AssertionError: Tensor Transformer-5-MultiHeadSelfAttention-Add/All:0 already processed

[RoacherM]

When I tried to convert A small version of RoBERT loading by keras to ONNX version, it ran out of this error. Here are the detail info: Traceback (most recent call last): File "/Users/wonbyron/Desktop/work/company-datapack/demo_robert.py", line 29, in onnx_model = keras2onnx.convert_keras(model=model, name=model.name, debug_mode=1) File "/Users/wonbyron/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras2onnx/main.py", line 80, in convert_keras parse_graph(topology, tf_graph, target_opset, output_names, output_dict) File "/Users/wonbyron/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras2onnx/parser.py", line 842, in parse_graph graph, keras_node_dict, topo, top_level, output_names) File "/Users/wonbyron/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras2onnx/parser.py", line 606, in _parse_graph_core _on_parsing_tf_nodes(graph, nodes, varset, topology.debug_mode) File "/Users/wonbyron/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras2onnx/parser.py", line 319, in _on_parsing_tf_nodes operator.add_output(out0) File "/Users/wonbyron/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras2onnx/common/intop.py", line 72, in add_output assert False, "Tensor {} already processed".format(var.full_name) AssertionError: Tensor Transformer-5-MultiHeadSelfAttention-Add/All:0 already processed

my environment

python3.6 virtualenv

tensorflow==1.15

keras==2.3.1

onnx==1.6.0

keras2onnx==1.7.0

onnxconverter-common==1.7.0