【Caffe】Check failed: status == CUDNN_STATUS_SUCCESS (3 vs. 0) CUDNN_STATUS_BAD_PARAM

如果是在编译caffe阶段 运行 make runtest 时报错，在Makefile.config文件里将USE_CUDNN=1注释掉，但此方法可能达不到cudnn加速的效果

---

如果是在训练阶段报错，则在卷积层convolution_param中添加engine: CAFFE

---

https://blog.csdn.net/goodxin_ie/article/details/89057095

https://blog.csdn.net/grief_of_the_nazgul/article/details/46361699

https://blog.csdn.net/sjh123223/article/details/79659544