Parameters36.0M
GFLOPs100.0
Input Size640px
Best mAP54.8%
LicenseApache-2.0
Architecture
Type
transformer
Backbone
PResNet
Neck
HybridEncoder
Head
DETR
Benchmark Results
Performance on COCO val2017 across different hardware configurations
| Hardware | Runtime | mAP@50-95 | FPS | Latency | VRAM |
|---|---|---|---|---|---|
| NVIDIA Jetson Orin Nano Super 8GB | ONNX Runtime FP32 | 54.8% | 5.4 | 184.3ms | - |
| NVIDIA Jetson Orin Nano Super 8GB | PyTorch FP32 | 54.8% | 5.1 | 196.8ms | 298 MB |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP16 | 54.8% | 29.7 | 33.6ms | - |
| NVIDIA Jetson Orin Nano Super 8GB | TensorRT FP32 | 54.8% | 15.8 | 63.3ms | - |
| NVIDIA RTX 5070 Ti | ONNX Runtime FP32 | 54.8% | 79.6 | 12.6ms | - |
| NVIDIA RTX 5070 Ti | PyTorch FP32 | 54.8% | 39.9 | 25.1ms | 302 MB |
| NVIDIA RTX 5070 Ti | TensorRT FP16 | 54.7% | 113.0 | 8.9ms | - |
| NVIDIA RTX 5070 Ti | TensorRT FP32 | 54.8% | 103.3 | 9.7ms | - |
Speed Breakdown(NVIDIA Jetson Orin Nano Super 8GB)
4.5ms
189.4ms
2.9ms
Preprocess
Inference
Postprocess (NMS)
Usage with LibreYOLO
from libreyolo import LibreYOLO
# Load model (auto-downloads from HuggingFace if not found locally)
model = LibreYOLO("LibreRTDETRv2r50m.pt")
# Run inference
result = model("image.jpg", conf=0.25, iou=0.45)
# Process results
print(f"Found {len(result)} objects")
print(result.boxes.xyxy) # bounding boxes (N, 4)
print(result.boxes.conf) # confidence scores (N,)
print(result.boxes.cls) # class IDs (N,)detrnms-freePaper: 51.9% mAP