Update custom layer's parameter every iteration

I have defined a custom regression layer as follows

classdef Sparse_Layer < nnet.layer.RegressionLayer
    % Example custom regression layer with mean-absolute-error loss.
    properties
        % (Optional) Layer properties.
        L1_Lambda   =   0;
        

        % Layer properties go here.
    end
    
    methods
        function layer = Sparse_Layer(L1_Lambda)
            		
            % Lambda should be 1xN, N-> Batch size       
            layer.L1_Lambda =   L1_Lambda;
            
            % Set layer description.
            layer.Description = 'Mean absolute error';
        end
        function layer_out = myFun(layer_in)
            
        end
        function loss = forwardLoss(layer, Y, T)
            % loss = forwardLoss(layer, Y, T) returns the MAE loss between
            % the predictions Y and the training targets T.

            % Calculate L1 error mean over batch.
            Dif         =   Y-T;
            Batch_sz    =   size(Y,2);
            L1_err      =   sum(abs(Dif),1);
            L2_err      =   sum(Dif.*Dif,1);
            loss        =   sum(layer.L1_Lambda(iteration_num)*L1_err + .5*L2_err  )/Batch_sz;
            
            
            
        end
    end
end

Now, i want to multiply Lambda(iteration_no), this iteration number is something that i am unable to update. I tried having a layer property that counts the iteration,but that cant work as this a Value Class and not a handler. The reason i want to do this is beacuse i want this lambda to change for every batch (the values i have pre-calcuated based on some feautures of the data in each batch).