vscode-pytorch-ffmpeg

An OpenShift AI Image running VSCode for PyTorch development.

• Based on the VS code server image by Red Hat Data Services.
• Uses cuda-toolkit-12-8.
• Uses flash-attention.

Base image: https://quay.io/repository/modh/codeserver?tab=tags&tag=codeserver-ubi9-python-3.11-20250212

Python packages	Description
flash-attention	Fast and Memory-Efficient Exact Attention with IO-Awareness

System packages	Description
cuda-toolkit-12-8	Meta-package containing all runtime library packages and the CUDA driver.

Build the container with podman

podman build -t nerc-images/vscode-pytorch-ffmpeg:latest .

Run the container with podman

podman run --rm -it --entrypoint /bin/bash nerc-images/vscode-pytorch-ffmpeg:latest

You can pull the latest vscode-pytorch-ffmpeg container image below:

podman pull quay.io/nerc-images/vscode-pytorch-ffmpeg:latest