All pull requests

Pull requests

• Fix broken documentation links

  #8732·atharva-mashalkar opened Oct 3, 2026·

  ·

• Fix audio array handling in preprocess_function

  #8731·causius0 opened Oct 2, 2026·

  ·

• Fix compression in Dataset.to_csv

  #8730·PDGGK opened Oct 2, 2026·

  ·
  1

• Fix docstrings that name parameters the functions do not have

  #8728·VenishPaneliya opened Oct 2, 2026·

  ·

• Fix repeated Arrow rebatch checkpoint restoration

  #8727·tryu-ai opened Oct 2, 2026·

  ·

• Honor Video stream_index for remote paths

  #8726·Arthur031221 opened Oct 2, 2026·

  ·

• Fix IterableDataset.filter with arrow, pandas and polars formats

  #8725·felipebridge opened Oct 2, 2026·

  ·
  1

• Fix concatenate_datasets(axis=1) on arrow-backed iterable datasets with different lengths

  #8723·felipebridge opened Oct 1, 2026·

  ·
  1

• Fix iflatmap_unordered with empty worker results

  #8722·Mnilax opened Oct 1, 2026·

  ·

• Load directories written by Dataset.save_to_disk in load_dataset

  #8721·behroozazarkhalili opened Oct 1, 2026Contributor·

  ·
  1

• Add DatasetBuilder.as_iterable_dataset to stream the Arrow cache

  #8720·behroozazarkhalili opened Oct 1, 2026Contributor·

  ·
  1

• Don't serve the cached dataset module when download_mode is force_redownload

  #8718·ebarkhordar opened Oct 1, 2026Contributor·

  ·
  1

• docs: add "Use with Hugging Face Jobs" guide

  #8717·davanstrien opened Oct 1, 2026Member·

  ·
  9

• Allow IterableDataset.push_to_hub num_shards above dataset.num_shards

  #8716·behroozazarkhalili opened Oct 1, 2026Contributor·

  ·
  1

  2

• Fix Json casts of mixed encoded and plain strings

  #8715·Marcuswang0824 opened Sep 30, 2026·

  ·
  1

• Keep Json fields inside a struct from being double-encoded when some struct rows are null

  #8712·breken-ai opened Sep 30, 2026·

  ·
  1

• Advance the user-supplied generator in Dataset.train_test_split even on cache hits

  #8711·breken-ai opened Sep 30, 2026·

  ·

• Fix string_to_dict with unrelated regex capture groups

  #8710·charan-rathore opened Sep 29, 2026·

  ·
  1

• Raise instead of hanging on a probability-0 source without replacement

  #8708·Arthur031221 opened Sep 29, 2026·

  ·

• Fix epoch iteration after skipping or taking a streaming subset

  #8707·galleonli opened Sep 28, 2026·

  ·
  1

• Cap lock filename so the total path stays under MAX_PATH on Windows

  #8703·825pranav opened Sep 27, 2026·

  ·
  1

• Implement __arrow_array__ on PandasArrayExtensionArray so ArrayXD columns can go back from pandas to Arrow

  #8701·JoeyTan21 opened Sep 27, 2026·

  ·
  1

  1

• Fix DatasetInfosDict metadata overwrite

  #8696·r-abaryan opened Sep 27, 2026·

  ·

• Keep take/skip counts when there are more DataLoader workers than shards

  #8695·yupengtang opened Sep 27, 2026·

  ·

• fix: use Sequence instead of List for PathLike type annotations (fixes #5354)

  #8693·CJstate opened Sep 27, 2026·

  ·
  1