url stringlengths 61 61 | repository_url stringclasses 1
value | labels_url stringlengths 75 75 | comments_url stringlengths 70 70 | events_url stringlengths 68 68 | html_url stringlengths 49 51 | id int64 1.11B 5.54B | node_id stringlengths 18 25 | number int64 3.59k 8.66k | title stringlengths 1 290 | user dict | labels listlengths 0 4 | state stringclasses 2
values | locked bool 1
class | assignees listlengths 0 4 | milestone dict | comments int64 0 49 | created_at timestamp[s]date 2022-01-18 11:22:53 2026-09-22 04:23:19 | updated_at timestamp[s]date 2022-01-19 11:22:57 2026-09-22 04:23:54 | closed_at timestamp[s]date 2022-01-18 19:10:22 2026-09-21 16:12:41 ⌀ | assignee dict | author_association stringclasses 4
values | issue_field_values listlengths 0 0 | type null | active_lock_reason null | draft bool 2
classes | pull_request dict | body stringlengths 1 59.9k ⌀ | closed_by dict | reactions dict | timeline_url stringlengths 70 70 | performed_via_github_app null | state_reason stringclasses 4
values | sub_issues_summary dict | issue_dependencies_summary dict | pinned_comment null | is_pull_request bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/huggingface/datasets/issues/8661 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8661/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8661/comments | https://api.github.com/repos/huggingface/datasets/issues/8661/events | https://github.com/huggingface/datasets/pull/8661 | 5,535,263,354 | PR_kwDODunzps8AAAABEha4Pw | 8,661 | Validate a split name against the whole string | {
"login": "Rodrigo-Palma",
"id": 86628053,
"node_id": "MDQ6VXNlcjg2NjI4MDUz",
"avatar_url": "https://avatars.githubusercontent.com/u/86628053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rodrigo-Palma",
"html_url": "https://github.com/Rodrigo-Palma",
"followers_url": "https://api.githu... | [] | open | false | [] | null | 0 | 2026-09-22T04:23:19 | 2026-09-22T04:23:19 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8661",
"html_url": "https://github.com/huggingface/datasets/pull/8661",
"diff_url": "https://github.com/huggingface/datasets/pull/8661.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8661.patch",
"merged_at": null
} | ## Summary
Eight checks in the library validate a split name with `re.match(_split_re, name)` and,
when it fails, raise `Split name should match '^\w+(\.\w+)*$' but got '<name>'`. The
message promises a full match, but `"$"` also matches right before a newline at the end
of the string, so a name such as `"train\n"` pa... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8661/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8661/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8660 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8660/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8660/comments | https://api.github.com/repos/huggingface/datasets/issues/8660/events | https://github.com/huggingface/datasets/pull/8660 | 5,532,130,631 | PR_kwDODunzps8AAAABEe8Egg | 8,660 | Fix empty translations with declared languages | {
"login": "HuaTNA",
"id": 126794161,
"node_id": "U_kgDOB465sQ",
"avatar_url": "https://avatars.githubusercontent.com/u/126794161?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/HuaTNA",
"html_url": "https://github.com/HuaTNA",
"followers_url": "https://api.github.com/users/HuaTNA/follower... | [] | open | false | [] | null | 0 | 2026-09-21T20:30:05 | 2026-09-21T20:30:05 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8660",
"html_url": "https://github.com/huggingface/datasets/pull/8660",
"diff_url": "https://github.com/huggingface/datasets/pull/8660.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8660.patch",
"merged_at": null
} | Fixes #8659.
With declared languages, `TranslationVariableLanguages` raises a `ValueError` for `{}`, `{"en": []}`, and other inputs containing no translations. Return `{"language": [], "translation": []}` when normalization produces no translation pairs, after validating the language names.
