RomiChess P3n released

[Guenther]

...

To see if Romi is learning you can start Romi from the command line and type pr 1 enter. Romi will print out for example, "Record 1 sib 487 chd 2 fs 12 ts 28 some other stuff score 17 w 283 l 264, d 191" which is 1.e2e4. Then type pr 487 in this example to get the record for the first sibling move. In this example it is d2d4, has a score of 19 w 23 l 22 d 13. anyway it is possible to examine the learn file starting with "pr 1".

The resigning problem is new as far as I know. It is an issue between the compiler I use and Windows 7/vista and windows 8/10. The resigning problem does not happen on w10 or as far as I know w8.

Thanks for the info! Was that already documented somewhere?
BTW exists a possibility to see the content as whole in an ascii export?
(something like PG dump for Black/White into ascii files)

Another question, I guess the format hasn't changed since you introduced
it and we can still experiment with the old and big learn files e.g. Marc Lacrosse and Pavel Hase made?

I also have Win7 here and will take a look at the resign problem, which is
yet unknown to me - I was also almost away from computerchess from 2011-2016 ;-)

Guenther

[carldaman]

Hi Michael,

I have downloaded Romichess for the first time to see what all the hype is about with the reinforcement learning.

• using RomiChess64P3n.exe
  created as a winboard 2 engine in arena
  learn.dat file autocreated on creation of the engine
  100 game tournament setup with a computer opponent
  tournament is ongoing but learn.dat is still 274KB and not growing

What did I do wrong?

So far Romichess is 0/33

Thanks!

To see if Romi is learning you can start Romi from the command line and type pr 1 enter. Romi will print out for example, "Record 1 sib 487 chd 2 fs 12 ts 28 some other stuff score 17 w 283 l 264, d 191" which is 1.e2e4. Then type pr 487 in this example to get the record for the first sibling move. In this example it is d2d4, has a score of 19 w 23 l 22 d 13. anyway it is possible to examine the learn file starting with "pr 1".

The resigning problem is new as far as I know. It is an issue between the compiler I use and Windows 7/vista and windows 8/10. The resigning problem does not happen on w10 or as far as I know w8.

The abortive resign problem also happens on my Win 8.1 system, albeit less frequently. Would it be possible to compile Romi locally from source? If so, what would I need to use to compile it?

By the way, I can verify that Romi's learning file updates normally after each game (using Arena 3.5). I created a very high adjudication threshold to give Romi a chance to resign in a lost position before GUI adjudication kicks in.

CL

[Guenther]

I also have Win7 here and will take a look at the resign problem, which is
yet unknown to me - I was also almost away from computerchess from 2011-2016 ;-)

Guenther

Yes it happens an awful lot here. Obviously it comes from the learn file.
May be a wrong value is read (used a learn file from 2007) which triggers resign?
All moves before the resign come from the learn file...

Code: Select all

[Event "GS-Test_2017D"]
[Site "RWBC::CAPPUCCINO"]
[Date "2017.12.15"]
[Round "1"]
[White "RomiChess_Proto3N2-64"]
[Black "ZMatt_X-64"]
[Result "0-1"]
[ECO "B15"]
[Opening "Caro-Kann defense"]
[PlyCount "11"]
[TimeControl "40/60"]

1. e4 {book} c6 {book} 2. d4 d5 {-0.34/14 0.94s} 3. Nc3 e6 {-0.43/14 1.4s}
4. Nf3 Bb4 {-0.43/14 1.0s} 5. Bd3 dxe4 {-0.48/16 1.4s} 6. Bxe4 {White resigns}

