search

dusty-nv / jetson-reinforcement

Deep reinforcement learning GPU libraries for NVIDIA Jetson TX1/TX2 with PyTorch, OpenAI Gym, and Gazebo robotics simulator.
MIT License
874 stars 225 forks source link

xavier support? #34

Open WERAQS opened 5 years ago

WERAQS commented 5 years ago

this whole setup doesn't work on Xavier. it generates random errors. I was able to make and run all the examples on Jetson TX2. What is the recommended setup procedure on xavier?

wappler99 commented 5 years ago

Same here, none of this is working on the Xavier. I understand this is all bleeding edge stuff, but it would be awesome if you could make your installs hardware independent! Otherwise this will not scale at all. Thanks!

uersoy commented 4 years ago

This does not work on Xavier NX either. Any plan to revise?

WERAQS commented 4 years ago

ngreedia must update their software to match their hardware.

2020-06-08 3:01 GMT+03:00, ugorski notifications@github.com:

This does not work on Xavier NX either. Any plan to revise?

-- You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: https://github.com/dusty-nv/jetson-reinforcement/issues/34#issuecomment-640299676

dusty-nv commented 4 years ago

The issue stems from versions of PyTorch newer than 0.3 breaking backwards compatibility, so the RL algorithm doesnt work on newer versions. I don't personally plan to update it, but I do keep jetson-inference updated. Since I made this example tutorial there have been more comprehensive RL libraries released - for example see:

https://discourse.ros.org/t/a-toolkit-for-reinforcement-learning-using-ros-and-gazebo/442

https://spinningup.openai.com/en/latest/

uersoy commented 4 years ago

The issue stems from versions of PyTorch newer than 0.3 breaking backwards compatibility, so the RL algorithm doesnt work on newer versions. I don't personally plan to update it, but I do keep jetson-inference updated. Since I made this example tutorial there have been more comprehensive RL libraries released - for example see:

https://discourse.ros.org/t/a-toolkit-for-reinforcement-learning-using-ros-and-gazebo/442

https://spinningup.openai.com/en/latest/

PERFECT! Thanks a lot for the links. I will take a look.