Reinforcement Learning for SBM Graphon Games with Re-Sampling