RuntimeError: CUDA error (10): invalid device ordinal

两张卡跑的train，在本机一张卡上test不了
结果很简单，直接将 model.module 拿出来就行了
因为之前双卡用的parallel

def __init__(self, module, device_ids=None, output_device=None, dim=0):
        super(DataParallel, self).__init__()

        if not torch.cuda.is_available():
            self.module = module
            self.device_ids = []
            return

        if device_ids is None:
            device_ids = list(range(torch.cuda.device_count()))
        if output_device is None:
            output_device = device_ids[0]

详见：parallel源码
https://pytorch.org/docs/stable/_modules/torch/nn/parallel/data_parallel.html#DataParallel