Code: Select all

183583 <RomiChess_Proto3N2-64(4): This is RomiChess by Michael J Sherwin
183584 <RomiChess_Proto3N2-64(4): Version P3n 64 bit -- 3/Oct/2017
183584 <RomiChess_Proto3N2-64(4): Copyright 2005-2017 Michael J Sherwin
183785 <ZMatt_X-64(5): zurichess master http://www.zurichess.xyz
183786 <ZMatt_X-64(5): build with devel +1c69384da4 Mon Nov 27 20:26:19 2017 +0000 at 2017-12-02 11:55:36, running on amd64
183787 <ZMatt_X-64(5): id name zurichess master
183787 <ZMatt_X-64(5): id author Alexandru Mo?oi
183788 <ZMatt_X-64(5): option name Hash type spin default 64 min 1 max 65536
183788 <ZMatt_X-64(5): option name MultiPV type spin default 1 min 1 max 16
183788 <ZMatt_X-64(5): option name Ponder type check default true
183789 <ZMatt_X-64(5): option name Handicap Level type spin default 0 min 0 max 20
183789 <ZMatt_X-64(5): option name UCI_AnalyseMode type check default false
183789 <ZMatt_X-64(5): uciok
183790 >ZMatt_X-64(5): setoption name Handicap Level value 0
183790 >ZMatt_X-64(5): setoption name Hash value 256
183790 >ZMatt_X-64(5): setoption name MultiPV value 1
183790 >ZMatt_X-64(5): isready
183799 <RomiChess_Proto3N2-64(4): feature analyze=1
183799 >RomiChess_Proto3N2-64(4): rejected analyze
183799 <RomiChess_Proto3N2-64(4): feature setboard=1
183799 >RomiChess_Proto3N2-64(4): accepted setboard
183800 <RomiChess_Proto3N2-64(4): feature done=1
183800 >RomiChess_Proto3N2-64(4): accepted done
183800 <RomiChess_Proto3N2-64(4): Error (unknown command): rejected
183800 <RomiChess_Proto3N2-64(4): Error (unknown command): accepted
183801 <RomiChess_Proto3N2-64(4): Error (unknown command): accepted
183833 <ZMatt_X-64(5): readyok
Started game 3 of 120 (RomiChess_Proto3N2-64 vs ZMatt_X-64)
183834 >RomiChess_Proto3N2-64(4): new
183834 >RomiChess_Proto3N2-64(4): force
183834 >RomiChess_Proto3N2-64(4): level 40 1 0
183835 >RomiChess_Proto3N2-64(4): post
183835 >RomiChess_Proto3N2-64(4): easy
183835 >ZMatt_X-64(5): ucinewgame
183835 >ZMatt_X-64(5): setoption name Ponder value false
183836 >ZMatt_X-64(5): position startpos
183836 >RomiChess_Proto3N2-64(4): e2e4
183836 >ZMatt_X-64(5): position startpos moves e2e4
183837 >ZMatt_X-64(5): position startpos moves e2e4 c7c6
183837 >RomiChess_Proto3N2-64(4): c7c6
183837 >RomiChess_Proto3N2-64(4): time 6000
otim 6000
183837 >RomiChess_Proto3N2-64(4): go
184064 <RomiChess_Proto3N2-64(4): Error (unknown command): post
184065 <RomiChess_Proto3N2-64(4): move d2d4
184065 >ZMatt_X-64(5): position startpos moves e2e4 c7c6 d2d4
184066 >ZMatt_X-64(5): isready
184281 <ZMatt_X-64(5): readyok
184282 >ZMatt_X-64(5): go wtime 59769 btime 60000 movestogo 39
