Skip to content

Latest commit

 

History

History
38 lines (30 loc) · 1.51 KB

README.md

File metadata and controls

38 lines (30 loc) · 1.51 KB

Jupyter Notebooks

The Jupyter Notebooks provide easy to use tutorials on how to use the xfDNN tools and deploy models within the Xilinx ML Suite.

Note for AWS users: The ml-suite AMI on the AWS marketplace already has Jupyter installed, and it will auto-start the notebooks shortly after you launch the EC2 instance. You just need to navigate via a web-browser to <public-dns>:8888

Installation

  1. Install Jupyter on the remote system (We recommend doing this inside your Anaconda environment).
$ source ~/.bashrc
$ source activate ml-suite
$ pip install jupyter`

Follow the instructions to complete the install. You do not need to install Microsoft VScode, when prompted.

  1. Navigate to the top level `ml-suite' dir and source the setup script
$ cd /ml-suite/
$ source ./overlaybins/setup.sh <platform>

Options for <platform>: aws nimbix 1525 alveo-u200 alveo-u250 alveo-u200-ml alveo-u250-ml

  1. Set initial Password for Jupyter Server
$ jupyter notebook --generate-config
$ jupyter notebook password 
$ Enter Password: 
$ Verify Password: 
$ [NotebookPasswordApp] Wrote hashed password to /root/.jupyter/jupyter_notebook_config.json

Launching Notebook Server

  1. Launch the jupyter notebook server
    jupyter notebook --no-browser --ip=*

  2. On local machine, open a broswer and navigate to http://localhost:8888 or youripaddress>:8888

  3. On a remote machine, open a broswer on your local machine navigate to yourpublicipaddress>:8888