setではなくOrderedDictを使用して順番が保持されるよう修正

python/src/nnabla/utils/converter/onnx/reader.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 
 import struct
+from collections import OrderedDict
 import nnabla.logger as logger
 from nnabla.utils import nnabla_pb2
 from onnx import (ModelProto, TensorProto, AttributeProto)
@@ -211,7 +212,9 @@ def onnx_graph_to_nnp_protobuf(pb, graph):
     network = pb.network.add()
     network.name = graph.name
 
-    all_vars = set()
+    # We use an OrderedDict and not a set
+    # to preserve order
+    all_vars = OrderedDict()
     func_counter = {}
     # convert nodes
     for n in graph.node:
@@ -221,13 +224,15 @@ def onnx_graph_to_nnp_protobuf(pb, graph):
         f = convert_to_function(n, graph.name, func_counter)
         #Gather all unique names for input and output
         for i in f.input:
-            all_vars.add(i)
+            all_vars[i] = None
         for o in f.output:
-            all_vars.add(o)
+            all_vars[o] = None
         network.function.extend([f])
 
     # convert parameters
-    param_vars = set()
+    # We use an OrderedDict and not a set
+    # to preserve order
+    param_vars = OrderedDict()
     for init in graph.initializer:
         if init.data_type != TensorProto.FLOAT:
             raise ValueError("Only floating point data is supported for parameters {} (Got {})"
@@ -242,7 +247,7 @@ def onnx_graph_to_nnp_protobuf(pb, graph):
         p.data.extend(data)
         p.need_grad = False
         # Keep the list of all initializer names
-        param_vars.add(init.name)
+        param_vars[init.name] = None
     # We need to distinguish constant parameters (which become 'Parameter' in NNabla)
     # from input/output variables (which become 'Buffer' in NNabla).
     # Contant parameters appear in the initializer list so we keep
@@ -267,19 +272,19 @@ def onnx_graph_to_nnp_protobuf(pb, graph):
             # This input is a variable
             v.type ="Buffer"
             in_list.append(v)
-        all_vars.remove(v.name)
+        del all_vars[v.name]
     for o in graph.output:
         v = onnx_value_info_proto_to_variable(o, network)
         v.type = "Buffer"
         out_list.append(v)
-        all_vars.remove(v.name)
+        del all_vars[v.name]
 
     for varg in all_vars:
         # We add all remaining variables as intermediate buffer
+        # We leave the buffer size of all intermediate buffers empty
         v = network.variable.add()
         v.type = "Buffer"
         v.name = varg
-        # We calculate the buffer size of all intermediate buffers here
 
     #pdb.set_trace()