Skip to main content

Bring Your Own Container (BYOC)

info

This is an advanced feature. If you are new to DL, it would be easier to choose one of the frameworks, such as PyTorch, TensorFlow, or FastAI.

With Jarvislabs, you can Bring Your Own Container (BYOC).

Jarvislabs.ai BYOC

To launch an instance from a public image, follow these steps:

  1. Choose BYOC under the framework.
  2. Enter a public image name, such as pytorch/pytorch (organization/repository), in the docker field.
  3. Ensure that you have configured your SSH key in the API keys section.

After the instance is launched, click on the SSH button ssh button to get the SSH string and connect to the instance. The SSH string would look something like this:

​ ssh -p port_number1 root@ssha.jarvislabs.ai -L port_number2:localhost:8888

Connecting to a JupyterLab inside the instance​

Install JupyterLab in the newly created instance:

​ pip install jupyterlab

danger

This step could change based on the image you choose. Some Docker images may have Jupyter installed, or you may need to activate a conda environment. So look at the specific Docker image docs on how it is configured to use.

Starting JupyterLab​

​ jupyter lab --allow-root

Open a browser, and enter the following URL:

​ http://localhost:port_number2/

info

Copy the port number from the SSH URL and replace the port number above. Use the token printed in the terminal while starting JupyterLab.

This feature is in beta. If you face any challenges, please reach out to us. We will be happy to adapt and make the process smoother.

tip

Nvidia provides a variety of optimized NGC Containers. You can choose the required container from here.