Deep Learning VM Images is a set of prepackaged virtual machine images
with a deep learning framework that are ready to be run out of the box. The
Deep Learning VM images take the hassle out of setting up an
environment for training deep learning models by preconfiguring dependencies,
preinstalling essential tools, and optimizing performance.

It is important to distinguish between *images* , *image families* , and
*instances*.

## Image families

An image family is a set of images that are preconfigured for a specific
purpose or using a specific architecture.

There are numerous Deep Learning VM image families. They are listed
here, according to each one's main installed framework:

- TensorFlow Enterprise family, with or without GPU
- TensorFlow family, with or without GPU
- PyTorch family, with or without GPU
- R experimental family
- A base, or "common" image family to which you can add your preferred framework, with or without GPU

[TensorFlow Enterprise](https://docs.cloud.google.com/tensorflow-enterprise/docs) is a distribution of
[TensorFlow](https://www.tensorflow.org/)
that has been optimized to run on Google Cloud and includes
[Long Term Version
Support](https://docs.cloud.google.com/tensorflow-enterprise/docs/overview#long_term_version_support).

If you know what framework you want to use and whether you want to enable one
or more GPUs, you then choose whether to use the latest version of your
framework or, if available, an older version. You can use
Deep Learning VM images for some previous versions of frameworks.
To find out more, see [Choosing an Image](https://docs.cloud.google.com/deep-learning-vm/docs/images).

## Images

An image is an individual virtual machine template. The Deep Learning VM
images are preconfigured [public Compute Engine virtual machine
images](https://docs.cloud.google.com/compute/docs/images#os-compute-support).

For more information about VM images, see the [Images](https://docs.cloud.google.com/compute/docs/images)
conceptual topic in Compute Engine documentation.

For information about choosing a Deep Learning VM image, see [Choosing
an Image](https://docs.cloud.google.com/deep-learning-vm/docs/images).

## Instances

An instance is a single virtual machine hosted on Google's infrastructure. A
Deep Learning VM instance is based on one of the preconfigured
Deep Learning VM images. You create an instance from an image using the
Google Cloud console or the Google Cloud CLI.

For information about choosing a Deep Learning VM image, see [Choosing
an Image](https://docs.cloud.google.com/deep-learning-vm/docs/images).

For information about creating instances of a Deep Learning VM image,
see:

- [Launching a Deep Learning VM Instance from the Command
  Line](https://docs.cloud.google.com/deep-learning-vm/docs/cli)

- [Launching a Deep Learning VM Instance from the Cloud
  Marketplace](https://docs.cloud.google.com/deep-learning-vm/docs/cloud-marketplace)

For general information about creating instances in Compute Engine, see [Virtual
Machine Instances](https://docs.cloud.google.com/compute/docs/instances).