We believe that powerful AI tools should be accessible to every researcher, not just those at mega-corporations. Today, we are officially releasing LabScale, our internal Python library for managing distributed training across heterogeneous hardware.
Whether you are working with a single workstation or a multi-node cluster, LabScale automates the orchestration of data parallelism, allowing you to focus on the architecture, not the infrastructure.
"Our goal is to lower the barrier to entry for high-performance deep learning." — Dr. Sarah Chen, Lead Researcher.