ufoym/deepo resource

Health v2 · maintenance only

Adoption not part of the score

6276 stars · 737 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Deepo is a Dockerfile generator and collection of pre-built Docker images for setting up deep learning research environments. It composes modular 'Lego-like' components (CUDA, cuDNN, PyTorch, TensorFlow, etc.) into custom images with automatic dependency resolution.

Use cases

• set up a deep learning environment in seconds
• generate a custom dockerfile with pytorch and tensorflow
• run jupyter with gpu acceleration in docker
• create a reproducible cuda and cudnn container
• build an all-in-one machine learning docker image
• resolve compatible versions of cuda python and deep learning frameworks

When to choose

• you want a ready-made or custom Docker image with multiple deep learning frameworks preinstalled
• you need GPU (CUDA/cuDNN) support in containers with minimal configuration
• you want automatic dependency and version compatibility resolution across frameworks

When to avoid

• you only need a single framework and prefer official framework images or pip/conda installs
• you don't use Docker or can't install the NVIDIA Container Toolkit
• you need lightweight CPU-only environments where a full image is overkill

Facets

infra-config · maturity active

deep-learning llm-training machine-learning container-runtime infrastructure-as-code gpu-computing deep-learning machine-learning gpu-computing developer-tools windows python dockerfile-generator docker-images cuda cudnn jupyter tensorflow pytorch environment-setup containers docker linux macos gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ufoym/deepo")