torch.distributed.elastic.multiprocessing.errors.ChildFailedError, when running the train_new.py

[Skywalker-Harrison]

I encountered a ChildFailedError,
I used the command python -m torch.distributed.run train_new.py
below is the error information.

The program was run on py3.10, and pip list is as below

[Skywalker-Harrison]

This code does not work as well when using the following command in the Readme
accelerate launch --config_file config/default_config.yaml train_new.py

[MonkeyTB]

This code does not work as well when using the following command in the Readme accelerate launch --config_file config/default_config.yaml train_new.py

解决了吗？老哥
和 gcc 版本有关系没