Using DVC Commands

DVC is a command line tool. Here we provide the specifications, complete descriptions, and comprehensive usage examples for dvc commands. For a list of commands, type dvc -h.

💡 To execute any DVC command in a different directory, use dvc --cd <path> ... before the actual command and it's options/arguments (this does not change directories in your terminal).

Typical DVC workflow

• In an existing Git repository, initialize a DVC project with dvc init.
• Copy data files or dataset directories for modeling into the repository, and track them with DVC using the dvc add command.
• Process the data with your own source code, using dvc.yaml and/or the dvc run command, specifying further outputs that should also be tracked by DVC after the code is executed.
• Sharing a DVC repository with the codified data pipeline will not include the project's cache. Use remote storage and dvc push to share this cache (data tracked by DVC).
• Use dvc repro to automatically reproduce your full pipeline iteratively as input data or source code change.