• W and amp;B Launch is a simple tool for automatically packaging code and launching a task into any target environment. It is currently available in public preview.
  • W and amp;B Models is a scalable method for governing ML model lifecycles in a central repository and facilitating cross-functional discovery and collaboration. It is currently generally available to all W and amp;B users.

Weights and amp; Biases Inc., a startup in machine learning development whose software is utilized by organizations like OpenAI LLC and Nvidia Corp. to create new artificial intelligence models, recently announced two significant upgrades to its platform.

Weights and Biases have developed a platform for MLOps (machine learning models and operations) development and collaboration. The tool allows teams to monitor machine learning projects. It also includes tools for analyzing the performance of various machine learning models, versioning datasets, and managing pipelines.

The W and amp;B platform is intended to increase the effectiveness of the iterative development process of creating AI software. The business claims that because it addresses one of the significant problems facing AI initiatives—the organization and processing of project data—it helps by enhancing developer productivity.

W and amp;B Launch and W and amp;B Models are two new platform features unveiled at Fully Connected, W and amp;B’s first user conference.

W and amp;B Launch offers a more straightforward approach for users to package code automatically and begin a new machine learning job on any target environment, and it is now accessible in public preview. This reduces infrastructure complexity while facilitating machine learning practitioners’ access to the required computing resources. Additionally, W and amp;B Launch makes it simpler for teams to reproduce runs and scale those activities up and down.

Principal AI researcher at ARQUIMEA Research Center, Orlando Avila-Garcia, stated that W and amp;B Launch would simplify his group’s research on a range of deep learning approaches, such as neural radial fields and graph neural networks. Orlando Avila-Garcia said, “Abstracting away the complexity of using the infrastructure for our researchers is very beneficial to our overall team. Launch greatly simplifies our work optimizing, experimenting with, and benchmarking ML methods, letting us focus on reliability and reproducibility of results.”

W and amp;B Models is now broadly accessible and provides a more scalable method for teams to regulate machine learning model lifecycles in a centralized repository while enabling cross-functional discovery and collaboration. W and amp;B Models may assist teams in maintaining higher-quality models over time due to its reproducibility and lineage tracking capabilities, which allow users to know precisely when a model was transferred from staging to production.

Weights and amp; Biases is one of the more established players in the MLOps market. According to Andy Thurai, vice president and lead analyst of Constellation Research Inc., it excels in model development; data set versioning and experiment monitoring. He continued that

W and amp;B Models is an excellent addition to the company’s solution set. “It offers ML model registry and model governance, allowing users to collaborate more easily by making models discoverable across large enterprises. This, combined with model lineage tracking, enables users to track models in the ML pipeline from inception to production.” He then added.

W and amp;B Vice President of Product Phil Gurbacki said, “We’re confident that W and amp;B users will quickly see the benefits of Launch and Models in accelerating model training, utilizing compute resources efficiently, managing models with more confidence, and having a more cohesive end-to-end ML workflow.”

The platform upgrades result from what W and amp;B deems a fruitful year, which has seen substantial traction and momentum. The firm, which most recently received USD 135 million at USD 1 billion in October 2021, claims to have quadrupled its personnel count over the past year, building new offices in Berlin, London, and Tokyo to enhance its worldwide footprint.