[Computer-go] Please keep the old mailing list up for a while

Thanks Michael and Petr for setting up computer-go at computer-go.org . Please keep the old mailing list computer-go at dvandva.org up for a while. We will investigate the issue of Gm

Re: [Computer-go] Teaching Deep Convolutional Neural Networks to Play Go

Chris Maddison also produced very good (in fact much better) results using a deep convolutional network during his internship at Google. Currently waiting for publication approval, I will post the paper once it is passed. Aja On Mon, Dec 15, 2014 at 2:59 PM, Erik van der Werf wrote: > > Thanks f

Re: [Computer-go] Teaching Deep Convolutional Neural Networks to Play Go

2014-12-15 21:31 GMT+00:00 Petr Baudis : > > Still, strong play makes sense for a strong predictor. I believe I > can also beat GNUGo >90% of time in blitz settings without doing pretty > much *any* concious sequence reading. So I would expect a module that's > supposed to mirror my intuition t

Re: [Computer-go] Teaching Deep Convolutional Neural Networks to Play Go

2014-12-15 23:29 GMT+00:00 Petr Baudis : > > Huh, aren't you? > > I just played quick two games GnuGoBot39 where I tried very hard not > to read anything at all, and had no trouble winning. (Well, one of my > groups had some trouble but mindless clicking saved it anyway.) That well explains

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional NeuralNetworks

2014-12-20 11:33 GMT+00:00 Hiroshi Yamashita : > > I don't understand minibatch. > Does CNN need 0.15sec for a positon, or 0.15sec for 128 positions? > 0.15 sec for 128 positions. > ABCDEFGHJ > 9. White(O) to move. > 8...OO Previous Black move is H5(X) > 7..XXXOO.. > 6.XXO.

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional NeuralNetworks

Hi Hiroshi, 2014-12-20 3:31 GMT+00:00 Hiroshi Yamashita : > > But it looks playing strength is similar to Clark's CNN. > Against GnuGo our 12-layer CNN is about 300 Elo stronger (97% winning rate against 86% based one the same KGS games). Against Fuego using their time setting (10 sec per move on

Re: [Computer-go] Fuego 1.1 vs current Fuego

Hi Martin, Thanks for the explanation. I'll write Fuego runs 2 (considering early stop) * 2.2 / 3.1 roughly 14200 playouts per move on the empty board using their time limits. Thanks, Aja On Sat, Dec 20, 2014 at 11:27 PM, Martin Mueller wrote: > > I ran Fuego (latest svn version) on our

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional Neural Networks

> >> > >> I am still fighting with the NN slang, but why do you zero-padd the > >> output (page 3: 4 Architecture & Training)? > >> > >> From all I read up to now, most are zero-padding the input to make the > >> output fit 19x1

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional NeuralNetworks

2014-12-21 3:02 GMT+00:00 Hiroshi Yamashita : > > I tried Fuego 1.1(2011, Windows version) on Intel Core i3 540, > 2 cores 4 thread. 3.07GHz. > Thanks. You remind me we should write Fuego's version as "1.1.SVN" rather than "1.1". In Clark's paper they tested against Fuego 1.1. So the reason why ou

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional Neural Networks

Hi Hugh, On Fri, Dec 26, 2014 at 9:49 AM, Hugh Perkins wrote: > Estimated total number of parameters > approx = 12 layers * 128 filters * 128 previous featuremaps * 3 * 3 > filtersize > = 1.8 million > > But you say 2.3 million. It's similar, so seems feature maps are > fully connected to lower

Re: [Computer-go] Representing Komi for neural network

2015-01-11 15:59 GMT+00:00 Detlef Schmicker : > > By the way: > Todays bot tournament nicego19n (oakfoam) played with a CNN for move > prediction. > It was mixed into the original gamma with some quickly optimized parameter > leading to >100ELO improvement for selfplay with 2000 playouts/move. I us

