Getting Started

Installing Neuro and Logging In #

pip install -U neuromation neuro login

Paste that in a macOS, Windows, or Linux terminal prompt. This command automatically installs Neuro CLI and brings you to the log you in screen of Neuro Platform.

Neuro Platform CLI requires Python 3.6 or higher installed. We suggest installing Anaconda Python 3.7 Distribution. On some distributions, you might have to run pip3 install -U neuromation.

Running Your First Job #

From your terminal or command prompt, with CLI installed and logged in, run:

neuro run --name jupyter --preset gpu-small --http 8888 --browse --volume ALL neuromation/base jupyter notebook --no-browser --ip=0.0.0.0 --allow-root --NotebookApp.token= --notebook-dir=/var/neuro
Click to see detailed explanation.
PartExplanation
neuro runa command that starts a job on Neuro Platform
--name jupyterunique IDs identify jobs, but you can also give a job a meaningful name, which you can use in all job-related commands
--preset gpu-smallyou can execute jobs with different amount of GPU/CPU/RAM resources allocated; you can run neuro config show to see the available configurations
--http 8888this job exposes this HTTP port, so that you can access Jupyter notebook web IU
--browsethis option enables Neuro CLI to open the web interface provided by the job in your default browser; you can always reopen it with neuro job browse jupyter
--volume ALLyour storage is a persistent entity, and to use the data from your storage, you have to attach the required folders to a job explicitly. ALL is a handy shortcut which attaches all the data from your storage, as well as all the content other users have shared with you
neuromation/basean environment in which this job runs; this one is our default environment which contains TensorFlow, Keras, PyTorch, TensorBoard, Jupyter Notebook, JupyterLab, and a bunch of other useful libraries and tools

jupyter notebook --no-browser --ip=0.0.0.0 --allow-root --NotebookApp.token= --notebook-dir=/var/neuro is a command executed by this job.

This command starts a Jupyter Notebook instance with all notebooks which are available to you. To create a new notebook, go to the folder called by your user name (you can check it out by running neuro config show).

Do not forget to terminate the job when you are done by running neuro kill jupyter or clicking a "KILL" button on your dashboard.