{"slug": "treeverse-dvc-data-versioning-and-ml-experiments", "title": "Treeverse/Dvc: Data Versioning and ML Experiments", "summary": "Treeverse's DVC (Data Version Control) is a command line tool and VS Code Extension that versions data and models in cloud storage while keeping version metadata in a Git repo, tracks ML experiments locally without servers, and connects code and data through lightweight pipelines. DVC stores data and model files in a cache outside Git and supports remotes on S3, Azure, Google Cloud and SSH-based on-premise storage, with a VS Code extension from Iterative offering experiment tracking and data management that requires core DVC to be installed separately.", "body_md": "[Website](https://dvc.org)\n• [Docs](https://dvc.org/doc)\n• [Blog](http://blog.dataversioncontrol.com)\n• [Tutorial](https://dvc.org/doc/get-started)\n• [Related Technologies](https://dvc.org/doc/user-guide/related-technologies)\n• [How DVC works](#how-dvc-works)\n• [VS Code Extension](#vs-code-extension)\n• [Installation](#installation)\n• [Contributing](#contributing)\n• [Community and Support](#community-and-support)\n\n**Data Version Control** or **DVC** is a command line tool and [VS Code Extension](#vs-code-extension) to help you develop reproducible machine learning projects:\n\n1. **Version** your data and models.\nStore them in your cloud storage but keep their version info in your Git repo.\n2. **Iterate** fast with lightweight pipelines.\nWhen you make changes, only run the steps impacted by those changes.\n3. **Track** experiments in your local Git repo (no servers needed).\n4. **Compare** any data, code, parameters, model, or performance plots.\n5. **Share** experiments and automatically reproduce anyone's experiment.\n\nPlease read our [Command Reference](https://dvc.org/doc/command-reference) for a complete list.\n\nA common CLI workflow includes:\n\n| Task | Terminal | \n|---|---|\n| Track data | `$ git add train.py params.yaml``$ dvc add images/` | \n| Connect code and data | `$ dvc stage add -n featurize -d images/ -o features/ python featurize.py``$ dvc stage add -n train -d features/ -d train.py -o model.p -M metrics.json python train.py` | \n| Make changes and experiment | `$ dvc exp run -n exp-baseline``$ vi train.py``$ dvc exp run -n exp-code-change` | \n| Compare and select experiments | `$ dvc exp show``$ dvc exp apply exp-baseline` | \n| Share code | `$ git add .``$ git commit -m 'The baseline model'``$ git push` | \n| Share data and ML models | `$ dvc remote add myremote -d s3://mybucket/image_cnn``$ dvc push` | \n\nWe encourage you to read our [Get Started](https://dvc.org/doc/get-started) docs to better understand what DVC\ndoes and how it can fit your scenarios.\n\nThe closest *analogies* to describe the main DVC features are these:\n\n1. **Git for data** : Store and share data artifacts (like Git-LFS but without a server) and models, connecting them with a Git repository. Data management meets GitOps!\n2. **Makefiles** for ML: Describes how data or model artifacts are built from other data and code in a standard format. Now you can version your data pipelines with Git.\n3. Local **experiment tracking** : Turn your machine into an ML experiment management platform, and collaborate with others using existing Git hosting (Github, Gitlab, etc.).\n\nGit is employed as usual to store and version code (including DVC meta-files as placeholders for data).\nDVC [stores data and model files](https://dvc.org/doc/start/data-management) seamlessly in a cache outside of Git, while preserving almost the same user experience as if they were in the repo.\nTo share and back up the *data cache*, DVC supports multiple remote storage platforms - any cloud (S3, Azure, Google Cloud, etc.) or on-premise network storage (via SSH, for example).\n\n[DVC pipelines](https://dvc.org/doc/start/data-management/data-pipelines) (computational graphs) connect code and data together.\nThey specify all steps required to produce a model: input dependencies including code, data, commands to run; and output information to be saved.