jetson-containers
Machine Learning Containers for NVIDIA Jetson and JetPack-L4T
What is jetson-containers?
Machine Learning Containers for NVIDIA Jetson and JetPack-L4T
Overview of jetson-containers
What it is
jetson-containers is a modular container build system that provides the latest AI/ML packages for NVIDIA Jetson and JetPack-L4T. It is hosted on GitHub at https://github.com/dusty-nv/jetson-containers and is categorized under AI. The license is not stated (NOASSERTION), and it is open source.
Key features
- Provides containers for ML frameworks: PyTorch, TensorFlow, JAX, ONNX Runtime, DeepStream, Holoscan, CTranslate2, JupyterLab - Includes LLM packages: SGLang, vLLM, MLC, AWQ, Transformers, text-generation-webui, text-generation-inference, Ollama, Optimum, OpenAI, GPTQModel, llama.cpp, llama-factory, ExLlama, ktransformers, FlashAttention, DeepSpeed, bitsandbytes, xformers, xgrammar - Supports VLM: LLaVA, llama-vision, VILA, LITA, NanoLLM, Prismatic, xtuner, gemma_vlm - Offers VIT tools: NanoOWL, NanoSAM, Segment Anything (SAM), Track Anything (TAM), clip_trt - Includes RAG components: llama-index, langchain, jetson-copilot, NanoDB, FAISS, RAFT - Covers robotics: ROS, LeRobot, OpenVLA, 3D Diffusion Policy, Crossformer, MimicGen, OpenDroneMap, ZED, openpi
Use cases
- Running machine learning inference and training on NVIDIA Jetson devices - Deploying large language models (LLMs) and vision-language models (VLMs) on edge hardware - Building robotics applications with simulation and real-world deployment - Developing computer vision pipelines using VIT and graphics tools - Creating retrieval-augmented generation (RAG) systems on Jetson
Getting started
See the repository README for installation and quick-start instructions.
License and pricing
The license is not stated (NOASSERTION). The project is open source.
Based on the project's public README and repository metadata.