[Streamline] handle edge case in AbsorbTransposeIntoMultiThreshold

need to do things different if MultiThreshold is the final node
in the graph (producing top-level output)

src/finn/transformation/streamline/absorb.py

@@ -321,11 +321,16 @@ class AbsorbTransposeIntoMultiThreshold(Transformation):
                 perms = list(get_by_name(n.attribute, "perm").ints)
                 if perms == [0, 3, 1, 2]:
                     mt_cand = model.find_consumer(n.output[0])
-                    if mt_cand.op_type == "MultiThreshold" and not model.is_fork_node(
-                        mt_cand
+                    if (
+                        mt_cand is not None
+                        and mt_cand.op_type == "MultiThreshold"
+                        and not model.is_fork_node(mt_cand)
                     ):
                         final_t_cand = model.find_consumer(mt_cand.output[0])
-                        if final_t_cand.op_type == "Transpose":
+                        if (
+                            final_t_cand is not None
+                            and final_t_cand.op_type == "Transpose"
+                        ):
                             perms = list(
                                 get_by_name(final_t_cand.attribute, "perm").ints
                             )
@@ -356,7 +361,15 @@ class AbsorbTransposeIntoMultiThreshold(Transformation):
                                 perm=[0, 3, 1, 2],
                             )
                             graph.node.insert(node_ind + 1, new_transpose)
-                            final_t_cand.input[0] = transpose_output
+                            if final_t_cand is not None:
+                                # rewire next node's input
+                                final_t_cand.input[0] = transpose_output
+                            else:
+                                # replace graph top-level output
+                                get_by_name(
+                                    model.graph.output, mt_cand.output[0]
+                                ).name = transpose_output
+                                model.set_tensor_shape(mt_cand.output[0], mt_ishape)
                             graph_modified = True
         if graph_modified:
             model = model.transform(InferDataTypes())