Error using digraph.va​lidateEdge​Properties (line 363) Edge properties must be a table.

I am training a LSTM network with training data of dimension 2x142. YTrain is a categorical input. i have ran the same code and got output yesterday. but today it suddenly shows the error

Error using trainNetwork (line 183)

Edge properties must be a table.

Error using digraph.validateEdgeProperties (line 363)

Edge properties must be a table.

my code is

Labels=categorical(Response);
    CnX = Signals(Labels=='1');
    CnY = Labels(Labels=='1');
    
    SzX = Signals(Labels=='0');
    SzY = Labels(Labels=='0');
    [trainIndCn,ValIndCn,testIndCn] = dividerand(75,0.7,0.0,0.3);
    [trainIndSz,ValIndSz,testIndSz] = dividerand(72,0.7,0.0,0.3);
    
    XTrainCn = CnX(trainIndCn);
    YTrainCn = CnY(trainIndCn);
    XTrainSz = SzX(trainIndSz);
    YTrainSz = SzY(trainIndSz);
    
    XTestCn = CnX(testIndCn);
    YTestCn = CnY(testIndCn);
    XTestSz = SzX(testIndSz);
    YTestSz = SzY(testIndSz);
    
    XTrain=[XTrainCn,XTrainSz];
    YTrain=[YTrainCn,YTrainSz];
    
    XTest=[XTestCn,XTestSz];
    YTest=[YTestCn,YTestSz];
     Val={XTest, YTest};
    
     numFeatures = 127;
    numResponses = 2;
    numHiddenUnits4 =200;
    layers = [ ...
        sequenceInputLayer(4)
         bilstmLayer(numHiddenUnits4,'OutputMode','sequence')
          fullyConnectedLayer(50)
             dropoutLayer(0.5)
          fullyConnectedLayer(2)
        softmaxLayer
        classificationLayer];
    opts=trainingOptions('adam',...
        'MiniBatchSize',40, ...
        'MaxEpochs',100, ...
        'Shuffle','every-epoch',...
        'InitialLearnRate',0.001, ...
         'SequenceLength', 'longest',...
         'ValidationData',Val, ...
        'ValidationFrequency',30, ...
        'ValidationPatience',Inf, ...
           'Verbose',true, ...
        'VerboseFrequency',30, ...
        'Plots','training-progress');
    net = trainNetwork(XTrain,YTrain,layers,opts);
    trainPred = classify(net,XTrain,'SequenceLength','longest');
    figure 
    plotconfusion(YTrain',trainPred','Training Accuracy');
    testPred = classify(net,XTest,'SequenceLength','longest');
    figure
    plotconfusion(YTest',testPred','Testing Accuracy');

[yanqi liu]

yes，sir，may be check your data on today，is it normal