Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into feature/add_reduce_op_handle

paddle/fluid/framework/details/reduce_op_handle.cc

@@ -91,7 +91,6 @@ void ReduceOpHandle::RunImpl() {
     if (paddle::platform::is_cpu_place(pre_place)) {
       ReduceLoDTensor func(lod_tensors, trg);
       VisitDataType(ToDataType(lod_tensors[0].type()), func);
-
     } else if (paddle::platform::is_gpu_place(pre_place)) {
 #ifdef PADDLE_WITH_CUDA
       auto out_p = out_var_handles[0]->place_;

paddle/fluid/operators/conv_mkldnn_op.cc

@@ -72,10 +72,12 @@ class ConvMKLDNNOpKernel : public paddle::framework::OpKernel<T> {
     auto dst_md = platform::MKLDNNMemDesc(
         dst_tz, mkldnn::memory::data_type::f32, mkldnn::memory::format::nchw);
 
-    auto src_memory = mkldnn::memory({src_md, mkldnn_engine},
-                                     reinterpret_cast<void*>(input_data));
-    auto weights_memory = mkldnn::memory({weights_md, mkldnn_engine},
-                                         reinterpret_cast<void*>(filter_data));
+    auto src_memory =
+        mkldnn::memory({src_md, mkldnn_engine},
+                       reinterpret_cast<void*>(const_cast<T*>(input_data)));
+    auto weights_memory =
+        mkldnn::memory({weights_md, mkldnn_engine},
+                       reinterpret_cast<void*>(const_cast<T*>(filter_data)));
     auto dst_memory = mkldnn::memory({dst_md, mkldnn_engine}, output_data);
 
     std::shared_ptr<mkldnn::convolution_forward::primitive_desc> conv_pd =
@@ -180,9 +182,9 @@ class ConvMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
         dst_tz, mkldnn::memory::data_type::f32, mkldnn::memory::format::nchw);
 
     // create memory
-    auto diff_dst_memory =
-        mkldnn::memory({diff_weights_md, mkldnn_engine},
-                       reinterpret_cast<void*>(output_grad_data));
+    auto diff_dst_memory = mkldnn::memory(
+        {diff_weights_md, mkldnn_engine},
+        reinterpret_cast<void*>(const_cast<T*>(output_grad_data)));
     // Retrieve conv_pd from device context
     auto conv_pd =
         std::static_pointer_cast<mkldnn::convolution_forward::primitive_desc>(
@@ -202,8 +204,9 @@ class ConvMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
       auto diff_weights_memory =
           mkldnn::memory({diff_weights_md, mkldnn_engine},
                          reinterpret_cast<void*>(filter_grad_data));
-      auto src_memory = mkldnn::memory({src_md, mkldnn_engine},
-                                       reinterpret_cast<void*>(input_data));
+      auto src_memory =
+          mkldnn::memory({src_md, mkldnn_engine},
+                         reinterpret_cast<void*>(const_cast<T*>(input_data)));
 
       // create backward conv primitive for weights
       auto conv_bwd_weights_prim = mkldnn::convolution_backward_weights(
@@ -222,11 +225,12 @@ class ConvMKLDNNGradOpKernel : public paddle::framework::OpKernel<T> {
                                    strides, paddings, *conv_pd, mkldnn_engine);
 
       // create memory
-      auto diff_src_memory =
-          mkldnn::memory({diff_src_md, mkldnn_engine},
-                         reinterpret_cast<void*>(input_grad_data));
-      auto weights_memory = mkldnn::memory(
-          {weights_md, mkldnn_engine}, reinterpret_cast<void*>(filter_data));
+      auto diff_src_memory = mkldnn::memory(
+          {diff_src_md, mkldnn_engine},
+          reinterpret_cast<void*>(const_cast<T*>(input_grad_data)));
+      auto weights_memory =
+          mkldnn::memory({weights_md, mkldnn_engine},
+                         reinterpret_cast<void*>(const_cast<T*>(filter_data)));
 
       // create backward conv primitive for data
       auto conv_bwd_data_prim = mkldnn::convolution_backward_data(

paddle/fluid/operators/softmax_mkldnn_op.cc

@@ -73,6 +73,15 @@ class SoftmaxMKLDNNKernel : public paddle::framework::OpKernel<T> {
                                    softmax_dst_memory);
     std::vector<primitive> pipeline{softmax};
     stream(stream::kind::eager).submit(pipeline).wait();
+
+    const bool is_test = ctx.Attr<bool>("is_test");
+    if (!is_test) {
+      T threshold = exp(-64);
+      for (size_t i = 0; i < dst_tz[0] * dst_tz[1]; ++i) {
+        output_data[i] =
+            output_data[i] < threshold ? threshold : output_data[i];
+      }
+    }
   }
 };
 

paddle/fluid/operators/softmax_op.cc

@@ -97,6 +97,9 @@ class SoftmaxOpMaker : public framework::OpProtoAndCheckerMaker {
     AddAttr<bool>("use_mkldnn",
                   "(bool, default false) Only used in mkldnn kernel")
         .SetDefault(false);
+    AddAttr<bool>("is_test",
+                  "Disable epsilon adding to softmax results. Used by MKLDNN.")
+        .SetDefault(false);
     AddComment(R"DOC(
 Softmax Operator.
 