Regression tests cover dir... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8660/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8660/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8659 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8659/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8659/comments | https://api.github.com/repos/huggingface/datasets/issues/8659/events | https://github.com/huggingface/datasets/issues/8659 | 5,529,142,676 | I_kwDODunzps8AAAABSZAFlA | 8,659 | TranslationVariableLanguages crashes on empty translations with declared languages | {
"login": "HuaTNA",
"id": 126794161,
"node_id": "U_kgDOB465sQ",
"avatar_url": "https://avatars.githubusercontent.com/u/126794161?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/HuaTNA",
"html_url": "https://github.com/HuaTNA",
"followers_url": "https://api.github.com/users/HuaTNA/follower... | [] | open | false | [] | null | 0 | 2026-09-21T15:38:25 | 2026-09-21T15:38:25 | null | null | NONE | [] | null | null | null | null | ## Describe the bug
With an explicit language list, `TranslationVariableLanguages.encode_example()`
raises `ValueError: not enough values to unpack (expected 2, got 0)` for `{}`,
`{"en": []}`, or `{"en": [], "fr": []}`. The same inputs also prevent
`Dataset.from_dict()` from constructing a dataset.
The equivalent alr... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8659/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8659/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | null | false |
https://api.github.com/repos/huggingface/datasets/issues/8658 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8658/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8658/comments | https://api.github.com/repos/huggingface/datasets/issues/8658/events | https://github.com/huggingface/datasets/pull/8658 | 5,528,884,336 | PR_kwDODunzps8AAAABEcT27w | 8,658 | Fix push_to_hub for IterableDataset with unknown features | {
"login": "davanstrien",
"id": 8995957,
"node_id": "MDQ6VXNlcjg5OTU5NTc=",
"avatar_url": "https://avatars.githubusercontent.com/u/8995957?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/davanstrien",
"html_url": "https://github.com/davanstrien",
"followers_url": "https://api.github.com/us... | [] | open | false | [] | null | 1 | 2026-09-21T15:15:52 | 2026-09-21T15:18:53 | null | null | MEMBER | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8658",
"html_url": "https://github.com/huggingface/datasets/pull/8658",
"diff_url": "https://github.com/huggingface/datasets/pull/8658.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8658.patch",
"merged_at": null
} | An IterableDataset can have unknown features (features=None), e.g. when built with IterableDataset.from_generator() without `features=`, or when streaming CSV files (or JSON files that require a full read) without declared features.
IterableDataset._push_parquet_shards_to_hub() then crashed at the decodable-columns ... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8658/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8658/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8657 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8657/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8657/comments | https://api.github.com/repos/huggingface/datasets/issues/8657/events | https://github.com/huggingface/datasets/pull/8657 | 5,528,122,157 | PR_kwDODunzps8AAAABEbsYug | 8,657 | [docs] Fix legacy hf CLI references (datasets) | {
"login": "Wauplin",
"id": 11801849,
"node_id": "MDQ6VXNlcjExODAxODQ5",
"avatar_url": "https://avatars.githubusercontent.com/u/11801849?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Wauplin",
"html_url": "https://github.com/Wauplin",
"followers_url": "https://api.github.com/users/Waupli... | [] | closed | false | [] | null | 1 | 2026-09-21T14:10:37 | 2026-09-21T16:12:43 | 2026-09-21T16:12:41 | null | CONTRIBUTOR | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8657",
"html_url": "https://github.com/huggingface/datasets/pull/8657",
"diff_url": "https://github.com/huggingface/datasets/pull/8657.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8657.patch",
"merged_at": "2026-09-21T16:12... | The docs and a few docstrings still use commands that are no longer the recommended `huggingface_hub` usage in 1.x: `hf` replaced `huggingface-cli`, `hf repos` replaced `hf repo`, and `--organization` was replaced by a namespaced repo id.
