| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2023-01-15 | new neural network arch (faster + better) & minor changes + fixes | tslil clingman | |
| Gone is the convolutional neural network, for it turns out not only is it more difficult to train, but all of the extra information about board layers didn't make much of a difference at this size. So cnn1986 has been replaced by nn1986, a standard, two-layer, dense nn configured as a binary classifier and (mis)used in that capacity. Note: total number of parameters is unchanged. HARK: this new nn exposes a bug somewhere in ctak. Run ctlm with self-play to see the completely borked board state at the end. | |||
| 2021-10-05 | Welcome geminict! | tslil clingman | |
| This is a special interface to negamax_cnn1986 which is designed to generate output for use in a CGI tak interface to be used over gemini. Also in this commit is a reformating of the various source files to use the traditional tab width of 8 spaces. | |||
| 2021-03-17 | Corrected generation of training data for 6s | tslil clingman | |
| 2021-02-11 | Don't generate header for training data + tweaks | tslil | |
| 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. | |||
| 2021-02-10 | Change the training data generation a little | tslil clingman | |
| 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. | |||
| 2021-02-05 | TEI interface working! | tslil clingman | |
| 2021-02-01 | Tried some naive iterative deepening. Work on TEI interface next | tslil clingman | |
| 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. | |||
| 2021-01-31 | Just some #weightgoals ;) | tslil clingman | |
| 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. | |||
| 2021-01-29 | Syntax errors, small tweak to training data generation | tslil clingman | |
| 2021-01-28 | Added license information! | tslil clingman | |
| 2021-01-18 | Added tunable search depth and self-play | tslil clingman | |
| 2021-01-18 | Toying with symmetrising data | tslil clingman | |
| 2021-01-16 | Went back up to 1986 weights | tslil | |
| 2021-01-15 | Changed function back to macro | tslil | |
| 2021-01-14 | Seems tolerable, but not great. Still wont win or lose ??? | tslil | |
| 2021-01-13 | Fixed minimax (!), fixed bugs in tak.c | tslil | |
| With minimax of depth 1 the evaluation function seems alright with the current method of training and generating weights | |||
| 2021-01-13 | Trying minimax | tslil | |
| 2021-01-13 | Weights? | tslil | |
| 2021-01-13 | Wider? | tslil | |
| 2021-01-13 | Trying various things to teach ti | tslil | |
| 2021-01-12 | Generate the correct format directly | tslil | |
| 2021-01-12 | Train on data that is close to the end of the game only | tslil | |
| 2021-01-11 | Hopefully removed all layer bugs | tslil | |
| 2021-01-11 | Trying again with correct data generation | tslil | |
| 2021-01-11 | Close in principle, but there;s a bug in ct1997 &/ pptdb! | tslil | |
| 2021-01-09 | Two inputs to model, stacks and flat counts | tslil | |
| 2021-01-09 | Shuffling is important | tslil | |
| 2021-01-08 | Fixed a small bug in PTN parsing, other stuff | tslil | |
| 2021-01-08 | Closing in on something reasonable for roads | tslil | |
| 2021-01-07 | Searching for a good representation | tslil | |
| 2021-01-06 | Small tweaks | tslil | |
| 2021-01-06 | On the hunt for a better representation | tslil | |
| 2021-01-05 | Experimenting with training nn | tslil | |
| 2021-01-05 | Separate concerns in pptdb, remove legacy in extract.sh | tslil | |
| 2021-01-05 | Gotta go fast | tslil | |
| 2021-01-05 | First attempt at training data + exclude draws + list overflows too | tslil | |
| Take the output and do grep -Fvxf output_file input_file to drop the bad games from consideration | |||
| 2021-01-04 | Tabs for indentation, spaces for alignment | tslil | |
| 2021-01-04 | Fixed indentation and some bugs, stats programme | tslil | |
 |
index : ctak | |
| An implementation of Tak and a computer opponent in C | git repository hosting |
| aboutsummaryrefslogtreecommitdiff |