Life In 19x19

he probably had reasons to walk like that, but unfortunately he won’t explain it
I continued the game after move 149 several times, ZEN (gtp4zen) wins more often than loses. ZEN original is the same

SAI #82 8ee88aa5 (10 sec/move, cpu version) http://sai.unich.it/networks/8ee88aa51f ... 526963b.gz
against Crazy Stone Deep Learning 5 dan (WINDOWS) - SAI B+19.5:

SAI #92 a520f71b (-v 1000, cpu version) http://sai.unich.it/networks/a520f71b10 ... 2c663ca.gz
against ZEN 7 6 dan (gtp4zen -t4 -r15 -s3000 -n3 -o4.4 -p0.6) - 1:19
zen surrendered in one game:

SAI developer:

"I believe SAI will get superhuman in a few months, with the support of the community. Anyhow we don't presume that it will ever be as strong as the latest LZ networks. That's not its purpose.

The aim of SAI is to get rid of the suboptimal moves LZ or AG may play when they are too far ahead of their opponent (i.e. almost always against humans) or when they are far behind (i.e. at the beginning of a game with handicap).

When SAI will be superhuman, it could show really how much handicap (in points or maybe in stones) it could give to a strong player. Or how many points difference it could score in a fair game.
In endgame, it may be able to show unknown smart plays or exhibit optimal choices in difficult situations.

This is presently from very difficult to impossible under the LZ/AG structure, because you would need to tailor positions in which the perfection of endgame moves really would matter for winning, and not just for score."

https://github.com/leela-zero/leela-zer ... -548264646

SAI #99 109c93bc (-v 1000, cpu version) http://sai.unich.it/networks/109c93bcf1 ... babe7ef.gz
against ZEN 7 7 dan (gtp4zen -t4 -r15 -s3500 -n3 -o2.8 -p0.65) - 1:19
zen surrendered in one game:

TIME_B: 999.5, TIME_W: 751.4

SAI #107 6542a4fd (-v 1600, cpu version) http://sai.unich.it/networks/6542a4fd58 ... 9fd5a11.gz
against ZEN 7 8 dan (gtp4zen -t4 -r20 -s4000 -n3 -o1.4 -p0.7) - 1:19
zen surrendered in one game:

TIME_B: 1248.9, TIME_W: 1074

SAI #109 ddcec541 (-v 1600, cpu version) http://sai.unich.it/networks/ddcec541a9 ... 0277f2b.gz
against ZEN 7 9 dan (gtp4zen -t4 -r20 -s6000 -n3 -o1.0 -p0.75) in the first game ZEN surrendered:

TIME_B: 1997.4, TIME_W: 1483.1

SAI #109 ddcec541 (-v 1600) http://sai.unich.it/networks/ddcec541a9 ... 0277f2b.gz
against KataGo (6x96) 2:48 (5 sec/move, gt 610)

TIME_B: 621.4 TIME_W: 644.8


TIME_B: 659 TIME_W: 643.4

perhaps the next after 6x128 will be the network 9x192. in test matches these networks appeared, for example a48ceb19 http://sai.unich.it/networks/a48ceb1926 ... 7a34235.gz

11 hours ago:
"We plan to wait
until we have a 9x192 network trained on the most recent data before switching.
It should happen in about 10 hours." https://github.com/sai-dev/sai/issues/3 ... -561051187
very interesting...

# 139 84910f22 - 9x192! http://sai.unich.it/networks/84910f22d5 ... b03523b.gz

And wrote:http://sai.unich.it

OK I'm stupid. How can I play it 9x9 with Sabaki, what's the command line? Thx.

download: https://github.com/sai-dev/sai/releases/

in Sabaki launch like Leela Zero:
path d:\sai\sai9x9.exe
--gtp --noponder -w 94619dea457de054503cec030269ce842c47055ba51e96db8fee841dfbaf05f9.gz
additionally: time_settings 0 5 1

SAI for android https://github.com/Grant-Tao/compiled-l ... oid-phones

100 games 9x9, KataGo 10x128-s1108 - SAI (gt 610, 5 sec/move) 30:70 (KataGo b 9:41 and KataGo w 21:29)
100 games 9x9, KataGo 10x128-s4588 - SAI (gt 610, 5 sec/move) 65:35 (KataGo b 23:27 and KataGo w 42:8)
in this game SAI gave up on move 30: