Week3: Setup ML Container

• The l4t-ml docker image contains TensorFlow, PyTorch, JupyterLab, and other popular ML and data science frameworks such as scikit-learn, scipy, and Pandas pre-installed in a Python 3 environment. You can find all information in NVIDIA L4T ML

Machine Learning Container

• The l4t-ml docker image contains TensorFlow, PyTorch, JupyterLab, and other popular ML and data science frameworks such as scikit-learn, scipy, and Pandas pre-installed in a Python 3.6 environment
  • Latest l4t-ml:r32.6.1-py3
    • TensorFlow 1.15.5
    • PyTorch v1.9.0
    • torchvision v0.10.0
    • torchaudio v0.9.0
    • onnx 1.8.0
    • CuPy 9.2.0
    • numpy 1.19.5
    • numba 0.53.1
    • OpenCV 4.5.0 (with CUDA)
    • pandas 1.1.5
    • scipy 1.5.4
    • scikit-learn 0.23.2
    • JupyterLab 2.2.9

Running the Container on your Jetson-Nano

• First pull one of the l4t-ml container:

nano@jetson-nano:~$ sudo docker pull nvcr.io/nvidia/l4t-ml:r32.6.1-py3
....

• Then to start an interactive session in the container, run the following command:

nano@jetson-nano:~$ sudo docker run -it --gpus all -e DISPLAY=:0 -v /tmp/.X11-unix:/tmp/.X11-unix --network host nvcr.io/nvidia/l4t-ml:r32.6.1-py3

Mounting Directories

• To mount scripts, data, ect. from your Jetson's filesystem to run inside the container, use Docker's -v flag when starting your Docker instance:

nano@jetson-nano:~$ sudo docker run -it --rm --runtime nvidia --network host -v /home/user/project:/location/in/container nvcr.io/nvidia/l4t-ml:r32.6.1-py3

• Note that you should change the path directories and adapt to your installation: -v flag adds the mecanism to share a folder between host and docker image running, the path "/home/user/project" corresponds to host path (adapt to your installation "/home/nano/l4t-ml-data"), and "/location/in/container" corresponds to docker path (you could change to "/home/l4t-ml-data")
• Now, you should then be able to start a Python3 interpreter

Connecting to JupyterLab Server

• A JupyterLab server instance is automatically started along with the container.
• You can connect http://localhost:8888 (or substitute the IP address of your Jetson device)
  • Password: nvidia

Please send a message to the professor as soon as you finished

Please send a message to the professor as soon as you finished