Hi all, I would very much like to do some changes to gnubg.bd. Of course, I'll try to keep backwards compability. Very shortly, gnubg will be able to read large one-sided databases and large two-side
For the record, I'm happy to try to fix all of this, I'm just not sure what the best solution is. I haven't spent much time thinking about makehyper because it almost never comes up. This happened in
About the training: Temporal difference learning was only used in the first stages of the program. With this kind of training the program reached a intermediate playing strength. Later the nets have
Yes! That was my initial idea for separating the two. I've already separated the gnubg_os0.bd into two databases, one which gives the bearoff distributions and one for giving the gammon distributions
Long story....... TD learning was the first thing. I'm not sure what version number this net had, but lets call this contact net 0.0. contact net 0.0 wasn't the best player in the world, and the thin
Just shot! That's what the list is for! Ehhh - did I miss something in last build, I think it should be 0.13-devel now. The manual is a bit sparse... we apologeeze. What kind of database? A bearoff
Thanks for your report. The makebearoff code was written last millennium, and the problem you're seeing is probably caused by some overflow problem in the code. These kind of stuff was hard to check
Jon's new build works. I have libintl-1.dllbut not libintl.dll, which is perhaps why Petr's build didn't work for me. Here are dumps from my working 12-point and broken 13-point databases. Sadly, th
1. Report of the 12th Computer Olympiade 2. Matches played -- 1. Report == Participants: MCGammon by François van Lishout and Guillaume Chaslot (Université de Liège, Belgium) BgBlitz by Frank Berg
There are some instructions in /usr/share/doc/gnubg/README.Debian about the bearoff databases in general: database from <ftp://ftp.demon.nl/pub/games/gnubg/databases/>. Save You can always just gener
The format of the file is (implicitely) described in the StartFromDatabase() function of makehyper.c : a 40 bytes header followed by 28 bytes records each encoding 10 values: 5 outputs and 5 equitie
The format of the file is (implicitely) described in the StartFromDatabase() function of makehyper.c : a 40 bytes header followed by 28 bytes records each encoding 10 values: 5 outputs and 5 equities
Yee ha! 15 chequers on the 13 point = 195 pips. epc - wastage = 202.052 - 7.052 = 195. See below. At present, people don't use the bearoffdatabase on disk because it's slow compared to the race net.
Hi Rick, What you need to do is download the databases you want and put them in the gnubg directory, unzipped of course. Then you rename them as listed below, to govern how gnubg uses each file. Gnub
The hash size is just given in bytes. For one sided bearoff databases each position requires 128 bytes. For the two sided bearoffdatabase it requires something different (can't remember). No. The da
Hi, now I'm with you for a couple of hours, too. new those C:\HB\_Projects\gnubg>grep -n '[{}]' br1.c 31:static unsigned char acBearoff1[] = { 185208:}; 185212:BearoffInitBuiltin ( void ) { 185216: i
Holger posted a game which he reports crashes which is in the bearoff stage on one side - gnubg is bearing off, holger has just come in from the bar. I think the other posted crashing match was also
The Windows builds crash: - Race position during game play - Analyse Kleinman Count - Analyse Thorp Count - Analyse One Chequer Race - Analyse One Sided Rollout (all other Analyse: OK) FYI: Analyse