python/paddle/fluid/__init__.py

@@ -37,6 +37,7 @@ from distribute_transpiler import DistributeTranspiler
 from distribute_transpiler_simple import SimpleDistributeTranspiler
 from concurrency import (Go, make_channel, channel_send, channel_recv,
                          channel_close, Select)
+from inference_transpiler import InferenceTranspiler
 import clip
 from memory_optimization_transpiler import memory_optimize, release_memory
 import profiler
@@ -66,6 +67,7 @@ __all__ = framework.__all__ + executor.__all__ + concurrency.__all__ + [
     'clip',
     'SimpleDistributeTranspiler',
     'DistributeTranspiler',
+    'InferenceTranspiler',
     'memory_optimize',
     'release_memory',
     'profiler',

python/paddle/fluid/inference_transpiler.py

+#   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.
+
+import numpy as np
+from framework import Program
+from executor import global_scope
+from . import core
+
+
+class InferenceTranspiler:
+    def transpile(self, program, place, scope=None):
+        '''
+        Transpile the program. Support only fuse batch normalization now.
+
+        :param program: program to transpile 
+        :type program: Program
+        :param place: inference place 
+        :type place: Place
+        :param scope: inference scope 
+        :type scope: Scope or None
+        '''
+        if not isinstance(program, Program):
+            raise TypeError("program should be as Program type")
+        if not isinstance(place, core.CPUPlace) and not isinstance(
+                place, core.CUDAPlace):
+            raise TypeError("place should be as CPUPlace/CUDAPlace type")
+        if scope is None:
+            scope = global_scope()
+        if not isinstance(scope, core.Scope):
+            raise TypeError("scope should be as Scope type or None")
+        self.fuse_batch_norm(program, place, scope)
+
+    def fuse_batch_norm(self, program, place, scope):
+        '''
+        Transpile the program by fused batch normalization.
+ 
+        The batch normalization followed the convolution or fully connected layer 
+        can be integrated with them. Doing so will give us a forward acceleration, 
+        especially in environments like mobile or embedded.
+                    
+        For input X:
+        - Conv process:        X = input * W + bias 
+        - Batch norm process:  X' = (X - mean) / std 
+        - Scale Process:       Y = a * X' + b
+
+        After fuse into one operation:
+
+        Y = (input * W + bias - mean) / std * a + b
+          = input * a * W / std + ((bias - mean) / std * a + b)
+
+        The operator transformation is: 
+        - before:
+          - conv->batch_norm->any_other_op (bias == 0)
+          - conv->elementwise_add->batch_norm->any_other_op (bias != 0)
+        - after: 
+          - conv->elementwise_add->any_other_op
+        
+        The transpile stages are:
+        1. insert elementwise_add op when bias == 0.
+        2. fuse the batch_norm's parameters to conv and elementwise_add operators.
+        3. remove batch_norm ops which are not used in any other ops.
+        4. adjust the input of any_other_op to be the output of elementwise_add operator.
+        5. remove unused variables.
+
+        :param program: program to transpile 
+        :type program: Program
+        :param place: inference place 
+        :type place: Place
+        :param scope: inference scope 
+        :type scope: Scope
+        '''
+        self.scope = scope
+        self.place = place
+        self.block = program.block(0)
+        self.input_map = {}  # store the input names should be adjusted 
+
+        i = 0
+        while i < len(self.block.ops):
+            current_op = self.block.ops[i]
+            # TODO(luotao1): consider only conv2d now. fc would be delt later.
+            if current_op.type in ['conv2d']:
+                # TODO(luotao1): consider single chain network now. 
+                # For branch network, we counldn't use block.ops[i + 1] as 
+                # the judgment condition.
+                next_op = self.block.ops[i + 1]
+                # conv2d without bias
+                if (next_op.type == 'batch_norm'):
+                    # insert bias op
+                    bias_op = self._insert_bias_op(i + 1, current_op, next_op)
+                    # fuse batch_norm
+                    self._fuse_param(current_op, next_op, bias_op, 0)
+                    # remove batch_norm_op
+                    self.block.remove_op(i + 2)
+                    i = i + 1
+                # conv2d with bias, the next_op.type is elementwise_add
+                elif (next_op.type == 'elementwise_add'):
+                    next_next_op = self.block.ops[i + 2]
+                    if (next_next_op.type == 'batch_norm'):
+                        # fuse batch_norm
+                        self._fuse_param(current_op, next_next_op, next_op, 1)
+                        # remove batch_norm_op
+                        self.block.remove_op(i + 2)
+                        i = i + 1
+            i = i + 1
+
+        self._adjust_input()
