openvino: [Good First Issue][TF FE]: Fix (AttributeError) NoneType object has no attribute shape or dtype
Context
OpenVINO component responsible for support of TensorFlow models is called as TensorFlow Frontend (TF FE). TF FE converts a model represented in TensorFlow opset to a model in OpenVINO opset.
We have 14 models from TensorFlow Hub that are failing due to issue AttributeError: NoneType object has no attribute shape or dtype. Your fix greatly helps to make OpenVINO better.
What needs to be done?
The task is to fix a problem AttributeError: NoneType object has no attribute shape or dtype.
How to reproduce the issue:
- Build OpenVINO following How to Build OpenVINO document
- Install built
openvino-*.whlpackage usingpip. You can find the package inopenvino/build/wheelsdirectory. Do not installopenvino-dev-*.whlpackage. Also, it is recommended to use new virtual environment to avoid mess with previously installed OV packages in the system. - Run the model conversion as follows:
import tensorflow_hub as hub
import openvino as ov
model_link = 'https://www.kaggle.com/models/google/gnews-swivel/frameworks/tensorFlow2/variations/tf2-preview-20dim/versions/1'
model = hub.load(model_link)
ov.convert_model(model)
Received error message:
File "<stdin>", line 1, in <module>
File "openvino/tools/ovc/convert.py", line 101, in convert_model
ov_model, _ = _convert(cli_parser, params, True)
File "openvino/tools/ovc/convert_impl.py", line 534, in _convert
raise e
File "openvino/tools/ovc/convert_impl.py", line 476, in _convert
ov_model = driver(argv, {"conversion_parameters": non_default_params})
File "openvino/tools/ovc/convert_impl.py", line 227, in driver
ov_model = moc_emit_ir(prepare_ir(argv), argv)
File "openvino/tools/ovc/convert_impl.py", line 177, in prepare_ir
ov_model = moc_pipeline(argv, moc_front_end)
File "openvino/tools/ovc/moc_frontend/pipeline.py", line 64, in moc_pipeline
input_model = moc_front_end.load(argv.input_model, share_weights)
RuntimeError: Check 'false' failed at src/frontends/common/src/frontend.cpp:47:
Loading input model
AttributeError: 'NoneType' object has no attribute 'dtype'
At:
openvino/frontend/tensorflow/node_decoder.py(149): get_attribute
openvino/tools/ovc/moc_frontend/pipeline.py(64): moc_pipeline
openvino/tools/ovc/convert_impl.py(177): prepare_ir
openvino/tools/ovc/convert_impl.py(227): driver
openvino/tools/ovc/convert_impl.py(476): _convert
openvino/tools/ovc/convert.py(101): convert_model
<stdin>(1): <module>
Resources
- What is OpenVINO?
- How to Build OpenVINO
- Developer documentation for TensorFlow Frontend
- Contribution guide - start here!
- Intel DevHub Discord channel - engage in discussions, ask questions and talk to OpenVINO developers
Contact points
About this issue
- Original URL
- State: closed
- Created 6 months ago
- Comments: 23 (16 by maintainers)
Commits related to this issue
- patch #22076 — committed to abhinav231-valisetti/openvino by abhinav231-valisetti 4 months ago
@mvafin @rkazants Most thanks for both of you!i have fix the problem under the environment of torch=2.0.1, openvino=2023.1.0
i want to try this issue can you please assign it to me ?