Search awesome repositories

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.

PyTorch native quantization for training and inference

LLM speculative inference server for heterogeneous hardware & consumer GPUs

cuGraph - RAPIDS Graph Analytics Library

Large-scale LLM inference engine

Large-scale LLM inference engine

ThunderSVM: A Fast SVM Library on GPUs and CPUs

Self-host the powerful Chatterbox TTS model. This server offers a user-friendly Web UI, flexible API endpoints (incl. OpenAI compatible), predefined voices, voice cloning, and large audiobook-scale text processing. Runs accelerated on NVIDIA (CUDA), AMD (ROCm), and CPU.

The FFmpeg build script provides an easy way to build a static FFmpeg on OSX and Linux with non-free codecs included.

ThunderGBM: Fast GBDTs and Random Forests on GPUs

Numerical linear algebra software package

A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.

CEED Library: Code for Efficient Extensible Discretizations

eBPF agent and MCP server for GPU causal observability

A dynamic binary instrumentation tool for tracing and analyzing GPU kernel instructions.

CUDA kernels in any language supported by LLVM

Prototype of an intelligent safety system for detecting driver drowsiness

Optimized vLLM deployment for NVIDIA Blackwell (RTX 5090) on Linux Kernel 6.14. Resolves SM_120 kernel incompatibilities, P2P deadlocks, and memory fragmentation for high-performance LLM inference.

GPU texture reservoir computing — fragment-shader cellular automata with linear readouts for image classification and PDE solving. Honest benchmarks, CPU/OpenGL/CUDA backends.