But someone starting work in that area can take the old frame for another spin. They would learn a lot, even if they don't improve anything.
They can start with the "relatively" low hanging fruit of the race net. (even though it is not as low hanging as some may think.) Oystein can add more on that. -Joseph On Mon, 19 Oct 2020 at 10:08, Philippe Michel <[email protected]> wrote: > On Fri, Oct 16, 2020 at 04:19:32PM +0000, Aaron Tikuisis wrote: > > > How often/recently is the gnubg NN retrained? Do we think it can be > > improved much? > > The current nets date back from 2012. > > They certainly can be improved, but to make a difference one would need > some useful change in either the inputs, the nets structure, the > training database (in case of supervised training) or the training method. > > >