184297 <ZMatt_X-64(5): info depth 0 seldepth 0 multipv 1 score cp -76 nodes 2 time 15 nps 133 pv
184332 <ZMatt_X-64(5): info depth 1 seldepth 1 multipv 1 score cp -36 nodes 57 time 50 nps 1140 pv d7d5
184332 <ZMatt_X-64(5): info depth 2 seldepth 2 multipv 1 score cp -45 nodes 283 time 51 nps 5549 pv d7d5 b1c3
184354 <ZMatt_X-64(5): info depth 3 seldepth 4 multipv 1 score cp -20 nodes 882 time 72 nps 12250 pv d7d5 e4e5 c8f5
184355 <ZMatt_X-64(5): info depth 4 seldepth 4 multipv 1 score cp -44 nodes 2156 time 73 nps 29534 pv d7d5 e4e5 c8f5 g1f3
184358 <ZMatt_X-64(5): info depth 5 seldepth 5 multipv 1 score cp -27 nodes 4872 time 76 nps 64105 pv d7d5 e4e5 e7e6 g1f3 f8e7
184362 <ZMatt_X-64(5): info depth 6 seldepth 7 multipv 1 score cp -50 nodes 9869 time 81 nps 121839 pv d7d5 e4e5 e7e6 g1f3 f8e7 c1d2
184366 <ZMatt_X-64(5): info depth 7 seldepth 8 multipv 1 score cp -27 nodes 13068 time 84 nps 155571 pv d7d5 e4e5 e7e6 g1f3 g8e7 c1d2 e7f5
184382 <ZMatt_X-64(5): info depth 8 seldepth 9 multipv 1 score cp -47 nodes 28003 time 101 nps 277257 pv d7d5 e4e5 e7e6 g1f3 c6c5 c2c4 c5d4 c4d5 d8d5 d1d4 d5d4 f3d4
184392 <ZMatt_X-64(5): info depth 9 seldepth 10 multipv 1 score cp -42 nodes 36432 time 110 nps 331200 pv d7d5 e4e5 e7e6 g1f3 c6c5 c2c4 c5d4 c4d5 f8b4 c1d2 b4d2 d1d2 d8d5 d2d4 d5d4 f3d4
184427 <ZMatt_X-64(5): info depth 10 seldepth 12 multipv 1 score cp -28 nodes 70381 time 145 nps 485386 pv d7d5 e4e5 c6c5 c2c3 e7e6 f1d3 c5d4 c3d4 f8b4 b1c3 b8c6
184505 <ZMatt_X-64(5): info depth 11 seldepth 12 multipv 1 score cp -20 nodes 141240 time 223 nps 633363 pv d7d5 e4d5 c6d5 g1f3 g8f6 f1e2 h7h6 e1g1 b8c6 f3e5 c8f5
184641 <ZMatt_X-64(5): info depth 12 seldepth 13 multipv 1 score cp -40 nodes 265735 time 359 nps 740208 pv d7d5 e4d5 c6d5 g1f3 g8f6 f1e2 e7e6 e1g1 f8d6 c2c4 e8g8 c4c5
185219 <ZMatt_X-64(5): info depth 14 seldepth 16 multipv 1 score cp -34 nodes 795540 time 938 nps 848123 pv d7d5 e4d5 c6d5 g1f3 g8f6 f1e2 e7e6 e1g1 f8d6 c2c4 e8g8 c4c5 d6c7 c1d2
185220 <ZMatt_X-64(5): bestmove d7d5 ponder e4d5
185220 >RomiChess_Proto3N2-64(4): time 5976
otim 5906
185221 >RomiChess_Proto3N2-64(4): d7d5
185221 <RomiChess_Proto3N2-64(4): move b1c3
185221 >ZMatt_X-64(5): position startpos moves e2e4 c7c6 d2d4 d7d5 b1c3
185221 >ZMatt_X-64(5): isready
185222 <ZMatt_X-64(5): readyok
185222 >ZMatt_X-64(5): go wtime 59769 btime 59062 movestogo 38
185222 <ZMatt_X-64(5): info depth 0 seldepth 0 multipv 1 score cp -45 nodes 4 time 0 nps 4000000 pv d5e4 c3e4
185222 <ZMatt_X-64(5): info depth 1 seldepth 1 multipv 1 score cp -15 nodes 68 time 0 nps 68000000 pv g8f6 f2f3
185223 <ZMatt_X-64(5): info depth 2 seldepth 2 multipv 1 score cp -38 nodes 349 time 0 nps 349000000 pv g8f6 e4e5 f6e4
