How to map two different graphs together when sampling with NeighborLoader

[SulafahKhaled]

Hello guys, I hope you are doing great.
I am trying to use two different graphs as input to a model, and I need to align the two graphs together. I have a neighborLoader, and two graphs with exactly the same nodes, ordered in the same way for creating the adjacency matrix, but with different edges. I want to make sure that the two graphs are aligned when I sample the batch. Meaning that each batch includes exactly the same nodes for each iteration. For my current logic. I am enforcing that by saying g1.n_id=g2.n_id
however, this doesn't seem to work.
I am using stratified cross-validation and Optuna hyperparameters optimizer.
thank you.