The gradient descent method for the convexification to solve boundary value problems of quasi-linear PDEs and a coefficient inverse problem.