from vibe coding to agentic engineering - practice makes claude perfect

Puzzles for learning Triton

🚀🚀🚀 This repository lists some awesome public CUDA, cuda-python, cuBLAS, cuDNN, CUTLASS, TensorRT, TensorRT-LLM, Triton, TVM, MLIR, PTX and High Performance Computing (HPC) projects.

LLM training parallelisms (DP, FSDP, TP, PP) in pure C

a minimal cache manager for PagedAttention, on top of llama3.

Nano vLLM

FlashMLA: Efficient Multi-head Latent Attention Kernels

Fully Open Language Models with Stellar Performance

🔥 A minimal training framework for scaling FLA models

Python API for writing multiprocessing pipelines

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

Machine Learning Engineering Open Book

Fully open reproduction of DeepSeek-R1

100 days of building GPU kernels!

This is the homepage of a new book entitled "Mathematical Foundations of Reinforcement Learning."

This repository is a curated collection of resources, tutorials, and practical examples designed to guide you through the journey of mastering CUDA programming. Whether you're just starting or look…

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

Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch

Official JAX implementation of Learning to (Learn at Test Time): RNNs with Expressive Hidden States

GPU Kernels

Inference Speed Benchmark for Learning to (Learn at Test Time): RNNs with Expressive Hidden States

Official PyTorch implementation of Learning to (Learn at Test Time): RNNs with Expressive Hidden States

A Self-adaptation Framework🐙 that adapts LLMs for unseen tasks in real-time!

What would you do with 1000 H100s...

Minimalistic 4D-parallelism distributed training framework for education purpose

A generic, composable multi-dimensional array library.

HunyuanVideo: A Systematic Framework For Large Video Generation Model

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.