Re: [Computer-go] Move Evaluation in Go Using Deep Convolutional NeuralNetworks

2015-01-09 23:04 GMT+00:00 Darren Cook : > Aja wrote: > >> I hope you enjoy our work. Comments and questions are welcome. > > I've just been catching up on the last few weeks, and its papers. Very > interesting :-) > > I think Hiroshi's questions got missed? > I did answer Hiroshi's questions. h

Re: [Computer-go] UEC Cup

Hi Hideki, 2015-03-16 11:34 GMT+00:00 Hideki Kato : > > #So far, Zen often lost territory and built thickness in early stages > and revesed in hte middle. However, rivals are quickly getting stronger > while top two are almost standing :-( these a few years, it's also > getting harder to reverse.

Re: [Computer-go] Fwd: Teaching Deep Convolutional Neural Networks to Play Go

Hello Oliver, 2015-03-16 11:58 GMT+00:00 Oliver Lewis : > > It's impressive that the same network learned to play seven games with > just a win/lose signal. It's also interesting that both these teams are in > different parts of Google. I assume they are aware of each other's work, > but maybe Aj

Re: [Computer-go] Seki in playouts

On Wed, Apr 15, 2015 at 12:24 PM, Igor Polyakov wrote: > > 5. But it misses this because squeezing leads to a few accidental losses > in playouts where black plays something that white doesn't respond to > My two cents: just prevent the accidental losses in playouts then the program will play B8

Re: [Computer-go] 25x25 experiment

On Wed, Apr 29, 2015 at 5:54 PM, Detlef Schmicker wrote: > I did not test 100k playouts, I was lower than 2k playouts in all my > tests. No I am not sure we are stronger than Fuego, I should run on KGS... > (possibly after the 19x19 tournament on sunday) > We won against hirabot (1d if playing w

Re: [Computer-go] Computer Go Turnament at EGC 2015: Call for Participation

Hi, Thanks. I'll attend but I'll play Go myself rather than entering a Go program. :) See you in Liberec, Aja On Tue, Jun 16, 2015 at 11:07 AM, Petr Baudis wrote: > Hi! > > On Mon, Jun 15, 2015 at 01:21:49PM +, Josef Moudrik wrote: > > The tournament will take place on 29th July 2015. Th

Re: [Computer-go] Report KGS Slow Tournament

Congrats to Abakus. Does anyone know who is Abakus` author? Aja On 24 Sep 2015 18:15, "Nick Wedd" wrote: > Thank you, Ingo, for noticing this. I did forget to post to the list about > it. > > And congratulations to Abakus! > > Nick > > On 24 September 2015 at 16:58, "Ingo Althöfer" <3-hirn-ver..

Re: [Computer-go] Report KGS Slow Tournament

2015-09-26 2:48 GMT+01:00 Hideki Kato : > Aja Huang: < > calqb9_kdkgkxsvrjksjmnvywe3tpjw1ct_n-tpjt_sdmmit...@mail.gmail.com>: > >Congrats to Abakus. Does anyone know who is Abakus` author? > > Tobias Graf is the author. > See http://www.computer-go.info/db/operson.p

Re: [Computer-go] Komi 6.5/7.5

2015-11-06 9:41 GMT+00:00 Robert Jasiek : > On 06.11.2015 10:35, Aja Huang wrote: > >> another very strong Chinese pro Shi Yue >> <http://www.goratings.org/players/449.html> (No.3 at Go ratings >> http://www.goratings.org/, current No.2 in China) said he likes to ta

[Computer-go] 7x7 Go is weakly solved

It's the work by Chinese pro Li Zhe 7p. http://blog.sina.com.cn/s/blog_53a2e03d0102vyt5.html His conclusions on 7x7 Go board: 1. Optimal komi is 9.0. 2. Optimal solution is not unique. But the first 3 moves are unique, and the first 7 moves generate 5 major optimal solutions. 3. There are many var

