How to use multi GPUs to train the model

[Xxiaoting]

When I was training the GATs model using all the Chinese stock market data (by setting instrument = 'all'), it turns out an error 'RuntimeError: CUDA out of memory'. Actually I have two GPUs but only one was using when I did the training. How can I use multi GPUs to do the training?

Thanks for the help in advance.

[gagaein]

Hello, I am facing the same problem. Have you solved it:)

[you-n-g]

Qlib provides a framework for quant research. But it does not manage the computation resource. You integrate https://pytorch.org/docs/stable/generated/torch.nn.DataParallel.html or something else into https://github.com/microsoft/qlib/blob/main/qlib/contrib/model/pytorch_gats.py to fully leverage your resource.

[github-actions[bot]]

This issue is stale because it has been open for three months with no activity. Remove the stale label or comment on the issue otherwise this will be closed in 5 days