RT-ODLab/models/detectors/yolov5

YOLOv5:

• For training, we train YOLOv5 series with 300 epochs on COCO.
• For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation, following the setting of YOLOv5.
• For optimizer, we use SGD with momentum 0.937, weight decay 0.0005 and base lr 0.01.
• For learning rate scheduler, we use linear decay scheduler.
• For YOLOv5's structure, we use decoupled head, following the setting of YOLOX.
• For YOLOv5-M and YOLOv5-L, increasing the batch size may improve performance. Due to my computing resources, I can only set the batch size to 16.

Train YOLOv5

Single GPU

Taking training YOLOv5-S on COCO as the example,

python train.py --cuda -d coco --root path/to/coco -m yolov5_s -bs 16 -size 640 --wp_epoch 3 --max_epoch 300 --eval_epoch 10 --no_aug_epoch 20 --ema --fp16 --multi_scale 

Multi GPU

Taking training YOLOv5 on COCO as the example,

python -m torch.distributed.run --nproc_per_node=8 train.py --cuda -dist -d coco --root /data/datasets/ -m yolov5_s -bs 128 -size 640 --wp_epoch 3 --max_epoch 300  --eval_epoch 10 --no_aug_epoch 20 --ema --fp16 --sybn --multi_scale --save_folder weights/ 

Test YOLOv5

Taking testing YOLOv5 on COCO-val as the example,

python test.py --cuda -d coco --root path/to/coco -m yolov5_s --weight path/to/yolov5.pth -size 640 -vt 0.4 --show 

Evaluate YOLOv5

Taking evaluating YOLOv5 on COCO-val as the example,

python eval.py --cuda -d coco-val --root path/to/coco -m yolov5_s --weight path/to/yolov5.pth 

Demo

Detect with Image

python demo.py --mode image --path_to_img path/to/image_dirs/ --cuda -m yolov5_s --weight path/to/weight -size 640 -vt 0.4 --show

Detect with Video

python demo.py --mode video --path_to_vid path/to/video --cuda -m yolov5_s --weight path/to/weight -size 640 -vt 0.4 --show --gif

Detect with Camera

python demo.py --mode camera --cuda -m yolov5_s --weight path/to/weight -size 640 -vt 0.4 --show --gif

© 2025 Gogs

Страница: 311ms Шаблон: 1ms

српски

српски English 简体中文 繁體中文（香港） 繁體中文（臺灣） Deutsch français Nederlands latviešu русский 日本語 español português do Brasil polski български italiano suomi Türkçe čeština svenska 한국어 galego українська English (United Kingdom) Magyar Slovenčina Indonesian Persian Vietnamese Português Монгол Română

Javascript Licenses Веб-страница

Model	Backbone	Batch	Scale	APval 0.5:0.95	APval 0.5	FLOPs (G)	Params (M)	Weight
YOLOv5-N	CSPDarkNet-N	1xb16	640	29.8	47.1	7.7	2.4	ckpt
YOLOv5-S	CSPDarkNet-S	1xb16	640	37.8	56.5	27.1	9.0	ckpt
YOLOv5-M	CSPDarkNet-M	1xb16	640	43.5	62.5	74.3	25.4	ckpt
YOLOv5-L	CSPDarkNet-L	1xb16	640	46.7	65.5	155.6	54.2	ckpt