1: \begin{abstract}

2: The stochastic Landau--Lifshitz--Gilbert (LLG) equation coupled with the Maxwell

3: equations (the so called stochastic MLLG system) describes the creation of

4: domain walls and vortices (fundamental objects for the novel nanostructured

5: magnetic memories).

6: We first reformulate the stochastic LLG equation into an equation 

7: with time-differentiable solutions.

8: We then propose a convergent $\theta$-linear scheme to approximate the

9: solutions

10: of the reformulated system. As a consequence, we prove convergence of the

11: approximate solutions, with no or minor conditions on time and space steps

12: (depending on the value of $\theta$).

13: Hence, we prove the existence of weak martingale solutions of the stochastic 

14: MLLG system. Numerical results are presented to show applicability of the method.

15: \end{abstract}