Commit graph

24 commits

Author SHA1 Message Date
Henri Bourcereau 0c58490f87 feat: bot sac & ppo save & load 2025-08-21 14:35:25 +02:00
Henri Bourcereau 778ac1817b script train bots 2025-08-10 15:35:12 +02:00
Henri Bourcereau fc9733b729 doc train bot results 2025-08-03 22:16:28 +02:00
Henri Bourcereau c0d42a0c45 réglages train bot dqn burnrl 2025-08-03 16:11:45 +02:00
Henri Bourcereau fd269b491d wip stackoverflow debug 2025-07-28 14:03:14 +02:00
Henri Bourcereau 3e1775428d action mask 2025-07-27 09:35:30 +02:00
Henri Bourcereau 6a7b1cbebc fix by gemini 2025-06-28 22:18:39 +02:00
Henri Bourcereau f05094b2d4 wip 2025-06-28 21:34:44 +02:00
Henri Bourcereau cf93255f03 claude not tested 2025-06-23 22:17:24 +02:00
Henri Bourcereau a06b47628e burn dqn trainer 2025-06-22 21:25:45 +02:00
Henri Bourcereau bae0632f82 use game state context to reduce actions space 2025-06-03 21:41:07 +02:00
Henri Bourcereau f7eea0ed02 extend actions space 2025-06-01 20:00:15 +02:00
Henri Bourcereau ab959fa27b train command 2025-05-26 20:44:35 +02:00
Henri Bourcereau 480b2ff427 remove python stuff & simple DQN implementation 2025-05-24 22:41:44 +02:00
Henri Bourcereau 8368b0d837 wip Gym : Claude AI suggestion 2025-03-02 11:01:24 +01:00
Henri Bourcereau 6478f5043d fix: allowed moves infinite loop 2025-01-29 20:07:30 +01:00
Henri Bourcereau 38100a61b2 todo 2025-01-25 23:51:30 +01:00
Henri Bourcereau e95b25a9bc maj doc 2025-01-24 18:04:44 +01:00
Henri Bourcereau fb5e954b85 passage intermédiaire sur coin de repos 2024-06-23 11:38:09 +02:00
Henri Bourcereau 3c3c6d8458 moves allowed : check if opponent corner can be filled #2 2024-05-18 21:46:26 +02:00
Henri Bourcereau 44c040b414 doc rng 2024-03-10 20:00:18 +01:00
Henri Bourcereau 782a6dce87 cli start game 2024-02-17 12:55:36 +01:00
Henri Bourcereau 7cf8cd1c46 reorga doc 2024-02-13 20:13:25 +01:00
Henri Bourcereau 6fe5a268da doc 2024-01-14 12:24:57 +01:00