Choose a virtual machine (VM) image

User-managed notebooks instances are Deep Learning VM Images instances with JupyterLab notebook environments enabled and ready for use. Specific user-managed notebooks images are available to suit your choice of framework and processor. To find the image that you want, see the following table.

Deciding on an image family

Deciding on which user-managed notebooks image family to use depends on your needs. The following table lists the default versions of image families, organized by framework type. Creating an instance by referencing an image family with -notebookin the name ensures that your instance is a supported image family. If you need a specific framework version that is not shown here, skip to Listing all available versions.

Framework Processor Image family name(s)
Base GPU common-cu100-notebooks
CPU common-cpu-notebooks
TensorFlow Enterprise 2.x GPU tf2-ent-2-1-cu110-notebooks
TensorFlow Enterprise 1.x GPU tf-ent-1-15-cu110-notebooks
PyTorch GPU pytorch-1-9-cu110-notebooks
PyTorch XLA TPU/GPU/CPU (experimental) pytorch-1-9-xla-notebooks
R CPU (experimental) r-4-0-cpu-experimental-notebooks
RAPIDS GPU (experimental) rapids-0-18-gpu-experimental-notebooks
Deprecated notebook images Mixed tf2-2-0-cu100-notebooks

Choosing an operating system

For most frameworks, Debian 10 is the default OS. Ubuntu 18.04 images are available for some frameworks. They are denoted by the -ubuntu-1804 suffixes in the image family name (see Listing all available versions). Debian 9 images have been deprecated.

PyTorch and TensorFlow Enterprise (tf-ent and tf2-ent) image families support A100 GPU accelerators.

TensorFlow Enterprise images

TensorFlow Enterprise image families provide you with a Google Cloud optimized distribution of TensorFlow, and specific versions of the TensorFlow Enterprise distribution also include Long Term Version Support. To learn more about TensorFlow Enterprise, read the TensorFlow Enterprise overview.

Use the following table of available TensorFlow images to help you select the image with the version of TensorFlow or TensorFlow Enterprise that you want.

Version of TensorFlow or TensorFlow Enterprise Processor Image family name Long Term Version Support
TensorFlow Enterprise 2.6 GPU tf2-ent-2-6-cu110-notebooks Included
TensorFlow Enterprise 2.5 GPU tf2-ent-2-5-cu110-notebooks Not included
TensorFlow 2.4 GPU tf2-2-4-cu110-notebooks Not included
TensorFlow Enterprise 2.3 GPU tf2-ent-2-3-cu110-notebooks Included
TensorFlow 2.2 GPU tf2-2-2-cu101-notebooks Not included
TensorFlow Enterprise 2.1 GPU tf2-ent-2-1-cu110-notebooks Included
TensorFlow 2.0 GPU tf2-2-0-cu100-notebooks Not included
TensorFlow Enterprise 1.15 GPU tf-ent-1-15-cu110-notebooks Included
TensorFlow 1.14 GPU tf-1-14-cu100-notebooks Not included
TensorFlow 1.13 GPU tf-1-13-cu100-notebooks Not included

Experimental images

Some user-managed notebooks image families are experimental, as indicated by the table of image families. Experimental images are supported on a best-effort basis, and may not receive refreshes on each new release of the framework.

Specifying an image version

Creating a new user-managed notebooks instance based on the image family name gives you the most recent image of that version of the framework. For example, if you create a user-managed notebooks instance based on the family name tf-ent-1-15-cu110-notebooks, the specific image name might look like tf-ent-1-15-cu110-notebooks-v20201016.

To create multiple user-managed notebooks instances based on the exact same image, use the image name instead of the image family name.

To determine the exact name of the most recent image, use the following command in the gcloud command-line tool with your preferred terminal or in Cloud Shell. Replace IMAGE_FAMILY with the image family name for which you want to find out the most recent version number.

gcloud compute images describe-from-family IMAGE_FAMILY \
        --project deeplearning-platform-release

Look for the name field in the output and use the image name given there when creating new instances.

Listing all available versions

If you need a specific framework, CUDA version, or operating system, you can search the complete list of available images. To list all available user-managed notebooks images, use the following gcloud tool command.

gcloud compute images list \
        --project deeplearning-platform-release | grep notebooks

Image families will be in the format FRAMEWORK-VERSION-CUDA_VERSION(-experimental)-notebooks, where FRAMEWORK is the target library, VERSION is the framework version, and CUDA_VERSION is the version of the CUDA stack, if present.

For example, an image from the family tf2-ent-2-3-cu110-notebooks has TensorFlow Enterprise 2.3 and CUDA 11.0.

What's next