185223 <ZMatt_X-64(5): info depth 3 seldepth 3 multipv 1 score cp -6 nodes 1097 time 1 nps 1097000 pv d5e4 c3e4 g8f6 e4f6
185224 <ZMatt_X-64(5): info depth 4 seldepth 5 multipv 1 score cp -20 nodes 2253 time 2 nps 1126500 pv d5e4 c3e4 g8f6 f1d3 f6e4
185224 <ZMatt_X-64(5): info depth 5 seldepth 6 multipv 1 score cp -21 nodes 3514 time 3 nps 1171333 pv d5e4 c3e4 g8f6 e4f6 e7f6 g1f3 f8d6
185225 <ZMatt_X-64(5): info depth 6 seldepth 7 multipv 1 score cp -23 nodes 4860 time 4 nps 1215000 pv d5e4 c3e4 g8f6 f1d3 f6e4 d3e4 e7e6
185227 <ZMatt_X-64(5): info depth 7 seldepth 9 multipv 1 score cp -34 nodes 7441 time 6 nps 1240166 pv d5e4 c3e4 g8f6 e4f6 e7f6 c2c3 f8d6 f1d3
185259 <ZMatt_X-64(5): info depth 8 seldepth 10 multipv 1 score cp -37 nodes 33581 time 38 nps 883710 pv d5e4 c3e4 g8f6 f1d3 e7e5 d4e5 d8a5 e4c3 a5e5 c1e3
185266 <ZMatt_X-64(5): info depth 9 seldepth 11 multipv 1 score cp -27 nodes 40055 time 45 nps 890111 pv d5e4 c3e4 g8f6 e4f6 e7f6 c1f4 f8d6 f4d6 d8d6 c2c3
185275 <ZMatt_X-64(5): info depth 10 seldepth 12 multipv 1 score cp -22 nodes 48680 time 54 nps 901481 pv d5e4 c3e4 g8f6 e4f6 e7f6 c1f4 f8d6 f4d6 d8d6 d1e2 c8e6 e1c1
185303 <ZMatt_X-64(5): info depth 11 seldepth 13 multipv 1 score cp -17 nodes 74276 time 83 nps 894891 pv d5e4 c3e4 g8f6 f1d3 f6e4 d3e4 e7e6 g1f3 f8d6 e1g1 e8g8
185464 <ZMatt_X-64(5): info depth 12 seldepth 14 multipv 1 score cp -36 nodes 209850 time 243 nps 863580 pv d5e4 c3e4 c8f5 f1d3 b8d7 c2c4 f5e4 d3e4 e7e6 g1f3 f8b4 c1d2 g8f6
185655 <ZMatt_X-64(5): info depth 13 seldepth 15 multipv 1 score cp -34 nodes 378167 time 435 nps 869349 pv d5e4 c3e4 g8f6 f1d3 f6e4 d3e4 e7e6 g1f3 b8d7 e1g1 d7f6 e4d3 f8d6
186572 <ZMatt_X-64(5): info depth 14 seldepth 15 multipv 1 score cp -43 nodes 1091780 time 1351 nps 808127 pv e7e6 f1d3 g8f6 g1f3 f8b4 e4e5 f6e4 d3e4 d5e4 f3g5 c6c5 e1g1 b4c3 b2c3
186573 <ZMatt_X-64(5): bestmove e7e6 ponder f1d3
186573 >RomiChess_Proto3N2-64(4): time 5976
otim 5771
186573 >RomiChess_Proto3N2-64(4): e7e6
186575 <RomiChess_Proto3N2-64(4): move g1f3
186575 >ZMatt_X-64(5): position startpos moves e2e4 c7c6 d2d4 d7d5 b1c3 e7e6 g1f3
186576 >ZMatt_X-64(5): isready
186576 <ZMatt_X-64(5): readyok
186576 >ZMatt_X-64(5): go wtime 59767 btime 57711 movestogo 37
186577 <ZMatt_X-64(5): info depth 0 seldepth 0 multipv 1 score cp -53 nodes 4 time 0 nps 4000000 pv f8b4
186577 <ZMatt_X-64(5): info depth 1 seldepth 1 multipv 1 score cp -28 nodes 83 time 0 nps 83000000 pv b8d7
186580 <ZMatt_X-64(5): info depth 2 seldepth 2 multipv 1 score cp -49 nodes 774 time 3 nps 258000 pv f8e7 a2a3
186582 <ZMatt_X-64(5): info depth 3 seldepth 3 multipv 1 score cp -35 nodes 1677 time 4 nps 419250 pv f8e7 c1f4 b8d7
