在创建docker时配置较大的共享内存，加入参数--shm-size="15g"，设置15g（根据实际情况酌量设置）的共享内存：
nvidia-docker run -it --name [container_name] --shm-size="15g" ...
这样可以在训练神经网络时把number of workers调大一些，也不会报错：RuntimeError: DataLoader worker (pid 57431) is killed by signal: Bus error. It is possible that dataloader's workers are out of shared memory. Please try to raise your shared memory limit.