Conversion from TF to ONNX stops, throwing KeyError 21

[jackhu3301]

Pls help me. Thank you very much. I am trying to convert tf2 model to onnx using graphfile.

But got this error:

2023-02-03 15:59:30.193108: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:7 with 30969 MB memory: -> device: 7, name: Tesla V100S-PCIE-32GB, pci bus id: 0000:41:00.0, compute capability: 7.0

2023-02-03 15:59:39,646 - INFO - Using tensorflow=2.11.0, onnx=1.13.0, tf2onnx=1.13.0/2c1db5
2023-02-03 15:59:39,646 - INFO - Using opset <onnx, 13>
Traceback (most recent call last):
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/convert.py", line 706, in
main()
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/convert.py", line 269, in main
model_proto, _ = _convert_common(
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/convert.py", line 164, in _convert_common
g = process_tf_graph(tf_graph, const_node_values=const_node_values,
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tfonnx.py", line 459, in process_tf_graph
main_g, subgraphs = graphs_from_tf(tf_graph, input_names, output_names, shape_override, const_node_values,
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tfonnx.py", line 474, in graphs_from_tf
ordered_func = resolve_functions(tf_graph)
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tf_loader.py", line 760, in resolve_functions
_, _, _, _, _, functions = tflist_to_onnx(tf_graph, {})
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tf_utils.py", line 441, in tflist_to_onnx
onnx_tensor = tf_to_onnx_tensor(value, name=port_name(node.name))
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tf_utils.py", line 52, in tf_to_onnx_tensor
np_data = get_tf_tensor_data(tensor)
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tf2onnx/tf_utils.py", line 70, in get_tf_tensor_data
np_data = tensor_util.MakeNdarray(tensor)
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tensorflow/python/framework/tensor_util.py", line 595, in MakeNdarray
dtype = tensor_dtype.as_numpy_dtype
File "/data/home/hujie/miniconda3/envs/py39/lib/python3.8/site-packages/tensorflow/python/framework/dtypes.py", line 98, in as_numpy_dtype
return _TF_TO_NP[self._type_enum]
KeyError: 21`

This section can be obtained by running get_version.py from util folder.

• Python version: 3.8.13
• ONNX version: 1.13.0
• ONNX-TF version: 1.13.0/2c1db5
• Tensorflow version: 2.11.0

Additional context

Have considered this issue maybe similar #299 , and also install onnx-tf, but not fix.

Add any other context about the problem here.