Python package built to ease deep learning on graph, on top of existing DL frameworks.

Robyn is a Super Fast Async Python Web Framework with a Rust runtime.

A PyTorch implementation of Dynamic Graph CNN for Learning on Point Clouds (DGCNN)

Kernel Point Convolution implemented in PyTorch

how to optimize some algorithm in cuda.

🦀 Small exercises to get you used to reading and writing Rust code!

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

"LightRAG: Simple and Fast Retrieval-Augmented Generation"

A self-learning tutorail for CUDA High Performance Programing.

Gaussian Haircut: Human Hair Reconstruction with Strand-Aligned 3D Gaussians

DyNet: The Dynamic Neural Network Toolkit

Self-hosted AI coding assistant

Efficient implementations of state-of-the-art linear attention models in Pytorch and Triton

Burn is a new comprehensive dynamic Deep Learning Framework built using Rust with extreme flexibility, compute efficiency and portability as its primary goals.

SGLang is a fast serving framework for large language models and vision language models.

A Collection of BM25 Algorithms in Python

🎮 A step-by-step guide to implementing SSAO, depth of field, lighting, normal mapping, and more for your 3D game.

A simple, easy-to-hack GraphRAG implementation

A Bulletproof Way to Generate Structured JSON from Language Models

DSPy: The framework for programming—not prompting—language models

nanoGPT style version of Llama 3.1

TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficie…

S-LoRA: Serving Thousands of Concurrent LoRA Adapters

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

Example models using DeepSpeed

An extremely fast Python package and project manager, written in Rust.

LLM101n: Let's build a Storyteller

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

Finetune Llama 3.3, Mistral, Phi, Qwen 2.5 & Gemma LLMs 2-5x faster with 80% less memory