#cuda Repositories
325 repositories total
NVIDIA's CUDA platform for Rust. Host runtime crates plus Tile (cutile-rs) and SIMT (cuda-oxide) kernel programming models in idiomatic Rust.
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
Self-hosted, local only NVR and AI Computer Vision software. With features such as object detection, motion detection, face recognition and more, it gives you the power to keep an eye on your home, office or any other place you want to monitor.
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.
AI Fundamentals - GPU architecture, CUDA programming, basics of large models, and knowledge related to AI Agents.
LYGIA, it's a granular and multi-language (GLSL, HLSL, Metal, WGSL, WEGL and CUDA) shader library designed for performance and flexibility
Open source neural network chess engine with GPU acceleration and broad hardware support.
Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.
Run GLM-5.3 (744B), DeepSeek-V4.1-Flash (748B), MiniMax-M3.1-Flash (428B), GLM-5.3-Flash (321B), and Hy4-Flash (220B) on your own GPUs โ starting from 8 GB of VRAM, with NVIDIA and AMD support, one dashboard, and an OpenAI- and Anthropic-compatible API.
RamaLama is an open-source developer tool that simplifies the local serving of AI models from any source and facilitates their use for inference in production, all through the familiar language of containers.
Minkowski Engine is an auto-diff neural network library for high-dimensional sparse tensors
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.
CV-CUDAโข is an open-source, GPU accelerated library for cloud-scale image processing and computer vision.
Computer Vision Projects | Fun AI Projects related to Computer Vision (Python, C++, embedded system)
Cross-vendor 3D Gaussian Splatting trainer - video to splat to mesh, Vulkan or CUDA.
LUPINE is a GPU over IP bridge allowing GPUs on remote machines to be attached to CPU-only machines.
Local AI song generator with an editable score โ YuE2 on your GPU: full songs with vocals, sheet music, covers, exact replay. Native Windows app, no Python, installer with auto-update.
Apache Mahout - an environment for quickly creating scalable, performant machine learning applications.
Official SGLang x Datawhale course on LLM inference: understand inference, build a mini-sglang from scratch, then read the real SGLang source and land your first PR. Available in English and Chinese.
Many studies have shown that the performance on deep learning is significantly affected by volume of training data. The MedicalNet project provides a series of 3D-ResNet pre-trained models and relative code.
label-smooth, amsoftmax, partial-fc, focal-loss, triplet-loss, lovasz-softmax. Maybe useful
Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".
NVIDIA curated collection of educational resources related to general purpose GPU programming.
Translate GPU shaders and compute kernels between GLSL, HLSL, Metal, SPIR-V, Slang, CUDA, and HIP, and to WGSL, through the CrossGL IR
Run AI models locally on your machine with node.js bindings for llama.cpp. Enforce a JSON schema on the model output on the generation level.
A deep learning package for many-body potential energy representation and molecular dynamics