Re: [Computer-go] 7x7 Go is weakly solved

calculates the value? > Did you mean 7x7 Go was weakly solved before? Aja > Best, > Erik > > > On Mon, Nov 30, 2015 at 2:10 AM, Aja Huang wrote: > >> It's the work by Chinese pro Li Zhe 7p. >> http://blog.sina.com.cn/s/blog_53a2e03d0102vyt5.html >> &g

Re: [Computer-go] Number of Go positions computed at last

Very interesting. Thanks John. :) Aja On Fri, Jan 22, 2016 at 4:18 AM, John Tromp wrote: > It's been a long journey, and now it's finally complete! > > http://tromp.github.io/go/legal.html > > has all the juicy details... > > regards, > -John > ___ >

Re: [Computer-go] Mastering the Game of Go with Deep Neural Networks and Tree Search

2016-01-27 18:46 GMT+00:00 Aja Huang : > Hi all, > > We are very excited to announce that our Go program, AlphaGo, has beaten a > professional player for the first time. AlphaGo beat the European champion > Fan Hui by 5 games to 0. We hope you enjoy our paper, published in Natu

Re: [Computer-go] Zen19X achieved stable KGS 7d

Congrats to Zen's achievements! :) Aja On Mon, Feb 1, 2016 at 11:44 AM, Petr Baudis wrote: > On Sun, Jan 31, 2016 at 10:16:25AM +0900, Hideki Kato wrote: > > Petr Baudis: <20160130150502.gf12...@machine.or.cz>: > > > Hi, > > > > > > it seems that Zen19X grabbed at KGS 7d and looks like it's g

Re: [Computer-go] Value Network

Good stuff, Hiroshi. Looks like I don't need to answer the questions regarding value network. :) Aja On Sat, Mar 19, 2016 at 9:23 PM, Hiroshi Yamashita wrote: > What are you using for loss? >> > > I use this, > > layers { > name: "loss" > type: EUCLIDEAN_LOSS > bottom: "fc14" > bottom: "lab

Re: [Computer-go] Nice graph

2016-03-26 2:48 GMT+00:00 Petr Baudis : > > The word covered by the speaker's head is "self". Bot results in > self-play are always(?) massively exaggerated. It's not uncommon to see > a 75% self-play winrate in selfplay to translate to 52% winrate against > a third-party reference opponent. c.f

Re: [Computer-go] BetaGo?

On Tue, Apr 19, 2016 at 4:11 PM, Rémi Coulom wrote: > Anybody knows who is the author of BetaGo? It is playing with account > GoBeta on KGS, and is 6d. > I can confirm it's not AlphaGo or a weaker version of AlphaGo. We haven't decided to play AlphaGo online yet, but when the decision is made we

Re: [Computer-go] Lee Sedol's reviews on AlphaGo games

Hi Ingo, 2016-04-14 14:21 GMT+01:00 "Ingo Althöfer" <3-hirn-ver...@gmx.de>: > > Is 102.R10 really so special? I analysed that position with CrazyStone2013 > about 30 times (independent runs). In 20+ % of the runs, CS2013 also wants > to play R12 (giving White 60+% chances of winning). But also in

Re: [Computer-go] Hajin Lee will play a live commented game against CrazyStone

On 23 May 2016 08:00, "Rémi Coulom" wrote: > > I think one of the main problems is that the network learns good replies to good moves. The training set does not have good replies to bad moves, but the search tree is full of bad moves that need to be punished. As the joseki that CS messed up was q

Re: [Computer-go] DarkForest is open-source now.