186583 <ZMatt_X-64(5): info depth 4 seldepth 5 multipv 1 score cp -43 nodes 3651 time 6 nps 608500 pv g8f6 f1d3 f8b4 e4e5 f6e4 d3e4 d5e4 f3g5 c6c5 e1g1 b4c3 b2c3
186584 <ZMatt_X-64(5): info depth 5 seldepth 5 multipv 1 score cp -43 nodes 4252 time 7 nps 607428 pv g8f6 f1d3 f8b4 e4e5 f6e4 d3e4 d5e4 f3g5 c6c5 e1g1 b4c3 b2c3
186586 <ZMatt_X-64(5): info depth 6 seldepth 5 multipv 1 score cp -43 nodes 5157 time 8 nps 644625 pv g8f6 f1d3 f8b4 e4e5 f6e4 d3e4 d5e4 f3g5 c6c5 e1g1 b4c3 b2c3
186593 <ZMatt_X-64(5): info depth 7 seldepth 7 multipv 1 score cp -51 nodes 12208 time 17 nps 718117 pv g8f6 e4e5 f6e4 c3e4 d5e4 f3g5 d8d5
186634 <ZMatt_X-64(5): info depth 8 seldepth 9 multipv 1 score cp -46 nodes 45942 time 57 nps 806000 pv f8b4 e4e5 g8e7 f1d3 b4c3 b2c3 e8g8 e1g1 e7f5
186659 <ZMatt_X-64(5): info depth 9 seldepth 11 multipv 1 score cp -34 nodes 67957 time 82 nps 828743 pv f8b4 e4e5 g8e7 f1d3 b4c3 b2c3 e8g8 e1g1 e7f5 d3f5
186690 <ZMatt_X-64(5): info depth 10 seldepth 11 multipv 1 score cp -46 nodes 96243 time 113 nps 851707 pv f8b4 e4e5 g8e7 f1d3 c6c5 e1g1 b8c6 a2a3 b4c3 b2c3
186764 <ZMatt_X-64(5): info depth 11 seldepth 13 multipv 1 score cp -45 nodes 158575 time 187 nps 847994 pv f8b4 e4e5 g8e7 f1e2 b4c3 b2c3 e8g8 e1g1 b8d7 a1b1 d7b6
186972 <ZMatt_X-64(5): info depth 12 seldepth 13 multipv 1 score cp -39 nodes 333262 time 395 nps 843701 pv f8b4 e4d5 c6d5 f1e2 g8f6 e1g1 e8g8 c1d2 b8c6 a2a3 b4d6 c3b5
187135 <ZMatt_X-64(5): info depth 13 seldepth 13 multipv 1 score cp -33 nodes 482935 time 559 nps 863926 pv f8b4 f1d3 d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b4d6 c1g5 h7h6 g5f6 d8f6
187597 <ZMatt_X-64(5): info depth 14 seldepth 14 multipv 1 score cp -43 nodes 852658 time 1018 nps 837581 pv f8b4 f1d3 d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187598 <ZMatt_X-64(5): bestmove f8b4 ponder f1d3
187598 >RomiChess_Proto3N2-64(4): time 5976
otim 5669
187598 >RomiChess_Proto3N2-64(4): f8b4
187598 <RomiChess_Proto3N2-64(4): move f1d3
187599 >ZMatt_X-64(5): position startpos moves e2e4 c7c6 d2d4 d7d5 b1c3 e7e6 g1f3 f8b4 f1d3
187599 >ZMatt_X-64(5): isready
187603 <ZMatt_X-64(5): readyok
187603 >ZMatt_X-64(5): go wtime 59766 btime 56691 movestogo 36
187604 <ZMatt_X-64(5): info depth 0 seldepth 0 multipv 1 score cp -43 nodes 2 time 0 nps 2000000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187605 <ZMatt_X-64(5): info depth 1 seldepth 0 multipv 1 score cp -43 nodes 3 time 1 nps 3000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187606 <ZMatt_X-64(5): info depth 2 seldepth 0 multipv 1 score cp -43 nodes 4 time 2 nps 2000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187607 <ZMatt_X-64(5): info depth 3 