+        self._remove_unused_var()
+        # TODO(luotao): use clone() method to flush the program.desc in force, 
+        # since some large program.desc will not be flushed immediately. 
+        # And a better solution will be considered later.
+        program = program.clone()
+
+    # ====================== private transpiler functions =====================
+    def _insert_bias_op(self, index, current_op, bn_op):
+        '''
+        Construct elementwise_add operator for adding bias 
+        and insert it into program.
+        
+        :param index: insert location of bias_op
+        :type index: Int
+        :param current_op: current operator (conv or fc)
+        :type current_op: Operator
+        :param bn_op: batch norm operator
+        :type bn_op: Operator
+        :return: bias_op
+        :rtype: Operator
+        '''
+        # The input of bias_op is current_op's output and Bias of bn_op
+        # The output of bias_op is bn_op's output
+        x_var = self.block.var(current_op.output("Output")[0])
+        y_var = self.block.var(bn_op.input("Bias")[0])
+        out_var = self.block.var(bn_op.output("Y")[0])
+
+        bias_op = self.block.insert_op(
+            index,
+            type="elementwise_add",
+            inputs={"X": x_var,
+                    "Y": y_var},
+            outputs={"Out": out_var},
+            attrs={"axis": 1})  # dim_start=1
+        return bias_op
+
+    def _fuse_param(self, current_op, bn_op, bias_op, with_bias):
+        '''
+        fuse the batch_norm_op' parameters to current_op (conv or fc)
+        
+        :param current_op: current operator (conv or fc)
+        :type current_op: Operator
+        :param bn_op: batch norm operator
+        :type bn_op: Operator
+        :param bias_op: elementwise_add operator for adding bias
+        :type bias_op: Operator
+        :param with_bias: If current operator has bias, with_bias = 1; otherwise 0. 
+        :type with_bias: Int
+        '''
+
+        def _update_param(op, old_param_name, new_param):
+            # For the sake of remaining the original variables the same as before,
+            # create new variables in scope to store the new parameters.
+            old_param_name = old_param_name[0]
+            old_var = self.block.vars[old_param_name]
+            new_param_name = old_param_name + '_fuse_bn'
+            new_var = self.block.create_parameter(
+                name=new_param_name.encode('ascii'),
+                type=old_var.type,
+                dtype=old_var.dtype,
+                shape=old_var.shape)
+            op.rename_input(old_param_name, new_param_name)
+            self.scope.var(new_param_name)
+
+            tensor = self.scope.find_var(new_param_name).get_tensor()
+            tensor.set(np.array(new_param), self.place)
+
+        def _load_param(param_name):
+            return np.array(self.scope.find_var(param_name[0]).get_tensor())
+
+        bias_bn = _load_param(bn_op.input("Bias"))  #Bias
+        scale_bn = _load_param(bn_op.input("Scale"))  #Scale
+        mean_bn = _load_param(bn_op.input("Mean"))  #Mean
+        var_bn = _load_param(bn_op.input("Variance"))  #Variance
+
+        # TODO(luotao1): consider only conv2d now. fc would be delt later.
+        current_param = _load_param(current_op.input("Filter"))
+        std_bn = np.float32(np.sqrt(np.add(var_bn, 1e-5)))
+        tmp = np.float32(np.divide(scale_bn, std_bn))
+
+        # add bias of batch_norm_op to conv2d
+        if with_bias:
+            bias = _load_param(bias_op.input("Y"))
+        else:
+            bias = np.zeros(bias_bn.shape)
+        bias = np.float32(
+            np.add(np.multiply(np.subtract(bias, mean_bn), tmp), bias_bn))
+
+        # re-compute weight of conv2d
+        tmp = tmp.reshape(tmp.shape[0], -1)
+        dst_param = current_param.reshape((tmp.shape[0], -1))
+        dst_param = np.float32(np.multiply(dst_param, tmp))
+        dst_param = dst_param.reshape(current_param.shape)
+
+        # update parameters
+        _update_param(current_op, current_op.input("Filter"), dst_param)
+        _update_param(bias_op, bias_op.input("Y"), bias)
+
+        # collect the renamed input
+        self.input_map[bn_op.output("Y")[0]] = bias_op.output("Out")[0]
+
+    def _adjust_input(self):
+        for i in range(len(self.block.ops)):
+            current_op = self.block.ops[i]
+            for input_arg in current_op.input_arg_names:
+                if input_arg in self.input_map:
+                    current_op.rename_input(input_arg,
+                                            self.input_map[input_arg])
+
+    def _remove_unused_var(self):
+        '''
+        remove unused varibles in program
+        '''
+        args = []
+        for i in range(len(self.block.ops)):
+            current_op = self.block.ops[i]
+            args += current_op.input_arg_names
+            args += current_op.output_arg_names
+        args = list(set(args))  # unique the input and output arguments
+
+        for var in self.block.vars.keys():
+            if var not in args:
+                self.block.remove_var(var)