Well done, Yuangdong. :) Aja 2016-06-10 18:31 GMT+01:00 Yuandong Tian : > Hi all, > > DarkForest Go engine is now public on the Github (pre-trained CNN models > are also public). Hopefully it will help the community. > > https://github.com/facebookresearch/darkforestGo > > Thanks! > > Best, > Yu

Re: [Computer-go] Timetable "Computers and Games 2016"

16/index.php?page=browseSessions&print=yes&doprint=yes&presentations=show > > At least 8 of the talks (including the keynote presentation by Aja Huang) > are directly related to Go and computer Go. > Thanks Ingo. In my talk I will focus on the source of AlphaGo's power:

Re: [Computer-go] Very interesting presentation of Fan Hui about AlphaGo

On Tue, Jul 26, 2016 at 12:55 PM, J. van der Steen < j.van.der.st...@gobase.org> wrote: > > Dear Alexander Kozlovsky, > > Are the sheets of this presentation available somewhere by any chance. The > sheets are a bit hard to read in the video on YouTube. > Hi, Thanks for the interest. We are post

Re: [Computer-go] Very interesting presentation of Fan Hui about AlphaGo

On Tue, Jul 26, 2016 at 1:47 PM, Aja Huang wrote: > > four AlphaGo v18 vs. AlphaGo v18 self-play games > Correction: There will be three v18 self-play games, not four. Regards, Aja ___ Computer-go mailing list Computer-go@computer-go

Re: [Computer-go] AlphaGo selfplay 3 games

Hi Richard, On Wed, Sep 14, 2016 at 2:54 PM, Richard Lorentz wrote: > After following the link and then playing out one of the games in my > browser (Chrome) I don't see the variations. Can anyone help me figure out > how to make them visible? > You can see the variations by clicking on labels

Re: [Computer-go] AlphaGo selfplay 3 games

; It was nice to see you, too. I'm glad you enjoy the commentary. :) Aja > > > > On 09/14/2016 07:41 AM, Aja Huang wrote: > > Hi Richard, > > On Wed, Sep 14, 2016 at 2:54 PM, Richard Lorentz > wrote: > >> After following the link and then playing out one

Re: [Computer-go] "Google event match" in October?

Hi Paweł, On Thu, Sep 15, 2016 at 9:12 AM, Paweł Morawiecki < pawel.morawie...@gmail.com> wrote: > Dear Aja, > > On the _official_ website of Korean Baduk Association, specifically on > their schedule (http://www.baduk.or.kr/info/schedule.asp), they announce > what is called "Google event match(e

Re: [Computer-go] NATURE: Top 10 people 2016

On Mon, Dec 19, 2016 at 6:46 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de> wrote: > Hi, > NATURE published a list of top 10 people in science 2016: > http://www.nature.com/news/nature-s-10-1.21157 > > Demis Hassabis, manager of the AlphaGo team, is in the list. > Thanks Ingo. Yes, I think Demis is w

Re: [Computer-go] Ke Jie vs. AlphaGo match

Thanks Hiroshi. I hope you will enjoy AlphaGo's games. Aja 2017-05-19 11:28 GMT+08:00 Hiroshi Yamashita : > Hi, > > It will be played in a week. > But there are few information about this. > Is there YouTube live available? > > I found a schedule in Panda-net site. > > Ke Jie vs. AlphaGo (3 hou

Re: [Computer-go] AlphaGo Retires

2017-05-28 23:14 GMT+08:00 Petr Baudis : > After winning against Ke Jie 3-0 > > https://deepmind.com/research/alphago/alphago-china/ > > AlphaGo retires from competitive Go. The team released 50 selfplay > games, will work on a teaching tool and publishing a second paper. > > ht

Re: [Computer-go] AlphaGo Zero SGF - Free Use or Copyright?

