Fix optimizer in remaining chapters with high level API (#11172)

python/paddle/fluid/tests/book/high-level-api/CMakeLists.txt

@@ -7,10 +7,10 @@ foreach(src ${TEST_OPS})
 endforeach()
 
 add_subdirectory(fit_a_line)
-#add_subdirectory(recognize_digits)
+add_subdirectory(recognize_digits)
-#add_subdirectory(image_classification)
+add_subdirectory(image_classification)
-#add_subdirectory(understand_sentiment)
+add_subdirectory(understand_sentiment)
 add_subdirectory(label_semantic_roles)
-#add_subdirectory(word2vec)
+add_subdirectory(word2vec)
-#add_subdirectory(recommender_system)
+add_subdirectory(recommender_system)
-#add_subdirectory(machine_translation)
+add_subdirectory(machine_translation)

python/paddle/fluid/tests/book/high-level-api/image_classification/test_image_classification_resnet.py

@@ -85,6 +85,10 @@ def train_network():
     return [avg_cost, accuracy]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adam(learning_rate=0.001)
+
+
 def train(use_cuda, train_program, params_dirname):
     BATCH_SIZE = 128
     EPOCH_NUM = 1
@@ -111,9 +115,7 @@ def train(use_cuda, train_program, params_dirname):
 
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
     trainer = fluid.Trainer(
-        train_func=train_program,
+        train_func=train_program, optimizer_func=optimizer_func, place=place)
-        optimizer=fluid.optimizer.Adam(learning_rate=0.001),
-        place=place)
 
     trainer.train(
         reader=train_reader,

python/paddle/fluid/tests/book/high-level-api/image_classification/test_image_classification_vgg.py

@@ -64,6 +64,10 @@ def train_network():
     return [avg_cost, accuracy]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adam(learning_rate=0.001)
+
+
 def train(use_cuda, train_program, params_dirname):
     BATCH_SIZE = 128
     train_reader = paddle.batch(
@@ -88,9 +92,7 @@ def train(use_cuda, train_program, params_dirname):
 
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
     trainer = fluid.Trainer(
-        train_func=train_program,
+        train_func=train_program, place=place, optimizer_func=optimizer_func)
-        place=place,
-        optimizer=fluid.optimizer.Adam(learning_rate=0.001))
 
     trainer.train(
         reader=train_reader,

python/paddle/fluid/tests/book/high-level-api/machine_translation/test_machine_translation.py

@@ -158,6 +158,13 @@ def train_program(is_sparse):
     return avg_cost
 
 
+def optimizer_func():
+    return fluid.optimizer.Adagrad(
+        learning_rate=1e-4,
+        regularization=fluid.regularizer.L2DecayRegularizer(
+            regularization_coeff=0.1))
+
+
 def train(use_cuda, is_sparse, is_local=True):
     EPOCH_NUM = 1
 
@@ -182,11 +189,8 @@ def train(use_cuda, is_sparse, is_local=True):
 
     trainer = fluid.Trainer(
         train_func=partial(train_program, is_sparse),
-        optimizer=fluid.optimizer.Adagrad(
+        place=place,
-            learning_rate=1e-4,
+        optimizer_func=optimizer_func)
-            regularization=fluid.regularizer.L2DecayRegularizer(
-                regularization_coeff=0.1)),
-        place=place)
 
     trainer.train(
         reader=train_reader,

python/paddle/fluid/tests/book/high-level-api/recognize_digits/test_recognize_digits_conv.py

@@ -57,14 +57,17 @@ def train_program():
     return [avg_cost, acc]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adam(learning_rate=0.001)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.Adam(learning_rate=0.001)
 
     trainer = fluid.Trainer(
         train_func=train_program,
         place=place,
-        optimizer=optimizer,
+        optimizer_func=optimizer_func,
         parallel=True)
 
     def event_handler(event):

python/paddle/fluid/tests/book/high-level-api/recognize_digits/test_recognize_digits_mlp.py

