Closed Coopyrightdmin closed 1 year ago
We've extended the configuration to allow users to specify both Linux libraries associated with specific Python packages and custom packages independent of any Python dependencies. Here's a brief overview of the YAML structure:
dependencies:
- name: numpy
packages:
- gcc
- g++
- musl-dev
- python3-dev
- gfortran
- name: scikit_learn
packages:
- gcc
- g++
- musl-dev
- python3-dev
- gfortran
- openblas-dev
- lapack-dev
custom_packages:
- libjpeg
- imagemagick
- jq
This update provides users with a more flexible and customizable container setup experience, allowing them to specify both Python-related and independent Linux packages.
Problem Statement
Currently, users do not have the capability to configure which Linux libraries should be pre-installed. This limitation may hinder the container's adaptability for specific requirements.
Benefits
Allowing users to select which Linux libraries to pre-install would:
Additional context
Many users have specific requirements that demand certain Linux libraries. This feature would enable them to produce a ready-to-use container without the need to manually install these libraries post-container creation.
Priority/Severity