I had installed tensorflow-gpu follow the Installing TensorFlow on Windows,and it seems work well when I import tensorflow, but today when I try a new program, it shows some errors:
`
2017-07-12 09:33:31.301985: E c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\stream_executor\cuda\cuda_dnn.cc:352] Loaded runtime CuDNN library: 6021 (compatibility version 6000) but source was compiled with 5105 (compatibility version 5100). 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.
2017-07-12 09:33:31.302851: F c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\kernels\conv_ops.cc:659] Check failed: stream->parent()->GetConvolveAlgorithms(&algorithms)
`
I knows it means that I should upgrade my cudnn5.1 to cudnn6, and I download cudnn6 and copy the files to my cuda install directory. Then in the directory cudn/v8.0/bin exist two cudnn dll:cudnn64_5.dll and cudnn64_6.dll.
When I rerun my program, the same errors still occur and when I delete cudnn64_5.dll，I even cannot run my program, some new errors happen:
`
builtins.ImportError: Traceback (most recent call last):
File "c:\Users\chenc\Anaconda3\envs\tensorflow-gpu\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in swig_import_helper
return importlib.import_module(mname)
File "c:\Users\chenc\Anaconda3\envs\tensorflow-gpu\lib\importlib\__init__.py", line 126, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "<frozen importlib._bootstrap>", line 986, in _gcd_import
File "<frozen importlib._bootstrap>", line 969, in _find_and_load
File "<frozen importlib._bootstrap>", line 958, in _find_and_load_unlocked
File "<frozen importlib._bootstrap>", line 666, in _load_unlocked
File "<frozen importlib._bootstrap>", line 577, in module_from_spec
`
I rename my cudnn64_6.dll to cudnn64_5..dll，errors above gone but still shows that my cudnn version is 5105.It seems like that I have to do other things to make the cudnn upgrade work,but I don't know.What should I do to resolve the error?
I am not sure how to solve your current problem, but I am sure you should not update to cuDNN_6.0. It is not currently supported with the latest version (1.2) and will not work.