Solve the Numerical applying back propagation. Calculate input $1,$ input $2,$ output then input $1$ error and input $2$ error then update the weights and do similarly for others and then find bias. Dont just give explanation or formula but solve it numerically need calculations.