# \[rllib\] Hierarchical Reinforcement Learning (HRL) implementation?

**URL:** <https://discuss.ray.io/t/rllib-hierarchical-reinforcement-learning-hrl-implementation/3525>\
**Category:** RLlib\
**Created:** [September 14, 2021, 5:53pm UTC](https://discuss.ray.io/t/rllib-hierarchical-reinforcement-learning-hrl-implementation/3525 "2021-09-14T17:53:01Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Oblimandias](https://avatars.discourse-cdn.com/v4/letter/o/d26b3c/32.png) [@Oblimandias](https://discuss.ray.io/u/Oblimandias)\
**Post date:** [September 14, 2021, 5:53pm UTC](https://discuss.ray.io/t/rllib-hierarchical-reinforcement-learning-hrl-implementation/3525/1 "2021-09-14T17:53:01Z")

</div>

Hello,

I don’t see any signs of Hierarchical Reinforcement Learning (HRL) being implemented in RLlib (or anything similar to HRL). Is this accurate?  
It seems like a useful method for solving complex environments, so I was wondering if there is interest in implementing it? There is a tensorflow 1.x implementation here, which may provide a good starting point: [https://github.com/tensorflow/models/tree/master/research/efficient-hrl](https://github.com/tensorflow/models/tree/master/research/efficient-hrl)  
The original paper can be found here: [https://arxiv.org/abs/1805.08296](https://arxiv.org/abs/1805.08296)

---

<div class="post-metadata">

**Author:** ![michaelzhiluo](https://avatars.discourse-cdn.com/v4/letter/m/41988e/32.png) [@michaelzhiluo](https://discuss.ray.io/u/michaelzhiluo)\
**Post date:** [September 30, 2021, 11:32pm UTC](https://discuss.ray.io/t/rllib-hierarchical-reinforcement-learning-hrl-implementation/3525/2 "2021-09-30T23:32:20Z")

</div>

Sorry for the late response. Here is a feudal style example in RLlib: [ray/hierarchical\_training.py at master · ray-project/ray · GitHub](https://github.com/ray-project/ray/blob/master/rllib/examples/hierarchical_training.py)