1: \begin{abstract}

2: Hybrid precoding plays a key role in realizing massive \ac{mimo} transmitters with controllable cost. \ac{mimo} precoders are required to frequently adapt based on the variations in the channel conditions. In hybrid \ac{mimo}, where precoding is comprised of digital and analog beamforming, such an adaptation  involves lengthy optimization and  depends on accurate \ac{csi}. This affects the spectral efficiency  when the channel varies rapidly and when operating with noisy \ac{csi}. In this work we employ deep learning techniques to learn how to {\em rapidly} and {\em robustly} optimize hybrid precoders, while being fully interpretable. 

3: We leverage data to learn iteration-dependent hyperparameter settings of projected gradient sum-rate optimization with a predefined number of iterations. The algorithm maps channel realizations into hybrid precoding settings while preserving the interpretable flow of the optimizer and improving its convergence speed. To cope with noisy \ac{csi}, we learn to optimize the minimal achievable sum-rate among all tolerable errors, proposing a robust hybrid precoding based on the projected conceptual mirror prox minimax optimizer. 

4: Numerical results  demonstrate that our approach allows using over ten times less iterations compared to that required by conventional optimization with shared hyperparameters, while achieving similar and even improved sum-rate performance.

5: \end{abstract}

6: