Dock Your Jupyter Notebook
summary
Builds a Dockerfile from a tensorflow-notebook:latest base, installs Python packages, clones the repo, and pulls data with DVC so experiments run identically on any OS. Creates and tags an image pneumonia-classification:vgg19, launches a Jupyter server, edits code, and uses docker exec to git and dvc push. Claim: reproducible ML workflow that moves from laptop to GPU cluster. Demonstration shows the container runs and the experiment history is versioned and comparable.
AI-written summary · about
chapters (12)
- 0:00DagsHub
- 3:00NIR BARAZIDA
- 4:04Docker for Machine Learning
- 5:56Docker Components
- 8:28Container Registry
- 9:25New Docker Project
- 10:29Docker File
- 15:10Useful Dockerfile Commands
- 15:39Build Docker Image
- 16:51Run Docker Container
- 18:31SSH into a Container
- 20:13Recap
related talks
36mHow to Roll your own Dockerized Jupyter Environment
01h 02mDocker for Data Science: Reproducibility and Deployment
01h 11mLearning to use Docker for development
26mTransforming a Jupyter Notebook into a reproducible pipeline for ML experiments
24mDocker for Machine Learning!
29mLaunch Jupyter to the Cloud: an example of using Docker and Terraform
42mMaking Docker & Python Play Nicely and Securely for Machine Learning & Data Science
01h 32mHow to easily set up and version control your Machine Learning Pipelines