Unify evaluation metrics across all algorithms - Githubissues

LucasAlegre / morl-baselines

Multi-Objective Reinforcement Learning algorithms implementations.

https://lucasalegre.github.io/morl-baselines

MIT License

295 stars 47 forks source link

Unify evaluation metrics across all algorithms #36

Closed ffelten closed 1 year ago

ffelten commented 1 year ago

Relates to #12

Algos checklist:

Multi-policy

[x] Envelope
[x] GPI
[x] GPI Continuous
[x] MPMOQL
[x] OLS
[x] PGMORL
[x] PCN
[x] PQL Here is the resulting dashboard:

Single-policy

[x] EUPG
[x] MOQL
[x] wait for MO-Gymnasium latest release with pareto_front computations