torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 90.00 MiB (GPU 0; 7.93 GiB total capacity; 758.34 MiB already allocated; 5.75 MiB free; 858.00 MiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
原因,我选的卡号选错了,
import osos.environ["CUDA_VISIBLE_DEVICES"] = "0"
确认好两件事:
1、本地文件和远程文件同步好了
2、代码中有没有指定哪块GPU的操作
他这个报错很反直觉的一个地方:如果你指定了2卡,2卡显存满了,他会说0卡显存满了,你去看0发现0根本没人用,这就很容易被绕进去
来源地址:https://blog.csdn.net/gongkeguo/article/details/127674524