TianhongDai / hindsight-experience-replay

This is the pytorch implementation of Hindsight Experience Replay (HER) - Experiment on all fetch robotic environments.
MIT License
396 stars 75 forks source link

get global standard deviation for plotting #13

Open stefanwanckel opened 3 years ago

stefanwanckel commented 3 years ago

To plot the training results I can simply use the mean value logged in .log. However, the standard deviation is not computed in _eval_agent. How did you get it?

TianhongDai commented 3 years ago

The std is computed using 5 different random seeds, and you also need to calculate the mean value by averaging the success rate across 5 different random seeds