Good point, Roel. Perhaps in the final layers one could make it predict a model of the expected score distribution (before combining with the komi and other rules specific adjustments for handicap stones, pass stones, last-play parity, etc.). Should be easy enough to back-propagate win/loss information (and perhaps even more) through such a model.
On Thu, Oct 26, 2017 at 3:55 PM, Roel van Engelen <gosuba...@gmail.com> wrote: > @Gian-Carlo Pascutto > > Since training uses a ridiculous amount of computing power i wonder if it > would > be useful to make certain changes for future research, like training the > value head > with multiple komi values <https://arxiv.org/pdf/1705.10701.pdf> > > On 26 October 2017 at 03:02, Brian Sheppard via Computer-go < > computer-go@computer-go.org> wrote: > >> I think it uses the champion network. That is, the training periodically >> generates a candidate, and there is a playoff against the current champion. >> If the candidate wins by more than 55% then a new champion is declared. >> >> >> >> Keeping a champion is an important mechanism, I believe. That creates the >> competitive coevolution dynamic, where the network is evolving to learn how >> to beat the best, and not just most recent. Without that dynamic, the >> training process can go up and down. >> >> >> >> *From:* Computer-go [mailto:computer-go-boun...@computer-go.org] *On >> Behalf Of *uurtamo . >> *Sent:* Wednesday, October 25, 2017 6:07 PM >> *To:* computer-go <computer-go@computer-go.org> >> *Subject:* Re: [Computer-go] Source code (Was: Reducing network size? >> (Was: AlphaGo Zero)) >> >> >> >> Does the self-play step use the most recent network for each move? >> >> >> >> On Oct 25, 2017 2:23 PM, "Gian-Carlo Pascutto" <g...@sjeng.org> wrote: >> >> On 25-10-17 17:57, Xavier Combelle wrote: >> > Is there some way to distribute learning of a neural network ? >> >> Learning as in training the DCNN, not really unless there are high >> bandwidth links between the machines (AFAIK - unless the state of the >> art changed?). >> >> Learning as in generating self-play games: yes. Especially if you update >> the network only every 25 000 games. >> >> My understanding is that this task is much more bottlenecked on game >> generation than on DCNN training, until you get quite a bit of machines >> that generate games. >> >> -- >> GCP >> _______________________________________________ >> Computer-go mailing list >> Computer-go@computer-go.org >> http://computer-go.org/mailman/listinfo/computer-go >> >> >> _______________________________________________ >> Computer-go mailing list >> Computer-go@computer-go.org >> http://computer-go.org/mailman/listinfo/computer-go >> > > > _______________________________________________ > Computer-go mailing list > Computer-go@computer-go.org > http://computer-go.org/mailman/listinfo/computer-go >
_______________________________________________ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