seldepth 0 multipv 1 score cp -43 nodes 5 time 2 nps 2500 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187607 <ZMatt_X-64(5): info depth 4 seldepth 0 multipv 1 score cp -43 nodes 6 time 2 nps 3000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187608 <ZMatt_X-64(5): info depth 5 seldepth 0 multipv 1 score cp -43 nodes 7 time 2 nps 3500 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187609 <ZMatt_X-64(5): info depth 6 seldepth 0 multipv 1 score cp -43 nodes 8 time 2 nps 4000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187609 <ZMatt_X-64(5): info depth 7 seldepth 0 multipv 1 score cp -43 nodes 9 time 2 nps 4500 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187610 <ZMatt_X-64(5): info depth 8 seldepth 0 multipv 1 score cp -43 nodes 10 time 2 nps 5000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187611 <ZMatt_X-64(5): info depth 9 seldepth 0 multipv 1 score cp -43 nodes 11 time 2 nps 5500 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187612 <ZMatt_X-64(5): info depth 10 seldepth 0 multipv 1 score cp -43 nodes 12 time 2 nps 6000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187612 <ZMatt_X-64(5): info depth 11 seldepth 0 multipv 1 score cp -43 nodes 13 time 2 nps 6500 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187613 <ZMatt_X-64(5): info depth 12 seldepth 0 multipv 1 score cp -43 nodes 14 time 2 nps 7000 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 c1f4 d7b6 a2a3 b4d6 d1d2
187703 <ZMatt_X-64(5): info depth 13 seldepth 13 multipv 1 score cp -37 nodes 72850 time 99 nps 735858 pv d5e4 d3e4 g8f6 e4d3 e8g8 e1g1 b8d7 f1e1 b4d6 f3e5 f6d5 c3d5 c6d5
187967 <ZMatt_X-64(5): info depth 14 seldepth 15 multipv 1 score cp -41 nodes 301196 time 363 nps 829741 pv d5e4 d3e4 g8f6 e4d3 c6c5 d4c5 e8g8 e1g1 b8d7 c5c6 b7c6 c1e3 b4c3 b2c3
188345 <ZMatt_X-64(5): info depth 15 seldepth 15 multipv 1 score cp -38 nodes 619997 time 741 nps 836703 pv d5e4 d3e4 g8f6 e4d3 c6c5 d4c5 e8g8 e1g1 b8d7 c1g5 d7c5 d1d2 c5d3 g5f6 d8f6 d2d3
188990 <ZMatt_X-64(5): info depth 16 seldepth 17 multipv 1 score cp -48 nodes 1194153 time 1386 nps 861582 pv d5e4 d3e4 g8f6 e4d3 c6c5 a2a3 b4c3 b2c3 b8c6 e1g1 e8g8 c1e3 f6d5 d1d2 d5e3 d2e3 c5d4 c3d4
188991 <ZMatt_X-64(5): bestmove d5e4 ponder d3e4
188991 >RomiChess_Proto3N2-64(4): time 5976
otim 5530
188992 >RomiChess_Proto3N2-64(4): d5e4
188992 <RomiChess_Proto3N2-64(4): move d3e4
188992 >ZMatt_X-64(5): position startpos moves e2e4 c7c6 d2d4 d7d5 b1c3 e7e6 g1f3 f8b4 f1d3 d5e4 d3e4
188992 >ZMatt_X-64(5): isready
188992 <RomiChess_Proto3N2-64(4): resign
188993 >RomiChess_Proto3N2-64(4): force
188993 >RomiChess_Proto3N2-64(4): result 0-1 {White resigns}

