[SIGGRAPH 2026] Pixal3D: Pixel-Aligned 3D Generation from Images

🙌 OpenHands: AI-Driven Development

Benchmark harness and code for "SWE-fficiency: Can Language Models Optimize Real World Repositories on Real World Workloads?"

KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

CUDA 算子手撕与面试指南

Samples for CUDA Developers which demonstrates features in CUDA Toolkit

BLISlab: A Sandbox for Optimizing GEMM

A high-throughput and memory-efficient inference and serving engine for LLMs

JittorInfer is a high-performance C++ inference framework designed for large language models on Huawei's Ascend AI processor.

LaTeX Thesis Template for Tsinghua University

LLM inference in C/C++

SGLang is a high-performance serving framework for large language models and multimodal models.

Transformer training code for sequential tasks

Longformer: The Long-Document Transformer

SoftVC VITS Singing Voice Conversion

Fast and memory-efficient exact attention

The simplest, fastest repository for training/finetuning medium-sized GPTs.

The official Meta Llama 3 GitHub site

Jittor is a high-performance deep learning framework based on JIT compiling and meta-operators.