- `docs/source/share.mdx`: `huggingface-cli login` → `hf auth login`, `huggingfa... | {
"login": "lhoestq",
"id": 42851186,
"node_id": "MDQ6VXNlcjQyODUxMTg2",
"avatar_url": "https://avatars.githubusercontent.com/u/42851186?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lhoestq",
"html_url": "https://github.com/lhoestq",
"followers_url": "https://api.github.com/users/lhoest... | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8657/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8657/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8656 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8656/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8656/comments | https://api.github.com/repos/huggingface/datasets/issues/8656/events | https://github.com/huggingface/datasets/pull/8656 | 5,527,926,378 | PR_kwDODunzps8AAAABEbicEA | 8,656 | Fix EXIF orientation for in-memory PIL image inputs | {
"login": "Yang1107-wzy",
"id": 193335484,
"node_id": "U_kgDOC4YQvA",
"avatar_url": "https://avatars.githubusercontent.com/u/193335484?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Yang1107-wzy",
"html_url": "https://github.com/Yang1107-wzy",
"followers_url": "https://api.github.com/use... | [] | open | false | [] | null | 0 | 2026-09-21T13:54:11 | 2026-09-21T13:54:11 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8656",
"html_url": "https://github.com/huggingface/datasets/pull/8656",
"diff_url": "https://github.com/huggingface/datasets/pull/8656.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8656.patch",
"merged_at": null
} | Serializing an in-memory PIL image through `Image` drops its EXIF orientation before decoding can apply the transpose added in #6739. The same JPEG with orientation 6 is decoded as `(4, 8)` when supplied as bytes, but `(8, 4)` when supplied as `PIL.Image.open(BytesIO(...))`. This also reproduces on the 5.0.1 release.
... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8656/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8656/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8655 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8655/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8655/comments | https://api.github.com/repos/huggingface/datasets/issues/8655/events | https://github.com/huggingface/datasets/pull/8655 | 5,527,145,904 | PR_kwDODunzps8AAAABEa6MSg | 8,655 | Format a zero size as "0 bytes" instead of "Unknown size" | {
"login": "Rodrigo-Palma",
"id": 86628053,
"node_id": "MDQ6VXNlcjg2NjI4MDUz",
"avatar_url": "https://avatars.githubusercontent.com/u/86628053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rodrigo-Palma",
"html_url": "https://github.com/Rodrigo-Palma",
"followers_url": "https://api.githu... | [] | open | false | [] | null | 0 | 2026-09-21T12:44:49 | 2026-09-21T12:44:49 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8655",
"html_url": "https://github.com/huggingface/datasets/pull/8655",
"diff_url": "https://github.com/huggingface/datasets/pull/8655.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8655.patch",
"merged_at": null
} | `size_str()` documents that it returns `"Unknown size"` when `size_in_bytes` is `None`:
```python
def size_str(size_in_bytes):
"""Returns a human readable size string.
If size_in_bytes is None, then returns "Unknown size".
...
"""
if not size_in_bytes:
return "Unknown size"
```
The test i... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8655/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8655/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8654 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8654/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8654/comments | https://api.github.com/repos/huggingface/datasets/issues/8654/events | https://github.com/huggingface/datasets/pull/8654 | 5,526,112,669 | PR_kwDODunzps8AAAABEaGD5g | 8,654 | Fix mutual exclusion for long lock filenames across processes | {
"login": "Yang1107-wzy",
"id": 193335484,
"node_id": "U_kgDOC4YQvA",
"avatar_url": "https://avatars.githubusercontent.com/u/193335484?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Yang1107-wzy",
"html_url": "https://github.com/Yang1107-wzy",
"followers_url": "https://api.github.com/use... | [] | open | false | [] | null | 0 | 2026-09-21T11:04:31 | 2026-09-21T11:18:46 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8654",
"html_url": "https://github.com/huggingface/datasets/pull/8654",
"diff_url": "https://github.com/huggingface/datasets/pull/8654.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8654.patch",
"merged_at": null
} | Fixes #8653.
