pytorch 容器

if result is None:
    result = temp
else:
    result = torch.cat([result, temp])