YOLO分类模型训练报错：期望路径对象而非NoneType

尝试用包含两类JPG图片的自定义数据集训练YOLO分类模型，启动训练时触发报错：

Traceback (most recent call last):
  File "C:\pythonProject\main.py", line 5, in <module>
    model.train(data='./training_dataset', epochs=1, imgsz=64)
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\engine\model.py", line 341, in train
    self.trainer.train()
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\engine\trainer.py", line 192, in train
    self._do_train(world_size)
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\engine\trainer.py", line 288, in _do_train
    self._setup_train(world_size)
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\engine\trainer.py", line 255, in _setup_train
    self.test_loader = self.get_dataloader(self.testset, batch_size=batch_size * 2, rank=-1, mode='val')
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\models\yolo\classify\train.py", line 88, in get_dataloader
    dataset = self.build_dataset(dataset_path, mode)
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\models\yolo\classify\train.py", line 83, in build_dataset
    return ClassificationDataset(root=img_path, args=self.args, augment=mode == 'train', prefix=mode)
  File "C:\pythonProject\venv\lib\site-packages\ultralytics\data\dataset.py", line 220, in __init__
    super().__init__(root=root)
  File "C:\pythonProject\venv\lib\site-packages\torchvision\datasets\folder.py", line 309, in __init__
    super().__init__(
  File "C:\pythonProject\venv\lib\site-packages\torchvision\datasets\folder.py", line 145, in __init__
    samples = self.make_dataset(self.root, class_to_idx, extensions, is_valid_file)
  File "C:\pythonProject\venv\lib\site-packages\torchvision\datasets\folder.py", line 189, in make_dataset
    return make_dataset(directory, class_to_idx, extensions=extensions, is_valid_file=is_valid_file)
  File "C:\pythonProject\venv\lib\site-packages\torchvision\datasets\folder.py", line 61, in make_dataset
    directory = os.path.expanduser(directory)
  File "C:\Users\User\AppData\Local\Programs\Python\Python310\lib\ntpath.py", line 293, in expanduser
    path = os.fspath(path)
TypeError: expected str, bytes or os.PathLike object, not NoneType

更换两个不同数据集后问题仍存在，完整代码如下：

from ultralytics import YOLO

model = YOLO('yolov8n-cls.pt')

model.train(data='./training_dataset', epochs=1, imgsz=64)

数据集结构为：training_dataset/train/cats、training_dataset/train/dogs

• 补全数据集结构：YOLO分类训练默认需要训练集和验证集，在training_dataset目录下添加val文件夹，内部创建与train一致的cats和dogs子目录，并放入对应类别的验证图片。最终结构应为：

  training_dataset/
  ├─ train/
  │  ├─ cats/
  │  └─ dogs/
  └─ val/
     ├─ cats/
     └─ dogs/
  

• 临时跳过验证集（不推荐，无法评估模型泛化能力）：修改训练代码，添加val=False参数，代码如下：

  from ultralytics import YOLO
  
  model = YOLO('yolov8n-cls.pt')
  
  model.train(data='./training_dataset', epochs=1, imgsz=64, val=False)
  

内容的提问来源于stack exchange，提问作者Max