This node describes the synthetic dataset used to train and evaluate the deep reinforcement learning (DRL) agent. Three simulators generated the interaction data: Half-Life Regression (HLR), the General Power-Law (GPL) forgetting curve, and the Exponential Forgetting Curve (EFC); EFC was specifically used to generate the training data for the LSTM. Each data point is represented as a tuple {q, a, t, d}, where q is the most recent item shown to the student, a is the [0,1] correctness of the answer, t is the timestamp, and d is the delay between the most recent item and the previous one (fixed at 5s). Data was generated in three ways: (1) Random Sample, where exercises and student answers were randomly collected; (2) Random Policy Tutor, where exercises were chosen at random by the policy but student answers were not; and (3) Supermemo Tutor, where the SuperMemo algorithm selected exercises.