When I want to train Fully convolutional neural network, I have the following error

Error using trainNetwork (line 154)

Unexpected error calling cuDNN: CUDNN_STATUS_BAD_PARAM.

Error in Essai_example (line 154)

net = trainNetwork(trainingData,lgraph,options);

Caused by:

Error using nnet.internal.cnngpu.convolveForward2D

Unexpected error calling cuDNN: CUDNN_STATUS_BAD_PARAM.

[Ahmed Zaafouri]

I try to use the answer in the mathworks community but the problem persists

[Ahmed Zaafouri]

Matlab R2018a

and gpuDevice has the following characteristics

CUDADevice with properties:

Name: 'GeForce MX110'

Index: 1

ComputeCapability: '5.0'

SupportsDouble: 1

DriverVersion: 11.2000

ToolkitVersion: 9

MaxThreadsPerBlock: 1024

MaxShmemPerBlock: 49152

MaxThreadBlockSize: [1024 1024 64]

MaxGridSize: [2.1475e+09 65535 65535]

SIMDWidth: 32

TotalMemory: 2.1475e+09

AvailableMemory: 1.6950e+09

MultiprocessorCount: 2

ClockRateKHz: 1006000

ComputeMode: 'Default'

GPUOverlapsTransfers: 1

KernelExecutionTimeout: 1

CanMapHostMemory: 1

DeviceSupported: 1

DeviceSelected: 1

Answer 1

[Joss Knight]

I suspect your best bet here is to upgrade MATLAB to a more recent version. Many of these bugs will be fixed with newer versions of the NVIDIA libraries.