diff --git a/fluid/DeepASR/tools/profile.py b/fluid/DeepASR/tools/profile.py
index 9d0b47694273345357726ad51062d01ff01b120d..49884844215b2eeec91143dc187e8336cc5f3129 100644
--- a/fluid/DeepASR/tools/profile.py
+++ b/fluid/DeepASR/tools/profile.py
@@ -49,7 +49,7 @@ def parse_args():
     parser.add_argument(
         '--learning_rate',
         type=float,
-        default=0.002,
+        default=0.00016,
         help='Learning rate used to train. (default: %(default)f)')
     parser.add_argument(
         '--device',
@@ -125,8 +125,7 @@ def profile(args):
         class_num=1749,
         parallel=args.parallel)
 
-    optimizer = fluid.optimizer.Momentum(
-        learning_rate=args.learning_rate, momentum=0.9)
+    optimizer = fluid.optimizer.Adam(learning_rate=args.learning_rate)
     optimizer.minimize(avg_cost)
 
     place = fluid.CPUPlace() if args.device == 'CPU' else fluid.CUDAPlace(0)
diff --git a/fluid/DeepASR/train.py b/fluid/DeepASR/train.py
index b5d2239e94dd5ddcd79d0245e4b980d6cf5bfbf4..1876927fa7dcc57811125b8936f184ebc0f87e07 100644
--- a/fluid/DeepASR/train.py
+++ b/fluid/DeepASR/train.py
@@ -58,7 +58,7 @@ def parse_args():
     parser.add_argument(
         '--learning_rate',
         type=float,
-        default=0.002,
+        default=0.00016,
         help='Learning rate used to train. (default: %(default)f)')
     parser.add_argument(
         '--device',
@@ -143,8 +143,7 @@ def train(args):
         class_num=1749,
         parallel=args.parallel)
 
-    optimizer = fluid.optimizer.Momentum(
-        learning_rate=args.learning_rate, momentum=0.9)
+    optimizer = fluid.optimizer.Adam(learning_rate=args.learning_rate)
     optimizer.minimize(avg_cost)
 
     # program for test