YOLO2 Pytorch: train and run YOLOv2 object detection

Understand the project

Practical overview

Use it when you need a modifiable PyTorch YOLOv2 codebase for detector training, evaluation or image, camera and video detection.

Train object detectors, including with multi-GPU support.
Run detection on images, camera input and video files.
Evaluate models using cached datasets and inspect training summaries and debugging images.
Convert models through ONNX into Caffe2 for a Caffe2-oriented deployment workflow.

Before you choose

The project is public open-source software released under LGPL-3.0.

The quick-start workflow uses Python 3 and requires aria2 to download its example datasets and models.
Dataset cache plugins include PASCAL VOC and MS COCO.
Multi-GPU training is supported; distributed training is marked unsupported.
The repository describes ONNX conversion to Caffe2 but does not establish compatibility with a specific deployment target.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the Python dependencies and use the included example script for detection and evaluation.

Install dependencies with sudo pip3 install -r requirements.txt in a terminal.
Run quick_start.sh.
Let the script download datasets and models, cache the datasets, evaluate the models and display detection results for an example image.

The example workflow evaluates models and displays object-detection results for an example image.

The README documents this as an example workflow; it does not specify hardware requirements.
Distributed training is marked as unsupported.

Evidence and transparency

Source record details

Source availability: Active
Sources checked: 17 October 2026
Original website title: ruiminshen/yolo2-pytorch · Official homepage title · retained in its source language