gradient_descent.py

Deriving the cost function is only half the battle. To actually "learn," a model requires an optimization engine that shifts its parameters toward the lowest possible error without manual intervention