CML · Continuous Machine Learning
CML is an open-source CI/CD tool for machine learning projects that integrates with platforms like GitHub, GitLab, and Bitbucket, and supports cloud services such as AWS, Azure, and GCP.
CML Services
CML (Continuous Machine Learning) is an open-source CI/CD tool tailored for machine learning projects. CML facilitates managing machine learning (ML) experiments by allowing users to track changes using GitFlow. It integrates seamlessly with GitHub, GitLab, and Bitbucket, and auto-generates comprehensive reports with metrics and plots for each Git Pull Request. By supporting self-hosted runners, including those on cloud or on-premise, CML provides a versatile and scalable solution for ML workflows.
CML Integrations
CML supports extensive integrations to enhance ML workflows. It works with popular version control platforms like GitHub, GitLab, and Bitbucket. CML also connects with Data Version Control (DVC) for data management, creating a streamlined MLOps environment along with tools like Studio and MLEM. Additionally, CML facilitates launching runners on cloud platforms such as AWS, Azure, GCP, and Kubernetes, and supports GPU-based runners to optimize ML training and reporting tasks.
CML Installation and Setup
CML is designed for easy installation and setup. It can be installed as a Node.js package using npm, simplifying the deployment process. CML does not require additional services or complex setups and can be effectively used with GitHub or GitLab and various cloud services like AWS, Azure, GCP, or Kubernetes. To further streamline the environment, CML provides Docker images with pre-installed dependencies and supports infrastructure as code through Terraform and Docker-Machine.
CML Reporting and Visualization
One of CML's key features is its ability to auto-generate detailed reports with metrics and plots for each Git Pull Request. This functionality enables teams to easily visualize the results of their ML experiments. CML also supports the creation of TensorBoard reports, providing a powerful visualization tool to track and analyze ML experiments. These reporting capabilities are integral for maintaining transparency and facilitating collaboration within ML teams.