search

sush1996 / DDPG_Fetch

Exploring the performance of Prioritized Experience Replay (PER) with the DDPG+HER scheme on the Fetch Robotics Environemnt
16 stars 2 forks source link

fixed some bugs for using cuda #2

Closed kaixindelele closed 3 years ago

kaixindelele commented 3 years ago

https://github.com/kaixindelele/DDPG_Fetch

And how did you draw this multi-bar graph? The script to draw the graph is not in this code base?

sush1996 commented 3 years ago

If you're referring to the single plot for multiple models, I've separately saved the mean success rates in each epoch, for each mode in an array and plotted all those in "viz.py". I've uploaded my plot data and the program for plotting it - it's not clean at the moment. But I'll get to it. Thanks for pointing out.

kaixindelele commented 3 years ago

Thanks, I got the viz.py. I was going to upload two files, but I can not pull new request the new branch. I modified the two files： https://github.com/kaixindelele/DDPG_Fetch/blob/master/ddpg_agent.py

https://github.com/kaixindelele/DDPG_Fetch/blob/master/models.py