训练ConvNeXt时触发TypeError：__init__()收到意外参数'split'

问题场景 #

使用mmpretrain仓库训练ConvNeXt模型，执行训练命令：

python tools/train.py configs/convnext/convnext-tiny_32xb128_custom.py

错误日志 #

Traceback (most recent call last):
  File "tools/train.py", line 162, in <module>
    main()
  File "tools/train.py", line 158, in main
    runner.train()
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\runner\runner.py", line 1728, in train
    self._train_loop = self.build_train_loop(
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\runner\runner.py", line 1527, in build_train_loop
    loop = EpochBasedTrainLoop(
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\runner\loops.py", line 44, in __init__
    super().__init__(runner, dataloader)
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\runner\base_loop.py", line 26, in __init__
    self.dataloader = runner.build_dataloader(
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\runner\runner.py", line 1370, in build_dataloader
    dataset = DATASETS.build(dataset_cfg)
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\registry\registry.py", line 570, in build
    return self.build_func(cfg, *args, **kwargs, registry=self)
  File "C:\ProgramData\Anaconda3\envs\open-mmlab\lib\site-packages\mmengine\registry\build_functions.py", line 121, in build_from_cfg
    obj = obj_cls(**args)  # type: ignore
  File "d:\working\ai\convnext\convnext\mmpretrain\mmpretrain\datasets\custom.py", line 207, in __init__
    super().__init__(
TypeError: __init__() got an unexpected keyword argument 'split'

使用的配置文件 #

_base_ = [
    '../_base_/models/convnext/convnext-tiny.py',
    '../_base_/datasets/imagenet_bs64_swin_224.py',
    '../_base_/schedules/imagenet_bs1024_adamw_swin.py',
    '../_base_/default_runtime.py',
]

# dataset setting
train_dataloader = dict(batch_size=128, dataset=dict(type="CustomDataset"))

# schedule setting
optim_wrapper = dict(
    optimizer=dict(lr=4e-3),
    clip_grad=None,
)

# runtime setting
custom_hooks = [dict(type='EMAHook', momentum=1e-4, priority='ABOVE_NORMAL')]

# NOTE: `auto_scale_lr` is for automatically scaling LR
# based on the actual training batch size.
# base_batch_size = (32 GPUs) x (128 samples per GPU)
auto_scale_lr = dict(base_batch_size=4096)

原因分析 #

你继承的基础数据集配置../_base_/datasets/imagenet_bs64_swin_224.py中，给数据集定义了split参数，但CustomDataset的初始化方法并不支持该参数，导致构建数据集实例时抛出参数不匹配的错误。

解决方法 #

方案1：覆盖split参数 #

在自定义配置的train_dataloader.dataset中显式设置split=None，覆盖基础配置中的该参数：

train_dataloader = dict(batch_size=128, dataset=dict(type="CustomDataset", split=None))

方案2：替换基础数据集配置 #

将_base_中的数据集配置替换为适用于CustomDataset的模板（比如仓库中提供的custom_dataset.py基础配置），避免继承到split参数。

注意：确保CustomDataset所需的其他必要参数（如data_root、ann_file等）已正确配置，否则后续可能会出现其他数据集加载错误。

内容的提问来源于stack exchange，提问作者WonderCreator