训练自己的数据集

[bai666ai]

课时名称	课时知识点
训练自己的数据集	实战篇包括：PyTorch环境安装、YOLOv5项目安装、准备自己的数据集、修改配置文件、训练自己的数据集、Grad-CAM热力图可视化

[2401_84728570]

教授，您好，自己的数据集训练，出现以下错误，如何处理。
train: Scanning 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\train' images and labels...1415 found, 0 missing, 0 empty, 0 corrupt: 100%|████████████████████████████████| 1415/1415 [00:02<00:00, 549.25it/s]
train: WARNING: Cache directory D:\yolov5-6.1-gradcam\VOCdevkit\labels is not writeable: [WinError 183] 当文件已存在时，无法创建该文件。: 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\train.cache.npy' -> 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\train.cache'
val: Scanning 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\val' images and labels...995 found, 0 missing, 0 empty, 0 corrupt: 100%|███████████████████████████████████████| 995/995 [00:05<00:00, 194.98it/s]
val: WARNING: Cache directory D:\yolov5-6.1-gradcam\VOCdevkit\labels is not writeable: [WinError 183] 当文件已存在时，无法创建该文件。: 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\val.cache.npy' -> 'D:\yolov5-6.1-gradcam\VOCdevkit\labels\val.cache'
Epoch gpu_mem box obj cls labels img_size
0/99 3.5G 0.109 0.02131 0.01933 14 640: 100%|██████████| 89/89 [00:22<00:00, 3.88it/s]
Class Images Labels P R mAP@.5 mAP@.5:.95: 0%| | 0/32 [00:00<?, ?it/s]
Traceback (most recent call last):
File "train.py", line 643, in
main(opt)
File "train.py", line 539, in main
train(opt.hyp, opt, device, callbacks)
File "train.py", line 370, in train
results, maps, _ = val.run(data_dict,
File "D:\Users\LENOVO\anaconda3\envs\v5\lib\site-packages\torch\autograd\grad_mode.py", line 28, in decorate_context
return func(*args, **kwargs)
File "D:\yolov5-6.1-gradcam\val.py", line 193, in run
out, train_out = model(im) if training else model(im, augment=augment, val=True) # inference, loss outputs
ValueError: too many values to unpack (expected 2)