num_labels in the output layer

[gehaniaarti]

Hello,

I am trying to use the same code for classifying 31 labels. However, while calculating the pre-training accuracy I am getting the error ValueError: Shapes (None, 31) and (None, 99, 31) are incompatible. The X.shape and Y.shape in my case are (67871, 99, 32) and (67871, 31), respectively. How may I fix the error?

Thanks!