0

How to fix the error. Also i dont see CuDNN7.0.03 version. So please any lead would be helpful. BELOW IS THE ERROR:

Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3093 MB memory) -> physical GPU (device: 0, name: GeForce 840M, pci bus id: 0000:04:00.0, compute capability: 5.0) 2019-10-14 01:12:58.508334: E T:\src\github\tensorflow\tensorflow\stream_executor\cuda\cuda_dnn.cc:396] Loaded runtime CuDNN library: 7501 (compatibility version 7500) 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. 2019-10-14 01:12:58.521420: F T:\src\github\tensorflow\tensorflow\core\kernels\conv_ops.cc:712] Check failed: stream->parent()->GetConvolveAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo(), &algorithms)

sinG20
  • 143
  • 2
  • 15

1 Answers1

0

First, you need to get rid of tensorflow1.7 installation if you want to use Tensorflow-gpu, so please uninstall it completely and keep only tensorflow-gpu.

Also, as I could see in your dependencies versions, it happens to require cuDNN version 7.0 with TF1.7, which is also specified in the error raised, so downgrade your cuDNN to 7.0 from here. For all the compatible versions, you can follow this list.

Rishabh Sahrawat
  • 2,437
  • 1
  • 15
  • 32
  • I unistalled tensorflow, and the error was tensorflow module is not found. So i reinstalled the same verison of tensorflow, and the issue fixed. I have no idea how it fixed. But i still have the tensorflow-gpu also. – sinG20 Oct 15 '19 at 08:26
  • it sometimes work if one has both versions, but it also conflicts many times. Good thing is your problem is solved. Cheers! – Rishabh Sahrawat Oct 15 '19 at 08:28
  • Thanks a lot @Rishabh Sahrawat – sinG20 Oct 15 '19 at 08:33