Added self-play support for tianshou DQN example, fixed bugs in training code Full changelog: https://github.com/elliottower/gobblet-rl/commits/
1.3.3
Added dynamic search depth for greedy agent (1-3 turns), refactored code for better readability, fixed minor bugs. Updated dependencies and pyproject.toml to fix installation issues. Full changelog: https://github.com/elliottower/gobblet-rl/commits/
1.3.0
Added greedy agent as strong baseline (wins with search depth 2), added testing for Tianshou & internal game states, fixed edge cases in game logic & UI/interactive play. Full changelog: https://github.com/elliottower/gobblet-rl/commits/
1.2.3
Minor update: fixed bugs in game logic with UI rendering, updated `example_tianshou_DQN.py`to support screen recording, cleaned up and fixed bugs in `manual_policy.py` Full changelog: https://github.com/elliottower/gobblet-rl/commits/
1.2.2
Added support for human interaction vs trained Tianshou agents. Full changelog: https://github.com/elliottower/gobblet-rl/commits/
1.2.1
Minor update: restructure repository and removed unnecessary files, fixed minor bugs. Full changelog: https://github.com/elliottower/gobblet-rl/commits/