fix api doc: Optimizer.ModelAverage (#17395)

paddle/fluid/API.spec

@@ -500,13 +500,13 @@ paddle.fluid.optimizer.AdadeltaOptimizer.backward (ArgSpec(args=['self', 'loss',
 paddle.fluid.optimizer.AdadeltaOptimizer.get_opti_var_name_list (ArgSpec(args=['self'], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.optimizer.AdadeltaOptimizer.minimize (ArgSpec(args=['self', 'loss', 'startup_program', 'parameter_list', 'no_grad_set', 'grad_clip'], varargs=None, keywords=None, defaults=(None, None, None, None)), ('document', 'b15cffad0903fc81af77a0580ceb2a9b'))
 paddle.fluid.optimizer.ModelAverage.__init__ (ArgSpec(args=['self', 'average_window_rate', 'min_average_window', 'max_average_window', 'regularization', 'name'], varargs=None, keywords=None, defaults=(10000, 10000, None, None)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
-paddle.fluid.optimizer.ModelAverage.apply (ArgSpec(args=['self', 'executor', 'need_restore'], varargs=None, keywords=None, defaults=(True,)), ('document', '46234a5470590feb336346f70a3db715'))
+paddle.fluid.optimizer.ModelAverage.apply (ArgSpec(args=['self', 'executor', 'need_restore'], varargs=None, keywords=None, defaults=(True,)), ('document', '648010d0ac1fa707dac0b89f74b0e35c'))
 paddle.fluid.optimizer.ModelAverage.apply_gradients (ArgSpec(args=['self', 'params_grads'], varargs=None, keywords=None, defaults=None), ('document', 'bfe7305918552aaecfdaa22411dbe871'))
 paddle.fluid.optimizer.ModelAverage.apply_optimize (ArgSpec(args=['self', 'loss', 'startup_program', 'params_grads'], varargs=None, keywords=None, defaults=None), ('document', '5c46d1926a40f1f873ffe9f37ac89dae'))
 paddle.fluid.optimizer.ModelAverage.backward (ArgSpec(args=['self', 'loss', 'startup_program', 'parameter_list', 'no_grad_set', 'callbacks'], varargs=None, keywords=None, defaults=(None, None, None, None)), ('document', 'ba3a113d0229ff7bc9d39bda0a6d947f'))
 paddle.fluid.optimizer.ModelAverage.get_opti_var_name_list (ArgSpec(args=['self'], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.optimizer.ModelAverage.minimize (ArgSpec(args=['self', 'loss', 'startup_program', 'parameter_list', 'no_grad_set', 'grad_clip'], varargs=None, keywords=None, defaults=(None, None, None, None)), ('document', 'b15cffad0903fc81af77a0580ceb2a9b'))
-paddle.fluid.optimizer.ModelAverage.restore (ArgSpec(args=['self', 'executor'], varargs=None, keywords=None, defaults=None), ('document', '18db9c70be9c4dd466f9844457b21bfe'))
+paddle.fluid.optimizer.ModelAverage.restore (ArgSpec(args=['self', 'executor'], varargs=None, keywords=None, defaults=None), ('document', '5f14ea4adda2791e1c3b37ff327f6a83'))
 paddle.fluid.optimizer.LarsMomentumOptimizer.__init__ (ArgSpec(args=['self', 'learning_rate', 'momentum', 'lars_coeff', 'lars_weight_decay', 'regularization', 'name'], varargs=None, keywords=None, defaults=(0.001, 0.0005, None, None)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.optimizer.LarsMomentumOptimizer.apply_gradients (ArgSpec(args=['self', 'params_grads'], varargs=None, keywords=None, defaults=None), ('document', 'bfe7305918552aaecfdaa22411dbe871'))
 paddle.fluid.optimizer.LarsMomentumOptimizer.apply_optimize (ArgSpec(args=['self', 'loss', 'startup_program', 'params_grads'], varargs=None, keywords=None, defaults=None), ('document', '5c46d1926a40f1f873ffe9f37ac89dae'))

python/paddle/dataset/mnist.py

@@ -90,7 +90,7 @@ def train():
     MNIST training set creator.
 
     It returns a reader creator, each sample in the reader is image pixels in
-    [0, 1] and label in [0, 9].
+    [-1, 1] and label in [0, 9].
 
     :return: Training reader creator
     :rtype: callable
@@ -107,7 +107,7 @@ def test():
     MNIST test set creator.
 
     It returns a reader creator, each sample in the reader is image pixels in
-    [0, 1] and label in [0, 9].
+    [-1, 1] and label in [0, 9].
 
     :return: Test reader creator.
     :rtype: callable

python/paddle/fluid/optimizer.py

@@ -2145,22 +2145,45 @@ class ModelAverage(Optimizer):
         regularization: A Regularizer, such as
                         fluid.regularizer.L2DecayRegularizer.
         name: A optional name prefix.
+
     Examples:
 
       .. code-block:: python
 
-        optimizer = fluid.optimizer.Momentum()
-        optimizer.minimize(cost)
-        model_average = fluid.optimizer.ModelAverage(0.15,
-                                                min_average_window=10000,
-                                                max_average_window=20000)
-        for pass_id in range(args.pass_num):
-            for data in train_reader():
-                exe.run(fluid.default_main_program()...)
+        import paddle.fluid as fluid
+        import numpy
+
+        # First create the Executor.
+        place = fluid.CPUPlace()  # fluid.CUDAPlace(0)
+        exe = fluid.Executor(place)
 
+        train_program = fluid.Program()
+        startup_program = fluid.Program()
+        with fluid.program_guard(train_program, startup_program):
+            # build net
+            data = fluid.layers.data(name='X', shape=[1], dtype='float32')
+            hidden = fluid.layers.fc(input=data, size=10)
+            loss = fluid.layers.mean(hidden)
+            optimizer = fluid.optimizer.Momentum(learning_rate=0.2, momentum=0.1)
+            optimizer.minimize(loss)
+
+            # build ModelAverage optimizer
+            model_average = fluid.optimizer.ModelAverage(0.15,
+                                                         min_average_window=10000,
+                                                         max_average_window=20000)
+
+            exe.run(startup_program)
+            x = numpy.random.random(size=(10, 1)).astype('float32')
+            outs = exe.run(program=train_program,
+                           feed={'X': x},
+                           fetch_list=[loss.name])
+
+            # apply ModelAverage
             with model_average.apply(exe):
-                for data in test_reader():
-                    exe.run(inference_program...)
+                x = numpy.random.random(size=(10, 1)).astype('float32')
+                exe.run(program=train_program,
+                        feed={'X': x},
+                        fetch_list=[loss.name])
     """
 
     def __init__(self,
@@ -2275,6 +2298,10 @@ class ModelAverage(Optimizer):
     @signature_safe_contextmanager
     def apply(self, executor, need_restore=True):
         """Apply average values to parameters of current model.
+
+        Args:
+            executor(fluid.Executor): current executor.
+            need_restore(bool): If you finally need to do restore, set it to True. Default is True.
         """
         executor.run(self.apply_program)
         try:
@@ -2285,6 +2312,9 @@ class ModelAverage(Optimizer):
 
     def restore(self, executor):
         """Restore parameter values of current model.
+        
+        Args:
+            executor(fluid.Executor): current executor.
         """
         executor.run(self.restore_program)