Re: `make example` fail with `IndexOutOfRange` error while installing ncsdk

idata · ‎05-17-2018

Here are the system specs

OS: Ubuntu 18.04

TF Version: 1.8.0 (tensorflow-gpu)

NCSDK Version: 2.0.4

I cloned the repo and ran make install without a problem.

While running make example I get the following error. This occurs on several nets from ncappzoo that I tried on as well my custom nets.

mvNCCheck -s 12 output/inception-v1.meta -in=input -on=InceptionV1/Logits/Predictions/Reshape_1 -i ../../data/images/cat.jpg -id 829 -S 2 -M 128 -cs 0,1,2 -metric top1
mvNCCheck v02.00, Copyright @ Movidius Ltd 2016

/home/gokul/.local/lib/python3.6/site-packages/tensorflow/python/pywrap_tensorflow_internal.py:15: DeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
  import imp
/home/gokul/.local/lib/python3.6/site-packages/tensorflow/python/framework/ops.py:939: DeprecationWarning: builtin type EagerTensor has no __module__ attribute
  EagerTensor = c_api.TFE_Py_InitEagerTensor(_EagerTensorBase)
/home/gokul/.local/lib/python3.6/site-packages/tensorflow/python/util/tf_inspect.py:55: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() or inspect.getfullargspec()
  if d.decorator_argspec is not None), _inspect.getargspec(target))
/usr/local/lib/python3.6/dist-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
/home/gokul/.local/lib/python3.6/site-packages/tensorflow/python/util/tf_inspect.py:55: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() or inspect.getfullargspec()
  if d.decorator_argspec is not None), _inspect.getargspec(target))
/usr/local/lib/python3.6/dist-packages/skimage/transform/_warps.py:84: UserWarning: The default mode, 'constant', will be changed to 'reflect' in skimage 0.15.
  warn("The default mode, 'constant', will be changed to 'reflect' in "
Traceback (most recent call last):
  File "/usr/local/bin/mvNCCheck", line 152, in <module>
    quit_code = check_net(args.network, args.image, args.inputnode, args.outputnode, args.nshaves, args.inputsize, args.weights, args)
  File "/usr/local/bin/mvNCCheck", line 130, in check_net
    net = parse_tensor(args, myriad_config, file_gen=True)
  File "/usr/local/bin/ncsdk/Controllers/TensorFlowParser.py", line 290, in parse_tensor
    if have_first_input(strip_tensor_id(node.outputs[0].name)):
IndexError: list index out of range
Makefile:55: recipe for target 'check' failed

There are other similar errors like this and I can share the entire output if anyone likes.

I think this might be due to compatibility issues between TF and NCS tools.

Thanks!

idata · ‎05-17-2018

Quick hack that _seems_ to solve the issue. I've put in the line numbers for reference and am too lazy to remove them one by one ;)

 289             if not inputfound:
 290                 # print("inside not inputfound")
 291                 # print(node.outputs)
 292                 if len(node.outputs) == 0:
 293                     continue
 294                 if have_first_input(strip_tensor_id(node.outputs[0].name)):
 295                     inputfound = True
 296                     if debug:
 297                         print('Starting to process')
 298                 continue