prepare context for grad test

contrib/Python/cntk/context.py

@@ -58,7 +58,8 @@ class AbstractContext(object, metaclass=ABCMeta):
                  graph=None,
                  device_id=-1,
                  root_node=None,
-                 clean_up=True):
+                 clean_up=True,
+                 node_unit_test=False):
         '''
         AbstractContext Constructer
 
@@ -68,6 +69,7 @@ class AbstractContext(object, metaclass=ABCMeta):
         :param root_node: the top node of the graph
         :param clean_up: whether the temporary directory should be removed when the context is left
         are the GPUs indices.        
+        :param node_unit_test: set to True if you want to output the gradient of a node (backward pass)
 
         '''
         if isinstance(name, str):
@@ -88,6 +90,7 @@ class AbstractContext(object, metaclass=ABCMeta):
         self.clean_up = clean_up
         self.input_nodes = set()
         self.root_node = root_node
+        self.node_unit_test= node_unit_test
 
     def __enter__(self):
         _CONTEXT[self.name] = self
@@ -206,6 +209,7 @@ class AbstractContext(object, metaclass=ABCMeta):
         output_filename = os.path.join(self.directory, CNTK_OUTPUT_FILENAME)
         tmpl_dict = {
             'DevideId': self.device_id,
+            'NodeUnitTest': self.node_unit_test,
             'OutputFile': output_filename,
             'ModelDescription': model_description,
             'Reader': '\n'.join(r.generate_config() for r in readers),

contrib/Python/cntk/templates/cntk_eval_template.cntk

@@ -6,6 +6,7 @@ deviceId=%(DevideId)s
 				   
 Eval=[
 	action="write"
+	nodeUnitTest=%(NodeUnitTest)s
 	run=BrainScriptNetworkBuilder
 
     BrainScriptNetworkBuilder=[