The module `datasets.AutomaticSpeechRecognition` is deprecated in 3.0.0, breaking some dataloaders

[sabilmakbar]

Describe the bug

The module datasets.AutomaticSpeechRecognition (from datasets.tasks) was deprecated since 3.0.0
Release Note on 3.0.0: (https://github.com/huggingface/datasets/releases/tag/3.0.0)

Steps to reproduce the bug

# Sample code to reproduce the bug
load_dataset(
    path="SEACrowd/titml_idn",
    name="titml_idn_source",
    trust_remote_code=True
)

Expected results

It should load the dataset using load_dataset. example dataset:

load_dataset(
    path="SEACrowd/asr_indocsc",
    name="asr_indocsc_source",
    trust_remote_code=True
)

Actual results

Specify the actual results or traceback.

2057 verification_mode = VerificationMode(
   2058     (verification_mode or VerificationMode.BASIC_CHECKS) if not save_infos else VerificationMode.ALL_CHECKS
   2059 )
   2061 # Create a dataset builder
-> [2062](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a225265766f6c61622d547261696e65722d312d5461696c7363616c65227d.vscode-resource.vscode-cdn.net/sabil/id_asr_benchmark/.venv/lib/python3.11/site-packages/datasets/load.py:2062) builder_instance = load_dataset_builder(
   2063     path=path,
   2064     name=name,
   2065     data_dir=data_dir,
   2066     data_files=data_files,
   2067     cache_dir=cache_dir,
   2068     features=features,
   2069     download_config=download_config,
   2070     download_mode=download_mode,
...
     85     citation=_CITATION,
---> [86](https://vscode-remote+ssh-002dremote-002b7b22686f73744e616d65223a225265766f6c61622d547261696e65722d312d5461696c7363616c65227d.vscode-resource.vscode-cdn.net/sabil/~/.cache/huggingface/modules/datasets_modules/datasets/SEACrowd--titml_idn/2e3a4c66685a6eb7d1b3c3062d399b7c77fefef8571aa6b6e19f896188ccffee/titml_idn.py:86)     task_templates=[datasets.AutomaticSpeechRecognition(audio_column="audio", transcription_column="text")],
     87 )

AttributeError: module 'datasets' has no attribute 'AutomaticSpeechRecognition'

Environment info

• datasets version: 3.6.0
• Platform: Linux-6.8.0-101-generic-x86_64-with-glibc2.39
• Python version: 3.11.14
• huggingface_hub version: 1.5.0
• PyArrow version: 23.0.1
• Pandas version: 3.0.1
• fsspec version: 2025.3.0

Suggestion: since only 9 datasets are still using deprecated class, we can remove it entirely the kwargs task_templates from the code:
https://github.com/search?q=repo%3ASEACrowd%2Fseacrowd-datahub%20AutomaticSpeechRecognition&type=code

https://github.com/search?q=repo%3ASEACrowd%2Fseacrowd-datahub+task_templates&type=code

[sabilmakbar]

update: this issue is also reproducible using SEACrowd test suites:

./test_example.sh titml_idn

/Users/sabilakbar/Documents/.venv/lib/python3.9/site-packages/urllib3/__init__.py:35: NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See: https://github.com/urllib3/urllib3/issues/3020
  warnings.warn(
INFO:__main__:args: Namespace(path='seacrowd/sea_datasets/titml_idn/titml_idn.py', schema=None, subset_id=None, data_dir=None, use_auth_token=None)
INFO:__main__:self.PATH: seacrowd/sea_datasets/titml_idn/titml_idn.py
INFO:__main__:self.SUBSET_ID: titml_idn
INFO:__main__:self.SCHEMA: None
INFO:__main__:self.DATA_DIR: None
INFO:__main__:Checking for _SUPPORTED_TASKS ...
module seacrowd.sea_datasets.titml_idn.titml_idn
INFO:__main__:Found _SUPPORTED_TASKS=[<Tasks.SPEECH_RECOGNITION: 'ASR'>]
INFO:__main__:_SUPPORTED_TASKS implies _MAPPED_SCHEMAS={'SPTEXT'}
INFO:__main__:schemas_to_check: {'SPTEXT'}
INFO:__main__:Checking load_dataset with config name titml_idn_source
The repository for titml_idn contains custom code which must be executed to correctly load the dataset. You can inspect the repository content at https://hf.co/datasets/titml_idn.
You can avoid this prompt in future by passing the argument `trust_remote_code=True`.

Do you wish to run the custom code? [y/N] y
E
======================================================================
ERROR: runTest (__main__.TestDataLoader)
Run all tests that check:
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/sabilakbar/Documents/seacrowd-datahub-sabil-clone/tests/test_seacrowd.py", line 134, in setUp
    self.dataset_source = datasets.load_dataset(
  File "/Users/sabilakbar/Documents/.venv/lib/python3.9/site-packages/datasets/load.py", line 2062, in load_dataset
    builder_instance = load_dataset_builder(
  File "/Users/sabilakbar/Documents/.venv/lib/python3.9/site-packages/datasets/load.py", line 1819, in load_dataset_builder
    builder_instance: DatasetBuilder = builder_cls(
  File "/Users/sabilakbar/Documents/.venv/lib/python3.9/site-packages/datasets/builder.py", line 354, in __init__
    info.update(self._info())
  File "/Users/sabilakbar/.cache/huggingface/modules/datasets_modules/datasets/titml_idn/2e3a4c66685a6eb7d1b3c3062d399b7c77fefef8571aa6b6e19f896188ccffee/titml_idn.py", line 86, in _info
    task_templates=[datasets.AutomaticSpeechRecognition(audio_column="audio", transcription_column="text")],
AttributeError: module 'datasets' has no attribute 'AutomaticSpeechRecognition'

----------------------------------------------------------------------
Ran 1 test in 1.880s

FAILED (errors=1)