- In this release, the observation values for the terminated state in Hex have been changed (in #1288). There is no impact on the agent's learning or performance. In accordance with the versioning policy, the Hex environment version remains unchanged.
player_id
argument inEnv.observe
is deprecated
What's Changed
- [Docs] Update hex.md by @KazukiOhta in #1285
- [AlphaZero] Fix minor typo in train.py by @KazukiOhta in #1286
- [Hex] Extract game specific attributes by @sotetsuk in #1287
- [Hex] Extract logic functions by @sotetsuk in #1288
- [Hex] Create
Game
class by @sotetsuk in #1289 - [Hex] Enhance terminal computation by @sotetsuk in #1290
- [Hex] Use
_player_order
by @sotetsuk in #1291 - [API] Deprecate
player_id
in observe by @sotetsuk in #1292 - [Docs] Update README by @sotetsuk in #1293
Full Changelog: v2.4.2...v2.5.0