diff --git a/paddle/gserver/dataproviders/PyDataProvider2.cpp b/paddle/gserver/dataproviders/PyDataProvider2.cpp
index e3e4457f9b72c5edb8082fdf378ae662b4aee42f..b4215bb307cc31ce64bb724986b88fdc20bbbf45 100644
--- a/paddle/gserver/dataproviders/PyDataProvider2.cpp
+++ b/paddle/gserver/dataproviders/PyDataProvider2.cpp
@@ -390,9 +390,7 @@ private:
 
       if (this->loadThread_) {  // wait poolActualSize < poolSize;
         std::unique_lock<std::mutex> l(mtx_);
-        pushCV_.wait(l, [this, additionalBatchSize] {
-          return this->poolActualSize_ < poolSize_;
-        });
+        pushCV_.wait(l, [this] { return this->poolActualSize_ < poolSize_; });
       }
 
       {
diff --git a/paddle/gserver/gradientmachines/MultiGradientMachine.cpp b/paddle/gserver/gradientmachines/MultiGradientMachine.cpp
index 3f46cc98cdef17d14c253c732814bcba005fd667..b8d4d28f0f309a5f7348605e8d35e160e7fd5552 100644
--- a/paddle/gserver/gradientmachines/MultiGradientMachine.cpp
+++ b/paddle/gserver/gradientmachines/MultiGradientMachine.cpp
@@ -52,7 +52,7 @@ MultiGradientMachine::MultiGradientMachine(const ModelConfig& config,
   } else {
     numDevices_ = 0;
   }
-  ParamInitCallback mainParamInitCb = [this](int paramId, Parameter* para) {
+  ParamInitCallback mainParamInitCb = [](int paramId, Parameter* para) {
     // only create buf for CPU parameters
     // GPU parameters will be created in each thread
     if (para->useGpu()) return;
diff --git a/paddle/gserver/layers/RecurrentLayerGroup.cpp b/paddle/gserver/layers/RecurrentLayerGroup.cpp
index 27e8b5868e6d85cf004945d7cb086d6d57487f9f..44b57185c5a5fa7703ca477b990a73cdad2c2aa1 100644
--- a/paddle/gserver/layers/RecurrentLayerGroup.cpp
+++ b/paddle/gserver/layers/RecurrentLayerGroup.cpp
@@ -72,7 +72,7 @@ void RecurrentLayerGroup::initSubNetwork(
   setNeedGradient(true);
 
   network_.reset(new RecurrentGradientMachine(config_.name(), rootNetwork));
-  ParamInitCallback cb = [this, rootNetwork](int paramId, Parameter* para) {
+  ParamInitCallback cb = [rootNetwork](int paramId, Parameter* para) {
     para->enableSharedType(
         PARAMETER_VALUE,
         rootNetwork->getParameters()[paramId]->getBuf(PARAMETER_VALUE),
diff --git a/paddle/parameter/Argument.cpp b/paddle/parameter/Argument.cpp
index cfdaf8998b04e0307bc442dec0df734452634c67..94522f718a0c19bfc704ca92eddef5c5a9cb6919 100644
--- a/paddle/parameter/Argument.cpp
+++ b/paddle/parameter/Argument.cpp
@@ -325,12 +325,12 @@ void Argument::concat(const std::vector<Argument>& args,
         ->copyFrom(*src->subVec(srcStartRow, size), stream);
   };
 
-  auto copyStrs = [batchSize, stream](SVectorPtr& dst,
-                                      const SVectorPtr& src,
-                                      int desStartRow,
-                                      int srcStartRow,
-                                      int size,
-                                      bool useGpu) {
+  auto copyStrs = [batchSize](SVectorPtr& dst,
+                              const SVectorPtr& src,
+                              int desStartRow,
+                              int srcStartRow,
+                              int size,
+                              bool useGpu) {
     if (!src) {
       dst.reset();
       return;
@@ -413,7 +413,7 @@ void Argument::concat(const std::vector<Argument>& args,
     dst->subVec(startRow, src->getSize())->copyFrom(*src, stream);
   };
 
-  auto copyStrs = [batchSize, stream](
+  auto copyStrs = [batchSize](
       SVectorPtr& dst, const SVectorPtr& src, int startRow, bool useGpu) {
     if (!src) {
       dst.reset();
diff --git a/paddle/parameter/AverageOptimizer.cpp b/paddle/parameter/AverageOptimizer.cpp
index 75998d81dd9c8be35fe45e903dc1cd69068f83c6..82a7fed6c6451b8908851f2d039f17b9dc513818 100644
--- a/paddle/parameter/AverageOptimizer.cpp
+++ b/paddle/parameter/AverageOptimizer.cpp
@@ -81,9 +81,9 @@ ParameterOptimizer::TraverseCallback AverageOptimizer::needSpecialTraversal(
   if (numUpdates_ % kMaxNumAccumulates == 0) {
     // Move the sum to a different buffer to avoid loss of precision
     // due to too many sums.
-    callbacks.emplace_back([this](const VectorPtr vecs[],
-                                  const ParameterConfig& config,
-                                  size_t sparseId) {
+    callbacks.emplace_back([](const VectorPtr vecs[],
+                              const ParameterConfig& config,
+                              size_t sparseId) {
       vecs[PARAMETER_SUM2]->add(*vecs[PARAMETER_SUM1]);
       vecs[PARAMETER_SUM1]->zeroMem();
     });
@@ -94,9 +94,9 @@ ParameterOptimizer::TraverseCallback AverageOptimizer::needSpecialTraversal(
     if (auto callback = this->startCatchUpWith()) {
       callbacks.emplace_back(callback);
     }
-    callbacks.emplace_back([this](const VectorPtr vecs[],
-                                  const ParameterConfig& config,
-                                  size_t sparseId) {
+    callbacks.emplace_back([](const VectorPtr vecs[],
+                              const ParameterConfig& config,
+                              size_t sparseId) {
       vecs[PARAMETER_SUM3]->add(*vecs[PARAMETER_SUM1], *vecs[PARAMETER_SUM2]);
       vecs[PARAMETER_SUM1]->zeroMem();
       vecs[PARAMETER_SUM2]->zeroMem();
diff --git a/paddle/parameter/FirstOrderOptimizer.cpp b/paddle/parameter/FirstOrderOptimizer.cpp
index 5e280bcac3389179181d2eda58c08e579e867ecc..182e833405e8f8bc3a4c9ffddbf628040f9cceaa 100644
--- a/paddle/parameter/FirstOrderOptimizer.cpp
+++ b/paddle/parameter/FirstOrderOptimizer.cpp
@@ -145,9 +145,9 @@ AdagradParameterOptimizer::needSpecialTraversal(
   if (numUpdates_ % kMaxNumAccumulates == 0) {
     // Move the sum to a different buffer to avoid loss of precision
     // due to too many sums.
-    return [this](const VectorPtr vecs[],
-                  const ParameterConfig& config,
-                  size_t sparseId) {
+    return [](const VectorPtr vecs[],
+              const ParameterConfig& config,
+              size_t sparseId) {
       vecs[PARAMETER_GRADIENT_SQURESUM]->add(
           *vecs[PARAMETER_GRADIENT_SQURESUM1]);
       vecs[PARAMETER_GRADIENT_SQURESUM1]->zeroMem();