Steepest descent least-squares optimisation - derivation explained (watch before conjugate gradient)
How to find the step size for the gradient of the least-squares objective or cost function
The steepest descent algorithm is important to understand well before seeking to understand and derive the conjugate gradient method