Rubicon-ML
capitalone.github.io
- Free plan
- Linux
- Web
- API
- Self-hosted
What Rubicon-ML is
Rubicon-ML is a free, open-source Python package for recording and reviewing model training and execution information. It stores model inputs, outputs, parameters, and outcomes in a repeatable, searchable form. Its Git integration can connect experiments to a repository origin, branch, and commit, supporting auditability and reproducibility. A dashboard and widgets let users explore, filter, compare, visualize, and share logged work with reviewers or stakeholders. Logging supports parallel experiments and asynchronous S3 communication. Documented storage options are a local filesystem, Amazon S3, in-memory storage, and Weights & Biases. Integrations include Git, Prefect, and Scikit-learn; RubiconPipeline can log parameters and metrics when Scikit-learn pipelines are fit and scored. Install through pip or conda-forge. Visualization tools are an optional pip extra, while conda installs optional extras by default. The dashboard is a Dash app that can be dockerized, and remote logging does not require a hosted server. The W&B backend is experimental and lacks project-level artifacts and dataframes, as well as get_projects() listing. The project is licensed under Apache License 2.0.
Who it suits
Rubicon-ML suits developers and science or research users who want searchable experiment records and a dashboard for reviewing model metadata. It offers local, S3, in-memory, and Weights & Biases storage options.
Good to know, and what to watch
- Logs model inputs, outputs, parameters, and outcomes.
- Git metadata can link experiments to code revisions.
- Supports parallel logging and asynchronous S3 communication.
- Integrates with Prefect and Scikit-learn.
- Remote logging does not require a hosted server.
- The Weights & Biases backend is experimental.
- W&B lacks project-level artifacts and dataframes.
- Visualization tools are an optional pip extra.
Plans
Every plan and pricerubicon-ml
Open source
The verdict
Rubicon-ML offers a free, open-source way to log and review model experiments, with Git links, multiple storage backends, and workflow integrations. Check the documented Weights & Biases limits and visualization installation options before choosing a setup.
What the maker publishes
- Auditability
- Its Git integration associates model inputs and outputs with the model code that produced them to support auditability and reproducibility. source
- Concurrent and asynchronous logging
- The documentation says rubicon-ml supports logging experiments in parallel and asynchronous communication with S3 so network reads and writes do not block. source
- Concurrent logging
- Rubicon-ML supports logging multiple experiments in parallel and asynchronous communication with S3. source
- Dashboard deployment
- The dashboard is a Dash app that can be dockerized like other Dash apps. source
- Experiment review
- The dashboard lets users explore, filter, visualize, and share logged work with reviewers and stakeholders. source
- Founded
- 1994 source
- Git audit trail
- Its Git integration can record repository origin, branch names, and commit hashes alongside projects and experiments. source
- Git integration
- With auto_git_enabled=True, projects can record the GitHub origin URL and experiments can record the active branch and last commit hash. source
- Headquarters
- McLean, Virginia, United States source
- Installation
- The Python package can be installed using pip or conda-forge. source
- Integrations
- The documentation includes integrations for Git, Prefect, and Scikit-learn. source
- Intended users
- The package metadata identifies developers and science or research users as its intended audience. source
- License
- The project is licensed under Apache License 2.0. source
- Logging
- The logging component organizes, stores, and retrieves model inputs and outputs using pluggable storage backends powered by fsspec. source
- No hosted server requirement
- The FAQ says rubicon-ml does not require a hosted server for remote logging and supports local, S3, or in-memory data storage. source
- Prefect integration
- The Prefect integration adds rubicon-ml logging tasks to Prefect workflows to persist experimentation metadata. source
- Purpose
- Rubicon-ML captures and stores model training and execution information, including parameters and outcomes, in a repeatable and searchable way. source
- Scikit-learn integration
- RubiconPipeline can automatically log parameters and metrics when Scikit-learn pipelines are fit and scored. source
- Security guidance
- For EC2 S3 logging, the documentation says the IAM role needs s3:GetObject, s3:PutObject, and s3:DeleteObject permissions for the bucket. source
- Storage backends
- Documented backends include local filesystem, Amazon S3, in-memory storage, and Weights & Biases. source
- Storage options
- Documented backends include local filesystem, Amazon S3, in-memory storage, and Weights & Biases. source
- Support
- The project README directs users to open a GitHub issue to report a bug or make a suggestion. source
- Visualization
- The dashboard and widgets let users explore and compare logged model metadata. source
- Visualization dependencies
- The visualization tools are an optional pip extra, while conda installs the optional extras by default. source
Checked against the maker's pages on 9 Oct 2026.