![]() ![]() This situation frustrates me to no end personally. Often times all that's available are marketing slides with fuzzy performance claims. However, unlike desktop and server CPUs, mobile CPU and GPU vendors tend to do very little architectural disclosure - a fact that we've been working hard to change over the past few years. OpenCV tests: Can't find data file: dnn/layers/shared_weights.As a programmer who wants to write decent performing code, I am very interested in understanding the architectures of CPUs and GPUs. OpenCV tests: Can't find data file: dnn/layers/net_input.prototxt Can't open "/dnn/bvlc_googlenet.prototxt" in function 'ReadProtoFromTextFile' Can't open "/dnn/gtsrb.prototxt" in function 'ReadProtoFromTextFile'Ĭ++ exception with description "OpenCV(4.1.0-pre) /Users/touqeera/work/opencv-master2/modules/dnn/src/caffe/caffe_io.cpp:1121: error: (-2:Unspecified error) FAILED: fs.is_open(). OpenCV tests: Can't find data file: dnn/bvlc_googlenet.prototxtĬ++ exception with description "OpenCV(4.1.0-pre) /Users/touqeera/work/opencv-master2/modules/dnn/src/caffe/caffe_io.cpp:1121: error: (-2:Unspecified error) FAILED: fs.is_open(). Most of test are now failing and the end it shows a message saying '144 FAILED TEST' My bad, ran the opencv_test_dnn - here is what I am getting at the top, the entire run is too big to share. Name = AMD Radeon R9 M395X Compute EngineĬl_APPLE_command_queue_select_compute_units opencv_perf_dnnĬompiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ (ver 9.039)ĬPU features: SSE SSE2 SSE3 SSSE3 SSE4.1 *SSE4.2 *FP16 *AVX *AVX2 *AVX512-SKX? Looking for help and feedback! Steps to please find below the result of running 'opencv_perf_dnn' test: Is this a bug in OpenCV or am I overlooking a step in building OpenCv with OpenCL. } //this works & i can see my video card name & opencl version # Code Snippet #Ĭout << "OpenCL is not available." << endl Ĭout << "OpenCL is AVAILABLE! :) " << endl //this is the outputĬv::ocl::Context ctx = cv::ocl::Context::getDefault() Ĭout << ctx.ndevices() << " GPU devices are detected." << endl Ĭout << "name: " << device.name() << endl Ĭout << "available: " << device.available() << endl Ĭout << "imageSupport: " << device.imageSupport() << endl Ĭout << "OpenCL_C_Version: " << device.OpenCL_C_Version() << endl Running the following code snippet produces the positive output which verifies OpenCL being available. Via OPENCV_OCL4DNN_CONFIG_PATH parameter.Īlso, numerically the prediction is different than what I get from running directly on CPU. "OpenCV(ocl4dnn): consider to specify kernel configuration cache directory Net.setPreferableTarget(cv::dnn::DNN_TARGET_OPENCL) I have build OpenCV with OpenCL target, however when I set the preferable target to OpenCL using Trying to see any speedup possible for my simple Tensorflow network. Operating System / Platform => macOS High Sierra, 10.13.3. ![]()
0 Comments
Leave a Reply. |