MIT-licensed YOLO implementation with full support for training and inference across YOLOv9 (t, s, m, c), YOLOX (nano, tiny, s, m, l, x), and RF-DETR (nano, small, base, medium, large).
pip install libreyolofrom libreyolo import LibreYOLO
# Auto-detect model version and size
model = LibreYOLO("LibreYOLOXs.pt")
results = model(image="https://raw.githubusercontent.com/LibreYOLO/libreyolo/main/libreyolo/assets/parkour.jpg", save=True)
print(f"Detected {results['num_detections']} objects")Full documentation at libreyolo.com/docs:
- Code: MIT License
- Weights: Pre-trained weights may inherit licensing from the original source