Why not just disable auto resign (until the cause is found)? It is an obsolete feature nowadays, because the GUIs have advanced resign features anyway.

Guenther

[Michael Sherwin]

Hi Michael,

I have downloaded Romichess for the first time to see what all the hype is about with the reinforcement learning.

• using RomiChess64P3n.exe
  created as a winboard 2 engine in arena
  learn.dat file autocreated on creation of the engine
  100 game tournament setup with a computer opponent
  tournament is ongoing but learn.dat is still 274KB and not growing

What did I do wrong?

So far Romichess is 0/33

Thanks!

To see if Romi is learning you can start Romi from the command line and type pr 1 enter. Romi will print out for example, "Record 1 sib 487 chd 2 fs 12 ts 28 some other stuff score 17 w 283 l 264, d 191" which is 1.e2e4. Then type pr 487 in this example to get the record for the first sibling move. In this example it is d2d4, has a score of 19 w 23 l 22 d 13. anyway it is possible to examine the learn file starting with "pr 1".

The resigning problem is new as far as I know. It is an issue between the compiler I use and Windows 7/vista and windows 8/10. The resigning problem does not happen on w10 or as far as I know w8.

The abortive resign problem also happens on my Win 8.1 system, albeit less frequently. Would it be possible to compile Romi locally from source? If so, what would I need to use to compile it?

By the way, I can verify that Romi's learning file updates normally after each game (using Arena 3.5). I created a very high adjudication threshold to give Romi a chance to resign in a lost position before GUI adjudication kicks in.

CL

I compiled with MSVS 2005. I used nameless structs in the code which seem to be an MS extension only. If you can find MSVC 6 it will compile 32 bits. I really like nameless structs because amazing relationships can be defined and then eloquently accessed.

Now that I know it is a learn file problem thanks to Guenther maybe I'll be able to find it.

[Michael Sherwin]

...

To see if Romi is learning you can start Romi from the command line and type pr 1 enter. Romi will print out for example, "Record 1 sib 487 chd 2 fs 12 ts 28 some other stuff score 17 w 283 l 264, d 191" which is 1.e2e4. Then type pr 487 in this example to get the record for the first sibling move. In this example it is d2d4, has a score of 19 w 23 l 22 d 13. anyway it is possible to examine the learn file starting with "pr 1".

The resigning problem is new as far as I know. It is an issue between the compiler I use and Windows 7/vista and windows 8/10. The resigning problem does not happen on w10 or as far as I know w8.

Thanks for the info! Was that already documented somewhere?
BTW exists a possibility to see the content as whole in an ascii export?
(something like PG dump for Black/White into ascii files)

Another question, I guess the format hasn't changed since you introduced
it and we can still experiment with the old and big learn files e.g. Marc Lacrosse and Pavel Hase made?

I also have Win7 here and will take a look at the resign problem, which is
yet unknown to me - I was also almost away from computerchess from 2011-2016

Guenther

Hi Guenther, I think that I saw this but was too tired to answer. Last three days I spent most of my time in bed trying to keep my strength up. Sorry for the late reply.

Was the pr command documented? Yes, before I got sick way back when and did not have the energy except to release the executables of P3k. However, it can be understood from io.c rather easily.

