add ParallelDo example for benchmark/fluid

benchmark/fluid/fluid_benchmark.py

@@ -69,6 +69,11 @@ def parse_args():
         type=int,
         default=1,
         help='If gpus > 1, will use ParallelExecutor to run, else use Executor.')
+    parser.add_argument(
+        '--cpus',
+        type=int,
+        default=1,
+        help='If cpus > 1, will use ParallelDo to run, else use Executor.')
     parser.add_argument(
         '--data_set',
         type=str,

benchmark/fluid/models/mnist.py

@@ -69,15 +69,30 @@ def get_model(args):
     images = fluid.layers.data(name='pixel', shape=[1, 28, 28], dtype=DTYPE)
     label = fluid.layers.data(name='label', shape=[1], dtype='int64')
 
+    if args.device == 'CPU' and args.cpus > 1:
+        places = fluid.layers.get_places(args.cpus)
+        pd = fluid.layers.ParallelDo(places)
+        with pd.do():
+            predict = cnn_model(pd.read_input(images))
+            label = pd.read_input(label)
+            cost = fluid.layers.cross_entropy(input=predict, label=label)
+            avg_cost = fluid.layers.mean(x=cost)
+            batch_acc = fluid.layers.accuracy(input=predict, label=label)
+
+            pd.write_output(avg_cost)
+            pd.write_output(batch_acc)
+
+        avg_cost, batch_acc = pd()
+        avg_cost = fluid.layers.mean(avg_cost)
+        batch_acc = fluid.layers.mean(batch_acc)
+    else:
         # Train program
         predict = cnn_model(images)
         cost = fluid.layers.cross_entropy(input=predict, label=label)
         avg_cost = fluid.layers.mean(x=cost)
 
         # Evaluator
-    batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
-    batch_acc = fluid.layers.accuracy(
-        input=predict, label=label, total=batch_size_tensor)
+        batch_acc = fluid.layers.accuracy(input=predict, label=label)
 
     # inference program
     inference_program = fluid.default_main_program().clone()

benchmark/fluid/models/resnet.py

@@ -132,18 +132,33 @@ def get_model(args):
 
     input = fluid.layers.data(name='data', shape=dshape, dtype='float32')
     label = fluid.layers.data(name='label', shape=[1], dtype='int64')
-    predict = model(input, class_dim)
+
+    if args.device == 'CPU' and args.cpus > 1:
+        places = fluid.layers.get_places(args.cpus)
+        pd = fluid.layers.ParallelDo(places)
+        with pd.do():
+            predict = model(pd.read_input(input), class_dim)
+            label = pd.read_input(label)
             cost = fluid.layers.cross_entropy(input=predict, label=label)
             avg_cost = fluid.layers.mean(x=cost)
+            batch_acc = fluid.layers.accuracy(input=predict, label=label)
 
-    batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
-    batch_acc = fluid.layers.accuracy(
-        input=predict, label=label, total=batch_size_tensor)
+            pd.write_output(avg_cost)
+            pd.write_output(batch_acc)
+
+        avg_cost, batch_acc = pd()
+        avg_cost = fluid.layers.mean(avg_cost)
+        batch_acc = fluid.layers.mean(batch_acc)
+    else:
+        predict = model(input, class_dim)
+        cost = fluid.layers.cross_entropy(input=predict, label=label)
+        avg_cost = fluid.layers.mean(x=cost)
+        batch_acc = fluid.layers.accuracy(input=predict, label=label)
 
     inference_program = fluid.default_main_program().clone()
     with fluid.program_guard(inference_program):
         inference_program = fluid.io.get_inference_program(
-            target_vars=[batch_acc, batch_size_tensor])
+            target_vars=[batch_acc])
 
     optimizer = fluid.optimizer.Momentum(learning_rate=0.01, momentum=0.9)
 

benchmark/fluid/models/stacked_dynamic_lstm.py

@@ -101,9 +101,8 @@ def get_model(args):
     loss = fluid.layers.mean(x=loss)
 
     # add acc
-    batch_size_tensor = fluid.layers.create_tensor(dtype='int64')
     batch_acc = fluid.layers.accuracy(input=logit, label=fluid.layers.data(name='label', \
-                shape=[1], dtype='int64'), total=batch_size_tensor)
+                shape=[1], dtype='int64'))
 
     inference_program = fluid.default_main_program().clone()
     with fluid.program_guard(inference_program):