use caffe2.python.onnx.backend instead of onnx-caffe2 (#418)

nnvm/tests/python/frontend/onnx/test_forward.py

@@ -7,9 +7,9 @@ import onnx
 from model_zoo import super_resolution, squeezenet1_1, lenet, resnet18_1_0
 
 def verify_onnx_forward_impl(graph_file, data_shape, out_shape):
-    import onnx_caffe2.backend
+    import caffe2.python.onnx.backend
     def get_caffe2_output(model, x, dtype='float32'):
-        prepared_backend = onnx_caffe2.backend.prepare(model)
+        prepared_backend = caffe2.python.onnx.backend.prepare(model)
         W = {model.graph.input[0].name: x.astype(dtype)}
         c2_out = prepared_backend.run(W)[0]
         return c2_out