Windows-Machine-Learning
Windows-Machine-Learning copied to clipboard
I downgraded the ONNX model version to 9 to fit 19041, but it still doesn't work!
###@martinb35 My code:
onnx.export(model,
md_input,
"out/eye.onnx",
export_params=True,
opset_version=9,
verbose=False,
do_constant_folding=True,
input_names=['In'],
output_names=['Out'],
dynamic_axes={'In':{0:'batch_size'},
'Out':{0:'batch_size'}},
# keep_initializers_as_inputs=True,
)
I exported the PyTorch model with the above code and set the OP version to 9, but still got an error:
Parameter error Failed to load model with error: Unknown model file format
I have tried all kinds of methods, but it still doesn't work. Is there any way to solve this problem? Thanks!
Environment
Windows Build Number:19041
OS Version (Server, IoT Core, Desktop, etc):windows 10
Visual Studio
- [16 ] 2017 (version: )
Hi @Lookforworld, do you mind sharing your model for reproduction purposes?
Closing due to inactivity.