using torchrun, save_to_disk suddenly shows SIGTERM

### Describe the bug

When I run my code using the "torchrun" command, when the code reaches the "save_to_disk" part, suddenly I get the following warning and error messages:
Because the dataset is too large, the "save_to_disk" function splits it into 70 parts for saving. However, an error occurs suddenly when it reaches the 14th shard.
WARNING: torch.distributed.elastic.multiprocessing.api: Sending process 2224968 closing signal SIGTERM
ERROR: torch.distributed.elastic.multiprocessing.api: failed (exitcode: -7). traceback: Signal 7 (SIGBUS) received by PID 2224967.

### Steps to reproduce the bug

ds_shard = ds_shard.map(map_fn, *args, **kwargs)
ds_shard.save_to_disk(ds_shard_filepaths[rank])

Saving the dataset (14/70 shards): 20%|██ | 875350/4376702 [00:19<01:53, 30863.15 examples/s]
WARNING:torch.distributed.elastic.multiprocessing.api:Sending process 2224968 closing signal SIGTERM
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -7) local_rank: 0 (pid: 2224967) of binary: /home/bingxing2/home/scx6964/.conda/envs/ariya235/bin/python
Traceback (most recent call last):
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/bin/torchrun", line 8, in
sys.exit(main())
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/lib/python3.10/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
return f(*args, **kwargs)
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/lib/python3.10/site-packages/torch/distributed/run.py", line 794, in main
run(args)
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/lib/python3.10/site-packages/torch/distributed/run.py", line 785, in run
elastic_launch(
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 134, in __call__
return launch_agent(self._config, self._entrypoint, list(args))
File "/home/bingxing2/home/scx6964/.conda/envs/ariya235/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 250, in launch_agent
raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError:
==========================================================
run.py FAILED
----------------------------------------------------------
Failures:

----------------------------------------------------------
Root Cause (first observed failure):
[0]:
time : 2023-12-08_20:09:04
rank : 0 (local_rank: 0)
exitcode : -7 (pid: 2224967)
error_file:
traceback : Signal 7 (SIGBUS) received by PID 2224967

### Expected behavior

I hope it can save successfully without any issues, but it seems there is a problem.

### Environment info

`datasets` version: 2.14.6
- Platform: Linux-4.19.90-24.4.v2101.ky10.aarch64-aarch64-with-glibc2.28
- Python version: 3.10.11
- Huggingface_hub version: 0.17.3
- PyArrow version: 14.0.0
- Pandas version: 2.1.2