Fail to Convert resnet from Caffe to onnx and Pytorch

[payal211]

Hi,

I am getting below error while converting Resnet Model From caffe to onnx:

I0217 14:18:15.896023 8924 net.cpp:196] Top shape: 1 24 14 14 (4704) I0217 14:18:15.896029 8924 layer_factory.hpp:77] Creating layer proposal ImportError: No module named 'rpn' Traceback (most recent call last): File "/home/ubuntu/anaconda3/envs/python_3_5/bin/mmconvert", line 8, in sys.exit(_main()) File "/home/ubuntu/anaconda3/envs/python_3_5/lib/python3.5/site-packages/mmdnn/conversion/_script/convert.py", line 102, in _main ret = convertToIR._convert(ir_args) File "/home/ubuntu/anaconda3/envs/python_3_5/lib/python3.5/site-packages/mmdnn/conversion/_script/convertToIR.py", line 16, in _convert transformer = CaffeTransformer(args.network, args.weights, "tensorflow", inputshape[0], phase = args.caffePhase) File "/home/ubuntu/anaconda3/envs/python_3_5/lib/python3.5/site-packages/mmdnn/conversion/caffe/transformer.py", line 325, in init graph = GraphBuilder(def_path, self.input_shape, self.is_train_proto, phase).build() File "/home/ubuntu/anaconda3/envs/python_3_5/lib/python3.5/site-packages/mmdnn/conversion/caffe/graph.py", line 454, in build graph.compute_output_shapes(self.model) File "/home/ubuntu/anaconda3/envs/python_3_5/lib/python3.5/site-packages/mmdnn/conversion/caffe/graph.py", line 263, in compute_output_shapes net = caffe.Net(tmp_prototxt, caffe.TEST) SystemError: <Boost.Python.function object at 0x2ccfd50> returned NULL without setting an error

Anyone can please help to figure it out? Thank you.

Below are system Specifications and command I used to convert: Platform (like ubuntu 16.04/win10): ubuntu 16.04

Python version: 3.5

Source framework with version (like Tensorflow 1.4.1 with GPU): CAFFE

Destination framework with version (like CNTK 2.3 with GPU): ONNX or PYTORCH

Running scripts: mmconvert --srcFramework caffe --inputWeight ./AMAR_MODELS/ctd_tloc_iter_91000_bkup_2_11_CW.caffemodel --inputNetwork ./AMAR_MODELS/test_ctd_tloc.prototxt --dstFramework onnx --outputModel ctd_tloc_iter_91000_bkup_2_11_CW.onnx --inputShape 1 3 224 224