Home

Studio.ML is a framework for tracking and distributing machine learning experiments. Without invasion in your code, it will make your experiments reproducible and easy to ship for a cloud or a remote execution. Being very non-intrusive, Studio.ML can help with experiment management for use case of any size - from several local experiments that need to be reproducible to thousands of experiments fanned out to hundreds of cloud worker machines. While default experiment repo (zoo.studio.ml) is public, Studio.ML can also be used with Google Storage or Amazon S3 to generate private experiment repos for models based on sensitive data.