@@ -44,12 +44,15 @@ def train_program():
     return [avg_cost, acc]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adam(learning_rate=0.001)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.Adam(learning_rate=0.001)
 
     trainer = fluid.Trainer(
-        train_func=train_program, place=place, optimizer=optimizer)
+        train_func=train_program, place=place, optimizer_func=optimizer_func)
 
     def event_handler(event):
         if isinstance(event, fluid.EndEpochEvent):

python/paddle/fluid/tests/book/high-level-api/recommender_system/test_recommender_system_newapi.py

@@ -155,12 +155,15 @@ def train_program():
     return [avg_cost, scale_infer]
 
 
+def optimizer_func():
+    return fluid.optimizer.SGD(learning_rate=0.2)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.SGD(learning_rate=0.2)
 
     trainer = fluid.Trainer(
-        train_func=train_program, place=place, optimizer=optimizer)
+        train_func=train_program, place=place, optimizer_func=optimizer_func)
 
     feed_order = [
         'user_id', 'gender_id', 'age_id', 'job_id', 'movie_id', 'category_id',

python/paddle/fluid/tests/book/high-level-api/understand_sentiment/test_understand_sentiment_conv.py

@@ -64,15 +64,18 @@ def train_program(word_dict):
     return [avg_cost, accuracy]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adagrad(learning_rate=0.002)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.Adagrad(learning_rate=0.002)
 
     word_dict = paddle.dataset.imdb.word_dict()
     trainer = fluid.Trainer(
         train_func=partial(train_program, word_dict),
         place=place,
-        optimizer=optimizer)
+        optimizer_func=optimizer_func)
 
     def event_handler(event):
         if isinstance(event, fluid.EndEpochEvent):

python/paddle/fluid/tests/book/high-level-api/understand_sentiment/test_understand_sentiment_dynamic_rnn.py

@@ -79,15 +79,18 @@ def train_program(word_dict):
     return [avg_cost, accuracy]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adagrad(learning_rate=0.002)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.Adagrad(learning_rate=0.002)
 
     word_dict = paddle.dataset.imdb.word_dict()
     trainer = fluid.Trainer(
         train_func=partial(train_program, word_dict),
         place=place,
-        optimizer=optimizer)
+        optimizer_func=optimizer_func)
 
     def event_handler(event):
         if isinstance(event, fluid.EndEpochEvent):

python/paddle/fluid/tests/book/high-level-api/understand_sentiment/test_understand_sentiment_stacked_lstm.py

@@ -71,15 +71,18 @@ def train_program(word_dict):
     return [avg_cost, accuracy]
 
 
+def optimizer_func():
+    return fluid.optimizer.Adagrad(learning_rate=0.002)
+
+
 def train(use_cuda, train_program, params_dirname):
     place = fluid.CUDAPlace(0) if use_cuda else fluid.CPUPlace()
-    optimizer = fluid.optimizer.Adagrad(learning_rate=0.002)
 
     word_dict = paddle.dataset.imdb.word_dict()
     trainer = fluid.Trainer(
         train_func=partial(train_program, word_dict),
         place=place,
-        optimizer=optimizer)
+        optimizer_func=optimizer_func)
 
     def event_handler(event):
         if isinstance(event, fluid.EndEpochEvent):

python/paddle/fluid/tests/book/high-level-api/word2vec/test_word2vec_new_api.py

@@ -80,6 +80,10 @@ def train_program(is_sparse):
     return avg_cost
 
 
+def optimizer_func():
+    return fluid.optimizer.SGD(learning_rate=0.001)
+
+
 def train(use_cuda, train_program, params_dirname):
     train_reader = paddle.batch(
         paddle.dataset.imikolov.train(word_dict, N), BATCH_SIZE)
@@ -104,9 +108,7 @@ def train(use_cuda, train_program, params_dirname):
                 sys.exit("got NaN loss, training failed.")
 
     trainer = fluid.Trainer(
-        train_func=train_program,
+        train_func=train_program, optimizer_func=optimizer_func, place=place)
-        optimizer=fluid.optimizer.SGD(learning_rate=0.001),
-        place=place)
 
     trainer.train(
         reader=train_reader,