Our Jobs offering includes advanced functionality like public IPs and ports, custom containers etc. that can be beneficial to a Jupyter Notebook/Lab environment.
Running a Notebook as a Job is simple. Here are the requirements:
- A container with Jupyter installed
- Specify --ports 8888:8888 as a Job parameter
- Use jupyter lab --allow-root --ip=0.0.0.0 as the Job entrypoint
paperspace jobs create --container paperspace/fastai:cuda9_pytorchJL --machineType P4000 --command 'jupyter lab --allow-root --ip=0.0.0.0' --ports 8888:8888 --project 'FastAI-Job-Notebook'
Once the Job is in the Running state, your Notebook will be available on the assigned IP address with port 8888 eg http://18.104.22.168:8888. Your Jupyter token (password) can be found in your Job logs:
Here is an example Public Job of Job with Jupyter Lab running inside the Fast.ai container: https://www.paperspace.com/console/jobs/jsl8bq42cfm70q
Stopping your Notebook
Just click Stop on the Jobs list to end the Job.