how can i fix this error Error using gpuArray/arrayfun Out of memory on device. To view more detail about available memory on the GPU, use 'gpuDevice()'. If the problem persists, reset the GPU by calling 'gpuDevice(1)'.

Question

hadeel manasrah 2017-9-22

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/357898-how-can-i-fix-this-error-error-using-gpuarray-arrayfun-out-of-memory-on-device-to-view-more-detail

评论： Joss Knight 2018-2-7

在 MATLAB Online 中打开

YTest = classify(cifar10Net, testDigitData);

I am using MATLAB 2017a

with GPU

Name: 'GeForce 920MX'
                     Index: 1
         ComputeCapability: '5.0'
            SupportsDouble: 1
             DriverVersion: 8
            ToolkitVersion: 8
        MaxThreadsPerBlock: 1024
          MaxShmemPerBlock: 49152
        MaxThreadBlockSize: [1024 1024 64]
               MaxGridSize: [2.1475e+09 65535 65535]
                 SIMDWidth: 32
               TotalMemory: 2.1475e+09
           AvailableMemory: 1.6472e+09
       MultiprocessorCount: 2
              ClockRateKHz: 993000
               ComputeMode: 'Default'
      GPUOverlapsTransfers: 1
    KernelExecutionTimeout: 1
          CanMapHostMemory: 1
           DeviceSupported: 1
            DeviceSelected: 1

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Matt J 2017-9-22

Use less memory?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Joss Knight 2017-10-16

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/357898-how-can-i-fix-this-error-error-using-gpuarray-arrayfun-out-of-memory-on-device-to-view-more-detail#answer_286090

Try reducing the number of digits you are trying to classify at once, to reduce the amount of memory needed to process them. See the 'MiniBatchSize' setting.

2 个评论
显示无隐藏无

Huawei Tian 2018-2-6

when I set MiniBatchSize as 20, only 20 image are feed to classify! How can I feed next 20 images?

Joss Knight 2018-2-7

If you passed all your data to classify, all of it will be classified. So you'll have to post the code you are using so we can see what you're actually doing.

请先登录，再进行评论。