model_garden/notebooks/official/tensorboard/README.md

Vertex AI TensorBoard Custom Training with custom container

Learn how to create a custom training job using custom containers, and monitor your training process on Vertex AI TensorBoard in near real time.

The steps performed include:

* Create docker repository & config.
* Create a custom container image with your customized training code.
* Setup service account and Google Cloud Storage buckets.
* Create & launch your custom training job with your custom container.

   Learn more about Vertex AI TensorBoard.

   Learn more about Custom training.

Vertex AI TensorBoard custom training with prebuilt container

Learn how to create a custom training job using prebuilt containers, and monitor your training process on Vertex AI TensorBoard in near real time.

The steps performed include:

* Setup service account and Google Cloud Storage buckets.
* Write your customized training code.
* Package and upload your training code to Google Cloud Storage.
* Create & launch your custom training job with Tensorboard enabled for near real time monitorning.

   Learn more about Vertex AI TensorBoard.

   Learn more about Custom training.

Vertex AI TensorBoard Hyperparameter Tuning with the HParams Dashboard

This tutorial shows you how to log hyperparameter experiment results in TensorFlow and visualize the results in TensorBoard's Hparams dashboard.

The steps performed include:

* Adapt TensorFlow runs to log hyperparameters and metrics.
* Start runs and log them all under one parent directory.
* Visualize the results in TensorBoard's HParams dashboard.

   Learn more about Vertex AI TensorBoard.

Profile model training performance using Vertex AI TensorBoard Profiler

Learn how to enable Vertex AI TensorBoard Profiler for custom training jobs.

The steps performed include:

- Setup a service account and a Cloud Storage bucket
- Create a TensorBoard instance
- Create and run a custom training job that enables TensorBoard Profiler
- View the TensorBoard Profiler dashboard to debug your model training performance

   Learn more about Vertex AI TensorBoard Profiler.

Profile model training performance using Vertex AI TensorBoard Profiler in custom training with prebuilt container

Learn how to enable the TensorBoard Profiler in Vertex AI for custom training jobs with a prebuilt container.

The steps performed include:

- Prepare your custom training code and load your training code as a Python package to a prebuilt container
- Create and run a custom training job that enables the TensorBoard Profiler
- View the TensorBoard Profiler dashboard to debug your model training performance

   Learn more about Vertex AI TensorBoard Profiler.

Vertex AI TensorBoard integration with Vertex AI Pipelines

Learn how to create a training pipeline using the KFP SDK, execute the pipeline in Vertex AI Pipelines, and monitor your training process on Vertex AI TensorBoard in near real time.

The steps performed include:

* Setup a service account and Google Cloud Storage buckets.
* Construct a KFP pipeline with your custom training code.
* Compile and execute the KFP pipeline in Vertex AI Pipelines with Tensorboard enabled for near real time monitorning.

   Learn more about Vertex AI TensorBoard.

   Learn more about Vertex AI Pipelines.