liveBook · Manning

Versioning

As explained in Build AI Drug Discovery Pipelines

Version management is the systematic tracking and control of changing versions of software, models, data, and configuration artifacts.

Versioning tracks the models and supporting artifacts behind an analysis so results can be reproduced as drug-discovery systems evolve. section 13.4.3

Track the components that can change an analysis:

  • Component model versions, training datasets, preprocessing pipelines, and inference parameters. section 13.4.3

Model updates

The same analysis using ESM-2 version 1.0 or 2.0 may produce different results. section 13.4.3

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Identifies models, training datasets, preprocessing, and inference parameters to track.
Examples
Contrasts analyses run with ESM-2 versions 1.0 and 2.0.
What you can do after
Apply version tracking to reproduce analyses across changing models and supporting artifacts.
How it connects
Connects model versioning with reproducibility, training datasets, preprocessing pipelines, and inference parameters.

Also explained in

Compare how 36 books explain versioning →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.