# RT-DETR-X

> RT-DETR-X: transformer object detector from the RT-DETR family. Available in LibreYOLO, the MIT-licensed open-source library (free for commercial use).

- Family: RT-DETR (Baidu)
- Parameters: 67.0M
- GFLOPs: 234.0
- Default input size: 640px
- Detection: detr · nms-free
- Architecture: transformer · HGNetv2-X backbone
- Weight license: Apache-2.0
- Paper-reported mAP@50-95: 54.8%
- Paper: https://arxiv.org/abs/2304.08069
- Original code: https://github.com/lyuwenyu/RT-DETR

## Benchmarks (COCO val2017)

| Hardware | Runtime | mAP@50-95 | FPS | Latency (ms) | VRAM (MB) |
|---|---|---|---|---|---|
| NVIDIA Jetson Orin Nano Super 8GB | ONNX Runtime FP32 | 58.0 | 3.0 | 331.4 | - |
| NVIDIA Jetson Orin Nano Super 8GB | PyTorch FP32 | 57.9 | 3.0 | 338.1 | 394 |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP16 | 57.9 | 17.0 | 58.9 | - |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP32 | 58.0 | 8.6 | 115.8 | - |
| NVIDIA RTX 5070 Ti | ONNX Runtime FP32 | 58.0 | 48.4 | 20.7 | - |
| NVIDIA RTX 5070 Ti | PyTorch FP32 | 57.9 | 21.8 | 46.0 | 401 |
| NVIDIA RTX 5070 Ti | TensorRT FP16 | 57.9 | 82.3 | 12.2 | - |
| NVIDIA RTX 5070 Ti | TensorRT FP32 | 57.9 | 68.5 | 14.6 | - |
| Raspberry Pi 5 | ONNX Runtime FP32 | 58.0 | 0.4 | 2801.5 | - |

## Usage with LibreYOLO

```python
from libreyolo import LibreYOLO

model = LibreYOLO("LibreRTDETRx.pt")
result = model("image.jpg", conf=0.25, iou=0.45)
```

Source: https://www.visionanalysis.org/model/rtdetr-x. Benchmarks produced with LibreYOLO (https://github.com/Libre-YOLO/libreyolo).