Using `torch.distributed.all_gather_object` returns error when using 1 GPU but works fine for multiple GPUs

[muellerzr]

@seanswyi

Open to any other opinions.

accelerator.gather() is guarenteed to always work even on single node