search

jurgisp / memory-maze

Evaluating long-term memory of reinforcement learning algorithms
MIT License
129 stars 13 forks source link

Add environment variant with continuous action space #7

Open jurgisp opened 1 year ago

jurgisp commented 1 year ago

It would be nice to have a version of the environment with the raw continuous action space instead of the discretized one.