For torch 2.13, we get the following warning about all-gather:
FutureWarning: `torch.distributed.all_gather_into_tensor` is deprecated. Please use `torch.distributed.all_gather_single` instead.
We have also been getting the warning:
UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
Neither of these are blocking issues; just cleanup that should be done at some time.
For torch 2.13, we get the following warning about all-gather:
We have also been getting the warning:
Neither of these are blocking issues; just cleanup that should be done at some time.