DigitsRecognitionNeuralNetwork.ipynb throws Exception of type 'System.OutOfMemoryException'

[alexhiggins732]

Running the default docker DigitsRecognitionNeuralNetwork.ipynb throws an System.OutOfMemoryException when calling var mnist = await MnistModelLoader.LoadAsync("mnist", true);

OS: Windows 10 version 2204 (build 19041.421) Docker version 19.03.8, build afacb8b Host: Memory 16 GB

Steps to reproduce:

1. docker run --name scisharp -it -p 8888:8888 scisharpstack/scisharpcube
2. Browse to local host to authorize token http://127.0.0.1:8888/?token=[token]
3. Browse to sample: http://127.0.0.1:8888/notebooks/samples/DigitsRecognitionNeuralNetwork.ipynb
4. Click run on each block. A System.OutOfMemoryException is thrown when clicking run on the blockvar mnist = await MnistModelLoader.LoadAsync("mnist", true);

Stack Trace: Exception of type 'System.OutOfMemoryException' was thrown. at System.Array.InternalCreate(Void* elementType, Int32 rank, Int32* pLengths, Int32* pLowerBounds) at System.Array.CreateInstance(Type elementType, Int32[] lengths) at NumSharp.Backends.TypedArrayStorage.Allocate(Shape shape, Type dtype) at NumSharp.Backends.DefaultEngine.Multiply(NDArray x, NDArray y) at Tensorflow.Hub.MnistDataSet..ctor(NDArray images, NDArray labels, Type dataType, Boolean reshape) at Tensorflow.Hub.MnistModelLoader.LoadAsync(ModelLoadSetting setting) at Tensorflow.Hub.MnistModelLoader.LoadAsync(String trainDir, Boolean oneHot, Nullable1 trainSize, Nullable1 validationSize, Nullable1 testSize) at Submission#10.

Additional Information: The Hello World and Linear Regression samples run without issue.