cuDNN error on zbook when running R-CNN Example

3 views (last 30 days)
Alexander Kelly
Alexander Kelly on 24 Nov 2021
Commented: yanqi liu on 15 Dec 2021
Whenever I run it I get the following error:
Failed to initialize the cuDNN handle. Return code was CUDNN_STATUS_INTERNAL_ERROR.
Error in rcnnObjectDetector.train (line 222)
detector.Network = trainNetwork(dispatcher, layers, opts);
Error in trainRCNNObjectDetector (line 197)
detector = rcnnObjectDetector.train(trainingData, layers, options, params);
Error in TrainRCNNStopSignDetectorExample (line 23)
rcnn = trainRCNNObjectDetector(stopSigns, layers, options, 'NegativeOverlapRange', [0 0.3]);
I did a little research and it looks like it may have something to do with my GPU? But my knowledge on the topic ends there as it seems that my drivers are up to date and such.
I am running Matlab R2017a, below are the specs for the GPU:
CUDADevice with properties:
Name: 'Quadro T1000 with Max-Q Design'
Index: 1
ComputeCapability: '7.5'
SupportsDouble: 1
DriverVersion: 11
ToolkitVersion: 8
MaxThreadsPerBlock: 1024
MaxShmemPerBlock: 49152
MaxThreadBlockSize: [1024 1024 64]
MaxGridSize: [2.1475e+09 65535 65535]
SIMDWidth: 32
TotalMemory: 4.2950e+09
AvailableMemory: 3.2722e+09
MultiprocessorCount: 14
ClockRateKHz: 1350000
ComputeMode: 'Default'
GPUOverlapsTransfers: 1
KernelExecutionTimeout: 1
CanMapHostMemory: 1
DeviceSupported: 1
DeviceSelected: 1

Answers (1)

yanqi liu
yanqi liu on 25 Nov 2021
yes,sir,may be check the cuda and cudnn config,such as
  3 Comments
yanqi liu
yanqi liu on 15 Dec 2021
yes,sir,i think may be use
train on cpu,such as
options = trainingOptions(your other config, ...
'ExecutionEnvironment', 'cpu');
or
update your version, such as R2021 version

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!