openai evolution-strategies-starter issues

openai / evolution-strategies-starter

Code for the paper "Evolution Strategies as a Scalable Alternative to Reinforcement Learning"

https://arxiv.org/abs/1703.03864

MIT License

1.57k stars 278 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Doesn't the gradient need to be rescaled with σ ?

#24 pzdkn opened 3 years ago
4
How to decide values for noise sd and number of samples per gradient estimation?

#23 ben-arnao opened 3 years ago
0
fix bug for tf.concat api

#22 harpsword opened 5 years ago
0
what does the parameter 'eval_prob‘ mean？

#21 kiminh closed 5 years ago
0
update README with repo status

#20 christopherhesse closed 6 years ago
0
Comment

#19 ghost opened 6 years ago
0
Why design ac_noise?

#18 fiberleif opened 6 years ago
1
Config for non-mujoco users

#17 danbri opened 6 years ago
1
Intuitions on implementing ES on text data for text classification use-cases

#16 goodrahstar opened 6 years ago
0
Adding Prioritized Experience Replay

#15 ghost opened 7 years ago
0
Load

#14 ErnstTmp opened 7 years ago
1
ATARI Policy

#13 ErnstTmp opened 7 years ago
0
EBS Optimised Instances

#12 ErnstTmp closed 7 years ago
1
Discrete actions

#11 ErnstTmp closed 7 years ago
1
Commands for MuJoCo installation

#10 DanielTakeshi closed 7 years ago
1
Doesn't work for continuous_mountain_car

#9 joyousrabbit opened 7 years ago
6
Virtual batch Normalization

#8 sahiliitm opened 7 years ago
3
why - 0.5 in es.py#L83 ?

#7 joyousrabbit closed 7 years ago
2
why batched_weighted_sum?

#6 joyousrabbit opened 7 years ago
2
Atari config file

#5 sahiliitm opened 7 years ago
0
non-EC2 version?

#4 benmodels opened 7 years ago
2
Problem with aws s3 cp

#3 cheremovsky closed 7 years ago
0
Missing license

#2 frankier closed 7 years ago
1
Problems with viz.py

#1 truell20 opened 7 years ago
2