Vertex AI TensorBoard is an enterprise-ready managedversion ofOpen Source TensorBoard(TB), which is a Google Open Source project for machine learning experimentvisualization.

With Vertex AI TensorBoard, you can track, visualize, and compareML experiments and share them with your team.

Vertex AI TensorBoard provides various detailed visualizations, thatincludes:

• tracking and visualizing metrics such as loss and accuracy over time,
• visualizing model computational graphs (ops and layers),
• viewing histograms of weights, biases, or other tensors as they change overtime,
• projecting embeddings to a lower dimensional space,
• and displaying image, text, and audio samples.

In addition to the powerful visualizations fromTensorBoard, Vertex AI TensorBoard provides:

• a persistent, shareable link to your experiment's,Vertex AI TensorBoard experiment,
• tight integrations with Vertex AI services for model training,
• enterprise-grade security, privacy, and compliance.

Integration with Vertex AI Experiments lets you:

• use a searchable and compare list of all experiments in a project,
• view time series metrics in the Google Cloud console,
• compare scalars across experiments and experiment runs,
• have direct access to Vertex AI TensorBoard.

Vertex AI TensorBoard and the Google Cloud console

The Google Cloud console is used to:

• create or delete Vertex AI TensorBoard instances,
• createVertex AI Experiments,
• view Vertex AI TensorBoard instance storage size → associated costs,
• view associatedVertex AI Experiments, Custom Jobs, andPipeline Runs,
• visualize sometime series metrics.

What's next

• Setup Vertex AI TensorBoard
• Check out theTensorBoard API documentation.