Long lock filenames are shortened with Python's process-randomized string hash. Independent processes can therefore acquire different lock files for the same original path, bypassing mutual exclusion. This can affect the cache-directory lock in `DatasetBuilder`, where the directory path is flattened into ... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8654/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8654/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8653 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8653/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8653/comments | https://api.github.com/repos/huggingface/datasets/issues/8653/events | https://github.com/huggingface/datasets/issues/8653 | 5,526,093,910 | I_kwDODunzps8AAAABSWGAVg | 8,653 | Long lock filenames do not provide mutual exclusion across Python processes | {
"login": "Yang1107-wzy",
"id": 193335484,
"node_id": "U_kgDOC4YQvA",
"avatar_url": "https://avatars.githubusercontent.com/u/193335484?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Yang1107-wzy",
"html_url": "https://github.com/Yang1107-wzy",
"followers_url": "https://api.github.com/use... | [] | open | false | [] | null | 0 | 2026-09-21T11:02:34 | 2026-09-21T11:02:34 | null | null | NONE | [] | null | null | null | null | ### Describe the bug
`FileLock.hash_filename_if_too_long()` uses Python's `hash(filename)` to shorten long lock filenames. String hashes vary between independently started Python processes, so two processes given the same long lock path can lock different files and both enter the protected section.
This affects paths... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8653/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8653/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | null | false |
https://api.github.com/repos/huggingface/datasets/issues/8652 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8652/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8652/comments | https://api.github.com/repos/huggingface/datasets/issues/8652/events | https://github.com/huggingface/datasets/pull/8652 | 5,522,884,515 | PR_kwDODunzps8AAAABEXfrSQ | 8,652 | Read JSON floats exactly in the pandas fallback | {
"login": "Linxiushen",
"id": 142557582,
"node_id": "U_kgDOCH9Bjg",
"avatar_url": "https://avatars.githubusercontent.com/u/142557582?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Linxiushen",
"html_url": "https://github.com/Linxiushen",
"followers_url": "https://api.github.com/users/Lin... | [] | open | false | [] | null | 0 | 2026-09-21T04:49:51 | 2026-09-21T04:49:51 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8652",
"html_url": "https://github.com/huggingface/datasets/pull/8652",
"diff_url": "https://github.com/huggingface/datasets/pull/8652.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8652.patch",
"merged_at": null
} | `pandas_read_json()` in `packaged_modules/json/json.py` calls `pd.read_json()` with pandas' default float parser, which is fast but not round-trip exact:
```python
>>> import io, pandas as pd
>>> pd.read_json(io.StringIO('{"v": 0.7}\n'), lines=True)["v"][0]
np.float64(0.7000000000000001)
>>> pd.read_json(io.StringIO('... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8652/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8652/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8651 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8651/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8651/comments | https://api.github.com/repos/huggingface/datasets/issues/8651/events | https://github.com/huggingface/datasets/pull/8651 | 5,521,224,562 | PR_kwDODunzps8AAAABEWMWvw | 8,651 | Fix Python 3.14 dict pickling compatibility | {
"login": "FranciscoMaxwell",
"id": 79481236,
"node_id": "MDQ6VXNlcjc5NDgxMjM2",
"avatar_url": "https://avatars.githubusercontent.com/u/79481236?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/FranciscoMaxwell",
"html_url": "https://github.com/FranciscoMaxwell",
"followers_url": "https://... | [] | open | false | [] | null | 0 | 2026-09-20T23:45:34 | 2026-09-20T23:45:34 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8651",
"html_url": "https://github.com/huggingface/datasets/pull/8651",
"diff_url": "https://github.com/huggingface/datasets/pull/8651.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8651.patch",
"merged_at": null
} | Closes #8373.
## Summary
- passes the new Python 3.14 `_batch_setitems` object argument when the custom datasets pickler is called through the old two-argument path
- keeps the existing key sorting behavior for stable fingerprints
- adds a regression test that simulates the Python 3.14 parent pickler signature
... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8651/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8651/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8650 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8650/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8650/comments | https://api.github.com/repos/huggingface/datasets/issues/8650/events | https://github.com/huggingface/datasets/pull/8650 | 5,519,644,209 | PR_kwDODunzps8AAAABEU_uIg | 8,650 | Fix remote dataset loads after earlier downloads | {
"login": "1fanwang",
"id": 21269208,
"node_id": "MDQ6VXNlcjIxMjY5MjA4",
"avatar_url": "https://avatars.githubusercontent.com/u/21269208?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/1fanwang",
"html_url": "https://github.com/1fanwang",
"followers_url": "https://api.github.com/users/1fa... | [] | open | false | [] | null | 0 | 2026-09-20T18:24:16 | 2026-09-20T18:24:16 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8650",
"html_url": "https://github.com/huggingface/datasets/pull/8650",
"diff_url": "https://github.com/huggingface/datasets/pull/8650.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8650.patch",
"merged_at": null
} | ## Summary
When datasets are stored in nested remote directories, loading a child directory first can make loading its parent fail with `FileNotFoundError` for `state.json`. Reloading a remote dataset after its contents change can also return the previous rows.