\n\nLast but not least, [DVC Experiment Versioning](https://dvc.org/doc/start/experiments) lets you prepare and run a large number of experiments.\nTheir results can be filtered and compared based on hyperparameters and metrics, and visualized with multiple plots.\n\nTo use DVC as a GUI right from your VS Code IDE, install the [DVC Extension](https://marketplace.visualstudio.com/items?itemName=Iterative.dvc) from the Marketplace.\nIt currently features experiment tracking and data management, and more features (data pipeline support, etc.) are coming soon!\n\nNote: You'll have to install core DVC on your system separately (as detailed\nbelow). The Extension will guide you if needed.\n\nThere are several ways to install DVC: in VS Code; using `snap`, `choco`, `brew`, `conda`, `pip`; or with an OS-specific package.\nFull instructions are [available here](https://dvc.org/doc/get-started/install).\n\n```\nsnap install dvc --classic\n```\n\nThis corresponds to the latest tagged release.\nAdd `--beta` for the latest tagged release candidate, or `--edge` for the latest `main` version.\n\n```\nchoco install dvc\nbrew install dvc\nconda install -c conda-forge mamba # installs much faster than conda\nmamba install -c conda-forge dvc\n```\n\nDepending on the remote storage type you plan to use to keep and share your data, you might need to install optional dependencies: dvc-s3, dvc-azure, dvc-gdrive, dvc-gs, dvc-oss, dvc-ssh.\n\n```\npip install dvc\n```\n\nDepending on the remote storage type you plan to use to keep and share your data, you might need to specify one of the optional dependencies: `s3`, `gs`, `azure`, `oss`, `ssh`. Or `all` to include them all.\nThe command should look like this: `pip install 'dvc[s3]'` (in this case AWS S3 dependencies such as `boto3` will be installed automatically).\n\nTo install the development version, run:\n\n```\npip install \"dvc @ git+https://github.com/treeverse/dvc\"\n```\n\nSelf-contained packages for Linux, Windows, and Mac are available.\nThe latest version of the packages can be found on the GitHub [releases page](https://github.com/treeverse/dvc/releases).\n\n```\nsudo wget https://dvc.org/deb/dvc.list -O /etc/apt/sources.list.d/dvc.list\nwget -qO - https://dvc.org/deb/iterative.asc | sudo apt-key add -\nsudo apt update\nsudo apt install dvc\nsudo wget https://dvc.org/rpm/dvc.repo -O /etc/yum.repos.d/dvc.repo\nsudo rpm --import https://dvc.org/rpm/iterative.asc\nsudo yum update\nsudo yum install dvc\n```\n\nContributions are welcome!\nPlease see our [Contributing Guide](https://dvc.org/doc/user-guide/contributing/core) for more details.\nThanks to all our contributors!\n\nThis project is distributed under the Apache license version 2.0 (see the LICENSE file in the project root).\n\nBy submitting a pull request to this project, you agree to license your contribution under the Apache license version 2.0 to this project.\n\nIterative, *DVC: Data Version Control - Git for Data & Models* (2020)\n[DOI:10.5281/zenodo.012345](https://doi.org/10.5281/zenodo.3677553).\n\nBarrak, A., Eghan, E.E. and Adams, B. [On the Co-evolution of ML Pipelines and Source Code - Empirical Study of DVC Projects](https://mcis.cs.queensu.ca/publications/2021/saner.pdf) , in Proceedings of the 28th IEEE International Conference on Software Analysis, Evolution, and Reengineering, SANER 2021. Hawaii, USA.", "url": "https://wpnews.pro/news/treeverse-dvc-data-versioning-and-ml-experiments", "canonical_source": "https://github.com/treeverse/dvc", "published_at": "2026-11-06 09:00:46+00:00", "updated_at": "2026-11-06 09:20:10.029597+00:00", "lang": "en", "topics": ["mlops", "machine-learning", "developer-tools", "ai-tools"], "entities": ["Treeverse", "DVC", "Git", "VS Code", "Iterative", "Amazon S3", "Azure", "Google Cloud"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/treeverse-dvc-data-versioning-and-ml-experiments", "markdown": "https://wpnews.pro/news/treeverse-dvc-data-versioning-and-ml-experiments.md", "text": "https://wpnews.pro/news/treeverse-dvc-data-versioning-and-ml-experiments.txt", "jsonld": "https://wpnews.pro/news/treeverse-dvc-data-versioning-and-ml-experiments.jsonld"}}