Error while executing bash tools/prepare_pascal.sh command

[guruvishnuvardan]

Please note, I am getting the following error while executing bash tools/prepare_pascal.sh command, Can you pls let me know the cause of the error, PLease find the screen shot.

[zhreshold]

im2rec.py was updated in mxnet. --shuffle 1 is not supported, use --shuffle instead in prepare_pascal.sh

[arunbuduri]

In tools/prepare_pascal.sh, the default value for --shuffle is set to False that is passed as an argument to tools/prepare_dataset.py

In tools/prepare_dataset.py, im2rec.py is called with "--shuffle" option

Since the "--shuffle" option in im2rec.py was renamed as "--no-shuffle", I fixed this issue by changing the file tools/prepare_dataset.py

By changing the line below "--no-shuffle", str(int(args.shuffle)), "--pack-label", "1"])

to

"--no-shuffle", "--pack-label"])

This is a temporary fix since it assumes --no-shuffle always. Ideally, you should read the value passed to tools/prepare_dataset.py and pass "--no-shuffle" or not depending on the value

[WijdanMTak]

Hi, Firstly, thanks very much for your work. Secondly, I had the same problem and I fixed as you said, but then when I was training it, I got this error: File "/usr/local/lib/python2.7/dist-packages/mxnet/base.py", line 210, in check_call raise MXNetError(py_str(_LIB.MXGetLastError())) mxnet.base.MXNetError: [11:52:31] src/io/input_split_base.cc:173: Check failed: files_.size() != 0U (0 vs. 0) Cannot find any files that matches the URI pattern I have to mention that when I was preparing, it gave me this output: assert os.path.exists(image_file), 'Path does not exist: {}'.format(image_file) AssertionError: Path does not exist: /home/eng-wijdan/mxnet-ssd/tools/../data/VOCdevkit/VOC2012/JPEGImages/2009_004217.jpg I checked if it exist, it's not, but then the record file was generated. sorry if I'm bothering you, but I am a bit new, and I just follow the instructions. looking forward to your reply. Best regards.

[zhreshold]

@WijdanMTak Fix the corrupted files and re-generate the record file should be fine

[WijdanMTak]

Thanks for your immediate reply, I really appreciate that. you are right, since I've posted the question I redownload the corrupted file, and It was solved. :) but now I'm countering another problem which is :
File "/usr/lib/python2.7/ctypes/__init__.py", line 362, in __init__ self._handle = _dlopen(self._name, mode) OSError: libcudart.so.8.0: cannot open shared object file: No such file or directory I tried a lot to solve it, one of them was as they refer here but it doesn't work : https://github.com/apache/incubator-mxnet/issues/7608 I would be grateful for any suggestion. Regards.

[zhreshold]

@WijdanMTak You can google the problem, it's due to the cuda runtime library not found in PATH.

[webpage-maram]

Sir, may I ask a little silly question just to make sure. Is your implementation could apply on cpu and specifically for non-Nvidia graphics card? Many thanks.

[zhreshold]

@maram992 just specify use --cpu

[webpage-maram]

Sir,Thank you for your immediate reply. I asked this question because I had the same issue while trying to train: ' lib = ctypes.CDLL(lib_path[0], ctypes.RTLD_LOCAL) File "/usr/lib/python2.7/ctypes/init.py", line 362, in init self._handle = _dlopen(self._name, mode) OSError: libcudart.so.8.0: cannot open shared object file: No such file or directory ' And searching on its solution leads me to know that cuda only supports the nvidia cards so I thought that I can't apply your work on my machine. I was before able to run the demo but not now, same error appears when running the demo. I thought I don't have to deal with it as I don't have a nvidia card. I've done some tries to diagnose the problem:

' ~$ which nvcc /usr/bin/nvcc '

' ~$ nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2017 NVIDIA Corporation Built on Fri_Sep__1_21:08:03_CDT_2017 Cuda compilation tools, release 9.0, V9.0.176 '

' ~$ ls /usr/local bin etc games include lib libexec man MATLAB sbin share src '

' ~$ echo $PATH /opt/ros/kinetic/bin:/home/maram/bin:/home/maram/.local/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin '

' ~$ locate cuda | grep /cuda$ /home/maram/mxnet-ssd/mxnet/mshadow/mshadow/cuda /home/maram/mxnet-ssd/mxnet/nnvm/tvm/src/runtime/cuda /home/maram/mxnet-ssd/mxnet/nnvm/tvm/topi/python/topi/cuda /opt/ros/kinetic/include/opencv-3.3.1-dev/opencv2/core/cuda /usr/include/flann/util/cuda /usr/include/thrust/system/cuda /usr/local/MATLAB/R2017b/toolbox/vision/builtins/src/ocvcg/opencv/include/opencv2/core/cuda /usr/local/lib/python3.5/dist-packages/external/local_config_cuda/cuda /usr/local/lib/python3.5/dist-packages/external/local_config_cuda/cuda/cuda /usr/local/lib/python3.5/dist-packages/tensorflow/include/tensorflow/stream_executor/cuda '

' ~$ locate libcupti.so /usr/lib/x86_64-linux-gnu/libcupti.so /usr/lib/x86_64-linux-gnu/libcupti.so.9.0 /usr/lib/x86_64-linux-gnu/libcupti.so.9.0.176 ' Sorry for wasting your time, but I couldn't find any satisfactory solution. could you help me? Regards.

[chensonglu]

@arunbuduri thanks, it works for me