TensorBoard Online in 1 Line
Keep all your existing TensorBoard code. Get all the power of Weights & Biases.
Created on March 30|Last edited on March 31
Comment
TensorBoard server running remotely in the cloud on Weights & Biases
Why use TensorBoard with Weights & Biases?
Keep your existing TensorBoard code
- With only 1 line of code you can log your TensorBoard charts to the cloud and share with everyone
Share with anyone
- By logging to Weights & Biases you can easily share links to TensorBoard charts and insights to anyone
System Metrics
- Automatically get telemetry on your GPU and CPU usage to understand if you're making the most out of your hardware
Free Forever
- Weights & Biases is free for personal use and academic users, now and forever
Get Started Now
Using TensorBoard with Weights and Biases just requires you to initialise a Weights & Biases run before your main training loop
# Install wandb!pip install wandbimport wandb# Start a W&B run, passing `sync_tensorboard=True`, to plot your TensorBoard fileswandb.init(project='my-wonderful-project', sync_tensorboard=True)# Your Keras, Tensorflow or PyTorch training code using TensorBoard...# If running in a notebook, finish the wandb run to upload the TensorBoard logs to W&Bwandb.finish()
👉 Run this colab to see how easy it is!
👉 See a live, online TensorBoard running in Weights & Biases here
You Might Also Like
Advanced Tensorboard Features: Graph Dashboard
An introduction to exploring Computation Graphs of our Machine Learning workflows with Tensorboard
Tensorboard With Accelerators: A Guide
In this article, we explore how to effortlessly integrate Weights & Biases into pre-existing accelerator-based workflows (both GPU and TPU) using TensorBoard.
Add a comment