Seems to be broken

#1
by j3m - opened

With the code

import datasets

dataset = datasets.load_dataset("ammarnasr/the-stack-rust-clean")

I get

Traceback (most recent call last):
  File "/home/jmillist/code/work/scripts/24-03-30@21:53:20.py", line 3, in <module>
    dataset = datasets.load_dataset("ammarnasr/the-stack-rust-clean")
  File "/home/jmillist/code/work/pyenv/lib/python3.7/site-packages/datasets/load.py", line 1815, in load_dataset
    storage_options=storage_options,
  File "/home/jmillist/code/work/pyenv/lib/python3.7/site-packages/datasets/builder.py", line 913, in download_and_prepare
    **download_and_prepare_kwargs,
  File "/home/jmillist/code/work/pyenv/lib/python3.7/site-packages/datasets/builder.py", line 1022, in _download_and_prepare
    verify_splits(self.info.splits, split_dict)
  File "/home/jmillist/code/work/pyenv/lib/python3.7/site-packages/datasets/utils/info_utils.py", line 100, in verify_splits
    raise NonMatchingSplitsSizesError(str(bad_splits))
datasets.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=3582248477.9086223, num_examples=806789, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='train', num_bytes=11063453735, num_examples=893792, shard_lengths=[47242, 47242, 47242, 44484, 47242, 47242, 47242, 44482, 47241, 47241, 47241, 44482, 47241, 47241, 47241, 44482, 47241, 47241, 47241, 7241], dataset_name='parquet')}, {'expected': SplitInfo(name='test', num_bytes=394048264.9973618, num_examples=88747, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='test', num_bytes=618790407, num_examples=49655, shard_lengths=[44828, 4827], dataset_name='parquet')}, {'expected': SplitInfo(name='valid', num_bytes=3982797.09401595, num_examples=897, shard_lengths=None, dataset_name=None), 'recorded': SplitInfo(name='valid', num_bytes=598090380, num_examples=49656, shard_lengths=[44828, 4828], dataset_name='parquet')}]

datasets 2.13.2

Sign up or log in to comment