Each load now gets a fresh directory inside the managed ... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8650/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8650/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8649 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8649/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8649/comments | https://api.github.com/repos/huggingface/datasets/issues/8649/events | https://github.com/huggingface/datasets/pull/8649 | 5,518,545,448 | PR_kwDODunzps8AAAABEUKrGg | 8,649 | Apply the bit-to-byte conversion to binary units too | {
"login": "Rodrigo-Palma",
"id": 86628053,
"node_id": "MDQ6VXNlcjg2NjI4MDUz",
"avatar_url": "https://avatars.githubusercontent.com/u/86628053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Rodrigo-Palma",
"html_url": "https://github.com/Rodrigo-Palma",
"followers_url": "https://api.githu... | [] | open | false | [] | null | 1 | 2026-09-20T14:54:17 | 2026-09-21T12:59:51 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8649",
"html_url": "https://github.com/huggingface/datasets/pull/8649",
"diff_url": "https://github.com/huggingface/datasets/pull/8649.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8649.patch",
"merged_at": null
} | `convert_file_size_to_int` already distinguishes bits from bytes: a unit ending in a lowercase `b` is read as bits and divided by 8. That check is only applied to the decimal branches, though, so the binary ones return a bit size as if it were bytes, eight times larger than asked for:
```python
>>> from datasets.utils... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8649/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8649/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8648 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8648/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8648/comments | https://api.github.com/repos/huggingface/datasets/issues/8648/events | https://github.com/huggingface/datasets/pull/8648 | 5,517,849,108 | PR_kwDODunzps8AAAABETo7Jg | 8,648 | Pin UTF-8 encoding when reading eval sample logs | {
"login": "dltsum",
"id": 196378929,
"node_id": "U_kgDOC7SBMQ",
"avatar_url": "https://avatars.githubusercontent.com/u/196378929?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dltsum",
"html_url": "https://github.com/dltsum",
"followers_url": "https://api.github.com/users/dltsum/follower... | [] | open | false | [] | null | 0 | 2026-09-20T12:37:44 | 2026-09-20T12:37:44 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8648",
"html_url": "https://github.com/huggingface/datasets/pull/8648",
"diff_url": "https://github.com/huggingface/datasets/pull/8648.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8648.patch",
"merged_at": null
} | ## Problem
The `eval` packaged module reads sample log JSON files with the locale's preferred encoding:
```python
for sample_file in sample_files:
with open(sample_file) as f:
sample = json.load(f)
```
On Windows with a non-UTF-8 locale (e.g. cp936/GBK), any sample containing non-ASCII text — CJK prompts... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8648/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8648/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8647 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8647/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8647/comments | https://api.github.com/repos/huggingface/datasets/issues/8647/events | https://github.com/huggingface/datasets/pull/8647 | 5,517,383,587 | PR_kwDODunzps8AAAABETSp6g | 8,647 | Fix IterableDataset.filter() with no function | {
"login": "YaoxinHuang",
"id": 174019195,
"node_id": "U_kgDOCl9Sew",
"avatar_url": "https://avatars.githubusercontent.com/u/174019195?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YaoxinHuang",
"html_url": "https://github.com/YaoxinHuang",
"followers_url": "https://api.github.com/users/... | [] | open | false | [] | null | 0 | 2026-09-20T10:54:15 | 2026-09-20T15:48:40 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8647",
"html_url": "https://github.com/huggingface/datasets/pull/8647",
"diff_url": "https://github.com/huggingface/datasets/pull/8647.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8647.patch",
"merged_at": null
} | `list(dataset.filter())` raises `TypeError` because the iterator tries to call `None`. The documented default is to keep all rows.
