TensorBoard as a CML Application

This repository demonstrates a minimal example of running TensorBoard as an Application on CML by visualizing the training of a simple neural network on the MNIST digits dataset. The example used in this repo has been adapted from this notebook.

Repository Structure

.
├── cml                     # This folder contains scripts that facilitate the project launch on CML.
├── images                  # Storage for the images in this README
├── logs                    # Storage for the TensorBoard logs
├── load_and_train.py       # Simple script to train a model and capture logs
├── .project-metadata.yaml  # Declarative specification of this project
├── LICENSE                 # This code has an Apache 2.0 License
├── README.md               # This file
└── requirements.txt        # Python 3 package requirements

Launching the project on CML

There are three ways to launch this project on CML:

1. From Prototype Catalog - Navigate to the AMPs tab on a CML workspace, select the "TensorBoard" tile, click "Launch as Project", click "Configure Project"
2. As ML Prototype - In a CML workspace, click "New Project", add a Project Name, select "ML Prototype" as the Initial Setup option, copy in the repo URL, click "Create Project", click "Configure Project"
3. Manual Setup - In a CML workspace, click "New Project", add a Project Name, select "Git" as the Initial Setup option, copy in the repo URL, click "Create Project". Launch a Python 3 Workbench Session and run !pip3 install -r requirements.txt to install requirements. Then create a CML Application as described in the CML documentation, using cml/launch_tensorboard.py as the script.

Using the App

Once the CML Application has been created (by any means), you can launch it from the Applications pane. This should open a browser window displaying the TensorBoard dashboard. To track your own custom model development, configure your training script to save logs to the logs directory. For more information on configuring TensorBoard and advanced features, see the official documentation.