GPT-3.5-instruct is a ~1800 ELO chess player. 150 games of GPT-3.5 vs. stockfish (twitter.com) 3 points by seraine 2y ago ↗ HN
[–] sinuhe69 2y ago ↗ I glanced at the code, but still not sure how did he prompt ChatGPT. Just feeding the pure PGN with no further instruction? [–] seraine 2y ago ↗ Correct. An example can be seen here:https://github.com/adamkarvonen/chess_gpt_eval/blob/master/g...
[–] seraine 2y ago ↗ Correct. An example can be seen here:https://github.com/adamkarvonen/chess_gpt_eval/blob/master/g...
2 comments
[ 6.1 ms ] story [ 226 ms ] threadhttps://github.com/adamkarvonen/chess_gpt_eval/blob/master/g...