Error while loading EdgeTPU-DeepLab checkpoint in model zoo with tf.import_graph_def

File "test_edgetpu_deeplab.py", line 44, in init
self.export_onnx_model()
File "/mnt/data-1/user/test/tensorflow_model/tensorflow_model_test.py", line 78, in export_onnx_model
tf.import_graph_def(graph_def, name='')
File "/home/users/user/.local/lib/python3.6/site-packages/tensorflow/python/util/deprecation.py", line 507, in new_func
return func(*args, **kwargs)
File "/home/users/user/.local/lib/python3.6/site-packages/tensorflow/python/framework/importer.py", line 431, in import_graph_def
raise ValueError(str(e))
ValueError: NodeDef mentions attr 'incompatible_shape_error' not in Op<name=Equal; signature=x:T, y:T -> z:bool; attr=T:type,allowed=[DT_BFLOAT16, DT_HALF, DT_FLOAT, DT_DOUBLE, DT_UINT8, ..., DT_QINT8, DT_QINT32, DT_STRING, DT_BOOL, DT_COMPLEX128]; is_commutative=true>; NodeDef: {{node pad_to_bounding_box/Equal}}. (Check whether your GraphDef-interpreting binary is up to date with your GraphDef-generating binary.).

It seems to be a tensorflow version issue. Then I tried 1.8 1.13 1.14 and 2.0 but none of those worked. Does anyone have any ideas? Is it possible to know the tf version during model checkpoint generation? Thanks a lot!