Model drawing: allow rendering models w/o specifying the dataset (#294)

This commit allows us to draw diagrams of models, even if we don't support the specific dataset used by the model.  All you need is to specify the dimensions of the model inputs.
* Added `input_shape` argument to `draw_img_classifier_to_file`
* updated docstring

distiller/model_summaries.py

@@ -418,8 +418,8 @@ def draw_model_to_file(sgraph, png_fname, display_param_nodes=False, rankdir='TB
         fid.write(png)
 
 
-def draw_img_classifier_to_file(model, png_fname, dataset, display_param_nodes=False,
-                                rankdir='TB', styles=None):
+def draw_img_classifier_to_file(model, png_fname, dataset=None, display_param_nodes=False,
+                                rankdir='TB', styles=None, input_shape=None):
     """Draw a PyTorch image classifier to a PNG file.  This a helper function that
     simplifies the interface of draw_model_to_file().
 
@@ -436,8 +436,10 @@ def draw_img_classifier_to_file(model, png_fname, dataset, display_param_nodes=F
                 styles['conv1'] = {'shape': 'oval',
                                    'fillcolor': 'gray',
                                    'style': 'rounded, filled'}
+        input_shape (tuple): List of integers representing the input shape. Used only if 'dataset' is None
     """
-    dummy_input = distiller.get_dummy_input(dataset, distiller.model_device(model))
+    dummy_input = distiller.get_dummy_input(dataset=dataset,
+                                            device=distiller.model_device(model), input_shape=input_shape)
     try:
         non_para_model = distiller.make_non_parallel_copy(model)
         g = SummaryGraph(non_para_model, dummy_input)