Python 如何获取本地_rank属性

发动机代码如下所示：

class Engine(object):
    def __init__(self, cfg, custom_parser=None):
        self.version = 0.01
        self.state = State()
        self.devices = None
        self.distributed = False
        self.logger = None
        self.cfg = cfg

        if custom_parser is None:
            self.parser = argparse.ArgumentParser()
        else:
            assert isinstance(custom_parser, argparse.ArgumentParser)
            self.parser = custom_parser

        self.inject_default_parser()
        self.args = self.parser.parse_args()

        self.continue_state_object = self.args.continue_fpath

        if 'WORLD_SIZE' in os.environ:
            self.distributed = int(os.environ['WORLD_SIZE']) > 1

        if self.distributed:
            self.local_rank = self.args.local_rank
            self.world_size = int(os.environ['WORLD_SIZE'])
            self.world_rank = int(os.environ['RANK'])
            torch.cuda.set_device(self.local_rank)
            dist.init_process_group(backend="nccl", init_method='env://')
            dist.barrier()
            self.devices = [i for i in range(self.world_size)]
        else:
            # todo check non-distributed training
            self.world_rank = 1
            self.devices = parse_torch_devices(self.args.devices)

    def setup_log(self, name='train', log_dir=None, file_name=None):
        if not self.logger:
            self.logger = get_logger(
                name, log_dir, self.args.local_rank, filename=file_name)
        else:
            self.logger.warning('already exists logger')
        return self.logger

我想获取local\u rank，但当我只使用Engine.local\u rank时，它返回AttributeError:type对象“Engine”没有属性“local\u rank” 这样地

---

您的

local\u排名取决于
self.distributed==True
或
self.distributed=0
这意味着
'WORLD\u SIZE'
需要在
os.environ
中，所以只需添加环境变量
WORLD\u SIZE
（应该是整数）
引擎。local\u排名与self.local\u排名不同。Engine.local_rank建议local_rank是类变量，而不是实例变量。谢谢您的回答。我想我已经讲清楚了