weiliu89 / caffe

Caffe: a fast open framework for deep learning.
http://caffe.berkeleyvision.org/
Other
4.77k stars 1.68k forks source link

error python examples/ssd/ssd_pascal.py (segmentation fault) #895

Open linthy opened 6 years ago

linthy commented 6 years ago

Please use the caffe-users list for usage, installation, or modeling questions, or other requests for help. Do not post such requests to Issues. Doing so interferes with the development of Caffe.

Please read the guidelines for contributing before submitting this issue.

Issue summary

Hello, thanks for your attention. I build the caffe environmnet follow the tutorial, and make runtest without error. I try to train the example module by python examples/ssd/ssd_Pascal.py. when the script arrives to conv1_1 ->con1-1 , it alway present the problem that "*Aborted at ** (unit time) try "data -d @" if you are using GNU date***" Here is my job I0522 12:55:37.963074 23676 layer_factory.hpp:77] Creating layer conv1_1 I0522 12:55:37.963090 23676 net.cpp:100] Creating Layer conv1_1 I0522 12:55:37.963095 23676 net.cpp:434] conv1_1 <- data_data_0_split_0 I0522 12:55:37.963104 23676 net.cpp:408] conv1_1 -> conv1_1 Aborted at 1526964938 (unix time) try "date -d @1526964938" if you are using GNU date PC: @ 0x7fb4611445bb (unknown) SIGSEGV (@0x4) received by PID 23676 (TID 0x7fb46086c700) from PID 4; stack trace: *** @ 0x7fb491be04b0 (unknown) @ 0x7fb4611445bb (unknown) @ 0x7fb4617123fd (unknown) @ 0x7fb461147426 (unknown) @ 0x7fb461125170 (unknown) @ 0x7fb461723f9d (unknown) @ 0x7fb46112bfd2 (unknown) @ 0x7fb46112e391 (unknown) @ 0x7fb461124aec __cuda_CallJitEntryPoint @ 0x7fb4694245c2 fatBinaryCtl_Compile @ 0x7fb3d357cf92 (unknown) @ 0x7fb3d357db13 (unknown) @ 0x7fb3d357e1e3 (unknown) @ 0x7fb3d3478277 (unknown) @ 0x7fb3d3478b1d (unknown) @ 0x7fb3d3477a9d (unknown) @ 0x7fb4895dcac9 (anonymous namespace)::opencl_fn6<>::switch_fn() @ 0x7fb489773c49 cv::ocl::Context::getDefault() @ 0x7fb48977548b cv::ocl::Device::getDefault() @ 0x7fb4897754ed cv::ocl::useOpenCL() @ 0x7fb487b9480c cv::resize() @ 0x7fb493cbfe8a caffe::ApplyResize() @ 0x7fb493d227c2 caffe::DataTransformer<>::Transform() @ 0x7fb493d25ac1 caffe::DataTransformer<>::Transform() @ 0x7fb493d2eabb caffe::DataTransformer<>::Transform() @ 0x7fb493d2ebb8 caffe::DataTransformer<>::Transform() @ 0x7fb493d2ec8e caffe::DataTransformer<>::Transform() @ 0x7fb493c2f679 caffe::AnnotatedDataLayer<>::load_batch() @ 0x7fb493bbcfdf caffe::BasePrefetchingDataLayer<>::InternalThreadEntry() @ 0x7fb493d32595 caffe::InternalThread::entry() @ 0x7fb486e2b5d5 (unknown) @ 0x7fb47f2f06ba start_thread

i guess that it is possible for my bad caffe-environment. So i try to run mnist and cifar train, and the result is success. i guess that the dataset maybe become mistake. So i make my datatset from VOC2007 follow the tutorial, and the result also be error in the same place.

Then i find many way from the internet

  1. sudo python examples/ssd/ssd_pascal.py it present NO module name caffe. but, i can import caffe without sudo.

2.someone say recompile the opencv without opencl i try to do that, but also not use. i put CMakefile --WITH_OPENCL = OFF; and WITH_OPENCLAMDFFT,WITH_OPENCLAMDBLAS,WITH_VA_INTEL also being OFF ( i cmake opencv and can see the opencl is NO )

Finally, i do not know how to solve the problem. Forgive my lame English, and hoping to get a suggestion.

Steps to reproduce

If you are having difficulty building Caffe or training a model, please ask the caffe-users mailing list. If you are reporting a build error that seems to be due to a bug in Caffe, please attach your build configuration (either Makefile.config or CMakeCache.txt) and the output of the make (or cmake) command.

Your system configuration

Operating system:ubuntu 16.04.3 Compiler: OpenCV:3.1.0 CUDA version (if applicable): 8.0 CUDNN version (if applicable):5.1.10 BLAS: Python or MATLAB version (for pycaffe and matcaffe respectively): Python

mantou22 commented 5 years ago

你好,请问你解决这个问题了吗? 我也遇到了相同的问题

linthy commented 5 years ago

你好,请问你解决这个问题了吗? 我也遇到了相同的问题

重装了caffe,就可以了。我猜应该是环境被破坏了。