Back to Leaderboard

YOLOv7

yolov7

one-stage detector with E-ELAN backbone

Parameters

36.9M

FLOPs

104.7G

Input Size

640px

License

GPL-3.0

Architecture

Type

one-stage

Backbone

E-ELAN

Neck

SPPCSPC

Head

RepConv

Benchmark Results
Performance on COCO val2017 across different hardware configurations
HardwaremAP@50-95FPSLatencyVRAM
NVIDIA A100 (TensorRT FP16)51.3%96.310.4ms1624 MB
NVIDIA T4 (TensorRT FP16)51.4%35.828.0ms1585 MB
CPU (ONNX Runtime)51.5%4.1243.0ms1656 MB
Speed Breakdown (A100 TensorRT)
End-to-end latency breakdown showing preprocessing, inference, and postprocessing times
1.3ms
6.7ms
2.4ms
Preprocess
Inference
Postprocess (NMS)
Usage with LibreYOLO
from libreyolo import YOLO

# Load model
model = YOLO.from_pretrained("https://huggingface.co/Libre-YOLO/yolov7")

# Run inference
results = model.predict("image.jpg")

# Process results
for box in results.boxes:
    print(f"Class: {box.cls}, Confidence: {box.conf:.2f}")
balancedwidely-used
Related Models (yolov7)