rwightman / pytorch-dpn-pretrained

Dual Path Networks (DPN) supporting pretrained weights converted from original MXNet implementation

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Convert ResNet from MXNet to PyTorch

ahkarami opened this issue · comments

Dear @rwightman,
Thank you for your nice repository. I have a pre-trained ResNet152 model on MXNet and I want to convert it to PyTorch. Would you please kindly guide me to do that?
In fact, I have used convert_from_mxnet.py to do that (by some little modification), however, I have faced the following error:
AssertionError: Unexpected token

dump the model.state_dict() keys and tensor shapes, get a feel for the network structure

set debug=True on line 34

run the conversion and look what the last '...=>...' mapping is that's dumped before the error, try to match that up with your state key->tensor shape dump and see what the pattern is that the conditional isn't handling, add the necessary logic

Thank you for your time and response.