标签：PyTorch

USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference

News: the 10k dataset is ready for download.

Bilingual (中文+EN) ML / LLM / diffusion / agent interview cheat sheets for AI 秋招 — generated by ARIS /interview-cheatsheet, rendered by /render-html into single-file HTML, reads anywhere — plus a CV→DBLP-fact-checked academic homepage generator and hand-authored long-form blogs 🌱

🎉 PILOT: A Pre-trained Model-Based Continual Learning Toolbox

Image augmentation for computer vision. AGPL-3.0-only or commercial licensing.

Crater is a cloud-native AI training & inference platform.

From teacher to tiles — a from-scratch LLM distillation & serving engine: custom Triton/CUDA kernels, FSDP distillation, paged-KV continuous batching, speculative decoding, a Rust gateway, a JAX oracle, and interpretability tooling.

A low-latency & high-throughput serving engine for LLMs

[CVPR 2023] Official repository for downloading, processing, visualizing, and training models on the ARCTIC dataset.

an edge-real-time anchor-free object detector with decent performance

本地优先的全栈 AI 创作工作站：AI对话 · AI漫画 · 漫剧创作 · 写作台 · 知识学习五大模块；vLLM / llama.cpp / Transformers 多引擎本地推理，显存智能调度、舱壁隔离自愈、仅供学习使用，欢迎交流指正。（最新进度（除插件系统）受到协议限制，暂时无法提交到仓库）

🚀🚀🚀 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.

Official repo for DAD-3DHeads: A Large-scale Dense, Accurate and Diverse Dataset for 3D Head Alignment from a Single Image (CVPR 2022).

[CVPR 2024✨Highlight] Official repository for HOLD, the first method that jointly reconstructs articulated hands and objects from monocular videos without assuming a pre-scanned object template and 3D hand-object training data.

Accelerate your Neural Architecture Search (NAS) through fast, reproducible and modular research.

Official implementation of AsymFlow, pi-Flow, GMFlow

The Qualcomm® AI Hub apps are a collection of state-of-the-art machine learning models optimized for performance (latency, memory etc.) and ready to deploy on Qualcomm® devices.

JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).

A model compilation solution for various hardware

Model Compression Toolkit (MCT) is an open source project for neural network model optimization under efficient, constrained hardware. This project provides researchers, developers, and engineers advanced quantization and compression tools for deploying state-of-the-art neural networks.

Seeed SenseCraft Model Assistant is an open-source project focused on embedded AI. 🔥🔥🔥

Self-hosted DeepSeek-OCR-2 WebUI & API — vLLM / transformers on NVIDIA, Apple Silicon & CPU, per-page quality checks, PDF streaming, MCP

A QGIS plugin tool using Segment Anything Model (SAM) to accelerate segmenting or delineating landforms in geospatial raster images.

RealScaler - image/video AI upscaler app (Real-ESRGAN)

[CVPR 2025 Highlight] Official code for "Olympus: A Universal Task Router for Computer Vision Tasks"

[IEEE Transactions on Medical Imaging/TMI 2023] This repo is the official implementation of "LViT: Language meets Vision Transformer in Medical Image Segmentation"

iSeeBetter: Spatio-Temporal Video Super Resolution using Recurrent-Generative Back-Projection Networks | Python3 | PyTorch | GANs | CNNs | ResNets | RNNs | Published in Springer Journal of Computational Visual Media, September 2020, Tsinghua University Press

[ICCV 2023] Official implementation of "SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance Fields"

A tiny yet powerful LLM inference system tailored for researching purpose. vLLM-equivalent performance with only 2k lines of code (2% of vLLM).

Generates 3D from 2D videos in multiple output formats using AI-powered depth mapping.

Phonon: open speech recognition models (Phonon-2, Phonon-1) — CLI, CPU and CUDA images

A Simple, Lightweight, and Extensible Serving Framework for X-AnyLabeling

ML model optimization product to accelerate inference.

🔥 [ICCV 2025 Highlight] Official ComfyUI native node supporting InfiniteYou with FLUX

Code release for "i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models"

Build a complete experiment pipeline for your PyTorch MIP model in 10 seconds.

Official implementation for Tetra-NeRF paper - NeRF represented as triangulation of input point cloud.

Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna TPE.

A hands-on course on modern LLM architectures, training, and inference, with step-by-step PyTorch implementations and runnable notebooks.

GPU Profiler for AI agents

AI Auto Annotation & YOLO Training Pipeline, End-to-end object detection auto-labeling and YOLO training platform. VLM-powered annotation with NVIDIA LocateAnything-3B, manual refinement, one-click YOLO training, video keyframe extraction, and model validation. Supports image and video.

Live-bending a foundation model’s output at neural network level.

FasterAI: Prune and Distill your models with FastAI and PyTorch

A PyTorch native library for training speculative decoding models

🚀🚀 A multimodal System One decision model that gives calibrated answers to typed questions about screens, photos, video and text in one forward pass.

PyTorch template for Deep Learning projects with support for scalable multi-GPU and multi-node training.

🚀 Multi-Stream, Multi-Model, Multi-GPU Video Processing in Python with CUDA

A multi-functional library for full-stack Deep Learning. Simplifies Model Building, API development, and Model Deployment.

Real-time object detection, keypoint estimation, instance and semantic segmentation. Train, export (TensorRT, ONNX, OpenVINO, CoreML), deploy. Apache-2.0

Iris-3B: pixel-space generation & general vision learner. A 3B diffusion transformer that generates every pixel directly (no VAE); the same prior, fine-tuned, does monocular depth estimation and 4x image restoration. Paper, weights, training code and demo.