Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

EC

Episodic control algorithms:

Default options for MFEC:

python main.py --kernel mean

Default options for NEC:

python main.py --algorithm NEC \
               --key-size 128 \
               --num-neighbours 50 \
               --dictionary-capacity 500000 \
               --episodic-multi-step 100 \
               --epsilon-final 0.001 \
               --discount 0.99 \
               --learn-start 50000

Used in Memory-efficient episodic control reinforcement learning with dynamic online k-means and Sample-Efficient Reinforcement Learning with Maximum Entropy Mellowmax Episodic Control, both presented at the Workshop on Biological and Artificial Reinforcement Learning, NeurIPS 2019.

Acknowledgements

  • Alex Pritzel for implementation details

About

Episodic Control

Topics

Resources

License

Releases

No releases published

Packages

No packages published

Languages