The partnership provides ML practitioners with a complete solution for the entire ML lifecycle, from MLOps and experiment tracking to resource management at scale
Weights & Biases (W&B) and Run:ai, have announced a joint partnership, focusing on developing integrations between the two platforms to provide ML developers and MLOps platform owners a seamless experience for MLOps and GPU orchestration for machine learning and deep learning workloads.
“We are excited to partner with Run.ai to provide data scientists and ML practitioners with the best tools for their ML development workflow. The combination of the developer tools provided by Weights and Biases, the dynamic allocation and orchestration of computing resources from Run.ai, and the optimised hardware and software provided by Nvidia gives ML teams everything they need to accelerate the development and deployment of AI in the enterprise,” said Seann Gardiner, VP of Business Development, Weights & Biases.
The partnership will streamline AI projects for AI researchers and the MLOps and IT teams managing AI infrastructure. ML practitioners can leverage the MLOps capabilities provided by Weights & Biases – such as experiment and artefact tracking, hyperparameter optimisation, and collaboration reports – and access Nvidia accelerated computing resources orchestrated by Run:ai’s Atlas Platform, all in one experience. ML developers will be able to monitor Nvidia GPU utilisation within the W&B dashboard, and then improve utilisation with Run:ai’s scheduling and orchestration capabilities.
The MLOps platform owner will be able to optimise GPU resource scheduling and consumption for the ML practitioners and provide a single ML system of record to keep an accurate history of all ML experiments, model history, and dataset versioning.
“Run:ai and Weights & Biases together give data scientists, and the IT teams that support them, a complete solution for the full ML lifecycle, from building models, to training and inference in production. Companies building their AI infrastructure and tooling from scratch can use Nvidia GPUs, Run:ai, and W&B and have everything they need to manage AI initiatives at scale,” said Omri Geller, CEO and Co-founder, Run:ai.
The Weights & Biases and Run:ai partnership will benefit ML researchers and enterprises leveraging Nvidia GPUs. Both Run:ai and W&B has been validated as Nvidia AI Accelerated and Nvidia DGX-Ready software partners. The partnership will build on integrations between W&B and Run.ai with Nvidia AI Enterprise software to help enterprises maximise their investment in Nvidia-accelerated systems for ML and AI.