search

RLE-Foundation / RLeXplore

RLeXplore provides stable baselines of exploration methods in reinforcement learning, such as intrinsic curiosity module (ICM), random network distillation (RND) and rewarding impact-driven exploration (RIDE).
https://docs.rllte.dev/
MIT License
352 stars 16 forks source link

training performance of intrinsic module different from mlagents icm module #11

Open mchivuku opened 1 year ago

mchivuku commented 1 year ago

Hello,

I used the example code provided: https://github.com/yuanmingqi/rl-exploration-baselines/blob/main/examples/ppo_re3_bullet.py to compute intrinsic rewards using ICM module. I found the results were different from what I have got from running mlagents icm module. This is against the custom unity game wrapped in gym wrapper. Can you please let me know what could be different that I must be missing? Thank you so much for the help. I can also share the code.

yuanmingqi commented 7 months ago

Hello! We've published a big update that provides more reasonable implementations of these intrinsic rewrads.

Now you may get the same performance of the ICM.

If you have any other questions, please don't hesitate to ask here.

@mchivuku

mchivuku commented 7 months ago

Thank you. I will work on taking the update and applying it on my problem.