You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Oct 13, 2021. It is now read-only.
I am having an issue when trying to convert my model to onnx.
I see this error: 'KerasTensor' object has no attribute 'graph'
while trying to run this: onnx_model = keras2onnx.convert_keras(model, model.name)
It worked fine with tf 2.3, but it was running on CPU. I upgraded to tf 2.5, and the model trains on GPU now, but my script fails when trying to export to Onnx. I am working with an LSTM model.
versions:
Python: 3.8.10 (default, May 19 2021, 13:12:57) [MSC v.1916 64 bit (AMD64)]
Tensorflow Gpu Version: 2.5.0
Keras Version: 2.5.0
Hub Version: 0.12.0
Datasets Version: 4.3.0
cuDNN version: 64_8
Cuda version: 64_112
The text was updated successfully, but these errors were encountered:
I am having an issue when trying to convert my model to onnx.
I see this error:
'KerasTensor' object has no attribute 'graph'
while trying to run this:
onnx_model = keras2onnx.convert_keras(model, model.name)
It worked fine with tf 2.3, but it was running on CPU. I upgraded to tf 2.5, and the model trains on GPU now, but my script fails when trying to export to Onnx. I am working with an LSTM model.
versions:
Python: 3.8.10 (default, May 19 2021, 13:12:57) [MSC v.1916 64 bit (AMD64)]
Tensorflow Gpu Version: 2.5.0
Keras Version: 2.5.0
Hub Version: 0.12.0
Datasets Version: 4.3.0
cuDNN version: 64_8
Cuda version: 64_112
The text was updated successfully, but these errors were encountered: