Weights & Biases simplifies machine learning generation and management

Machine learning development startup Weights & Biases Inc., whose software is used by OpenAI LLC and Nvidia Corp. to develop new artificial intelligence models, today announced two major enhancements to its platform.

Weights and Biases has created a platform for teams to build and collaborate on machine learning models and operations, or MLOs. The platform enables teams to track their machine learning experiments. It also provides tools for evaluating the performance of various machine learning models, database differentiation, and pipeline management.

The W&B platform is designed to improve the efficiency of the trial-and-error process through which AI software is developed. According to the startup, it helps increase developer productivity as it solves one of the main challenges of AI initiatives: organizing and processing project data.

New additions to its platform were announced at Fully Connected, W&B’s inaugural user conference, and include W&B Launch and W&B Models.

Available in public preview starting today, W&B Launch provides an easier way for users to automatically package code and run new machine learning jobs on any target environment. In this way, machine learning practitioners gain easier access to the computing resources they need while simplifying infrastructure complexity. In addition, W&B Launch makes it easy for teams to replicate runs and scale and expand those operations.

Orlando Avila-Garcia, principal AI researcher at ARQUIMEA Research Center, said W&B Launch will make it easier for his organization to accelerate research into various deep learning techniques, including neural beam fields and graph neural networks. “Removing the complexity of using the infrastructure for our researchers is very beneficial for our overall team,” he said. “The startup greatly simplifies our work optimizing, testing and benchmarking ML methods, allowing us to focus on the reliability and reproducibility of results.”

As for W&B models, this is now generally available and provides a more scalable way for teams to manage machine learning model lifecycles in a centralized repository while enabling cross-functional discovery and collaboration. With its reproducibility and lineage tracking features that enable users to track exactly when a model moves from staging to production, W&B Models can help teams maintain higher quality models over time.

Andy Turay, vice president and principal analyst at Constellation Research Inc., told SiliconANGLE that Weights & Biases is an old player in the MLOps segment and is particularly strong in model building, dataset differentiation and experiment tracking. He added that W&B Models looks like it will be a good addition to the company’s suite of solutions. “It offers an ML model registry and model management that allows users to collaborate more easily, making models discoverable across large enterprises,” he said. “This, combined with model lineage tracking, enables users to track ML pipeline models from inception to production.”

“We are confident that W&B users will quickly see the benefits of Launch and Models in accelerating model training, efficient use of computing resources, managing models with more confidence, and a more consistent end-to-end ML workflow,” W&B said. Vice. Product President Phil Gurback.

The improvements to the platform follow what W&B says has been a very promising 12 months, during which it has gained significant traction and momentum. The company, which last raised $135 million at a $1 billion valuation in October 2021, claims to have doubled its employee base in the past year, opening new offices in Berlin, London and Tokyo to help boost its international presence.

Image: weights and biases

Show your support for our mission by joining our Cube Club and our community of Cube Event experts. Join a community that includes Amazon Web Services and Amazon.com CEO Andy Jassy, ​​Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many other stars and experts.

Source link