ctak/src/pptdb.c, branch sixesAn implementation of Tak and a computer opponent in Chttps://git.l-3.space/cgit/ctak/atom?h=sixes2026-08-28T18:37:41ZPlaying around with huge nn's, poorly trained.2026-08-28T18:37:41Ztsliltslil@posteo.de2021-03-28T03:59:33Zurn:sha1:35dc86f8476250f693aa419f404b2b4c8c381d92
Not great at shallow depths
Don't generate header for training data + tweaks2026-08-28T18:37:41Ztsliltslil@posteo.de2021-02-12T02:04:42Zurn:sha1:00a04c2929bdc8f8f4bf7d5d8cf413ebfb3cd006
For some reason it would seem that moving flats to a lower value and
increasing the proximity between caps and top flats improves
acquisition. Still not great, but every bit counts.
Change the training data generation a little2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-02-11T01:07:24Zurn:sha1:890eb8d7f4a8c46eae18283ca5ac9c61fd41ed97
Although it pains me to say it, ``label smoothing'' appears to be
actually work. I'm also currently experimenting with training simply
against _all_ games, instead of only bot matches. Once the training
finishes i'll pit cttei against itself with old and new weights,
hopefully there'll be a noticeable improvement.
TEI interface working!2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-02-03T00:23:54Zurn:sha1:40c6b1dafab4de169bac8a799e7953e85061218eTried some naive iterative deepening. Work on TEI interface next2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-02-02T02:46:24Zurn:sha1:11956b2e940f5e1839efab187d898092e819a766
If TEI is implemented, then i could make use of Morten's
racetrack (https://github.com/MortenLohne/racetrack) and develop a
quantitative measure of the bot's performance. This is the current
priority.
Just some #weightgoals ;)2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-02-01T04:33:39Zurn:sha1:328c8d1e3094a942d6a2edd933c9cc4ab09daab1
It turns out that while i was training on a 0/1 classification
problem, i was using 2*eval - 1. Training using this function instead,
and on bot-dominated game choices (chosen_player in extract.sh) seems
to have given a better evaluation function. At the least, Morten's
swindle doesn't work anymore.
Syntax errors, small tweak to training data generation2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-01-29T19:05:06Zurn:sha1:fcface1079b1c44c7ccca3d117c698347c978f20Added license information!2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-01-29T02:30:30Zurn:sha1:a77062ce3dc9a99ad0662010cc7eeb00dd5036a6Added tunable search depth and self-play2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-01-18T19:52:14Zurn:sha1:9c56a2c16bce2a7011994bb9f4a62c3ad2fef2dfToying with symmetrising data2026-08-28T18:37:41Ztslil clingmantslil@posteo.de2021-01-18T18:43:49Zurn:sha1:a8f69f1262bd76ab311ebe760f314fa6e4526d15