OnComplete callbacks in Deepleaning frameworks
I was thinking the other day, it would be very nice to have callback in various time-consuming functions. Especially its helpful for indie or hobby researchers, who have to spend they own money on cloud VM with GPU. Something like
would be great to have.