The best ChatGPT that $100 can buy.

Multi-Threaded FP32 Matrix Multiplication on x86 CPUs

Tensor library for machine learning

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

Nano vLLM

LLM inference in C/C++

📚 从零开始的大语言模型原理与实践教程

Silero VAD: pre-trained enterprise-grade Voice Activity Detector

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

FULL v0, Cursor, Manus, Same.dev, Lovable, Devin, Replit Agent, Windsurf Agent, VSCode Agent, Dia Browser & Trae AI (And other Open Sourced) System Prompts, Tools & AI Models.

Inference Llama 2 in one file of pure C

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

DeepEP: an efficient expert-parallel communication library

Fully open reproduction of DeepSeek-R1

s1: Simple test-time scaling

LLM101n: Let's build a Storyteller

Reverse Engineering: Decompiling Binary Code with Large Language Models

Cross-platform, graphics API agnostic, "Bring Your Own Engine/Framework" style rendering library.

ncnn is a high-performance neural network inference framework optimized for the mobile platform

The Autograd Engine

LLM training in simple, raw C/CUDA

UNet diffusion model in pure CUDA

OpenDiT: An Easy, Fast and Memory-Efficient System for DiT Training and Inference

VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs

llama3.cuda is a pure C/CUDA implementation for Llama 3 model.