ERROR - Tensorflow op [cudnn_lstm/rnn/multi_rnn_cell/cell_0/cudnn_compatible_lstm_cell/BlockLSTM: BlockLSTM] is not supported

[justzmq]

Ask a Question

I have just started learning how to convert a TensorFlow model to ONNX. How should I start fix this error?

ERROR - Tensorflow op [cudnn_lstm/rnn/multi_rnn_cell/cell_0/cudnn_compatible_lstm_cell/BlockLSTM: BlockLSTM] is not supported ERROR - Unsupported ops: Counter({'BlockLSTM': 1})

Further information

tensorflow version:1.14.0 onnx version:1.6.0 tf2onnx version:1.14.0

[fatcat-z]

This error is telling that the tf op 'BlockLSTM' could not be converted to proper ONNX op.

One suggestion is to update the tf model code to replace BlockLSTM with other tf ops which are supported by tf2onnx. Here is a list of supported tf ops so far.