# WongKinYiu/yolor

implementation of paper - You Only Learn One Representation: Unified Network for Multiple Tasks (https://arxiv.org/abs/2105.04206)

Repository: https://github.com/WongKinYiu/yolor
Canonical: https://ross.abutalabs.com/products/yolor
Language: Python
License: GPL-3.0
License Family: copyleft
Topics: yolov4, scaled-yolov4, yolov4-csp, darknet, pytorch, implicit, explicit, yolor, yolo, representation, unified-network
Last push: 2024-11-03T00:14:38+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1970, "days_push": 669, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2003, forks 507 (observed 2026-08-28T04:06:04.510509+00:00)

## What it is
PyTorch implementation of the YOLOR paper 'You Only Learn One Representation: Unified Network for Multiple Tasks', a real-time object detection architecture that unifies implicit and explicit knowledge in a single network. It includes training and inference code, model configs, and pretrained weights with state-of-the-art COCO results.

## Use cases
- train a YOLO-family object detection model in PyTorch
- run real-time object detection on images with pretrained weights
- reproduce YOLOR paper results on the COCO benchmark
- fine-tune an object detection model on a custom dataset
- compare YOLOR against YOLOv4 and Scaled-YOLOv4 performance
- study unified implicit and explicit representation learning for detection

## When to choose
- You specifically need the YOLOR architecture, its configs, or its pretrained weights for research or reproduction
- You want a high-accuracy real-time detector benchmarked on COCO with published AP/throughput numbers
- You are exploring unified representation learning combining implicit and explicit knowledge in vision models

## When to avoid
- You want the newest YOLO models with active development and rich tooling (e.g., YOLOv7, YOLOv8/Ultralytics)
- You need production-grade deployment support such as ONNX/TensorRT export pipelines or extensive documentation
- You need a general-purpose detection framework rather than a specific paper implementation

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning, artificial-intelligence, image-processing
- platform: python, cross-platform
- tags: object-detection, yolo, pytorch, paper-implementation, coco, darknet, real-time-detection, unified-network, implicit-knowledge, pretrained-weights, gpu, linux

## Member repositories
- WongKinYiu/yolor (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:04.510509+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:01:53.210730+00:00, confidence not recorded.
  - readme: https://github.com/WongKinYiu/yolor (fetched 2026-08-28T04:06:04.510509+00:00, sha 5709666acba6)
- Data as of 2026-08-30T08:39:29.467469+00:00.