PyTorch to IR error

[uhvardhan]

Platform (like ubuntu 16.04/win10): Ubuntu 16.04

Python version: 3.5

Source framework with version (like Tensorflow 1.4.1 with GPU): Pytorch 0.3.1 with gpu

Destination framework with version (like CNTK 2.3 with GPU): Tensorflow

Pre-trained model path (webpath or webdisk path): https://drive.google.com/file/d/1Y2ritTA6PXosQ9u66f2JVpP_CMTt9fxl/view?usp=sharing

Running scripts:
mmtoir -f pytorch -d out -n ResNet-2018-09-18T21:50:42.pth --inputShape 3,224,224
Traceback (most recent call last):
File "/usr/local/bin/mmtoir", line 11, in
sys.exit(_main())
File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/convertToIR.py", line 192, in _main
ret = _convert(args)
File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/convertToIR.py", line 91, in _convert
from mmdnn.conversion.pytorch.pytorch_parser import PytorchParser
File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/pytorch/pytorch_parser.py", line 12, in
from mmdnn.conversion.pytorch.pytorch_graph import PytorchGraph
File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/pytorch/pytorch_graph.py", line 12, in
from torch.jit import _unique_state_dict
ImportError: cannot import name '_unique_state_dict'

[muzi-8]

@uhvardhan 　I think the major problem is torch' version.
You should update to the torch 0.4.0 ; torchvision 0.2.1
Maybe you can refer to the same issue#274 、#413
Good Luck!

[uhvardhan]

I have tried updating it 0.4.0 and 0.2.1 torchvision but couldn't solve the problem. I will have a look at the issues now

Thanks for the help.

[namizzz]

Hi @uhvardhan ,from the stack trace:

File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/pytorch/pytorch_graph.py", line 12, in 
from torch.jit import _unique_state_dict
ImportError: cannot import name '_unique_state_dict'

I try to reproduce your problem, but it can pass. Please try again, thanks!

[uhvardhan]

Hi @namizzz

I am not sure if it worked because I have been getting another error.

Traceback (most recent call last):
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/mmdnn/conversion/pytorch/pytorch_parser.py", line 74, in init
model = torch.load(model_file_name)
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/torch/serialization.py", line 358, in load
return _load(f, map_location, pickle_module)
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/torch/serialization.py", line 542, in _load
result = unpickler.load()
ModuleNotFoundError: No module named 'agents'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/home/harshvardhan/anaconda3/bin/mmtoir", line 11, in
sys.exit(_main())
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/mmdnn/conversion/_script/convertToIR.py", line 192, in _main
ret = _convert(args)
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/mmdnn/conversion/_script/convertToIR.py", line 92, in _convert
parser = PytorchParser(args.network, inputshape[0])
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/mmdnn/conversion/pytorch/pytorch_parser.py", line 76, in init
model = torch.load(model_file_name, map_location='cpu')
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/torch/serialization.py", line 358, in load
return _load(f, map_location, pickle_module)
File "/home/harshvardhan/anaconda3/lib/python3.6/site-packages/torch/serialization.py", line 542, in _load
result = unpickler.load()
ModuleNotFoundError: No module named 'agents'

Can you please have a look ?

Thanks

[namizzz]

Hi @uhvardhan ,It is a class named as 'agents' defined by you. Please refer to

#241

#281