Skip to content

modern-yolonas

A clean, minimal Python reimplementation of YOLO-NAS object detection. No factory patterns, no registries, no OmegaConf — just PyTorch.

Features

  • Drop-in pretrained weights — loads super-gradients COCO checkpoints directly
  • Simple API — Detector("yolo_nas_s") → call with an image → get boxes
  • CLI — yolonas detect, yolonas train, yolonas export, yolonas eval
  • ONNX / OpenVINO export — including Frigate-compatible graph surgery
  • Training — full training loop with DDP, AMP, EMA, cosine LR
  • All 3 variants — S (~12M), M (~31M), L (~44M)

Quick install

pip install modern-yolonas

Minimal example

from modern_yolonas import Detector

det = Detector("yolo_nas_s", device="cuda")
result = det("image.jpg")
result.save("output.jpg")

Next steps