RuntimeWarning: overflow encountered in exp - Neural networks - Sigmoid

Question

I am working with Neural Networks and I completed all my code and it is time to train my network and when I stated training I see the following error

and the progress of my model stayed constant at 980/10,000 is there any way that I can fix it. I am following a tutorial and it is on digit_recognisation. And it worked for him so well......

and my code for the sigmoid function is just:

def sigmoid(z):

    return 1.0/1.0+np.exp(-z)

Thank You