MxNet to TensorRT


I have a trained RetinaFace model on MxNet now I want to convert it to TensorRT. I tried:
MxNet to Onnx
But its failing in conversion.
Can you help me in conversion.
Also what are the other possible ways for achieving the same task.
Thank you.

Error :
“SoftmaxActivation” function is not supported.
Some forums suggested to change “SoftmaxActivation” to “softmax & axis=1” in mxnet.json. I tried this approach but it didn’t work.

Link to the MxNet Model.


TensorRT Version 7.x:
GPU Type k80:
**Nvidia Driver Version **:
CUDA Version 11.x:
CUDNN Version:
Operating + Version Ubuntu 18.04:
Python Version (if applicable):
TensorFlow Version (if applicable):
PyTorch Version (if applicable):
Baremetal or Container (if container which image + tag):

Relevant Files

Please attach or include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.)

Steps To Reproduce

Please include:

  • Exact steps/commands to build your repro
  • Exact steps/commands to run your repro
  • Full traceback of errors encountered

Hi @yashkhokarale,
You can follow the below conversion process
MxNet<< ONNX << TRT
You can take referenece from the link

In case of issues while MxNet to ONNX conversion, kindly reach out to MxNet-Onnx Team.