Mergind and shuffling two datasets

[a-ism256]

Hello,

I'm wondering if there's a way to train on two datasets but not sequentially.

So let's say I have Dataset1 containing graphs A, B and C and Dataset2 containing graphs D and E. I would like to be able to generate a random sequence like B, D, A, C, E to train my model on.

Is there a way to currently do this without writing a third Dataset class that processes both sets of graphs?

Thanks.

[rusty1s]

You can use torch.utils.data.ConcatDataset for that, see here.

[a-ism256]

Alright I'll try it thanks! That was a really quick reply 😄.