DagsHub Client for connecting ML projects to DagsHub

How it works

The client provides interfaces for repository-related operations and integrations for tracking or working with data. The service supplies the hosted side of those connections.

What you need

Use a scoped credential for a test repository and choose which client integration to try. Review whether data files and metrics travel through different endpoints.

What to try first

Try a small project with synthetic data and a scoped credential. Verify which artifacts and metadata appear in the hosted repository after one run.

Test read and write access with separate credentials. This reveals which client functions need repository access and which need access to an artifact store.

Where to look for failures

Hosted storage and tracking involve separate access and data handling choices. Review what the client uploads automatically before pointing it at a real dataset.

Related Franklin pages

Choose what needs versioning and what needs tracking. A run's curves are easier to reproduce when they link to the exact data, configuration and saved artifact.

• Track large artifacts and dependency-aware stages alongside a Git project. DVC for versioning datasets and ML pipelines
• Record parameters, metrics and artifacts alongside model runs. MLflow for tracking experiments and model artifacts

For the steps that prepare raw data for training or evaluation, see Franklin’s glossary. Data preprocessing

Project documentation

• Current project README
• README at the reviewed revision