Suppose an RNN is used to predict next character in the input string $$abracadabra$($no upper case needs to be considered$).$ If one hot encoding is used for representing input and output, how many nodes will be needed in the input $[$ Select $]["5","11","26",$ "None of the other"$],$ output $[$ Select $]["11","5",$ "None of the other", $"2"]$ layer?