【cherrypick】support cuda11 for heterps; add profiler in oneps (#32957)

* cherrypick for #32640 :add profile and fix dataset hang in heterps;test=develop * cherrypick for #32640 :add profile and fix dataset hang in heterps;test=develop * cherrypick for #32640 :add profile and fix dataset hang in heterps;test=develop
PaddlePaddle · May 19, 2021 · ab1a4df · ab1a4df
1 parent 4639f5d
commit ab1a4df
Show file tree

Hide file tree

Showing 6 changed files with 65 additions and 4 deletions.
diff --git a/paddle/fluid/framework/device_worker.h b/paddle/fluid/framework/device_worker.h
@@ -266,6 +266,9 @@ class HogwildWorker : public CPUWorkerBase {
   HogwildWorkerParameter param_;
   std::vector<std::string> skip_ops_;
   std::map<std::string, int> stat_var_name_map_;
+#ifdef PADDLE_WITH_HETERPS
+  platform::DeviceContext* dev_ctx_ = nullptr;
+#endif
 };
 
 class DownpourWorker : public HogwildWorker {

diff --git a/paddle/fluid/framework/hogwild_worker.cc b/paddle/fluid/framework/hogwild_worker.cc
@@ -39,6 +39,9 @@ void HogwildWorker::Initialize(const TrainerDesc &desc) {
   for (int i = 0; i < param_.stat_var_names_size(); ++i) {
     stat_var_name_map_[param_.stat_var_names(i)] = 1;
   }
+#ifdef PADDLE_WITH_HETERPS
+  dev_ctx_ = platform::DeviceContextPool::Instance().Get(place_);
+#endif
 }
 
 void HogwildWorker::CreateThreadOperators(const ProgramDesc &program) {
@@ -150,6 +153,9 @@ void HogwildWorker::TrainFilesWithProfiler() {
       VLOG(3) << "Going to run op " << op_name[i];
       if (!need_skip) {
         ops_[i]->Run(*thread_scope_, place_);
+#ifdef PADDLE_WITH_HETERPS
+        dev_ctx_->Wait();
+#endif
       }
       VLOG(3) << "Op " << op_name[i] << " Finished";
       timeline.Pause();
@@ -167,6 +173,16 @@ void HogwildWorker::TrainFilesWithProfiler() {
     total_inst += cur_batch;
     ++batch_cnt;
     PrintFetchVars();
+#ifdef PADDLE_WITH_HETERPS
+    dev_ctx_->Wait();
+    VLOG(1) << "GpuPs worker " << thread_id_ << " train cost " << total_time
+            << " seconds, ins_num: " << total_inst;
+    for (size_t i = 0; i < op_name.size(); ++i) {
+      VLOG(1) << "card:" << thread_id_ << ", op: " << op_name[i]
+              << ", mean time: " << op_total_time[i] / total_inst
+              << "s, totol time:" << op_total_time[i] << "sec";
+    }
+#else
     if (thread_id_ == 0) {
       if (batch_cnt > 0 && batch_cnt % 100 == 0) {
         for (size_t i = 0; i < ops_.size(); ++i) {
@@ -178,6 +194,7 @@ void HogwildWorker::TrainFilesWithProfiler() {
         fprintf(stderr, "%6.2f instances/s\n", total_inst / total_time);
       }
     }
+#endif
     thread_scope_->DropKids();
     timeline.Start();
   }
@@ -195,7 +212,10 @@ void HogwildWorker::TrainFilesWithProfiler() {
 
 void HogwildWorker::TrainFiles() {
   platform::SetNumThreads(1);
+  platform::Timer timeline;
+  timeline.Start();
 
+  int total_ins_num = 0;
   // how to accumulate fetched values here
   device_reader_->Start();
   int cur_batch;
@@ -213,9 +233,13 @@ void HogwildWorker::TrainFiles() {
       }
     }
 
+    total_ins_num += cur_batch;
     PrintFetchVars();
     thread_scope_->DropKids();
   }
+  timeline.Pause();
+  VLOG(3) << "worker " << thread_id_ << " train cost " << timeline.ElapsedSec()
+          << " seconds, ins_num: " << total_ins_num;
 #if defined PADDLE_WITH_PSCORE
   if (thread_barrier_) {
     paddle::distributed::Communicator::GetInstance()->BarrierTriggerDecrement();

diff --git a/python/paddle/distributed/fleet/dataset/dataset.py b/python/paddle/distributed/fleet/dataset/dataset.py
@@ -31,6 +31,7 @@ def __init__(self):
         self.dataset = core.Dataset("MultiSlotDataset")
         self.thread_num = 1
         self.filelist = []
+        self.use_ps_gpu = False
 
     def init(self,
              batch_size=1,
@@ -212,6 +213,14 @@ def _prepare_to_run(self):
         self.dataset.set_data_feed_desc(self._desc())
         self.dataset.create_readers()
 
