Learning Rate (η):
0.1
Convergence Tolerance (ε):
0.01
Start Gradient Descent