这段时间在liunx服务器上训练模型代码,在pycharm中输入运行命令后总是出现各种关于内存的问题,像下面这样,代码第一轮还没跑完就直接被杀死,搞的头很大
RuntimeError: DataLoader worker (pid 12845) 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.
RuntimeError: DataLoader worker (pid(s) 15095) exited unexpectedly
RuntimeError: unable to write to file </torch_16319_2728251979_5>: No space left on device (28)
查了资料发现是liunx系统下的dev/shm虚拟内存太小,查了一下只有64M,完全不够运行代码
解决办法
查看大小
df -h /dev/shm
vi进去文件后,加入第二行代码
vi /etc/fstab
tmpfs /dev/shm tmpfs defaults,size=4096M 0 0
输入两行代码,重新挂载
umount /dev/shm
mount /dev/shm
最后查看修改的内存是否成功
df -h /dev/shm
贴一下我解决后的图