Return a new dataset wrapper when no function is supplied. Returning a new object, rather than `self`, keeps changes to its epoch from affecting the input dataset.
Tests cover the def... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8647/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8647/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8646 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8646/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8646/comments | https://api.github.com/repos/huggingface/datasets/issues/8646/events | https://github.com/huggingface/datasets/pull/8646 | 5,515,453,129 | PR_kwDODunzps8AAAABER02Gw | 8,646 | Fix map(drop_last_batch=True) returning the input unchanged when all batches are dropped | {
"login": "SIDDARTHAREDDY8",
"id": 75976672,
"node_id": "MDQ6VXNlcjc1OTc2Njcy",
"avatar_url": "https://avatars.githubusercontent.com/u/75976672?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SIDDARTHAREDDY8",
"html_url": "https://github.com/SIDDARTHAREDDY8",
"followers_url": "https://api... | [] | open | false | [] | null | 2 | 2026-09-20T03:18:53 | 2026-09-21T03:25:54 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8646",
"html_url": "https://github.com/huggingface/datasets/pull/8646",
"diff_url": "https://github.com/huggingface/datasets/pull/8646.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8646.patch",
"merged_at": null
} | Fixes #8386.
When `drop_last_batch=True` and a shard is smaller than `batch_size`, `_map_single` iterates zero batches, so the map function never runs and `update_data` stays `None`. The final branch then yielded the original shard back unchanged (3 rows instead of 0), disagreeing with `IterableDataset.batch` and `Dat... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8646/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8646/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8645 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8645/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8645/comments | https://api.github.com/repos/huggingface/datasets/issues/8645/events | https://github.com/huggingface/datasets/pull/8645 | 5,507,080,140 | PR_kwDODunzps8AAAABELOooA | 8,645 | Support class encoding list-valued columns | {
"login": "lindicaphxag-tech",
"id": 237745757,
"node_id": "U_kgDODiu2XQ",
"avatar_url": "https://avatars.githubusercontent.com/u/237745757?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lindicaphxag-tech",
"html_url": "https://github.com/lindicaphxag-tech",
"followers_url": "https://api... | [] | open | false | [] | null | 0 | 2026-09-19T02:11:35 | 2026-09-19T02:11:35 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8645",
"html_url": "https://github.com/huggingface/datasets/pull/8645",
"diff_url": "https://github.com/huggingface/datasets/pull/8645.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8645.patch",
"merged_at": null
} | Fixes #6267
class_encode_column currently rejects list-valued columns even when the feature is List(Value(...)). That makes multi-label datasets fall back to manually collecting labels and calling cast_column.
This change accepts List(Value(...)) columns, collects class names from the list elements, and maps each ele... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8645/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8645/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8644 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8644/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8644/comments | https://api.github.com/repos/huggingface/datasets/issues/8644/events | https://github.com/huggingface/datasets/pull/8644 | 5,500,093,871 | PR_kwDODunzps8AAAABEFgEDA | 8,644 | fix(py_utils): save/restore NPU RNG state in temp_seed | {
"login": "li-lizhe",
"id": 82501526,
"node_id": "MDQ6VXNlcjgyNTAxNTI2",
"avatar_url": "https://avatars.githubusercontent.com/u/82501526?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/li-lizhe",
"html_url": "https://github.com/li-lizhe",
"followers_url": "https://api.github.com/users/li-... | [] | open | false | [] | null | 2 | 2026-09-18T13:24:23 | 2026-09-19T23:45:43 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8644",
"html_url": "https://github.com/huggingface/datasets/pull/8644",
"diff_url": "https://github.com/huggingface/datasets/pull/8644.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8644.patch",
"merged_at": null
} | ## Description
When `set_pytorch=True` is passed to `temp_seed`, the context manager saves and restores the PyTorch CPU and CUDA RNG states. On non-CUDA accelerators (e.g. Ascend NPU), the device RNG state is silently left unseeded during the context, breaking reproducibility across seeded dataset transforms.
