Skip to main content

YOLO26 App

基于 Ultralytics YOLO 的桌面端标注、训练和推理应用,支持 YOLO26 与 YOLOv8。

Python PyQt6 License

English · 完整文档 · 环境与模型 · 常见问题

安装

Windows 一键安装

要求 Python 3.10-3.12。克隆项目后运行统一安装引导器:

git clone https://github.com/only-one-over/yolo26_app.git
cd yolo26_app
install.bat --profile auto

auto 仅在 nvidia-smi 可用时选择 CUDA 12.1 配置,否则选择 CPU 配置;不会根据 CUDA 字符串安装未经测试的 PyTorch 版本。安装器会创建 venv、使用锁定依赖、执行 pip check,并在 Windows 上创建“YOLO26 App”桌面快捷方式。

可用选项:

install.bat --profile cpu
install.bat --profile cuda --with-sam --with-dino --with-realsense
install.bat --dry-run

Conda 环境安装

如果已经安装 Anaconda 或 Miniconda,也可以使用 Conda 创建隔离环境;项目运行依赖仍通过锁定的 pip 文件安装,以保持 CPU/CUDA 版本组合一致。

CPU 环境:

conda create -n yolo26-app python=3.10 -y
conda activate yolo26-app
python -m pip install -r requirements-lock-cpu.txt
python -m pip install --no-deps .

NVIDIA CUDA 12.1 环境:

conda create -n yolo26-app python=3.10 -y
conda activate yolo26-app
python -m pip install -r requirements-lock-cuda-cu121.txt
python -m pip install --no-deps .

安装完成后运行 python main.py;Windows 也可以运行 python -m pip install --no-deps . 生成的 yolo26-app 图形启动命令。不要在同一个 Conda 环境中同时安装 onnxruntime 和 onnxruntime-gpu。

Windows 便携包

普通用户也可以从 GitHub Releases 下载 CPU 或 CUDA 便携包。完整解压后启动对应的 EXE;便携包默认不会自动创建快捷方式,可右键 EXE 选择“发送到 -> 桌面快捷方式”。不要单独移动 EXE 文件。

手动安装

git clone https://github.com/only-one-over/yolo26_app.git
cd yolo26_app
python -m venv venv

Windows:

venv\Scripts\python.exe -m pip install -r requirements-lock-cpu.txt
venv\Scripts\python.exe -m pip install --no-deps .

Linux 或 macOS:

source venv/bin/activate
python -m pip install -r requirements-lock-cpu.txt
python -m pip install --no-deps .

GPU、SAM2、Grounding DINO、RealSense 与 TensorRT 的安装方式见环境与模型。

构建与安装 wheel

发布或验证安装产物时,使用标准 wheel:

python -m pip install build
python -m build
python -m pip install --force-reinstall dist/yolo26_app-*.whl

构建命令会同时生成 wheel 与源码分发包。wheel 已包含界面 SVG 图标和 YAML 模板。

启动

python main.py

Windows 使用一键安装脚本后,也可以直接执行:

venv\Scripts\python.exe main.py

安装 wheel 后,Windows 可以直接运行图形启动命令:

yolo26-app

PyPI 安装

已发布的 Python 包可在新虚拟环境中安装:

python -m venv .venv

Windows:

.venv\Scripts\python.exe -m pip install "yolo26-app[onnx]"
.venv\Scripts\yolo26-app.exe

Linux 或 macOS:

.venv/bin/python -m pip install "yolo26-app[onnx]"
.venv/bin/yolo26-app

GPU ONNX 推理可将 [onnx] 改为 [onnx-gpu];SAM2、Grounding DINO、RealSense 和 TensorRT 也可按需追加对应 extra。PyPI 包不包含模型权重或 Windows 便携运行时。

基本使用

  1. 通过“文件 -> 新建项目”创建工作区。
  2. 在“标注”页导入图片、视频或素材目录,并添加类别。
  3. 使用矩形框、多边形、关键点或 OBB 工具完成标注。
  4. 点击“导出数据集”,生成 YOLO 格式的数据集。
  5. 在“训练”页选择 data.yaml,设置模型与训练参数后开始训练。
  6. 在“测试”页加载 best.pt,对图片、视频或相机执行推理,必要时导出 ONNX、TensorRT 等模型。

标注会自动保存;重新打开项目或异常退出后可恢复。更详细的标注、数据集和训练说明见标注、数据集与训练。

文档

主题 说明
标注、数据集与训练 素材导入、标注工具、辅助标注、数据集导出和训练
推理与模型导出 图片、视频、相机推理与部署模型导出
环境与模型 GPU、可选依赖、模型权重和 TensorRT
开发指南 架构、测试和扩展方式
诊断与排障 日志、诊断报告和常见异常

许可证

本项目采用 MIT License。项目依赖 Ultralytics YOLO,请同时遵守其许可证要求。

Metadata

Release files for yolo26-app 1.0.5

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for yolo26-app 1.0.5
File Size Uploaded Provenance Actions
yolo26_app-1.0.5.tar.gz 153.3 kB With provenance Details for yolo26_app-1.0.5.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for yolo26-app 1.0.5
File Interpreter ABI Platform Actions
yolo26_app-1.0.5-py3-none-any.whl Python 3 none any Details for yolo26_app-1.0.5-py3-none-any.whl

Total release size: 286.3 kB

Release files / yolo26_app-1.0.5.tar.gz

Download URL yolo26_app-1.0.5.tar.gz
Size 153.3 kB
Tags Source
SHA-256 checksum
How to use checksums
cd006a9c2f6bfa829c85673cc602d699db3b4178589be10e18ffffa564cd6562
BLAKE2b-256 checksum
How to use checksums
d9c8b7ebeedca3cfa22f874eb4e2f3ac3e69bba4436ec70834a76cbf64732e44
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 7, 2026.

Transparency log

Release files / yolo26_app-1.0.5-py3-none-any.whl

Download URL yolo26_app-1.0.5-py3-none-any.whl
Size 133.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
3b37c06e41b1f87b9d84311f4009e52d77474e02e5d86d33d4ef39fe6e7e60a2
BLAKE2b-256 checksum
How to use checksums
d7e4e1b06e7dff983609784578f6b42af9e92c18b7a09e8c7e7ec2b2e96d9ad7
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 7, 2026.

Transparency log

Security for yolo26-app 1.0.5

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 2 files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for yolo26_app-1.0.5.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 7, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page