When you import the datasets library using python 3.11 you get the following error:
Traceback (most recent call last):
File "/kaggle/working/import_datasets.py", line 1, in <module>
import datasets
File "/opt/conda/envs/myenv/lib/python3.11/site-packages/datasets/__init__.py", line 45, in <module>
from .builder import ArrowBasedBuilder, BeamBasedBuilder, BuilderConfig, DatasetBuilder, GeneratorBasedBuilder
File "/opt/conda/envs/myenv/lib/python3.11/site-packages/datasets/builder.py", line 91, in <module>
@dataclass
^^^^^^^^^
File "/opt/conda/envs/myenv/lib/python3.11/dataclasses.py", line 1221, in dataclass
return wrap(cls)
^^^^^^^^^
File "/opt/conda/envs/myenv/lib/python3.11/dataclasses.py", line 1211, in wrap
return _process_class(cls, init, repr, eq, order, unsafe_hash,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/conda/envs/myenv/lib/python3.11/dataclasses.py", line 959, in _process_class
cls_fields.append(_get_field(cls, name, type, kw_only))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/conda/envs/myenv/lib/python3.11/dataclasses.py", line 816, in _get_field
raise ValueError(f'mutable default {type(f.default)} for field '
ValueError: mutable default <class 'datasets.utils.version.Version'> for field version is not allowed: use default_factory
One of the changes in the dataclasses library in Python 3.11 (Python 3.10 works fine) is causing backward compatibility issues with the datasets library by hugging face.
When you import the datasets library using python 3.11 you get the following error:
To reproduce the error, run my kaggle kernels notebook