xavier support?

[WERAQS]

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]

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]

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

[WERAQS]

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: #34 (comment)

[dusty-nv]

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]

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.