Using Notebooks
Notebooks can be created by clicking Create Notebook button on the Notebooks tab. There are three types of base containers:
These are the templates that are most popular on Gradient. They include the most common ML frameworks and libraries. Click here to discover which Notebook Base Container will best suit your use case, then choose it from the options below:
These are additional templates that are made available in Gradient as a 1-click option. Click here to discover which Notebook Base Container will best suit your use case, then choose it from the options below:
If you need to import an existing container, visit the Custom Containers section. Learn more here.
This is instance type that your notebook will run on. Our hardware is configured to work with the tools and packages in each of the provided containers.
Provide a friendly name for your Notebook in the text field
If you have a Gradient subscription, you will have the ability to set a custom auto-shutdown interval.
If you do not have a Gradient subscription, your notebook will default to an auto-shutdown interval of 12 hours, to help avoid unexpected runtime charges. Once created, you can stop, start, fork, and swap out the instance type anytime.
Workspaces
Additionally, you have the option to select a remote Github repository to serve as workspace storage for your Notebook. If the repository is private, you need to provide access credentials (username and a password/apiKey). Also, you can provide an optional Github ref that points to a specific version of the repository.
That's it! Your Notebook will start once it has been created 🚀
Once created, your Notebook will automatically start. Just click Open and your Notebook will appear in a new browser tab.
gradient notebooks create [options...]
Machine Type: Such as
--P100or--C7or--TPUContainer: Such as
--tensorflow/tensorflow:1.5.1-gpu
Usage: gradient notebooks create \--name "my job" \--container "http://dockerhub.com/mycontainer" \--machineType "P5000" \--command "/paperspace/run.sh"--projectId "someProjectID"
Argument | Description |
| Notebook name |
| Project ID |
| An optional machine type to run the job on: either 'GPU+', 'P4000', 'P5000', 'P6000', 'V100', 'K80', 'P100', or 'TPU'. [Required] |
| A reference to a docker image in a public or private docker registry. Docker image repository references must be in lowercase and may include a tag and a hostname prefix followed by a slash; if committed the hostname defaults to that of the public Docker Hub registry. An example docker image reference: |
| Container user |
| Command (executed as |
| ID of the cluster to run the notebook. |
| Registry username |
| Registry password |
| Shutdown timeout in hours |
| Use a preemptible instance type |
| One or multiple tags that you want to add to the notebook |
| S3 url or git repository. Directory uploads
are not yet supported |
| Workspace password |
| Git commit hash, branch name or tag |
| Environment variables |
| Shutdown timeout in hours |
| API key to use this time only |
Start a stopped notebook again. The notebook can be started on a different cluster or machine type.
gradient notebooks start [options...]
Options:
Option | Description |
| Notebook ID [required] |
| Virtual Machine type label e.g. P5000 [required] |
| Cluster ID |
| Notebook name |
| Shutdown timeout in hours |
You can rename any Notebook, whether it is running or stopped, in order to help organize your work. Simply click on the notebook name in the list and it will become an editable text field; press the enter key when you are done to save it:
