|
11 | 11 |  |
12 | 12 | [](https://github.com/automl/Mighty/actions/workflows/test.yaml) |
13 | 13 | [](https://github.com/automl/Mighty/actions/workflows/docs_test.yaml) |
| 14 | +[](https://doi.org/10.21105/joss.10439) |
14 | 15 |
|
15 | 16 | </div> |
16 | 17 |
|
|
28 | 29 | # Mighty |
29 | 30 |
|
30 | 31 | Welcome to Mighty, hopefully your future one-stop shop for everything cRL. |
31 | | -Currently Mighty is still in its early stages with support for normal gym envs, DACBench and CARL. |
| 32 | +Mighty supports standard Gymnasium environments as well as DACBench and CARL for contextual RL. |
32 | 33 | The interface is controlled through hydra and we provide DQN, PPO and SAC algorithms. |
33 | 34 | We log training and regular evaluations to file and optionally also to wandb. |
34 | 35 | If you have any questions or feedback, please tell us, ideally via the GitHub issues! |
@@ -177,15 +178,29 @@ Mighty is meant to be a platform to build upon and not a large collection of met |
177 | 178 | - **Meta Components**: RND, NovelD, SPaCE, PLR |
178 | 179 | - **Runners**: online RL runner, ES runner |
179 | 180 |
|
| 181 | +## Contributions |
| 182 | + |
| 183 | +- **Aditya Mohan:** Implemented the core agent stack (training loop, model interfaces, evaluation tools), ran the experiments and evaluation, and maintained the codebase, including integration of new modules and cross-component consistency. |
| 184 | +- **Theresa Eimer:** Designed the system architecture, Built the environment interfaces and data pipelines, contributed to architectural decisions, and implemented the meta-learning components and their evaluation setups. |
| 185 | +- **Carolin Benjamins:** Reviewed the architecture, conducted focused code audits, and improved interface design and modular structure. |
| 186 | +- **Marius Lindauer:** Guided the overall research direction, supervised the project, and refined the methodological setup. |
| 187 | +- **André Biedenkapp:** Provided project supervision, contributed to research design, and advised on algorithmic methodology and experimentation. |
| 188 | + |
180 | 189 | ## Cite Us |
181 | 190 |
|
182 | | -If you use Mighty in your work, please cite us: |
| 191 | +If you use Mighty in your work, please cite our JOSS paper: |
183 | 192 |
|
184 | 193 | ```bibtex |
185 | | -@misc{mohaneimer24, |
186 | | - author = {A. Mohan and T. Eimer and C. Benjamins and M. Lindauer and A. Biedenkapp}, |
187 | | - title = {Mighty}, |
188 | | - year = {2024}, |
189 | | - url = {https://github.com/automl/mighty} |
| 194 | +@article{mohan2026mighty, |
| 195 | + author = {Aditya Mohan and Theresa Eimer and Carolin Benjamins and Marius Lindauer and Andr\'{e} Biedenkapp}, |
| 196 | + title = {Mighty: A Comprehensive Tool for Studying Generalization, Meta-RL and AutoRL}, |
| 197 | + journal = {Journal of Open Source Software}, |
| 198 | + year = {2026}, |
| 199 | + volume = {11}, |
| 200 | + number = {123}, |
| 201 | + pages = {10439}, |
| 202 | + publisher = {The Open Journal}, |
| 203 | + doi = {10.21105/joss.10439}, |
| 204 | + url = {https://doi.org/10.21105/joss.10439} |
190 | 205 | } |
191 | 206 | ``` |
0 commit comments