dog-qiuqiu / MobileNet-Yolo

MobileNetV2-YoloV3-Nano: 0.5BFlops 3MB HUAWEI P40: 6ms/img, YoloFace-500k:0.1Bflops 420KB:fire::fire::fire:
Other
1.72k stars 278 forks source link

关于cudnn编译的问题 #30

Open syxrz789 opened 3 years ago

syxrz789 commented 3 years ago

因为CUDA Version: 11.0,所以装的cudnn为 v8.0.4,设置cuda=1,cudnn=1,然后编译,然后报错了,如下 error:CUDNN_CONVOLUTION_FWD_PREFER_FASTEST undeclared (first use in this function); did you mean CUDNN_CONVOLUTION_BWD_FILTER_ALGO_3? int forward_algo = CUDNN_CONVOLUTION_FWD_PREFER_FASTEST; ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ CUDNN_CONVOLUTION_BWD_FILTER_ALGO_3 compilation terminated due to -Wfatal-errors.

网上有找原因,可能是cudnn版本过高导致,只能cudnnv3-v7版本,但是对应cuda11.0的cudnn只有v8版本的

然后同样的配置编译AlexeyAB/darknet的版本能够通过,但我想做mobilenet版本的yolo

KingofLong commented 3 years ago

我是吧cuda11卸了然后用cuda10.0配7.6.5 的cudnn,然后用cmake做的,不卸载cuda11的话cmake的CMAKE_CUDA_COMPILER总是会检测到cuda11,这个mobilenet-yolo的代码应该是要看AlexeyAB/darknet上个yolov3 release的readme来配置