Deep Learning Base Development Environment

This repository contains a Dockerfile to set up a base development environment tailored for deep learning projects. It includes essential tools such as Python, Poetry, and Miniconda, with CUDA support for GPU-accelerated computation. This setup is optimized for flexibility and efficiency in development.

Dockerfile Explanation

1. Base Image: The Docker image starts from nvidia/cuda:12.1.0-devel-ubuntu20.04, which includes the necessary CUDA toolkit for GPU-related tasks. This image does not come with Python pre-installed, so we manually install Python 3.8 to ensure compatibility with the development environment.

2. Python Installation: We update the package list and install essential utilities like wget, curl, git, and python3, followed by python3-pip for managing Python packages.

3. Poetry Installation: Poetry is used for Python dependency management and packaging. We install Poetry version 1.8.0 by fetching its installation script and adding its executable to the PATH.

4. Miniconda Installation: To manage environments more efficiently, we install Miniconda, a lightweight version of Anaconda. It allows the creation of isolated environments using conda and is ideal for deep learning workflows. Miniconda is installed by downloading and running the installer script.

5. TINI Installation: To handle proper signal handling for the container, we install TINI, a tiny init system, which ensures that the container shuts down gracefully.

6. Conda Environment Integration with Poetry: We configure Poetry to use a conda environment for virtual environments rather than Poetry’s default virtual environment system. This is achieved by setting the CONDA_ENV_PATH to /opt/conda/envs/ and disabling Poetry’s internal virtualenv creation.

Key Features:

• CUDA Toolkit: Enables GPU-accelerated computing for deep learning.
• Python 3.8: Installed for compatibility with deep learning libraries.
• Poetry: For efficient Python package management.
• Miniconda: A lightweight package manager for managing isolated environments.
• TINI: Ensures graceful shutdown of the container.
• Conda Environment Integration: Uses conda environments for virtual environments, managed by Poetry.

Setup Instructions (If Not Using a Development Container)::

1. Clone the repository:

   git clone https://github.com/yourusername/deep-learning-base-dev-env.git
   cd deep-learning-base-dev-env

2. Build the Docker image:

   docker build -t deep-learning-dev-env .

3. Run the Docker container:

   docker run -it deep-learning-dev-env bash

4. Verify Poetry Installation:

   poetry --version

5. Configure Conda environment (if needed): Poetry is already configured to use conda for environment management.