There is no way in Romi to dump the information. However, the structure of the learn file can be easily understood so one can create a standalone program to mine whatever is required.

Yes, the old learn files are valid, zero changes since then.

Glad to see you back! It would be easy to disable the outdated resign feature for the book. However, I'd also want to figure out why there is apparently wrong value getting into the learn file. And why it never happens in Windows 10. If I can not figure it out I will just not retrieve a record with a faulty value in it.

Sorry again for the late reply.

[Graham Banks]

Hi Graham,

this is still P3n with only a minor bug fix to stop Romi from resigning from a learn file glitch.

Thanks!
Mike

Download link - http://kirill-kryukov.com/chess/discuss ... p?id=40984

[Michael Sherwin]

Hi Graham,

this is still P3n with only a minor bug fix to stop Romi from resigning from a learn file glitch.

Thanks!
Mike

Download link - http://kirill-kryukov.com/chess/discuss ... p?id=40984

Thank you Graham for making RomiChess available to everyone!

[Ovyron]

Thanks Graham and Micheal!

Now it'll be harsh times for Romi as she'll begin training against the muses, how many games until she manages to get a win?

[Michael Sherwin]

Thanks Graham and Micheal!

Now it'll be harsh times for Romi as she'll begin training against the muses, how many games until she manages to get a win?

That is a good question!

[Ovyron]

Hi Michael.

What I tried was training Romi "bottom-top" instead of "top-bottom".

That is, instead of having Romi play a stronger opponent, she plays a weaker opponent (that she is always able to defeat), and then I increase the strength of the opponent until they're at the same level (i.e. increase the opponent's Depth until Romi draws or loses a game, then Romi trains against the opponent at this depth until she can beat them, and I increase its depth again.)

I was playing Romi at 10 seconds/move against Houdini Typhoon (Tactical=1, Contempt=10) and Romi won all her games up until Depth=7, at Depth=8 the game was drawn.

However, I'm unable to merge this PGN to the learn file, even though I hadn't problems adding 1.2 million games.

Romi just hangs when I try to merge the PGN, can you check it out? Are you able to merge it to a learn file? Am I doing something wrong?

Here's the game I'm unable to merge:

[Event "RomiGO"]
[Site "Mexico"]
[Date "2018.01.07"]
[Round "8"]
[White "Romi 10 Seconds"]
[Black "Houdini Typhoon Depth 8"]
[WhiteElo "2000"]
[BlackElo "2000"]
[ECO "A07"]
[Result "1/2-1/2"]

1. Nf3 d5 2. g3 Nf6 3. Bg2 e6 4. O-O c5 5. d4 cxd4 6. Nxd4
e5 7. Nf3 Nc6 8. c4 d4 9. Bg5 h6 10. Bxf6 Qxf6 11. Qa4 Bd7
12. Nbd2 Be7 13. Ne1 a6 14. Nd3 O-O 15. Ne4 Qg6 16. Nec5
Bg4 17. Be4 Qf6 18. Nxb7 Bxe2 19. Rfe1 Bxd3 20. Bxd3 Rab8
21. Qxa6 Rfc8 22. a3 Rc7 23. Na5 Rxb2 24. Qa8+ Bf8 25. Rf1
Rb8 26. Qa6 Qd6 27. Nxc6 Rxc6 28. Qa5 Rc5 29. Qa4 g6
30. Rfb1 f5 31. Rxb8 Qxb8 32. Rb1 Qc8 33. f3 Kh7 34. Rb6
Kg7 35. Qa6 Qxa6 36. Rxa6 Rc7 37. g4 Rb7 38. gxf5 gxf5
39. Bxf5 Rb3 40. Be4 d3 41. Ra7+ Kf6 42. Ra6+ Kf7 43. Ra7+
Ke6 44. Ra6+ Kf7 45. Ra7+ Ke6 46. Ra6+ Kf7 1/2-1/2