In numpy you can raise floating points error using seterr([all, divide, over, under, invalid]) . Is there something similar in Pytorch?
In numpy you can raise floating points error using seterr([all, divide, over, under, invalid]) . Is there something similar in Pytorch?