Something wrong with the code

Nerual Net.py

import math, random

#Def XOR
def XOR(x1,x2):
    if(x1 != x2):
        return 1;
    else:
        return 0;

#Sigmoid old school
def sigmoid(x):
    return 1/ (1+math.exp(-x) );

#The perceptron class
class perceptron(object):
    def __init__(self):
        self.weight = [random.uniform(-1,1), random.uniform(-1,1), random.uniform(-1,1)];
        self.value = 0;

#Creating the perceptrons
p = [];
for i in range(0,2):
    p.append(perceptron());
o = perceptron();

#Forwarding in the neural network
def forward():
    #Zeroing the values so we dont add old values with the new ones
    o.value = 0;
    for i in range(0,2):
        p[i].value = 0;

    #Forwarding
    for y in range(0,2):
        for x in range(0,3):
            p[y].value += inputs[x] * p[y].weight[x];

    for i in range(0,2):
        p[i].value = sigmoid(p[i].value);
    
    for i in range(0,2):
        o.value += p[i].value * o.weight[i];

    o.value += o.weight[2] * 1;
    
    o.value = sigmoid(o.value);

#As the name of the function states this is the backpropagation
def backPropagation():

    error = XOR(inputs[0], inputs[1]) - o.value;
    
    errorh = [None,None];
    errorh[0] = (o.weight[0] / (o.weight[0] + o.weight[1])) * error;
    errorh[1] = (o.weight[1] / (o.weight[0] + o.weight[1])) * error;

    for i in range(0,2):
        o.weight[i] += error * p[i].value * 0.1;

    o.weight[2] += error * 0.1;

    for y in range(0,2):
        for x in range(0,2):
            p[y].weight[x] += errorh[y] * inputs[x] * 0.1;

    for i in range(0,2):
        p[i].weight[2] += errorh[i] * 0.1;


inputs = [None,None,None];
#Supposed to be the training   
for i in range(0,10000):
    inputs = [random.randint(0,1), random.randint(0,1), 1];
    forward();
    backPropagation();
    
print("done");

#This is the manual testing
while True:
    inputs = [float(input()),float(input()),1]
    forward();
    backPropagation();
    print(o.value);