open-mmlab/mmyolo

Health v2 · maintenance only

Adoption not part of the score

3468 stars · 630 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

MMYOLO is the OpenMMLab toolbox and benchmark for the YOLO series of object detection models, implemented on PyTorch. It provides unified implementations of RTMDet, YOLOv5-v8, YOLOX, PPYOLOE, and rotated-box variants with a model zoo, training, and deployment tooling.

Use cases

• train a custom YOLOv8 object detection model
• benchmark YOLO model variants on my dataset
• detect rotated objects like oriented bounding boxes
• deploy YOLO models to production
• compare YOLOv5 vs YOLOX accuracy and speed
• fine-tune a pretrained object detection model in PyTorch

When to choose

• you want a unified PyTorch framework covering many YOLO variants
• you need rotated/oriented object detection support
• you want modular config-driven training with a model zoo of pretrained weights

When to avoid

• you need a lightweight single-model inference library
• you want the newest models or frequent updates, as development has slowed
• you prefer non-PyTorch frameworks or avoid GPL-3.0 licensing

Facets

library · maturity maintenance

machine-learning deep-learning computer-vision image-processing benchmarking computer-vision deep-learning machine-learning python yolo object-detection pytorch rotated-object-detection model-zoo openmmlab linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="open-mmlab/mmyolo")