradVarName("Out"));
+    auto* d_in = ctx.Output<framework::Tensor>(framework::GradVarName("Input"));
+    d_in->mutable_data<paddle::platform::float16>(ctx.GetPlace());
+
+    auto out_dims = d_out->dims();
+    auto in_dims = d_in->dims();
+    int rank = out_dims.size();
+    std::vector<int> offsets(rank, 0);
+    auto axes = ctx.Attr<std::vector<int>>("axes");
+    auto starts = ctx.Attr<std::vector<int>>("starts");
+
+    for (size_t i = 0; i < starts.size(); ++i) {
+      if (starts[i] < 0) {
+        starts[i] += in_dims[axes[i]];
+      }
+      offsets[axes[i]] = std::max(starts[i], 0);
+    }
+
+    math::SetConstant<paddle::platform::CUDADeviceContext,
+                      paddle::platform::float16>
+        set_zero;
+    auto& dev_ctx =
+        ctx.template device_context<paddle::platform::CUDADeviceContext>();
+    set_zero(dev_ctx, d_in, static_cast<paddle::platform::float16>(0));
+
+    int64_t numel = d_out->numel();
+    dim3 blocks((numel - 1) / PADDLE_CUDA_NUM_THREADS + 1, 1, 1);
+    dim3 threads(PADDLE_CUDA_NUM_THREADS, 1, 1);
+    auto stream = ctx.cuda_device_context().stream();
+
+    auto out_shape = framework::vectorize2int(out_dims);
+    thrust::device_vector<int> out_dims_vec(out_shape.begin(), out_shape.end());
+    auto in_shape = framework::vectorize2int(in_dims);
+    thrust::device_vector<int> in_dims_vec(in_shape.begin(), in_shape.end());
+    thrust::device_vector<int> offsets_vec(offsets.begin(), offsets.end());
+    const int* out_dims_ptr = thrust::raw_pointer_cast(out_dims_vec.data());
+    const int* in_dims_ptr = thrust::raw_pointer_cast(in_dims_vec.data());
+    const int* offsets_ptr = thrust::raw_pointer_cast(offsets_vec.data());
+
+    switch (rank) {
+      case 1:
+        Padding<1><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+      case 2:
+        Padding<2><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+      case 3:
+        Padding<3><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+      case 4:
+        Padding<4><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+      case 5:
+        Padding<5><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+      case 6:
+        Padding<6><<<blocks, threads, 0, stream>>>(
+            d_out->data<paddle::platform::float16>(), out_dims_ptr, in_dims_ptr,
+            offsets_ptr, numel, d_in->data<paddle::platform::float16>());
+        break;
+    }
+  }
+};
+
+}  // namespace operators
+}  // namespace paddle
 
 namespace ops = paddle::operators;
+namespace plat = paddle::platform;
 REGISTER_OP_CUDA_KERNEL(
     slice, ops::SliceKernel<paddle::platform::CUDADeviceContext, float>,
     ops::SliceKernel<paddle::platform::CUDADeviceContext, double>,
     ops::SliceKernel<paddle::platform::CUDADeviceContext, int>,
-    ops::SliceKernel<paddle::platform::CUDADeviceContext, int64_t>);
+    ops::SliceKernel<paddle::platform::CUDADeviceContext, int64_t>,
+    ops::SliceKernel<paddle::platform::CUDADeviceContext, plat::float16>);
 
 REGISTER_OP_CUDA_KERNEL(
     slice_grad,
     ops::SliceGradKernel<paddle::platform::CUDADeviceContext, float>,
     ops::SliceGradKernel<paddle::platform::CUDADeviceContext, double>,
     ops::SliceGradKernel<paddle::platform::CUDADeviceContext, int>,
-    ops::SliceGradKernel<paddle::platform::CUDADeviceContext, int64_t>);
+    ops::SliceGradKernel<paddle::platform::CUDADeviceContext, int64_t>,
+    ops::SliceGradKernel<paddle::platform::CUDADeviceContext, plat::float16>);

