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

Patches needed to make this work on L4T 4.4.38-tegra #32

Open chrislgarry opened 5 years ago

chrislgarry commented 5 years ago

The installation instructions do not hold for current out of the box Jetson TX2, which at this time is 4.4.38-tegra. There are some slight changes that need to be made. The first is enabling the universe repositories in /etc/apt/sources.list so the pre-build script can, at the least, successfully sudo apt-get install python-pip:

https://gist.github.com/chrislgarry/40e9cd3687eb3803e190ce3aeea05edb

The second is to pass the pyzmq version that works nicely with jupyter notebooks to pip:

https://gist.github.com/chrislgarry/af7be9ec17fbf0cd8f6883873f4bdbeb

I believe there is a PR for the latter fix. Running pip install --upgrade may work as well instead of that patch. See https://github.com/jupyter/notebook/issues/3579

chrislgarry commented 5 years ago

Further, avahi-autoipd needs to be updated/reinstalled else python notebooks may fail when attempting to use dhclient with tegra-ubuntu root: /etc/dhcp/dhclient-enter-hooks.d/avahi-autoipd returned non-zero exit status 1 in syslog.

sudo apt-get install avahi-autoipd