Datasets:

Modalities:
Audio
Text
ArXiv:
Libraries:
Datasets
License:

Got `KeyError: 'sentence_id'` when loading the yue data

#14
by laubonghaudoi - opened

I am trying to load the yue dataset with the codes:

from datasets import load_dataset, DatasetDict

common_voice = DatasetDict()

common_voice["train"] = load_dataset(
    "mozilla-foundation/common_voice_17_0", "yue", split="train+validation"
)
common_voice["test"] = load_dataset(
    "mozilla-foundation/common_voice_17_0", "yue", split="test"
)

print(common_voice)

However, I got this error:

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:1687, in GeneratorBasedBuilder._prepare_split_single(self, gen_kwargs, fpath, file_format, max_shard_size, split_info, check_duplicate_keys, job_id)
   1678     writer = writer_class(
   1679         features=writer._features,
   1680         path=fpath.replace("SSSSS", f"{shard_id:05d}").replace("JJJJJ", f"{job_id:05d}"),
   (...)
   1685         embed_local_files=embed_local_files,
   1686     )
-> 1687 example = self.info.features.encode_example(record) if self.info.features is not None else record
   1688 writer.write(example, key)

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1866, in Features.encode_example(self, example)
   1865 example = cast_to_python_objects(example)
-> 1866 return encode_nested_example(self, example)

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1243, in encode_nested_example(schema, obj, level)
   1241         raise ValueError("Got None but expected a dictionary instead")
   1242     return (
-> 1243         {
   1244             k: encode_nested_example(sub_schema, sub_obj, level=level + 1)
   1245             for k, (sub_schema, sub_obj) in zip_dict(schema, obj)
   1246         }
   1247         if obj is not None
   1248         else None
   1249     )
   1251 elif isinstance(schema, (list, tuple)):

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1243, in <dictcomp>(.0)
   1241         raise ValueError("Got None but expected a dictionary instead")
   1242     return (
-> 1243         {
   1244             k: encode_nested_example(sub_schema, sub_obj, level=level + 1)
   1245             for k, (sub_schema, sub_obj) in zip_dict(schema, obj)
   1246         }
   1247         if obj is not None
   1248         else None
   1249     )
   1251 elif isinstance(schema, (list, tuple)):

File /usr/local/lib/python3.10/dist-packages/datasets/utils/py_utils.py:323, in zip_dict(*dicts)
    321 for key in unique_values(itertools.chain(*dicts)):  # set merge all keys
    322     # Will raise KeyError if the dict don't have the same keys
--> 323     yield key, tuple(d[key] for d in dicts)

File /usr/local/lib/python3.10/dist-packages/datasets/utils/py_utils.py:323, in <genexpr>(.0)
    321 for key in unique_values(itertools.chain(*dicts)):  # set merge all keys
    322     # Will raise KeyError if the dict don't have the same keys
--> 323     yield key, tuple(d[key] for d in dicts)

KeyError: 'sentence_id'

The above exception was the direct cause of the following exception:

DatasetGenerationError                    Traceback (most recent call last)
Cell In[13], line 5
      1 from datasets import load_dataset, DatasetDict
      3 common_voice = DatasetDict()
----> 5 common_voice["train"] = load_dataset(
      6     "mozilla-foundation/common_voice_17_0", "yue", split="train+validation"
      7 )
      8 common_voice["test"] = load_dataset(
      9     "mozilla-foundation/common_voice_17_0", "yue", split="test"
     10 )
     12 print(common_voice)

