1: \begin{abstract}

2: Quantum information quantities play a substantial role in characterizing operational quantities in various quantum information-theoretic problems. 

3: We consider the numerical computation of four quantum information quantities: Petz-Augustin information, sandwiched Augustin information, conditional sandwiched R\'enyi entropy, and sandwiched R\'enyi information. 

4: To compute these quantities requires minimizing some order-\( \alpha \) quantum R\'enyi divergences over the set of quantum states. 

5: Whereas the optimization problems are obviously convex, they violate standard bounded gradient/Hessian conditions in literature, so existing convex optimization methods and their convergence guarantees do not directly apply. 

6: In this paper, we propose a new class of convex optimization methods called mirror descent with the Polyak step size. 

7: We prove their convergence under a weak condition, showing that they provably converge for minimizing quantum R\'enyi divergences. 

8: Numerical experiment results show that entropic mirror descent with the Polyak step size converges fast in minimizing quantum R\'enyi divergences. 

9: \end{abstract}

10: