2023年11月29日发(作者:)

调试YOLOv3YOLOv5过程中遇到的问题汇总

今⽇⽴冬

yolov1

yolo1论⽂

/video/BV15w411Z7LG

yolo2论⽂

/video/BV1Q64y1s74K

yolo3论⽂

/video/BV1Vg411V7bJ

yolov3

pycococtools的安装包/s/1nWQdPRtGwNnOO2DkxRuGuA提取码:i5d7 安装完直接解压复制其中pycocotools两个⽂件夹到

conda环境…Libsite-packages之中

bug2: Cant get attribute SPPF on module from

/RooKichenn/article/details/120866650

bug3:RuntimeError: Given groups=1, weight of size 512, 1024, 1, 1, expected input1, 512, 8, 8 to have

1024 channels, but got 512 channels instead

_argument('--cfg', type=str, default='models/', help=' path')

bug4: coco128

运⾏的时候,数据集⾃动下载之后,⼯程⽂件下会出现⽂件,报未解压报错,解决⽅法: 解压到(yolov5-5.0⽂件夹

同⼀⽬录下⽂件夹叫coco128),因为需要去train: ../coco128/images/train2017/ # 128 images

val: ../coco128/images/train2017/ # 128 images

bug5:BrokenPipeError: Errno 32 Broken pipe

根据⽂章/qq_33666011/article/details/81873217,我把workers改⼩了,暂时不报错了。

bug6RuntimeError: [enforce fail at …:72] tCPUAllocator: not enough

memory: you tried to allocate xxx bytes.

私以为设置的张量过⼤,导致内存分配不⾜,在中把batchsize改⼩了