File /usr/local/lib/python3.10/dist-packages/datasets/load.py:2152, in load_dataset(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, verification_mode, ignore_verifications, keep_in_memory, save_infos, revision, token, use_auth_token, task, streaming, num_proc, storage_options, **config_kwargs)
   2149 try_from_hf_gcs = path not in _PACKAGED_DATASETS_MODULES
   2151 # Download and prepare data
-> 2152 builder_instance.download_and_prepare(
   2153     download_config=download_config,
   2154     download_mode=download_mode,
   2155     verification_mode=verification_mode,
   2156     try_from_hf_gcs=try_from_hf_gcs,
   2157     num_proc=num_proc,
   2158     storage_options=storage_options,
   2159 )
   2161 # Build dataset for splits
   2162 keep_in_memory = (
   2163     keep_in_memory if keep_in_memory is not None else is_small_dataset(builder_instance.info.dataset_size)
   2164 )

File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:948, in DatasetBuilder.download_and_prepare(self, output_dir, download_config, download_mode, verification_mode, ignore_verifications, try_from_hf_gcs, dl_manager, base_path, use_auth_token, file_format, max_shard_size, num_proc, storage_options, **download_and_prepare_kwargs)
    946     if num_proc is not None:
    947         prepare_split_kwargs["num_proc"] = num_proc
--> 948     self._download_and_prepare(
    949         dl_manager=dl_manager,
    950         verification_mode=verification_mode,
    951         **prepare_split_kwargs,
    952         **download_and_prepare_kwargs,
    953     )
    954 # Sync info
    955 self.info.dataset_size = sum(split.num_bytes for split in self.info.splits.values())

File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:1711, in GeneratorBasedBuilder._download_and_prepare(self, dl_manager, verification_mode, **prepare_splits_kwargs)
   1710 def _download_and_prepare(self, dl_manager, verification_mode, **prepare_splits_kwargs):
-> 1711     super()._download_and_prepare(
   1712         dl_manager,
   1713         verification_mode,
   1714         check_duplicate_keys=verification_mode == VerificationMode.BASIC_CHECKS
   1715         or verification_mode == VerificationMode.ALL_CHECKS,
   1716         **prepare_splits_kwargs,
   1717     )

File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:1043, in DatasetBuilder._download_and_prepare(self, dl_manager, verification_mode, **prepare_split_kwargs)
   1039 split_dict.add(split_generator.split_info)
   1041 try:
   1042     # Prepare split will record examples associated to the split
-> 1043     self._prepare_split(split_generator, **prepare_split_kwargs)
   1044 except OSError as e:
   1045     raise OSError(
   1046         "Cannot find data file. "
   1047         + (self.manual_download_instructions or "")
   1048         + "\nOriginal error:\n"
   1049         + str(e)
   1050     ) from None

File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:1549, in GeneratorBasedBuilder._prepare_split(self, split_generator, check_duplicate_keys, file_format, num_proc, max_shard_size)
   1547 job_id = 0
   1548 with pbar:
-> 1549     for job_id, done, content in self._prepare_split_single(
   1550         gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
   1551     ):
   1552         if done:
   1553             result = content

File /usr/local/lib/python3.10/dist-packages/datasets/builder.py:1706, in GeneratorBasedBuilder._prepare_split_single(self, gen_kwargs, fpath, file_format, max_shard_size, split_info, check_duplicate_keys, job_id)
   1704     if isinstance(e, SchemaInferenceError) and e.__context__ is not None:
   1705         e = e.__context__
-> 1706     raise DatasetGenerationError("An error occurred while generating the dataset") from e
   1708 yield job_id, True, (total_num_examples, total_num_bytes, writer._features, num_shards, shard_lengths)

DatasetGenerationError: An error occurred while generating the dataset

I got no issues when loading the older versions of of Common Voice. How to resolve this?

Mozilla Foundation org

Hi @laubonghaudoi - It appears the dataset did not download correctly. Can you try the same with streaming mode too?

