The large number of antennas in 5G massive MIMO systems allows the base station to communicate with multiple users at the same time and frequency resource with multi-user beamforming. However, highly correlated user channels could drastically impede the spectral efficiency that multi-user beamforming can achieve. As such, it is critical for the base station to schedule a suitable group of users in each time and frequency resource block to achieve maximum spectral efficiency while adhering to fairness constraints among the users. Additionally, in dynamic wireless environments characterized by rapidly changing channel conditions, selecting the appropriate transmission link parameters, such as modulation and coding scheme (MCS), presents another formidable task for maximizing spectral efficiency. Inspired by recent achievements in machine learning to solve problems in wireless communication, we propose SMART, a dynamic scheduler for massive MIMO combining the state-of-the-art Soft Actor-Critic (SAC) DRL model with the K-Nearest Neighbors (KNN) algorithm, and a Convolutional Neural Networks and Long Short-Term Memory networks (CNN-LSTM) based adaptive MCS technique. Through comprehensive simulations using realistic massive MIMO channel models as well as real-world datasets from channel measurement experiments on the RENEW platform, we demonstrate the effectiveness of our proposed models in various channel conditions.