error: Tensorflow to caffe "caffe_emitter:emit_Constant" doesn't handle one dim tensor.

[Alnlll]

Enviroment

Platform (like ubuntu 16.04/win10): ubuntu 16.04

Python version: python3.5

Source framework with version (like Tensorflow 1.4.1 with GPU): Tensorflow 1.13

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

Pre-trained model path (webpath or webdisk path):

Running scripts:

mmconvert\
  -sf tensorflow\
  -in ../facenet_0330/facenet_0330-eval-ckpt.meta\
  -iw ../facenet_0330/facenet_0330-eval-ckpt\
  --dstNodeName InceptionResnetV1/Bottleneck/BatchNorm/Reshape_1\
  -df caffe\
  -om tf_facenet_0330_se_inception_resnetv1

Problem

Trying to convert a se_inception_resnetv1 from tensorflow to caffe(using master mmdnn), got an IndexError like this:

Traceback (most recent call last):
  File "/usr/local/bin/mmconvert", line 11, in <module>
    sys.exit(_main())
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/convert.py", line 108, in _main
    ret = IRToCode._convert(code_args)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/IRToCode.py", line 61, in _convert
    emitter.run(args.dstModelPath, args.dstWeightPath, args.phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 144, in run
    super(CaffeEmitter, self).run(dstNetworkPath, dstWeightPath, phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/common/DataStructure/emitter.py", line 22, in run
    self.save_code(dstNetworkPath, phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/common/DataStructure/emitter.py", line 59, in save_code
    code = self.gen_code(phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 132, in gen_code
    func(current_node)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 434, in emit_Constant
    shape[1],
IndexError: list index out of range

and got such shape info by adding some print code:

nodename: InceptionResnetV1/Logits/Flatten/flatten/Reshape/shape/1
<class 'graph_pb2.TensorShape'> dim {
  size: 2
}

<class 'list'> [2]

Any idea about this?

[rainLiuplus]

Hi @Alnlll , please use the latest code to have a try, thanks!

[Alnlll]

@rainLiuplus thanks for responding!!
Tried the latest 0.2.4 version, got the same error:

shape1:  dim {
  size: 2
}

shape2:  [2]
Traceback (most recent call last):
  File "/usr/local/bin/mmconvert", line 11, in <module>
    load_entry_point('mmdnn==0.2.4', 'console_scripts', 'mmconvert')()
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/convert.py", line 108, in _main
    ret = IRToCode._convert(code_args)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/_script/IRToCode.py", line 61, in _convert
    emitter.run(args.dstModelPath, args.dstWeightPath, args.phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 144, in run
    super(CaffeEmitter, self).run(dstNetworkPath, dstWeightPath, phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/common/DataStructure/emitter.py", line 22, in run
    self.save_code(dstNetworkPath, phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/common/DataStructure/emitter.py", line 59, in save_code
    code = self.gen_code(phase)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 132, in gen_code
    func(current_node)
  File "/usr/local/lib/python3.5/dist-packages/mmdnn/conversion/caffe/caffe_emitter.py", line 433, in emit_Constant
    shape[1],
IndexError: list index out of range

And checked the emit_Constant in caffe_emmiter, seems unmodified.

[Alnlll]

@rainLiuplus Thanks for helping.
Tried this patch, good for my case.