以下是我的训练命令行： nohup python train.py --cfg yolov5xP2CBAM-TPH-BiFPN-SPP.yaml --weights 'weights/yolov5x.pt' --img 640 --data myVisDrone.yaml --hyp data/hyps/hyp.visdrone.yaml --epochs 81 --device 0,1 --batch-size 2 --name '5x_tph_ori_640_526' &

然后我训练了81轮我的metrics/mAP_0.5:0.95只有0.024101 以下是我的train_batch.jpg

这是我的数据集文件

Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]

path: ../datasets/VisDrone # dataset root dir

path: /home/ff/WangZF/datasets/visdrone # dataset root dir

train: /home/ff/WangZF/datasets/visdrone/YOLOFormat/train/images # train images (relative to 'path') 6471 images val: /home/ff/WangZF/datasets/visdrone/YOLOFormat/val/images # val images (relative to 'path') 548 images test: /home/ff/WangZF/datasets/visdrone/YOLOFormat/val/images # test images (optional) 1610 images

Classes

nc: 10 # number of classes names: [ 'pedestrian', 'people', 'bicycle', 'car', 'van', 'truck', 'tricycle', 'awning-tricycle', 'bus', 'motor' ] 我不知道是哪里出错了

Gumpest / YOLOv5-Multibackbone-Compression

你好，我尝试复现您的yolov5xP2CBAM-TPH-BiFPN-SPP.yaml模型文件的结果，但是最终精度和您复现的效果相差比较大。 #84

Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]

path: ../datasets/VisDrone # dataset root dir

path: /home/ff/WangZF/datasets/visdrone # dataset root dir

Classes