Case Studies

Hear from our customers about how they use Weights & Biases for machine learning models in production use cases.
Quickstart guide to ML tracking
W&B REPORT
How Cohere Trains Business-Critical LLMs with W&B
Learn how Cohere leverages W&B to customize unique large language models for individual customers with nuanced datasets
Quickstart guide to ML tracking
W&B REPORT
How MARZ Uses W&B to Supercharge Visual FX
MARZ is an Emmy nominated visual FX studio. Learn how they streamline their workflows with machine learning and W&B
Quickstart guide to ML tracking
W&B REPORT
How Microsoft Leveraged Weights & Biases to Build the Models Behind Ink
A deep dive into how the team behind Microsoft Journal created a delightful freeform note taking experience
Quickstart guide to ML tracking
W&B REPORT
Democratizing AI for Biology with OpenFold
Learn how the researchers at OpenFold recreated AlphaFold and helped bring cutting-edge research to a wider audience
Quickstart guide to ML tracking
W&B REPORT
How Socure Fights Fraud with Machine Learning
Discover how the team at Socure use computer vision to keep everyone's wallet a little more secure
ML experiments tracking
PYTORCH OFFICIAL BLOG
Wadhwani AI Empowering Cotton Farmers
How Wadhwani identifies the most impactful hyperparameters, visualize predictions, and organize complex ML projects with W&B
Quickstart guide to ML tracking
YOUTUBE VIDEO
OpenAI's Robotics Team Visualizing Model Training
How the robotics team at OpenAI visualizes real time model training, optimizes performance, and identifies regressions
Quickstart guide to ML tracking
PYTORCH OFFICIAL BLOG
Production Machine Learning at Blue River
How John Deere's robotics team set up their ML tools stack to optimize for  efficiency, flexibility, and model performance
Quickstart guide to ML tracking
W&B REPORT
OpenAI Learning Dexterity End-to-End
How OpenAI runs a massive machine learning project – like training a robot hand to manipulate physical objects e.g. the Rubik's cube
Quickstart guide to ML tracking
WOVEN PLANET BLOG
Improving ML Models for Autonomous Vehicles
Discover how the team at Woven Planet Level 5 uses W&B to evaluate, introspect, and debug their AV models
Quickstart guide to ML tracking
W&B REPORT
The Science of Debugging ML with Latent Space
Building an AI-rendered 3D engine, pushing the frontier of generative modeling, and keeping a distributed team focused with W&B reports
Quickstart guide to ML tracking
YOUTUBE VIDEO
Toyota Research Institute and W&B
How ML engineers building self-driving car models rely on W&B for reliable tracking and flexible experimentation
Quickstart guide to ML tracking
W&B REPORT
Scaling BERT on Graphcore Hardware with W&B
Learn how W&B helped Graphcore drive 50 to 100 times more experiments on their advanced IPU hardware 
Quickstart guide to ML tracking
W&B REPORT
Capella Space’s Cutting Edge Satellite Data 
How Capella Space’s ML team seamlessly collaborates to provide the highest quality synthetic aperture radar data available
Quickstart guide to ML tracking
W&B REPORT
Lyft’s End-to-End Camera-Lidar Fusion
A deep dive into how Lyft marries multiple perception sensors to accelerate autonomous driving research