scvi-tools (single-cell variational inference tools) is a package for probabilistic modeling and analysis of single-cell omics data, built on top of PyTorch and AnnData.

Docker images

IMPORTANT: This repository is deprecated. Please see the updated images in the GitHub Container Registry

This repository contains the source files for our Docker images.

Tags

py{PYTHON_VERSION}-cu{CUDA_VERSION}-base: CUDA-based PyTorch and JAX installations

• Updated weekly Sunday at 05:00 PST
• We currently build images for PYTHON_VERSION=[3.9, 3.10, 3.11] and CUDA_VERSION=[11, 12]

py{PYTHON_VERSION}-cu{CUDA_VERSION}-runtime-{latest/stable/semver}: base image with runtime dependencies

• latest: Updated daily at 12:00 PST
• stable: Alias for the most recent semver
• semver: Updated with each new stable release on the main repository

py{PYTHON_VERSION}-cu{CUDA_VERSION}-dev-{latest/stable/semver}: base image with development dependencies

• latest: Updated daily at 12:00 PST
• stable: Alias for the most recent semver
• semver: Updated with each new stable release on the main repository

py{PYTHON_VERSION}-cu{CUDA_VERSION}-tutorials-{latest/stable/semver}: base image with tutorial dependencies

• latest: Updated daily at 12:00 PST
• stable: Alias for the most recent semver
• semver: Updated with each new stable release on the main repository