Skip to content
Edit on GitHub

Project Structure

Using dvc init in your workspace will initialize a DVC project, including the internal .dvc/ directory. From there on, you will create and manage different DVC metafiles (below), and populate the cache with data artifacts as you work on your ML experiments.

  • dvc.yaml files define stages that form the pipeline(s) of a project. All stage-based features such as dvc params, dvc metrics, and dvc plots are specified here.

  • .dvc files ("dot DVC files") are placeholders to track data files and directories.

  • .dvcignore files (optional) contain a list of paths for DVC to ignore, which can dramatically increase its operational performance.

  • Internal files and directories in .dvc/ contains the local configuration file(s), default local cache location, and other utilities that DVC needs to operate.

These metafiles are typically versioned with Git, as DVC does not replace its distributed version control features, but rather extends on them.

🐛 Found an issue? Let us know! Or fix it:

Edit on GitHub

Have a question? Join our chat, we will help you:

Discord Chat