ctak - An implementation of Tak and a computer opponent in C

Age	Commit message (Expand)	Author
42 hours	It works, but it's too slowmcu	tslil
42 hours	Starting work on MCU port	tslil
42 hours	Went back up to 1986 weights	tslil
42 hours	Quick sketch of non-determinism	tslil
42 hours	Another road bug	tslil
42 hours	Dragon is not separate, fixed concession, allowing playing as b/w	tslil
42 hours	Changed function back to macro	tslil
42 hours	There was a bug in the minimax!	tslil
42 hours	Better trained weights, an important oversight in minimax (steps <=)	tslil
42 hours	So after a small tweak, it actually plays well!	tslil
42 hours	Seems tolerable, but not great. Still wont win or lose ???	tslil
42 hours	Trying alpha-beta pruning. There's a bug somewhere ...	tslil
42 hours	There are still some bugs somewhere in minimax ...	tslil
42 hours	Fixed minimax (!), fixed bugs in tak.c	tslil
42 hours	Trying minimax	tslil
42 hours	Crass error	tslil
42 hours	Weights?	tslil
42 hours	Wider?	tslil
42 hours	Trying various things to teach ti	tslil
42 hours	Have to account for CPS	tslil
42 hours	Well it works, and it's bad. Many bugs fixed in interim	tslil
42 hours	Closing in!	tslil
42 hours	Auto-generating weights	tslil
42 hours	Generate the correct format directly	tslil
42 hours	New weights	tslil
42 hours	one output (of course), Pade approximant of logistic for activation	tslil
42 hours	Train on data that is close to the end of the game only	tslil
42 hours	Working version of this script	tslil
42 hours	This finally works!	tslil
42 hours	Hopefully removed all layer bugs	tslil
42 hours	Trying again with correct data generation	tslil
42 hours	Also this was wrong	tslil
42 hours	Close in principle, but there;s a bug in ct1997 &/ pptdb!	tslil
42 hours	This is the general idea, but i'm sure order is incorrect somewhere	tslil
42 hours	Starting on importing network!	tslil
42 hours	Two inputs to model, stacks and flat counts	tslil
42 hours	Fixed comments during extraction	tslil
42 hours	Shuffling is important	tslil
42 hours	Magic	tslil
42 hours	Fixed a small bug in PTN parsing, other stuff	tslil
42 hours	Closing in on something reasonable for roads	tslil
42 hours	Check for road wins before flat wins	tslil
42 hours	Searching for a good representation	tslil
42 hours	Small tweaks	tslil
42 hours	On the hunt for a better representation	tslil
42 hours	Experimenting with training nn	tslil
42 hours	Separate concerns in pptdb, remove legacy in extract.sh	tslil
42 hours	Gotta go fast	tslil
42 hours	First attempt at training data + exclude draws + list overflows too	tslil
42 hours	Tabs for indentation, spaces for alignment	tslil