标签：ONNX

PyTorch 生态里最省事的检测/分割基线，用来快速验证数据和导出部署模型

模型文件的可视化查看器，替代自己写脚本解析 ONNX/TFLite 图结构

不依赖 BLAS 的纯 C++ 推理运行时，用来把训练好的模型塞进 Android/iOS/嵌入式设备里跑

把训练好的模型从框架里剥离出来，用统一运行时跑在 CPU/GPU/NPU 上，替代各框架自带推理引擎

跨框架模型交换的中间表示层，导出链路的公共契约，不是推理引擎

按内容而非扩展名判断文件真实类型，替代 libmagic 在流水线里做第一道分拣

用 Rust 写训练代码，同一份代码直接跑推理，省掉导出模型这一步

离线语音全栈推理库，用 onnxruntime 顶替 Kaldi 部署链路，一套 C++ 核心供 12 种语言调用

用小型专用模型在本地跑完整 RAG 流水线，顶替 LangChain 加一堆解析器拼出来的自研栈

归档的 ONNX 端侧多语言 TTS，用来替掉云端 TTS API 或塞不进移动端的重型模型

JVM 上统一各家 LLM 与向量库 API 的抽象层，用来替代手写各家 SDK 的胶水代码

无锚框 YOLO 检测器，用一套训练代码同时覆盖科研复现和 TensorRT/ncnn 边缘部署

在 PyTorch 里训练和导出 YOLOv3 系列模型的参考实现，顶替 Darknet 原始仓库那条工具链

在音频流水线最前面切掉静音段的轻量 VAD，替掉 WebRTC VAD 那套老实现

把 Whisper + NLLB 塞进安卓手机本地跑，替代联网翻译 API 的离线对话翻译应用

ONNX 预训练模型与参考输入输出的对照样本库，用来验证推理后端而不是训练

在 Go 服务里直接调 OpenCV，省掉 CGO 桥接层和 Python 推理进程

在浏览器或 Node 进程内本地跑抠图模型，替代把图片上传到第三方 API 的流程

用模块化 Dockerfile 生成器把框架、CUDA、Python 版本组合成可直接运行的深度学习镜像

在 Caffe/Keras/MXNet 等老框架之间搬运模型时，用中间表示做一次性转换和结构可视化

在 Spark 作业里直接调用 LightGBM、视觉和文本模型，省去把数据搬出集群再搬回来的那一步

一个用 PyQt6 悬浮层和 ONNX 语音把「看屏幕主动搭话」做成可跑原型的桌面伴侣，顶替的是你自己拼感知-决策-表达三段胶水代码

An OBS plugin for removing background in portrait images (video), making it easy to replace the background when recording or streaming.

RuVector provides High Performance, Real-Time decisions and agent memory , Self-Learning Ai, Vector GNN DB built in Rust.

A 100M-parameter multilingual TTS model for real-time CPU inference, voice cloning, and 48 kHz stereo generation

A repository for storing models that have been inter-converted between various frameworks. Supported frameworks are TensorFlow, PyTorch, ONNX, OpenVINO, TFJS, TFTRT, TensorFlowLite (Float32/16/INT8), EdgeTPU, CoreML.

《Pytorch实用教程》（第二版）无论是零基础入门，还是CV、NLP、LLM项目应用，或是进阶工程化部署落地，在这里都有。相信在本书的帮助下，读者将能够轻松掌握 PyTorch 的使用，成为一名优秀的深度学习工程师。

A lite C++ AI toolkit: 100+ models with MNN, ORT and TRT, including Det, Seg, Stable-Diffusion, Face-Fusion.

Simplify your onnx model

State of the Art Natural Language Processing

A retargetable MLIR-based machine learning compiler and runtime toolkit.

Effortless AI-assisted data labeling with AI support from YOLO, Segment Anything (SAM+SAM2/2.1+SAM3), MobileSAM!!

🚀 Accelerate inference and training of 🤗 Transformers, Diffusers, TIMM and Sentence Transformers with easy to use hardware optimization tools

ONNX-TensorRT: TensorRT backend for ONNX

QualityScaler - image/video AI upscaler app

Sparsity-aware deep learning inference runtime for CPUs

Tiny, no-nonsense, self-contained, Tensorflow and ONNX inference

AudioMuse-AI uses sonic analysis to rediscover forgotten songs, uncover hidden connections in your music library, and generate intelligent playlists for Navidrome, Jellyfin, LMS, Lyrion, Emby and Plex: no metadata or external services required.

Local AI API Platform

Fast ML inference & training for ONNX models in Rust

Convert TensorFlow, Keras, Tensorflow.js and Tflite models to ONNX

Turn any computer or edge device into a command center for your computer vision projects.

Machine learning on FPGAs using HLS

Libraries for applying sparsification recipes to neural networks with a few lines of code, enabling faster and smaller models

Robust realtime face and facial landmark tracking on CPU with Unity integration

Lightweight inference library for ONNX files, written in C++. It can run Stable Diffusion XL 1.0 on a RPI Zero 2 (or in 298MB of RAM) but also Mistral 7B on desktops and servers. ARM, x86, WASM, RISC-V supported. Accelerated by XNNPACK. Python, C# and JS(WASM) bindings available.

Use AnimeGANv3 to make your own animation works, including turning photos or videos into anime.

YOLOv8 using TensorRT accelerate !

1M param OCR model - runs on CPU, 15x faster than tesseract, close to 1B+ param models in accuracy

A tool to modify ONNX models in a visualization fashion, based on Netron and Flask.