Vectice + Git: Documenting ML Beyond the Code
Git is the best tool for keeping track of code changes…
But AI projects are more than just code — they rely heavily on datasets, models, and computational graphs. While Git excels at managing code, it falls short when it comes to tracking these other critical components in the data science lifecycle.
Vectice Provides End-to-End ML Auto-Documentation, including Models and Datasets
That’s where Vectice comes in. Vectice works seamlessly with Git to provide end-to-end versioning and documentation for AI projects. Git handles source code tracking and versioning, while Vectice auto-documents the entire data preparation and model development pipeline, up to production deployment. Vectice also catalogs and versions datasets and models, keeping track of the complete lineage.
Vectice Integrates with Git Effortlessly
Vectice integrates effortlessly with any flavor of Git. Even better, it works automatically in the background. When data scientists use the Vectice library in their code, Vectice detects the linked Git repository and connects the created assets to the appropriate Git commits. This ensures that the code lineage is maintained accurately and assets can be traced back to the exact code version that generated them.
No Extra Work for Data Scientists
Data scientists don’t need to do any extra work. They simply use Vectice in their code to auto-document their models and datasets. When they check the Vectice web app, they’ll see the relevant Git commit linked to each asset — no manual tracking required.
Complete Version Control for AI Projects
In summary, Vectice is the perfect complement to Git for AI projects. While Git manages code changes, Vectice expands version control and documentation across datasets, models, and graphs. This enables AI projects to be fully tracked and auto-documented from end to end. Now you can maintain oversight of not just code but all the other critical components of your AI projects.