ClearML

ClearML is an open-source MLOps stack that automates experiment tracking, resource orchestration, and model deployment for production-grade AI.

ClearML delivers a unified platform to manage the entire machine learning lifecycle (from initial R&D to scaling on remote clusters). It integrates with two lines of code to log metrics, capture environment snapshots, and version data automatically. The platform includes a robust orchestrator (ClearML Agent) that turns any machine into a compute node, plus a serving layer for low-latency model inference. By decoupling code from infrastructure, it allows teams to scale from a single workstation to thousands of GPUs without rewriting their pipelines.

https://clear.ml