2017-10-23 0:29 GMT+01:00 Lucas Baker : > Hi Robert, > > The AlphaGo Zero games are free to use with proper attribution, so please > use them as you like for commentaries as long as you credit DeepMind. > Yes, Lucas is right. We hope you enjoy AlphaGo Zero games. :) Best regards, Aja > Best,

Re: [Computer-go] November KGS bot tournament

2017-10-27 8:15 GMT+01:00 Rémi Coulom : > Hi, > > I would like to thank Nick very much, too. When I was working on Crazy > Stone, these tournaments were a great source of motivation and enjoyment. > I'll keep particularly good memories of the KGS-tournament parties in Tokyo. > > So, thanks Nick fo

Re: [Computer-go] action-value Q for unexpanded nodes

2017-12-06 9:23 GMT+00:00 Gian-Carlo Pascutto : > On 03-12-17 17:57, Rémi Coulom wrote: > > They have a Q(s,a) term in their node-selection formula, but they > > don't tell what value they give to an action that has not yet been > > visited. Maybe Aja can tell us. > > FWIW I already asked Aja this

Re: [Computer-go] action-value Q for unexpanded nodes

2017-12-06 13:52 GMT+00:00 Gian-Carlo Pascutto : > On 06-12-17 11:47, Aja Huang wrote: > > All I can say is that first-play-urgency is not a significant > > technical detail, and what's why we didn't specify it in the paper. > > I will have to disagree here. Of c

Re: [Computer-go] 9x9 is last frontier?

2018-03-02 6:50 GMT+00:00 "Ingo Althöfer" <3-hirn-ver...@gmx.de>: > Von: "David Doshay" > > Go is hard. > > Programming is hard. > > > > Programming Go is hard squared. > > ;^) > > And that on square boards. > Mama mia! > Go is hard for humans, but in my own opinion I think Go seems to be too ea

Re: [Computer-go] 9x9 is last frontier?

n. Regards, Aja > > > > On Fri, Mar 2, 2018 at 4:52 AM, Aja Huang wrote: > >> >> >> 2018-03-02 6:50 GMT+00:00 "Ingo Althöfer" <3-hirn-ver...@gmx.de>: >> >>> Von: "David Doshay" >>> > Go is hard. >>> &

Re: [Computer-go] Leela Superstar!

Congrats for Leela's significant improvements. :) Aja On Tue, Feb 21, 2017 at 1:25 PM, Gian-Carlo Pascutto wrote: > On 19-02-17 17:00, "Ingo Althöfer" wrote: > > Hi, > > the rank graph of LeelaX on KGS looks impressive: > > > > http://www.dgob.de/yabbse/index.php?action=dlattach; > topic=6048.0

Re: [Computer-go] Zen lost to Mi Yu Ting

On Tue, Mar 21, 2017 at 10:48 AM, Hideki Kato wrote: > The value network has been trained with Chinese rules and 7.5 > pts komi. Using this for Japanese and 6.5, there will be some > error in close games. We knew this issue and thought such > chances would be so small that postponed correcting

Re: [Computer-go] Welcome to Download the "computer go dataset" (160K 9D vs 9D SGFs)

On Mon, Apr 17, 2017 at 4:19 PM, Yu Yuan wrote: > 2015.11.02 - 2016.12.31 > TYGEM "9D vs 9D" dataset (1,516,031 games). > > 2003.09.25 - 2011.12.28 > TOM "9D vs 9D" dataset (50,956 games). > Hi Yu Yuan, I was wondering did you check with TYGEM and TOM about the copyright of these SGFs? Aja >

Re: [Computer-go] Ke Jie vs. AlphaGo match

; > > That seems unbelievable. I'd like to set up live comment by Crazy Stone on > the Crazy Sensei web site. > > Rémi > > - Mail original - > De: "Aja Huang" > À: computer-go@computer-go.org > Envoyé: Vendredi 19 Mai 2017 05:52:41 > Objet: Re:

Re: [Computer-go] AlphaGo Zero

On Thu, Oct 19, 2017 at 11:04 AM, Hiroshi Yamashita wrote: > I have two questions. > > 2017 Jan, Master , defeat 60 pros in a row. > 2017 May, Master?, defeat Ke Jie 3-0. > > Master is Zero method with rollout. > Zero is Zero method without rollout. > > Did AlphaGo that played with Ke Jie use r