Notebooks
W
Weights and Biases
TensorBoard And Weights And Biases

TensorBoard And Weights And Biases

tensorboardwandb-examplescolabs

Open In Colab

Weights & Biases

Weights & Biases

By the end of this colab you will have a TensorBoard server running in Weights & Biases, just like this:

W&B in TensorBoard

This code is modified from the offical TensorBoard getting started code

🪴 Start a Weights & Biases run

When using Weights & Biases for the first time you will need to:

1️⃣. Sign-up for a free W&B account here

2️⃣. Create a new W&B API key at your settings page and store it securely. API keys can only be viewed once when created.

3️⃣. Initialise a W&B run with wandb.init and you will be prompted to enter your API key to log in

[ ]

Initialising a Weights & Biases run with sync_tensorboard=True will enable wandb to pick up

[ ]

🌿 Create Your Dataset and Model

[ ]

🌲 Train Your Model and Log to TensorBoard AND Weights & Biases

The tensorboard logs will be automatically picked up by Weights & Biases and logged

[ ]

(Notebook only) Finish the Weights & Biases Run

[ ]

Documentation

You can find additional documentation of how to use Weights & Biases with Tensorboard here