search

jr-robotics / robo-gym

An open source toolkit for Distributed Deep Reinforcement Learning on real and simulated robots.
https://sites.google.com/view/robo-gym
MIT License
426 stars 75 forks source link

Not expected training results #47

Closed Papul-Ghosh closed 3 years ago

Papul-Ghosh commented 3 years ago

Hey, I am trying to train the EndEffectorPositioningURSim-v0 by just replacing your example code. After 1000 episodes I didn't get any promising results. Around 990 times the final status was maximum steps exceeded. Although according to your research paper I was supposed to get something positive after 200 iterations. Does the performance vary from system configuration?

Thanks, Papul

matteolucchi commented 3 years ago

Hi @Papul-Ghosh,

I am sorry but we don't have the resources to support you with the training process, some other users obtained results training on our environments, I'll leave this open, maybe some of them is willing to help. Best of luck with your effort.

Cheers,

Matteo