cherry pick, fix dygraph api doc, test=release/1.5

BackwardStrategy
dygraph.nn
dygraph.checkpoint

paddle/fluid/API.spec

@@ -731,8 +731,8 @@ paddle.fluid.dygraph.Tracer.train_mode (ArgSpec(args=['self'], varargs=None, key
 paddle.fluid.dygraph.start_gperf_profiler (ArgSpec(args=[], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.dygraph.stop_gperf_profiler (ArgSpec(args=[], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.dygraph.prepare_context (ArgSpec(args=['strategy'], varargs=None, keywords=None, defaults=(None,)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
-paddle.fluid.dygraph.save_persistables (ArgSpec(args=['model_dict', 'dirname', 'optimizers'], varargs=None, keywords=None, defaults=('save_dir', None)), ('document', 'bdeefe733228f5f2d4a8f8c61a5956cf'))
-paddle.fluid.dygraph.load_persistables (ArgSpec(args=['dirname'], varargs=None, keywords=None, defaults=('save_dir',)), ('document', 'fb79b050b5eb52fa9c5fdccefe521aa1'))
+paddle.fluid.dygraph.save_persistables (ArgSpec(args=['model_dict', 'dirname', 'optimizers'], varargs=None, keywords=None, defaults=('save_dir', None)), ('document', '7f526f879139a14cda8e0b5a9171f264'))
+paddle.fluid.dygraph.load_persistables (ArgSpec(args=['dirname'], varargs=None, keywords=None, defaults=('save_dir',)), ('document', '2574d50a7a9f89fb0d74ddf73d8128f0'))
 paddle.fluid.dygraph.NoamDecay.__init__ (ArgSpec(args=['self', 'd_model', 'warmup_steps', 'begin', 'step', 'dtype'], varargs=None, keywords=None, defaults=(1, 1, 'float32')), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
 paddle.fluid.dygraph.NoamDecay.create_lr_var (ArgSpec(args=['self', 'lr'], varargs=None, keywords=None, defaults=None), ('document', '013bc233558149d0757b3df57845b866'))
 paddle.fluid.dygraph.NoamDecay.step (ArgSpec(args=['self'], varargs=None, keywords=None, defaults=None), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))

paddle/fluid/pybind/imperative.cc

@@ -162,6 +162,7 @@ void BindImperative(pybind11::module *m_ptr) {
     1. :code:`sort_sum_gradient`, which will sum the gradient by the reverse order of trace.
 
     Examples:
+
         .. code-block:: python
 
           import numpy as np

python/paddle/fluid/dygraph/checkpoint.py

@@ -43,9 +43,12 @@ def save_persistables(model_dict, dirname='save_dir', optimizers=None):
         optimizers(fluid.Optimizer|list(fluid.Optimizer)|None): The optimizers to be saved
 
     Returns:
+        None
 
     Examples:
+
         .. code-block:: python
+
           ptb_model = PtbModel(
                 hidden_size=hidden_size,
                 vocab_size=vocab_size,
@@ -95,7 +98,9 @@ def load_persistables(dirname='save_dir'):
         optimizer dict: The optimizer
 
     Examples:
+
          .. code-block:: python
+
            my_layer = layer(fluid.Layer)
            param_path = "./my_paddle_model"
            sgd = SGDOptimizer(learning_rate=1e-3)