paddle/fluid/operators/softmax_with_cross_entropy_op.cu

@@ -439,7 +439,8 @@ class SoftmaxWithCrossEntropyGradCUDAKernel : public framework::OpKernel<T> {
         context.Input<Tensor>(framework::GradVarName("Loss"))->data<T>();
     Tensor* logit_grad =
         context.Output<Tensor>(framework::GradVarName("Logits"));
-    logit_grad->ShareDataWith(*context.Input<Tensor>("Softmax"));
+    framework::TensorCopy(*context.Input<Tensor>("Softmax"), context.GetPlace(),
+                          context.device_context(), logit_grad);
     T* logit_grad_data = logit_grad->data<T>();
 
     const int batch_size = logit_grad->dims()[0];

paddle/fluid/operators/squeeze_op.cc

@@ -94,6 +94,7 @@ class SqueezeOpInferShape : public framework::InferShapeBase {
   }
 };
 
+// TODO(paddle-dev): Should use OpKernel.
 class SqueezeOp : public framework::OperatorBase {
  public:
   using OperatorBase::OperatorBase;

paddle/fluid/operators/sync_batch_norm_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 "paddle/fluid/operators/batch_norm_op.h"
+
+namespace ops = paddle::operators;
+REGISTER_OPERATOR(sync_batch_norm, ops::BatchNormOp, ops::BatchNormOpMaker,
+                  ops::BatchNormOpInferVarType, ops::BatchNormGradMaker);
+REGISTER_OPERATOR(sync_batch_norm_grad, ops::BatchNormGradOp);

paddle/fluid/platform/device_context.cc

@@ -57,7 +57,6 @@ DeviceContextPool::DeviceContextPool(
   for (auto& p : places) {
     set.insert(p);
   }
-
   for (auto& p : set) {
     if (platform::is_cpu_place(p)) {
 #ifdef PADDLE_WITH_MKLDNN
@@ -317,6 +316,9 @@ CUDADeviceContext::~CUDADeviceContext() {
   eigen_stream_.reset();
   eigen_device_.reset();
   PADDLE_ENFORCE(cudaStreamDestroy(stream_));
+#if !defined(_WIN32)
+  PADDLE_ENFORCE(dynload::ncclCommDestroy(nccl_comm_));
+#endif
 }
 
 Place CUDADeviceContext::GetPlace() const { return place_; }

paddle/fluid/platform/device_context.h

@@ -265,6 +265,14 @@ class CUDADeviceContext : public DeviceContext {
   /*! \brief  Return cuda stream in the device context. */
   cudaStream_t stream() const;
 
+#if !defined(_WIN32)
+  /*! \brief  Return nccl communicators. */
+  ncclComm_t nccl_comm() const { return nccl_comm_; }
+
+  /*! \brief  Set nccl communicators. */
+  void set_nccl_comm(ncclComm_t comm) { nccl_comm_ = comm; }
+#endif
+
   template <typename Callback>
   void RecordEvent(cudaEvent_t ev, Callback callback) {
     callback();
@@ -289,6 +297,15 @@ class CUDADeviceContext : public DeviceContext {
   std::unique_ptr<CublasHandleHolder> cublas_handle_;
   std::unique_ptr<CublasHandleHolder> cublas_tensor_core_handle_;
 
+#if !defined(_WIN32)
+  // NCCL communicator (single process version) for NCCL collective operations.
+  // NCCL collective operations provides fast collectives over multiple GPUs
+  // both within and across nodes.
+  // But, this collectives is used for collectives over multiple GPUs within
+  // nodes.
+  ncclComm_t nccl_comm_{nullptr};
+#endif
+
   int compute_capability_;
   int runtime_version_;
   int driver_version_;

paddle/fluid/platform/init.cc

@@ -13,6 +13,8 @@ See the License for the specific language governing permissions and
 limitations under the License. */
 #include <string.h>  // for strdup
 #include <algorithm>
+#include <memory>
+#include <set>
 #include <stdexcept>
 #include <string>
 
