Skip to content
WorkshopIn personPast event

Training Workshop: An Introduction to Docker

Event artwork for Training Workshop: An Introduction to Docker

This event has ended

This listing is kept as an archive. Explore upcoming events to find your next one.

See upcoming events in Barcelona
About this event

What to expect

Reproducibility is one of the most persistent challenges in modern research software development, particularly when working with Python-based machine learning projects. This hands-on training workshop introduces Docker as a powerful solution to eliminate versioning headaches and environment inconsistencies that plague scientific computing.

Research teams working with Python often struggle with fundamental questions: which Python version should be used? How do you manage conflicting dependencies across PyTorch, Scikit-Learn, and NumPy? Should you standardize on Conda, venv, or Poetry for environment management? What about GPU dependencies like CUDA Toolkit versions? These seemingly simple questions create significant barriers to reproducing experimental results, even when using identical data and methodology.

This workshop, organized by the Cambridge Accelerate Programme, demonstrates how Docker containers can resolve nearly all of these reproducibility challenges by packaging your entire computational environment — Python version, dependencies, system libraries, and configurations — into portable, version-controlled units. You'll learn practical techniques for containerizing research workflows, ensuring that experiments can be reliably reproduced by collaborators and future researchers.

What You'll Learn

  • Core Docker concepts and how containers differ from traditional virtual environments
  • Building Docker images for Python-based research projects
  • Managing dependencies, package versions, and system requirements within containers
  • Controlling randomness and ensuring deterministic results
  • Integrating Docker with Git and GitHub workflows for a streamlined development experience
  • Best practices for minimizing your technical stack and reducing environment management overhead

Who Should Attend

This workshop is open to all postgraduate students and research staff at the University of Cambridge working with computational research, particularly those building machine learning models or data analysis pipelines in Python. Whether you're struggling with reproducibility issues in your current projects or want to establish robust practices from the start, this session provides practical solutions you can implement immediately.

The session runs from 09:30 to 13:00, giving you a full morning to work through hands-on exercises and ask questions about your specific use cases. By com

Running an AI event in Barcelona?

List it here, free.

Meetups, workshops and conferences go in front of thousands of builders after a quick review.

List your event