This PR... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8644/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8644/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8643 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8643/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8643/comments | https://api.github.com/repos/huggingface/datasets/issues/8643/events | https://github.com/huggingface/datasets/pull/8643 | 5,496,656,185 | PR_kwDODunzps8AAAABECvV0A | 8,643 | Forward Dataset.cast arguments from DatasetDict.cast | {
"login": "RizgarOzan",
"id": 236759479,
"node_id": "U_kgDODhyptw",
"avatar_url": "https://avatars.githubusercontent.com/u/236759479?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RizgarOzan",
"html_url": "https://github.com/RizgarOzan",
"followers_url": "https://api.github.com/users/Riz... | [] | open | false | [] | null | 0 | 2026-09-18T06:57:01 | 2026-09-18T06:57:01 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8643",
"html_url": "https://github.com/huggingface/datasets/pull/8643",
"diff_url": "https://github.com/huggingface/datasets/pull/8643.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8643.patch",
"merged_at": null
} | ## What
`DatasetDict.cast` now accepts the same processing arguments as `Dataset.cast` (`batch_size`, `keep_in_memory`, `load_from_cache_file`, `writer_batch_size`, `num_proc`) plus `cache_file_names` as a per-split dict, and forwards them to each split.
## Why
`Dataset.cast` runs through `map` and exposes these kno... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8643/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8643/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8642 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8642/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8642/comments | https://api.github.com/repos/huggingface/datasets/issues/8642/events | https://github.com/huggingface/datasets/pull/8642 | 5,494,902,554 | PR_kwDODunzps8AAAABEBWK3A | 8,642 | Do not rebuild a finalized ArrowWriter, so a failing stream close is not masked | {
"login": "antnewman",
"id": 13312614,
"node_id": "MDQ6VXNlcjEzMzEyNjE0",
"avatar_url": "https://avatars.githubusercontent.com/u/13312614?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/antnewman",
"html_url": "https://github.com/antnewman",
"followers_url": "https://api.github.com/users/... | [] | open | false | [] | null | 0 | 2026-09-18T02:23:58 | 2026-09-18T02:23:58 | null | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8642",
"html_url": "https://github.com/huggingface/datasets/pull/8642",
"diff_url": "https://github.com/huggingface/datasets/pull/8642.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8642.patch",
"merged_at": null
} | Follows the discussion on #8558. Separate PR because that one fixes call sites in the builders and this fixes the writer itself, so they can be reviewed or reverted independently. @lhoestq happy to fold them together if you would rather have one change.
Line numbers below are against `main`.
## The problem
`finalize... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8642/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8642/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8641 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8641/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8641/comments | https://api.github.com/repos/huggingface/datasets/issues/8641/events | https://github.com/huggingface/datasets/pull/8641 | 5,493,864,489 | PR_kwDODunzps8AAAABEAg3vQ | 8,641 | Convert a dense ArrayXD column in one pass instead of per row | {
"login": "behroozazarkhalili",
"id": 80390531,
"node_id": "MDQ6VXNlcjgwMzkwNTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/80390531?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/behroozazarkhalili",
"html_url": "https://github.com/behroozazarkhalili",
"followers_url": "ht... | [] | open | false | [] | null | 0 | 2026-09-17T23:53:08 | 2026-09-17T23:53:08 | null | null | CONTRIBUTOR | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8641",
"html_url": "https://github.com/huggingface/datasets/pull/8641",
"diff_url": "https://github.com/huggingface/datasets/pull/8641.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8641.patch",
"merged_at": null
} | Towards #4964. This reduces the construction overhead rather than removing it, and the numbers below say by how much.