@@ -140,6 +142,7 @@ void InitDevices(bool init_p2p, const std::vector<int> devices) {
   places.emplace_back(platform::CPUPlace());
   platform::DeviceContextPool::Init(places);
   platform::DeviceTemporaryAllocator::Init();
+
 #ifndef PADDLE_WITH_MKLDNN
   platform::SetNumThreads(FLAGS_paddle_num_threads);
 #endif

paddle/fluid/platform/nccl_helper.h

@@ -16,9 +16,11 @@
 #pragma once
 
 #include <stdio.h>
+#include <memory>
 #include <string>
 #include <thread>  // NOLINT
 #include <typeindex>
+#include <unordered_map>
 #include <vector>
 #include "paddle/fluid/framework/data_type.h"
 #include "paddle/fluid/platform/dynload/nccl.h"
@@ -78,6 +80,8 @@ struct NCCLContext {
 
   cudaStream_t stream() const { return ctx_->stream(); }
 
+  ncclComm_t comm() const { return comm_; }
+
   int device_id() const {
     return boost::get<platform::CUDAPlace>(ctx_->GetPlace()).device;
   }

paddle/fluid/pybind/CMakeLists.txt

 set(PYBIND_DEPS pybind python proto_desc memory executor async_executor prune
   feed_fetch_method pass_builder parallel_executor profiler layer scope_pool
-  tracer analysis_predictor)
+  tracer analysis_predictor imperative_profiler)
 
 if(WITH_PYTHON)
   list(APPEND PYBIND_DEPS py_func_op)

paddle/fluid/pybind/imperative.cc

@@ -42,6 +42,7 @@ void BindTracer(pybind11::module* m) {
               framework::AttributeMap attrs_map,
               const platform::CPUPlace expected_place,
               const bool stop_gradient = false) {
+             pybind11::gil_scoped_release release;
              return self.Trace(op, inputs, outputs, attrs_map, expected_place,
                                stop_gradient);
            })
@@ -52,6 +53,7 @@ void BindTracer(pybind11::module* m) {
               framework::AttributeMap attrs_map,
               const platform::CUDAPlace expected_place,
               const bool stop_gradient = false) {
+             pybind11::gil_scoped_release release;
              return self.Trace(op, inputs, outputs, attrs_map, expected_place,
                                stop_gradient);
            })

paddle/fluid/pybind/pybind.cc

@@ -36,6 +36,7 @@ limitations under the License. */
 #include "paddle/fluid/framework/selected_rows.h"
 #include "paddle/fluid/framework/version.h"
 #include "paddle/fluid/imperative/layer.h"
+#include "paddle/fluid/imperative/profiler.h"
 #include "paddle/fluid/memory/allocation/allocator_strategy.h"
 #include "paddle/fluid/memory/allocation/legacy_allocator.h"
 #include "paddle/fluid/operators/activation_op.h"
@@ -162,6 +163,11 @@ PYBIND11_MODULE(core, m) {
   m.def("print_mem_usage",
         []() { return memory::allocation::GPUMemMonitor.PrintMemUsage(); });
 
+  m.def("start_imperative_gperf_profiler",
+        []() { imperative::StartProfile(); });
+
+  m.def("stop_imperative_gperf_profiler", []() { imperative::StopProfile(); });
+
   py::class_<imperative::VarBase>(m, "VarBase", R"DOC()DOC")
       .def(
           py::init<const std::string &, paddle::framework::proto::VarType::Type,
@@ -1248,6 +1254,21 @@ All parameter, weight, gradient are variables in Paddle.
                       it will save GPU memory and may make the execution faster.
                       This options is only available in GPU devices.
                       Default False)DOC")
+      .def_property(
+          "sync_batch_norm",
+          [](const BuildStrategy &self) { return self.sync_batch_norm_; },
+          [](BuildStrategy &self, bool b) {
+            PADDLE_ENFORCE(!self.IsFinalized(), "BuildStrategy is finlaized.");
+            self.sync_batch_norm_ = b;
+          },
+          R"DOC(The type is BOOL, sync_batch_norm indicates whether to use
+                synchronous batch normalization which synchronizes the mean
+                and variance through multi-devices in training phase.
+
+                Current implementation doesn't support FP16 training and CPU.
+                And only synchronous on one machine, not all machines.
+
+                Default False)DOC")
       .def_property(
           "memory_optimize",
           [](const BuildStrategy &self) { return self.memory_optimize_; },

