Save dummy_input in quantizer metadata (#333)

And use it when calling prepare_model when loading from a checkpoint

distiller/apputils/checkpoint.py

@@ -161,7 +161,7 @@ def load_checkpoint(model, chkpt_file, optimizer=None, model_device=None, *,
         msglogger.info('Loaded quantizer metadata from the checkpoint')
         qmd = checkpoint['quantizer_metadata']
         quantizer = qmd['type'](model, **qmd['params'])
-        quantizer.prepare_model()
+        quantizer.prepare_model(qmd['dummy_input'])
 
     if normalize_dataparallel_keys:
             checkpoint['state_dict'] = {normalize_module_name(k): v for k, v in checkpoint['state_dict'].items()}

distiller/quantization/quantizer.py

@@ -213,6 +213,7 @@ class Quantizer(object):
         """
         msglogger.info('Preparing model for quantization using {0}'.format(self.__class__.__name__))
 
+        self.model.quantizer_metadata["dummy_input"] = dummy_input
         if dummy_input is not None:
             summary_graph = distiller.SummaryGraph(self.model, dummy_input)
             self.adjacency_map = summary_graph.adjacency_map(dedicated_modules_only=False)