python/paddle/fluid/layers/nn.py

@@ -88,6 +88,7 @@ def fc(input,
        bias_attr=None,
        use_mkldnn=False,
        act=None,
+       is_test=False,
        name=None):
     """
     **Fully Connected Layer**
@@ -134,6 +135,7 @@ def fc(input,
         bias_attr (ParamAttr|list of ParamAttr, default None): The parameter attribute for the bias
             of this layer. If it is set to None, no bias will be added to the output units.
         act (str, default None): Activation to be applied to the output of this layer.
+        is_test(bool): A flag indicating whether execution is in test phase.
         use_mkldnn(bool): Use mkldnn kernel or not, it is valid only when the mkldnn
             library is installed. Default: False
         name (str, default None): The name of this layer.
@@ -177,8 +179,11 @@ def fc(input,
             inputs={"Input": input,
                     "W": w},
             outputs={"Out": tmp},
-            attrs={"use_mkldnn": use_mkldnn,
-                   "bias_attr": bias_attr})
+            attrs={
+                "use_mkldnn": use_mkldnn,
+                "is_test": is_test,
+                "bias_attr": bias_attr
+            })
         return helper.append_activation(tmp)
     else:
         for input_var, param_attr in helper.iter_inputs_and_params():

python/paddle/fluid/tests/book/test_image_classification.py

@@ -22,10 +22,17 @@ import sys
 import numpy
 import unittest
 import os
+import numpy as np
 
 
 def resnet_cifar10(input, depth=32):
-    def conv_bn_layer(input, ch_out, filter_size, stride, padding, act='relu'):
+    def conv_bn_layer(input,
+                      ch_out,
+                      filter_size,
+                      stride,
+                      padding,
+                      act='relu',
+                      bias_attr=False):
         tmp = fluid.layers.conv2d(
             input=input,
             filter_size=filter_size,
@@ -33,7 +40,7 @@ def resnet_cifar10(input, depth=32):
             stride=stride,
             padding=padding,
             act=None,
-            bias_attr=False)
+            bias_attr=bias_attr)
         return fluid.layers.batch_norm(input=tmp, act=act)
 
     def shortcut(input, ch_in, ch_out, stride):
@@ -44,7 +51,7 @@ def resnet_cifar10(input, depth=32):
 
     def basicblock(input, ch_in, ch_out, stride):
         tmp = conv_bn_layer(input, ch_out, 3, stride, 1)
-        tmp = conv_bn_layer(tmp, ch_out, 3, 1, 1, act=None)
+        tmp = conv_bn_layer(tmp, ch_out, 3, 1, 1, act=None, bias_attr=True)
         short = shortcut(input, ch_in, ch_out, stride)
         return fluid.layers.elementwise_add(x=tmp, y=short, act='relu')
 
@@ -219,11 +226,26 @@ def infer(use_cuda, save_dirname=None):
         batch_size = 1
         tensor_img = numpy.random.rand(batch_size, 3, 32, 32).astype("float32")
 
+        # Use inference_transpiler to speedup
+        inference_transpiler_program = inference_program.clone()
+        t = fluid.InferenceTranspiler()
+        t.transpile(inference_transpiler_program, place)
+
         # Construct feed as a dictionary of {feed_target_name: feed_target_data}
         # and results will contain a list of data corresponding to fetch_targets.
         results = exe.run(inference_program,
                           feed={feed_target_names[0]: tensor_img},
                           fetch_list=fetch_targets)
+
+        transpiler_results = exe.run(inference_transpiler_program,
+                                     feed={feed_target_names[0]: tensor_img},
+                                     fetch_list=fetch_targets)
+
+        assert len(results[0]) == len(transpiler_results[0])
+        for i in range(len(results[0])):
+            np.testing.assert_almost_equal(
+                results[0][i], transpiler_results[0][i], decimal=6)
+
         print("infer results: ", results[0])