如何使用 YOLOv3 修复"pickle_module.load(f, **pickle_load_args) _pickle.UnpicklingError: invalid load key,

Question

我正在尝试使用这个带有自定义数据集的存储库来训练YOLOv3

https://github.com/ultralytics/yolov3

我使用了自述文件yolov3-pp-ultralytics.pt 中推荐的重量

当我训练时，我得到以下错误：

Traceback (most recent call last):
File "yolov3/train.py", line 431, in <module>
train(hyp)  # train normally
File "yolov3/train.py", line 119, in train
ckpt = torch.load(weights, map_location=device)
File "/opt/conda/lib/python3.7/site-packages/torch/serialization.py", line 593, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "/opt/conda/lib/python3.7/site-packages/torch/serialization.py", line 763, in _legacy_load
magic_number = pickle_module.load(f, **pickle_load_args)
_pickle.UnpicklingError: invalid load key, '<'.

你知道Yolov3是怎么把文件腌出来的吗？

Answer 1

我认为您的yolov3-pp-ultralytics.pt不正确。请用记事本打开它，我猜内容是一个404 HTML页面。"<quot；是HTML页面的开头。