Build GPU-enabled Docker containers with NGC base images and uv.
npx skills add https://github.com/jayll1303/AIEKit --skill docker-gpu-setup
Please help me install this Agent Skill. Skill: docker-gpu-setup Source: https://github.com/jayll1303/AIEKit/tree/main/.kiro/skills/docker-gpu-setup Command: npx skills add https://github.com/jayll1303/AIEKit --skill docker-gpu-setup
Building and running GPU-enabled Docker containers for ML workloads is error-prone and slow without standard patterns, resulting in broken GPU passthrough, huge unreproducible images, and long rebuild times. This Skill provides clear, production-ready Dockerfile patterns, Compose snippets, and troubleshooting steps to ensure containers see GPUs, dependencies are cached, and images remain reproducible.
Create a multi-stage Dockerfile using an appropriate NGC base, copy the uv binary from the official image, build with BuildKit enabled, run the container with GPU passthrough, and verify GPU access by running nvidia-smi inside the container.
High-intent search queries and answers about installing and using this skill.
To enable GPU passthrough in Docker Compose, configure the runtime and GPU device access in your Compose v2 file. This Skill provides ready-to-use snippets and verification steps to confirm the container sees the GPU correctly.
The best way to build slim GPU Docker images is using multi-stage Dockerfile templates with NGC base images. This isolates dependency installs and produces a slim runtime image, ensuring reproducible ML builds.
Yes, you need the NVIDIA Container Toolkit and a working host NVIDIA driver to run CUDA workloads in Docker. This Skill includes a troubleshooting checklist for verifying runtime config and driver compatibility.
To cache dependencies during a Docker build, use BuildKit cache mounts alongside uv-based dependency installs. This Skill demonstrates how to copy the uv binary and apply BuildKit patterns to speed up builds.
nvidia-smi fails inside a GPU Docker container due to missing NVIDIA Container Toolkit configuration or host driver mismatch. This Skill provides a troubleshooting checklist covering driver compatibility and runtime errors.
To choose the right NGC base image, match the nvcr.io image with your host driver and CUDA versions. This Skill provides NGC guidance for selecting appropriate training, inference, or runtime scenarios.
NVIDIA-TAORun Docker containers on GPU hosts with NVIDIA Container Toolkit.
BagelHoleConfigure NVIDIA GPU servers with drivers, CUDA, and MIG partitioning.
NovaAI-innovationManage Python dependencies and virtual environments with uv.
brittaniebuffiecsuDeploy and execute machine learning models on Modal's serverless GPU cloud.
