Match output_name with keras model (#282)

* Match output_name with keras

keras2onnx/parser.py

@@ -708,12 +708,9 @@ def _parse_graph_core(graph, keras_node_dict, topology, top_scope, output_names)
     :return: The whole topology of the intermediate objects.
     """
     input_nodes = set()
-    raw_model_container = topology.raw_model
 
     # build the node in the working scope.
     varset = topology.declare_scope('curr_', top_scope)
-    for name in output_names:
-        raw_model_container.add_output_name(name)
 
     model_outputs = []
     for name in output_names:
@@ -797,4 +794,35 @@ def parse_graph(topo, graph, target_opset, output_names):
             op.add_output(var1)
         topo.raw_model.add_input_name(str_value)
 
+    output_name_dict = {}
+    for idx_, ts_ in enumerate(topo.raw_model.model.outputs):
+        op = top_level.declare_local_operator(TYPES.Identity)
+        output_ts = topo.raw_model.model.outputs[idx_]
+        var_type = _adjust_input_batch_size(_infer_variable_type(output_ts, target_opset))
+        str_value = output_ts.name
+        use_ts_name = False
+        if hasattr(topo.raw_model.model, 'output_names'):
+            str_value = topo.raw_model.model.output_names[idx_]
+        elif topo.raw_model.model.outputs[idx_].name.endswith(':0'):
+            str_value = topo.raw_model.model.outputs[idx_].name[:-2]
+        else:
+            # if there is no difference between output tensor name and model output name
+            # skip it.
+            use_ts_name = True
+
+        if str_value in output_name_dict:
+            cur_count = output_name_dict[str_value]
+            output_name_dict[str_value] = cur_count + 1
+            str_value = str_value + ':' + str(cur_count)
+        else:
+            output_name_dict[str_value] = 1
+
+        if not use_ts_name:
+            var0 = top_level.get_local_variable_or_declare_one(str_value, var_type)
+            var1 = top_level.get_local_variable_or_declare_one(topo.raw_model.model.outputs[idx_].name, var_type)
+            op.add_input(var1)
+            op.add_output(var0)
+
+        topo.raw_model.add_output_name(str_value)
+
     return _parse_graph_core(graph, keras_layer_ts_map, topo, top_level, output_names)