Weights & Biases Launches Best-in-Class Experiment Tracking for OpenAI's API to Power the Next Generation of Natural Language Machine Learning

SAN FRANCISCO, Feb. 8, 2022 /PRNewswire/ -- Today, Weights & Biases, the leading developer-first MLOps platform, announced a collaboration with OpenAI, an AI research and deployment company known for its powerful language model, GPT-3. This collaboration enables OpenAI developers to access GPT-3 projects on Weights & Biases' best-in-class platform, bringing experiment tracking, dataset versioning, and model management into their workflows and building powerful AI applications faster.

OpenAI's API gives practitioners access to GPT-3, which can be applied to virtually any task that involves understanding or generating natural language. The newly launched feature allows machine learning teams to easily log, track, and visualize their GPT-3 experiments, helping them create models that can be pushed to production quickly.

"OpenAI is one of the earliest Weights & Biases customers and we're proud that we're rolling out a powerful collaboration to the broader community together," said Lukas Biewald, CEO and Co-Founder of Weights & Biases. "We've seen GPT-3 power everything from code writing to creative fiction to text-driven design and we're proud that machine learning teams will now get the necessary insights our platform provides while they're building the next generation of AI-powered applications GPT-3 makes possible."

OpenAI just recently removed the waitlist for GPT-3 and tens of thousands of developers are already building powerful apps with OpenAI's cutting-edge technology as their backbone. The OpenAI API can be applied to virtually any task that involves understanding or generating natural language or code. OpenAI's API offers a wide spectrum of models suitable for different tasks that can be used for everything from content generation to semantic search and classification.

"We're big W&B fans here at OpenAI. Since W&B enables tracking progress over time, this integration allows our users who are fine-tuning GPT-3 to iterate faster toward increased accuracy of their production models," said Peter Welinder, VP of Product and Partnerships at OpenAI.

"We are excited to use W&B capabilities to monitor the fine-tuning of our OpenAI models that help improve how we understand client feedback at Stitch Fix," Reza Sohrabi, Staff ML Engineer, Stitch Fix Algorithms.

"With the new W&B and OpenAI integration, we're able to consolidate experiment tracking of our in-house ML projects and OpenAI fine-tuning projects in one place. This gives us huge time savings towards delivering insights from unstructured data for our customers," Prem Viswanathan, ML Lead at Artifact.io

The addition of OpenAI's API to Weights and Biases' MLOps platform promises to accelerate the production of state-of-the-art models in myriad industries and verticals. Teams training models through this feature will enjoy the power of GPT-3 with built-in tracking, versioning, data and model lineage, and the ability to collaborate on live dashboards that provide the valuable insights often hidden from practitioners using legacy tracking systems.

About Weights & Biases

Weights & Biases is the leading developer-first MLOps platform to build better models faster. Used by top researchers including teams at Qualcomm, NVIDIA, OpenAI, Lyft, Pfizer, Toyota, Github, and MILA, W&B is part of the new standard of best practices for machine learning.

For Media Inquiries, Please Contact [email protected].

SOURCE Weights & Biases