Add cuda set_device for local run

[rayg1234]

I was hoping to remove this way of assigning devices to GPUs but FSDP (https://pytorch.org/docs/main/fsdp.html#torch.distributed.fsdp.FullyShardedDataParallel) still requires it so we need to keep this code and to make local mode work with >1 GPU, we need to set it here as well

[codecov]

Codecov Report

Attention: Patch coverage is 50.00000% with 1 line in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/fairchem/core/common/distutils.py	50.00%	1 Missing ⚠️

Files with missing lines	Coverage Δ
src/fairchem/core/common/distutils.py	56.46% <50.00%> (-0.09%)	⬇️

[misko]

LGTM!