Thanks for the tips! I tried streaming mode like this:

from datasets import load_dataset

dataset = load_dataset(
    "mozilla-foundation/common_voice_17_0", "yue", split="train", streaming=True
)
print(next(iter(dataset)))

and got this error

Reading metadata...: 3150it [00:00, 11802.04it/s]
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
Cell In[5], line 10
      3 dataset = load_dataset(
      4     "mozilla-foundation/common_voice_17_0", "yue", split="train", streaming=True
      5 )
      6 # common_voice["test"] = load_dataset(
      7 #     "mozilla-foundation/common_voice_17_0", "yue", split="test"
      8 # )
---> 10 print(next(iter(dataset)))
     12 # print(common_voice)

File /usr/local/lib/python3.10/dist-packages/datasets/iterable_dataset.py:1383, in IterableDataset.__iter__(self)
   1379 for key, example in ex_iterable:
   1380     if self.features:
   1381         # `IterableDataset` automatically fills missing columns with None.
   1382         # This is done with `_apply_feature_types_on_example`.
-> 1383         example = _apply_feature_types_on_example(
   1384             example, self.features, token_per_repo_id=self._token_per_repo_id
   1385         )
   1386     yield format_dict(example) if format_dict else example

File /usr/local/lib/python3.10/dist-packages/datasets/iterable_dataset.py:1075, in _apply_feature_types_on_example(example, features, token_per_repo_id)
   1073         example[column_name] = None
   1074 # we encode the example for ClassLabel feature types for example
-> 1075 encoded_example = features.encode_example(example)
   1076 # Decode example for Audio feature, e.g.
   1077 decoded_example = features.decode_example(encoded_example, token_per_repo_id=token_per_repo_id)

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1866, in Features.encode_example(self, example)
   1855 """
   1856 Encode example into a format for Arrow.
   1857 
   (...)
   1863     `dict[str, Any]`
   1864 """
   1865 example = cast_to_python_objects(example)
-> 1866 return encode_nested_example(self, example)

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1243, in encode_nested_example(schema, obj, level)
   1240     if level == 0 and obj is None:
   1241         raise ValueError("Got None but expected a dictionary instead")
   1242     return (
-> 1243         {
   1244             k: encode_nested_example(sub_schema, sub_obj, level=level + 1)
   1245             for k, (sub_schema, sub_obj) in zip_dict(schema, obj)
   1246         }
   1247         if obj is not None
   1248         else None
   1249     )
   1251 elif isinstance(schema, (list, tuple)):
   1252     sub_schema = schema[0]

File /usr/local/lib/python3.10/dist-packages/datasets/features/features.py:1243, in <dictcomp>(.0)
   1240     if level == 0 and obj is None:
   1241         raise ValueError("Got None but expected a dictionary instead")
   1242     return (
-> 1243         {
   1244             k: encode_nested_example(sub_schema, sub_obj, level=level + 1)
   1245             for k, (sub_schema, sub_obj) in zip_dict(schema, obj)
   1246         }
   1247         if obj is not None
   1248         else None
   1249     )
   1251 elif isinstance(schema, (list, tuple)):
   1252     sub_schema = schema[0]

File /usr/local/lib/python3.10/dist-packages/datasets/utils/py_utils.py:323, in zip_dict(*dicts)
    320 """Iterate over items of dictionaries grouped by their keys."""
    321 for key in unique_values(itertools.chain(*dicts)):  # set merge all keys
    322     # Will raise KeyError if the dict don't have the same keys
--> 323     yield key, tuple(d[key] for d in dicts)

File /usr/local/lib/python3.10/dist-packages/datasets/utils/py_utils.py:323, in <genexpr>(.0)
    320 """Iterate over items of dictionaries grouped by their keys."""
    321 for key in unique_values(itertools.chain(*dicts)):  # set merge all keys
    322     # Will raise KeyError if the dict don't have the same keys
--> 323     yield key, tuple(d[key] for d in dicts)

KeyError: 'sentence_id'

is there a way to delete the downloaded data and re-download it? I am guessing that it's a download corruption problem.

@laubonghaudoi Maybe you found the issue? or work around? I have the same problem with this version of dataset. Older ones are working well.

Sign up or log in to comment