Implementing SER

[AlboAlby00]

Hello.
I am planning to implement Strong Experience Replay, here is the paper describing it (https://arxiv.org/abs/2305.13622).
It is a modified version of Der++ that claims to improve it by a noticeable margin.
Would it be appreciated as contribution?