error retrieving driver version: Unimplemented: kernel reported driver version not implemented on Wi

关注Tesorflow很久了,但是一直处于搭建环境阶段,说实话国内搞Tesorflow难度还是有点坑的,资料需要fq,下载更新需要fq,搜索看教程还需要fq。关键是Tesorflow本身也有很多问题,版本必须搭配才可以,这次我就遇到了版本的问题。

以前是在mac上跑了一个小例子。跑通了,但是很慢,而且是我公司开发的mac,被我跑的呼呼响,我还是挺心疼的。据说GPU版本的比较快,所以我就在自己的神船上搭建了一下环境

设备:
window10
i7
GTX960M

官方安装教程看这里:https://www.tensorflow.org/install/install_windows
CUDA9.0
cudnn7.1
然后就出了下面的问题:

2018-04-08 21:49:36.244256: E T:\src\github\tensorflow\tensorflow\stream_executor\cuda\cuda_dnn.cc:396] Loaded runtime CuDNN library: 7102 (compatibility version 7100) but source was compiled with 7003 (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.
2018-04-08 21:49:36.256412: F T:\src\github\tensorflow\tensorflow\core\kernels\conv_ops.cc:712] Check failed: stream->parent()->GetConvolveAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo(), &algorithms)

我看了**释,感觉到是版本不匹配,提示说源码是在7003版本编译的,但是我的是7102,所以我又去https://developer.nvidia.com/rdp/cudnn-download这里,下载了一个

error retrieving driver version: Unimplemented: kernel reported driver version not implemented on Wi

老版本的cudnn,然后顺利编译,速度确实比我用mac的cpu跑得快,大概几分钟就完成了,mac当时好像跑了一两个小时