7301 (compatibility version 7300) but source was compiledwith 7004 (compatibility version 7000)

最新推荐文章于 2021-03-17 19:26:38 发布

南有木兮木不知

最新推荐文章于 2021-03-17 19:26:38 发布

阅读量854

点赞数

分类专栏： AI

本文链接：https://blog.csdn.net/mhsszm/article/details/88350935

版权

AI 专栏收录该内容

36 篇文章 0 订阅

订阅专栏

错误报告如下：

2019-03-08 11:35:37.373619: E tensorflow/stream_executor/cuda/cuda_dnn.cc:378]
 Loaded runtime CuDNN library: 7301 (compatibility version 7300) but source was compiled
  with 7004 (compatibility version 7000).  If using a binary install, upgrade your CuDNN library to match.  
  If building from sources, make sure the library loaded at runtime matches a compatible version
   specified during compile configuration.
2019-03-08 11:35:37.374470: 
F tensorflow/core/kernels/conv_ops.cc:717] Check failed: stream->parent()->GetConvolveAlgorithms( 
conv_parameters.ShouldIncludeWinogradNonfusedAlgo<T>(), &algorithms) 
Aborted (core dumped)

说下我的linux环境：
GTX1060 + NVIDIA-Linux-x86_64-384.111 + cuda_9.0.176_384.81_linux +cudnn-9.0-linux-x64-v7.3.1.20.tgz+tensorflow-gpu 1.6.0
处理方法：
跑测试程序出现如上问题，意思是cudnn版本高了，降到cudnn-9.0-linux-x64-v7.tgz就好了。