YOLOX

yoloxIn LibreYOLO

Megvii · arXiv 2021

Paper Code

Variants6

Parameters0.9M - 99.1M

Best Paper mAP51.1%

LicenseApache-2.0

Model Variants

All variants benchmarked on COCO val2017. Click any variant for the full hardware breakdown.

Model	Params	GFLOPs	Paper mAP	Best FPS
YOLOX-Nano	0.9M	1.3	25.8%	100 FPS NVIDIA RTX 5070 Ti · TensorRT FP16	Details
YOLOX-Tiny	5.1M	7.7	32.8%	98 FPS NVIDIA RTX 5070 Ti · TensorRT FP16	Details
YOLOX-S	9.0M	13.5	40.5%	75 FPS NVIDIA RTX 5070 Ti · TensorRT FP32	Details
YOLOX-M	25.3M	37.0	46.9%	83 FPS NVIDIA RTX 5070 Ti · TensorRT FP16	Details
YOLOX-L	54.2M	78.0	49.7%	79 FPS NVIDIA RTX 5070 Ti · TensorRT FP16	Details
YOLOX-X	99.1M	141.2	51.1%	72 FPS NVIDIA RTX 5070 Ti · TensorRT FP16	Details

Architecture

Type

one-stage

Backbone

CSPDarknet

Neck

PAFPN

Head

Decoupled

anchor-freenmsTrained on COCO train2017

Run any model with one line

LibreYOLO has the best catalogue of state-of-the-art detectors, all behind one MIT-licensed Python API.

from libreyolo import LibreYOLO, SAMPLE_IMAGE # LibreYOLO has the best catalogue of state-of-the-art models. model = LibreYOLO("LibreRFDETRl.pt") # RF-DETR-L (transformer flagship) results = model(SAMPLE_IMAGE, save=True) # run inference, save the annotated image # Swap in any other model, same one-line API (weights auto-download): # LibreYOLO("LibreYOLO9c.pt") # YOLO9-C # LibreYOLO("LibreYOLOXx.pt") # YOLOX-X # LibreYOLO("LibreDFINEx.pt") # D-FINE-X # LibreYOLO("LibreRTDETRr50.pt") # RT-DETR-R50