[FEA]: Enable RMM PyTorch allocator for cugraph_dgl

[VibhuJawa]

Is this a new feature, an improvement, or a change to existing functionality?

New Feature

How would you describe the priority of this feature request

Critical (currently preventing usage)

Please provide a clear description of problem this feature solves

We should enable RMM PyTorch allocator for cugraph_dgl by default for Pytorch. This should prevent memory issues as well as enable speedups (2-3x) when using POOL .

RMM PR: rapidsai/rmm#1168

Describe your ideal solution

NA

Describe any alternatives you have considered

No response

Additional context

We will probably need to do the same for PYG

Code of Conduct

• I agree to follow cuGraph's Code of Conduct
• I have searched the open feature requests and have found no duplicates for this feature request

[alexbarghi-nv]

Done for now, this works on single GPU but we do not want to always have it enabled.