+    def _set_use_ps_gpu(self, use_ps_gpu):
+        """
+        set use_ps_gpu flag
+        Args:
+            use_ps_gpu: bool
+        """
+        self.use_ps_gpu = use_ps_gpu
+
     def _finish_to_run(self):
         self.dataset.destroy_readers()
 
@@ -529,12 +538,18 @@ def _prepare_to_run(self):
 
     def _dynamic_adjust_before_train(self, thread_num):
         if not self.is_user_set_queue_num:
-            self.dataset.dynamic_adjust_channel_num(thread_num, False)
+            if self.use_ps_gpu:
+                self.dataset.dynamic_adjust_channel_num(thread_num, True)
+            else:
+                self.dataset.dynamic_adjust_channel_num(thread_num, False)
         self.dataset.dynamic_adjust_readers_num(thread_num)
 
     def _dynamic_adjust_after_train(self):
         if not self.is_user_set_queue_num:
-            self.dataset.dynamic_adjust_channel_num(self.thread_num, False)
+            if self.use_ps_gpu:
+                self.dataset.dynamic_adjust_channel_num(self.thread_num, True)
+            else:
+                self.dataset.dynamic_adjust_channel_num(self.thread_num, False)
         self.dataset.dynamic_adjust_readers_num(self.thread_num)
 
     def _set_queue_num(self, queue_num):

diff --git a/python/paddle/fluid/dataset.py b/python/paddle/fluid/dataset.py
@@ -74,6 +74,7 @@ def __init__(self):
         self.dataset = core.Dataset("MultiSlotDataset")
         self.thread_num = 1
         self.filelist = []
+        self.use_ps_gpu = False
 
     def set_pipe_command(self, pipe_command):
         """
@@ -300,6 +301,14 @@ def _prepare_to_run(self):
         self.dataset.set_data_feed_desc(self.desc())
         self.dataset.create_readers()
 
+    def _set_use_ps_gpu(self, use_ps_gpu):
+        """
+        set use_ps_gpu flag
+        Args:
+            use_ps_gpu: bool
+        """
+        self.use_ps_gpu = use_ps_gpu
+
     def _finish_to_run(self):
         self.dataset.destroy_readers()
 
@@ -391,7 +400,10 @@ def _prepare_to_run(self):
     )
     def _dynamic_adjust_before_train(self, thread_num):
         if not self.is_user_set_queue_num:
-            self.dataset.dynamic_adjust_channel_num(thread_num, False)
+            if self.use_ps_gpu:
+                self.dataset.dynamic_adjust_channel_num(thread_num, True)
+            else:
+                self.dataset.dynamic_adjust_channel_num(thread_num, False)
         self.dataset.dynamic_adjust_readers_num(thread_num)
 
     @deprecated(
@@ -400,7 +412,10 @@ def _dynamic_adjust_before_train(self, thread_num):
     )
     def _dynamic_adjust_after_train(self):
         if not self.is_user_set_queue_num:
-            self.dataset.dynamic_adjust_channel_num(self.thread_num, False)
+            if self.use_ps_gpu:
+                self.dataset.dynamic_adjust_channel_num(self.thread_num, True)
+            else:
+                self.dataset.dynamic_adjust_channel_num(self.thread_num, False)
         self.dataset.dynamic_adjust_readers_num(self.thread_num)
 
     @deprecated(

diff --git a/python/paddle/fluid/executor.py b/python/paddle/fluid/executor.py
@@ -1507,6 +1507,9 @@ def _run_from_dataset(self,
         trainer._gen_trainer_desc()
 
         self._dump_debug_info(program=program, trainer=trainer)
+        # in case of calling _set_use_ps_gpu explicitly
+        if dataset.use_ps_gpu is False:
+            dataset._set_use_ps_gpu(trainer.proto_desc.use_ps_gpu)
         dataset._dynamic_adjust_before_train(trainer.proto_desc.thread_num)
 
         trainer_instance = self._default_executor.init_for_dataset(

diff --git a/python/paddle/fluid/tests/unittests/test_communicator_ps_gpu.py b/python/paddle/fluid/tests/unittests/test_communicator_ps_gpu.py
@@ -73,6 +73,7 @@ def test_communicator_ps_gpu(self):
         dataset.init(
             batch_size=32, thread_num=1, pipe_command="cat", use_var=slots_vars)
         dataset.set_filelist(["test_communicator_ps_gpu.txt"])
+        dataset._set_use_ps_gpu(1)
         dataset.load_into_memory()
 
         os.environ["TEST_MODE"] = "1"