python/paddle/fluid/__init__.py

@@ -132,7 +132,8 @@ def __bootstrap__():
         'allocator_strategy', 'reader_queue_speed_test_mode',
         'print_sub_graph_dir', 'pe_profile_fname', 'warpctc_dir',
         'inner_op_parallelism', 'enable_parallel_graph',
-        'multiple_of_cupti_buffer_size', 'enable_subgraph_optimize'
+        'multiple_of_cupti_buffer_size', 'enable_subgraph_optimize',
+        'tracer_profile_fname'
     ]
     if 'Darwin' not in sysstr:
         read_env_flags.append('use_pinned_memory')

python/paddle/fluid/compiler.py

@@ -230,6 +230,9 @@ class CompiledProgram(object):
                 tps), "num_trainers == len(end_points)"
             self._build_strategy.trainers_endpoints = tps
 
+        if self._build_strategy.sync_batch_norm:
+            self._build_strategy.enable_sequential_execution = True
+
         self._persistable_vars = []
         for node in self._graph.nodes():
             if node.is_var() and node.var() is not None and node.var().persistable() and \

python/paddle/fluid/contrib/quantize/quantize_transpiler.py

@@ -84,7 +84,8 @@ class QuantizeTranspiler(object):
                  activation_bits=8,
                  activation_quantize_type='abs_max',
                  weight_quantize_type='abs_max',
-                 window_size=10000):
+                 window_size=10000,
+                 moving_rate=0.9):
         """
         Convert and rewrite the fluid Program according to weight and
         activation quantization type.
@@ -117,23 +118,27 @@ class QuantizeTranspiler(object):
         """
         self.weight_bits = weight_bits
         self.activation_bits = activation_bits
-        quant_type = ['abs_max', 'range_abs_max']
+        quant_type = ['abs_max', 'range_abs_max', 'moving_average_abs_max']
         if weight_quantize_type not in quant_type:
             raise ValueError(
                 "Unknown weight_quantize_type: '%s'. It can only be ",
-                "'abs_max' or 'range_abs_max'.", str(weight_quantize_type))
+                "'abs_max' or 'range_abs_max' or 'moving_average_abs_max'.",
+                str(weight_quantize_type))
         if activation_quantize_type not in quant_type:
             raise ValueError(
                 "Unknown activation_quantize_type : '%s'. It can only be ",
-                "'abs_max' or 'range_abs_max'.", str(activation_quantize_type))
+                "'abs_max' or 'range_abs_max' or 'moving_average_abs_max'.",
+                str(activation_quantize_type))
 
         self.weight_quantize_type = weight_quantize_type
         self.activation_quantize_type = activation_quantize_type
 
         self.window_size = window_size
+        self.moving_rate = moving_rate
         self.helper = LayerHelper(self.__class__.__name__)
         self.fake_quant_op_types = [
-            'fake_quantize_abs_max', 'fake_quantize_range_abs_max'
+            'fake_quantize_abs_max', 'fake_quantize_range_abs_max',
+            'fake_quantize_moving_average_abs_max'
         ]
         self.fake_dequant_op_types = ['fake_dequantize_max_abs']
         self.is_test = None
@@ -168,6 +173,7 @@ class QuantizeTranspiler(object):
             block_id = block.idx
             # insert quant op and dequant op
             for name in op.input_arg_names:
+                #if share input between ops
                 if name in dequanted_vars[block_id]:
                     dequant_var = dequanted_vars[block_id][name]
                 else:
@@ -261,6 +267,7 @@ class QuantizeTranspiler(object):
             max_range = None
             scale_var = None
             for name in op.input_arg_names:
+                #rename input name of the op to the input name of last op which has be removed
                 if name in op_in_rename_map[block_id]:
                     op._rename_input(name, op_in_rename_map[block_id][name])
 
