Any possibilities to include A3C implementation?

higgsfield / RL-Adventure

Pytorch Implementation of DQN / DDQN / Prioritized replay/ noisy networks/ distributional values/ Rainbow/ hierarchical RL

2.99k stars 587 forks source link

Any possibilities to include A3C implementation? #4

Closed hohoCode closed 6 years ago

hohoCode commented 6 years ago

Thank you for your great tutorial.

higgsfield commented 6 years ago

Thanks! Actor-Critic will be in RL-adventure-2 :)

hohoCode commented 6 years ago

Thanks!

I am also sure you have noticed this very recent work, ApeX: 'Distributed Prioritized Experience Replay', which pushed everything to a whole new level (beating Rainbow a lot...). I know RayLIB has implemented a basic ApeX framework to do so, probably just need to connect the dots. If this distributed-DQN work also can be included in your great tutorial that would be amazing. Thanks.