DagsHub

dagsHub is a machine learning collaboration tool designed to help you manage your data, code, models, and experiment results on one platform and collaborate with your team.

5.0
Register Now!
Preview Image
Launch Date
2020
Monthly Visitors
736K
Country of Origin
united States
Platform
Web
Language
english

Keywords

  • DagsHub
  • Data Science Platform
  • DVC
  • Git integration
  • experiment tracking
  • MLops
  • data versioning
  • multimodal annotations
  • model registry
  • Jupyter notebook comparison
  • collaboration platform
  • open source integration
  • labeling automation
  • labeling automation
  • experiment reproducibility
  • enterprise MLOps

Platform Description

dagsHub is a platform for efficiently managing and collaborating on the code, data, models, and experimental records needed to develop AI models in one place. Built on top of Git and Data Version Control (DVC), you can systematically version control your data science projects the same way software developers do. you can consistently reproduce experimental results, even when dealing with complex data pipelines, and transparently compare and share changes and model performance differences between team members. with Jupyter notebooks, experiment comparison dashboards, multimodal data annotation capabilities, and integrations with open source tools like MLflow, Jenkins, and Label Studio, DagsHub is a great way to automate your entire machine learning operations (MLOps) workflow. whether you want to annotate experimental results, have collaborators label datasets, or visually compare model performance with teammates, the web interface is naturally embedded with the features you need in real-world AI development. best of all, open-source projects like GitHub are free for anyone to use, and individual users can run up to 100 experiments in a private repository. enterprise plans also offer security and scalability tailored to real-world conditions, including the number of collaborators, storage capacity, SSO login, and on-premises installation. this makes it a great platform for organizations of all sizes, from research teams to startups to enterprises, to build a production-ready AI collaboration environment.

Core Features

  • code-data-model integrated versioning

    Consistent tracking of your entire ML estate with Git + DVC

  • experiment tracking and comparison dashboard

    MLflow-based experiment recording and numerical-parameter comparison

  • compare notebook versions and changes

    Support for Jupyter notebook diff and history management

  • multimodal annotations and automatic labeling

    built-in annotation of image, audio, and text data

  • managing the Model Registry

    store trained models and manage deployment history

  • CI/CD integration

    Rerun experiments and automate pipelines with Git/Jenkins/MLflow integration

  • manage public/private repositories

    unlimited public projects, 100 private experiments for free

  • on-premises installation and SSO support for organizations

    provide security and authentication for large enterprises

Use Cases

  • data versioning
  • code versioning
  • tracking experiments
  • Compare Jupyter notebooks
  • multimodal data annotations
  • model Registry
  • DVC integration
  • MLflow Integration
  • automate your data pipeline
  • on-premises installations
  • team collaboration
  • hosting public projects
  • experiment reproducibility
  • Git Support
  • automate labeling
  • compare dashboards

How to Use

1

create a project

2

initialize local Git+DVC

3

view and compare experiment traces

4

dataset annotations - code review

Plans

Plan Price Key Features
Individual $0 • Unlimited public storage (with unlimited collaborators)
• Unlimited private storage for non-commercial use only
• Unlimited experiment tracking for public repositories
• Track up to 100 experiments in a private repository
• Up to 2 collaborators on private projects
• 20 GB of DagsHub storage
• Data versioning and tracking
• Annotation workspace for public repositories
• Notebook versioning and comparison
• CI/CD/CT integration
• Interactive pipeline
• Community support
Team $119/mo • Includes all features of the Individual plan
• Unlimited private storage
• Multimodal annotations and automatic labeling
• Ability to connect user storage
• Label Studio compatibility
• Team role-based access control (RBAC)
• Prioritized technical support
• Store up to 1 TB of data or 2 million files
Enterprise Contact us • Includes all features of the Team plan
• Manage petabytes of data
• Deploy models directly to clusters
• VPC or air-gapped on-premises installations
• SSO/LDAP/OIDC-based RBAC
• OpenShift compatible
• Control organizational resources
• Enterprise-grade support with SLAs

FAQs

  • DagsHub is a collaboration-centric web platform for data scientists and machine learning engineers that helps them unify, version control, and visually collaborate on code, data, models, and experimental results.
  • Git struggles with versioning large files and doesn't keep track of data pipeline changes. DagsHub is built on top of Git and DVC to efficiently version control even large data and models, detect pipeline changes, and automatically update only the necessary work.
  • yes. Just like GitHub, you can manage repositories, make pull requests (PRs), track issues, and synchronize two-way with your GitHub repository.
  • You can use Git as it is, and DVC uses a command structure similar to Git. existing Git users will find it easy to adapt, and DagsHub provides visualization tools to help those unfamiliar with the CLI.
  • no, DagsHub is completely language and library independent. Whether you're using Python, R, PyTorch, TensorFlow, Keras, or any other tool, it's compatible.
  • yes, you can. By connecting your GitHub repository with DagsHub, you can automatically synchronize with push events from GitHub and view PRs and issues in both directions, allowing you to use both platforms in parallel.
  • The DagsHub tutorial makes it easy to get started, and after installing Git and DVC, you just need to connect your project.
  • the open source project is completely free. personal repositories are also free for up to two collaborators, and if you need more features or people, there are paid plans available. see our pricing page for more information.
  • yes, DagsHub also offers secure options for enterprise use. you can use your own storage, integrate external storage, or even install it on-premises on a fully in-house infrastructure.
Select a rating for DagsHub.

Recommended Platforms

Weights&Biases

Weights&Biases (W&B) is an experiment tracking and MLOps platform that unifies all aspects of AI…

View More

ROK Solution

rOK Solution is a web-based platform that automatically designs and deploys business processes in minutes…

View More

MLflow

MLflow is an open source-based MLOps platform that unifies experiment tracking, model deployment, and project…

View More