@@ -272,8 +279,7 @@ class QuantizeTranspiler(object):
                     max_range = param_range * act_range / scale_v
                 else:
                     assert isinstance(scale_v, Variable)
-                    scale_var = var_scale_map[block_id][_original_var_name(
-                        name)]
+                    scale_var = scale_v
 
             if len(op.output_arg_names) != 1:
                 raise ValueError("Only support one output, but op %s has"
@@ -309,7 +315,7 @@ class QuantizeTranspiler(object):
                 op_type = op.type
 
                 # insert dequant_op after fc/conv, need to rename
-                # input of the followed ops
+                # input of the followed ops(of fc/conv) to the dquant_op
                 for name in op.input_arg_names:
                     if name in op_out_rename_map[block_id]:
                         op._rename_input(name,
@@ -389,8 +395,8 @@ class QuantizeTranspiler(object):
             for op in block.ops:
                 args += op.input_arg_names
                 args += op.output_arg_names
-            args = list(set(args))
-            var_names = block.vars.keys()
+            args = list(set(args))  #vals of all left ops
+            var_names = block.vars.keys()  # all vals
             sub_block_remove_vars = []
             for var in var_names:
                 if var not in args:
@@ -471,6 +477,61 @@ class QuantizeTranspiler(object):
 
         return quant_var, scale
 
+    def _insert_quant_moving_average_abs_max_op(self, block, idx, var,
+                                                quant_bits):
+        """Insert fake_quantize_moving_average_abs_max
+        """
+        quant_var = block.create_var(
+            name=_quantized_var_name(var.name),
+            type=var.type,
+            shape=var.shape,
+            dtype=var.dtype)
+        state = self.helper.create_global_variable(
+            name=unique_name.generate('state'),
+            persistable=True,
+            dtype=var.dtype,
+            shape=[1])
+        self.helper.set_variable_initializer(
+            state, initializer=Constant(value=1))
+        accum = self.helper.create_global_variable(
+            name=unique_name.generate('accum'),
+            persistable=True,
+            dtype=var.dtype,
+            shape=[1])
+        self.helper.set_variable_initializer(
+            accum, initializer=Constant(value=1))
+        scale = self.helper.create_parameter(
+            attr=ParamAttr(
+                name=_quantized_scale_name(var.name),
+                initializer=Constant(0.001),
+                trainable=False),
+            shape=[1],
+            dtype=var.dtype)
+        scale.stop_gradient = True
+
+        ins = {'X': var, 'InScale': scale}
+        outs = {'Out': quant_var, 'OutScale': scale}
+        if not self.is_test:
+            ins['InState'] = state
+            ins['InAccum'] = accum
+            outs['OutState'] = state
+            outs['OutAccum'] = accum
+
+        attrs = {
+            'bit_length': quant_bits,
+            'moving_rate': self.moving_rate,
+            'is_test': self.is_test
+        }
+
+        quant_op = block._insert_op(
+            idx,
+            type='fake_quantize_moving_average_abs_max',
+            attrs=attrs,
+            inputs=ins,
+            outputs=outs)
+
+        return quant_var, scale
+
     def _insert_quant_op(self, block, idx, var, quant_bits, quant_type):
         """
         Insert fake_quantize_op
@@ -480,6 +541,9 @@ class QuantizeTranspiler(object):
         elif quant_type == 'range_abs_max':
             return self._insert_quant_range_abs_max_op(block, idx, var,
                                                        quant_bits)
+        elif quant_type == 'moving_average_abs_max':
+            return self._insert_quant_moving_average_abs_max_op(block, idx, var,
+                                                                quant_bits)
 
     def _insert_dequant_op(self, block, idx, var, scale, quant_bits):
         """