Text this: Integrating Reinforcement Learning into M/M/1/K Retry Queueing Models for 6G Applications