Release 0.8.0

What's Changed

• Imitation learning support by @Ivan-267 in #169
• Adds downfall env to tests and hub by @edbeeching in #176
• Add SB3 script instructions link to readme by @Ivan-267 in #178
• Update sb3_imitation.py by @Ivan-267 in #180
• Fix typo "aims" by @batikanor in #182
• Updated WORKING_WITH_CSHARP.md with CSharpAll variant details by @LorinczAdrien in #185
• [WIP] Add working with c sharp doc by @Ivan-267 in #184
• Ready for testing 🧪 Multi-policy training support by @Ivan-267 in #181
• release 0.8.0 by @edbeeching in #190

New Contributors

• @batikanor made their first contribution in #182
• @LorinczAdrien made their first contribution in #185

Full Changelog: v0.7.0...v0.8.0

v0.7.0

What's Changed

• Fix rllib by @edbeeching in #140
• Adds nparallel multi-process support to CleanRL example by @Ivan-267 in #145
• Update ADV_STABLE_BASELINES_3.md by @Ivan-267 in #146
• Require sb3 version 2 or newer by @Ivan-267 in #148
• Fix a small spelling error in the Readme by @Curtiss02 in #151
• updates version to v0.6.2 by @edbeeching in #153
• Update readme by @metinc in #154
• fixing broken tests on main by @edbeeching in #156
• Refactor godot_env.py by @Ivan-267 in #129
• Fix inference when action names are not alphabetic by @Ivan-267 in #159
• Adds optional linear LR schedule to the example by @Ivan-267 in #157
• Adds general tips to the docs. by @Ivan-267 in #162
• Adds saving model on ctrl + c in Python during training by @Ivan-267 in #160
• Adds the ability to use more algorithms from sbx and sb3-contrib (with e.g. MlpPolicy) by @Ivan-267 in #163
• Adds a node reference to the documentation by @Ivan-267 in #165
• [fix]Ray 2.9.0 failed tests by @Ivan-267 in #164
• Adds new example envs to test suite by @edbeeching in #166
• Update troubleshooting by @Ivan-267 in #167
• fix dtype of image data by @Ivan-267 in #158
• Adds onnx export to cleanrl example by @Ivan-267 in #161
• Adds code quality by @edbeeching in #168
• Release v0.7.0 by @edbeeching in #171

New Contributors

• @Curtiss02 made their first contribution in #151
• @metinc made their first contribution in #154

Full Changelog: v0.6.0...v0.7.0

Release 0.6.0

What's Changed

• Adds multi-training backend support by @edbeeching in #111
• feat(sb3): also track reward in tensorboard by @visuallization in #113
• feat(multi discrete action spaces): added support for multi discrete action spaces by @visuallization and @Ivan-267 in #110
• Fixes clean_rl_example.py default env path and related CUSTOM_ENV documentation section by @Ivan-267 in #116
• Update setup.cfg to fix sb3 not training after installing godot-rl by @Ivan-267 in #120
• Update CUSTOM_ENV.md - Small spelling correction by @Ivan-267 in #122
• Adds multiple env support to the SB3 wrapper by @edbeeching in #123
• Feature: experiment_dir + experiment_name by @visuallization in #126
• Experiment name typo correction by @Ivan-267 in #131
• chore(docs): added short entry regarding training statistics and tensorboard by @visuallization in #128
• Feature/hp tuning by @visuallization in #130
• Updates support of sb3 to latest version by @edbeeching in #134
• Adds resuming and saving model to sb3 example by @Ivan-267 in #135
• Update ADV_STABLE_BASELINES_3.md by @Ivan-267 in #136

Full Changelog: v0.5.0...v0.6.0

Beta Release 0.5.0

• ONNX Support
• Greatly improved Documentation
• Adds ONNX models for BallChase, JumperHard and FlyBy
• Bugfixes

What's Changed

• Doc updates by @edbeeching in #85
• Improved FAQ and Fixed Typos by @AleryBerry in #87
• Feat(rllib): added possibility to set the experiment name by @visuallization in #82
• Test updates by @edbeeching in #88
• chore(readme): updated clean rl title by @visuallization in #93
• chore(debugging): added more debug information on import error by @visuallization in #91
• updates deps and wrappers for version 0.4.8 by @edbeeching in #94
• Update CUSTOM_ENV.md by @edbeeching in #96
• Fixed error in _set_platform_suffix by @AustereTriceratops in #97
• Update utils.py by @yizhangliu in #99
• Added note to CUSTOM_ENV.md that addresses issue #100 by @kasmith11 in #102
• Adds sb3 onnx export by @edbeeching in #101

New Contributors

• @AleryBerry made their first contribution in #87
• @AustereTriceratops made their first contribution in #97
• @yizhangliu made their first contribution in #99
• @kasmith11 made their first contribution in #102

Full Changelog: v0.4.7...v0.5.0

Release v0.4.7

What's Changed

• Godot4 by @edbeeching in #27
• Fixed issues mentioned in issue#33 by @ryash072007 in #38
• Created clean rl example by @edbeeching in #42
• Remove = from --fixed fps by @jtatusko in #43
• chore(readme): added missing code formatting + deleted redundant download snippet by @visuallization in #46
• Adds multi dist support to all envs by @edbeeching in #50
• chore(readme): added a disclaimer about the plugin's compatibility by @visuallization in #57
• Update CUSTOM_ENV.md by @ryash072007 in #67
• Add export option to RayRLLib by @yaelatletl in #68
• Feature: MacOs Support by @visuallization in #74
• MacOS M1 support for rllib by @visuallization in #76
• Feature: Allow Users to select version by @yaelatletl in #75
• Docs custom envs by @edbeeching in #80
• Feature: convert action space + make rllib run on wsl2 by @visuallization in #79

New Contributors

• @ryash072007 made their first contribution in #38
• @jtatusko made their first contribution in #43

Full Changelog: v0.2.2...v0.4.7

Release 0.2.2 fixes issue where linux builds were including in the environment.yml file.

Fixes issue where linux builds were including in the environment.yml file.

Release 0.2.1 fixes a bug in the installation instructions

Merges PR #13

Sensor suite and improved docs

Implemented a number of features, bug fixes and improvements to the documentation.

• Including an updated sensor suite.
• New checkpoints for the updated sensors.
• The conda environment should now work out of the box and support GPUs. #8 #9
• Fixed a bug with the reward function in the BallChase env #11
• Improved documentation #7

v0.1.0

version 0.1.0 of Godot RL agents
Includes 4 example games and wrappers for Ray RLlib and StableBaselines3