3 ms·
So we could see Adam optimization with hyper-parameter tuning by Adam? But IIRC this is never guaranteed to always find the global minima/maxima, so why would i
by skunkworker 7y ago
So we could see Adam optimization with hyper-parameter tuning by Adam? But IIRC this is never guaranteed to always find the global minima/maxima, so why would it in this scenario? Or is this just finding a way to let your optimizer self optimize?
- myrryr 7y agoPretty much. pick something stupid, but good enough. What IS interesting, is that the problems all end up in the same place after a few levels of recursion.