dockerized-deep-learning

🐛️🧠️🐋️ Debugging Deep Learning Docker Containers

This repo goes hand in hand with this Medium post I wrote.

Getting started

Clone this project: git clone git@github.com:StefanieStoppel/dockerized-deep-learning.git.

Prerequisites

You need to have the following things installed on your system:

• Python (≥3.6, I used version 3.8.6)
• Conda or miniconda
• Docker
• Docker Compose
• VSCode
• (optional) If your machine comes with a GPU and you want to use it for training the network, you need to install nvidia-docker as well.

Create and activate the conda environment

Open a terminal and cd into the directory you cloned the project into. Execute the following:

conda env create --file=environment.yml
conda activate docker-deep-learning

Build the Docker image

In the root directory execute the following:

docker-compose build ddl

Debug the application

• Open the project in VSCode.
• Set a breakpoint in a Python file.
• Open the terminal in the project root and run docker-compose up ddl to run the Docker container.
• Click on the ▶️play icon in the left hand sidebar of VSCode to launch the debugger.
• Wait until it stops at you breakpoint.

🐛️ Happy debugging! 🐛️