yolov10m.pt 报错 raise pickle.UnpicklingError(_get_wo_message(str(e))) from None_pickle.Unpicklin

使用yolov10m.onnx 可以正常使用，但是onnx 不可以自己训练，所以改用.pt ，出现报错
    raise pickle.UnpicklingError(_get_wo_message(str(e))) from None
_pickle.UnpicklingError: Weights only load failed. This file can still be loaded, to do so you have two options, do those steps only if you trust the source of the checkpoint. 
    (1) In PyTorch 2.6, we changed the default value of the `weights_only` argument in `torch.load` from `False` to `True`. Re-running `torch.load` with `weights_only` set to `False` will likely succeed, but it can result in arbitrary code execution. Do it only if you got the file from a trusted source.
    (2) Alternatively, to load with `weights_only=True` please check the recommended steps in the following error message.
    WeightsUnpickler error: Unsupported global: GLOBAL ultralytics.nn.tasks.YOLOv10DetectionModel was not an allowed global by default. Please use `torch.serialization.add_safe_globals([ultralytics.nn.tasks.YOLOv10DetectionModel])` or the `torch.serialization.safe_globals([ultralytics.nn.tasks.YOLOv10DetectionModel])` context manager to allowlist this global if you trust this class/function.

在 PyTorch 2.6 中，torch.load 的 weights_only 参数的默认值已经从 False 改为了 True，这意味着默认情况下，PyTorch 只会加载模型的权重，而不加载其他可能包含恶意代码的内容。因此，当你尝试加载 YOLO 模型时，可能会遇到 pickle.UnpicklingError，并且模型的全局引用 DetectionModel 不被默认允许

我使用的YOLOv10来加载模型，但是换汤 不换药，找到tasks.py

原本代码是 

ckpt = torch.load(file, map_location="cpu",weights_only=False)

在后面加上参数 

weights_only=False

就ok了