Hi, I run train.py, but every time after the validation of the first epoch, there is an error out of memory when I proceed to the second epoch. With 3080/3090, the memory suddenly increases at the beginning of the second epoch
Is there any solution to this problem
Hi, I run train.py, but every time after the validation of the first epoch, there is an error out of memory when I proceed to the second epoch. With 3080/3090, the memory suddenly increases at the beginning of the second epoch Is there any solution to this problem