### What is slow
Building a `Dataset` from a whole ndarray converts it row by row and then concatenates. On this machine, 3200 rows of 64x64 float32, a 50 MB payload:
| | peak RSS over baseline | rat... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8641/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8641/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8640 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8640/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8640/comments | https://api.github.com/repos/huggingface/datasets/issues/8640/events | https://github.com/huggingface/datasets/pull/8640 | 5,493,466,279 | PR_kwDODunzps8AAAABEAM5bg | 8,640 | Name the column in map() type mismatch errors and restore the overflow hint | {
"login": "behroozazarkhalili",
"id": 80390531,
"node_id": "MDQ6VXNlcjgwMzkwNTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/80390531?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/behroozazarkhalili",
"html_url": "https://github.com/behroozazarkhalili",
"followers_url": "ht... | [] | open | false | [] | null | 0 | 2026-09-17T22:47:22 | 2026-09-17T23:24:34 | null | null | CONTRIBUTOR | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8640",
"html_url": "https://github.com/huggingface/datasets/pull/8640",
"diff_url": "https://github.com/huggingface/datasets/pull/8640.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8640.patch",
"merged_at": null
} | Closes #4352.
### The problem
When a `map()` function returns values that do not match the declared features, the error never says which column is at fault. On a wide dataset that leaves you bisecting your own function to find it.
Measured on `main`, all four mismatch shapes:
| case | before |
|---|---|
| declared ... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8640/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8640/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8639 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8639/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8639/comments | https://api.github.com/repos/huggingface/datasets/issues/8639/events | https://github.com/huggingface/datasets/pull/8639 | 5,493,344,146 | PR_kwDODunzps8AAAABEAGwkQ | 8,639 | Document the formatter dtype coercion and pin it with tests | {
"login": "behroozazarkhalili",
"id": 80390531,
"node_id": "MDQ6VXNlcjgwMzkwNTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/80390531?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/behroozazarkhalili",
"html_url": "https://github.com/behroozazarkhalili",
"followers_url": "ht... | [] | open | false | [] | null | 0 | 2026-09-17T22:28:49 | 2026-09-17T22:28:49 | null | null | CONTRIBUTOR | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8639",
"html_url": "https://github.com/huggingface/datasets/pull/8639",
"diff_url": "https://github.com/huggingface/datasets/pull/8639.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8639.patch",
"merged_at": null
} | Closes #5517.
### What users hit
A column stored as `float64` comes back as `float32`:
```python
>>> ds = Dataset.from_dict({"v": [[3.141592653589793, 1e300]]},
... features=Features({"v": List(Value("float64"))}))
>>> ds.with_format("numpy")[0]["v"]
array([3.1415927e+00, inf], dtype... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8639/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8639/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8638 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8638/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8638/comments | https://api.github.com/repos/huggingface/datasets/issues/8638/events | https://github.com/huggingface/datasets/pull/8638 | 5,492,212,128 | PR_kwDODunzps8AAAABD_MSeA | 8,638 | Add HF_DATASETS_USE_SOFT_FILELOCK for filesystems without OS-level locking | {
"login": "behroozazarkhalili",
"id": 80390531,
"node_id": "MDQ6VXNlcjgwMzkwNTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/80390531?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/behroozazarkhalili",
"html_url": "https://github.com/behroozazarkhalili",
"followers_url": "ht... | [] | open | false | [] | null | 0 | 2026-09-17T20:09:58 | 2026-09-17T21:03:59 | null | null | CONTRIBUTOR | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8638",
"html_url": "https://github.com/huggingface/datasets/pull/8638",
"diff_url": "https://github.com/huggingface/datasets/pull/8638.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8638.patch",
"merged_at": null
} | Closes #6744.
### The problem
`load_dataset` takes cache locks through `filelock`, which uses `flock`/`fcntl`. Some network filesystems do not implement those calls, so the lock fails and the dataset never loads. The reporter hit this on an NFS-mounted Weka mount. Moving `cache_dir` to local disk is not a way out, be... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8638/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8638/timeline | null | null | null | null | null | true |
https://api.github.com/repos/huggingface/datasets/issues/8637 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8637/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8637/comments | https://api.github.com/repos/huggingface/datasets/issues/8637/events | https://github.com/huggingface/datasets/issues/8637 | 5,492,155,234 | I_kwDODunzps8AAAABR1ujYg | 8,637 | FileLock wrapper drops timeout and every other constructor argument | {
"login": "behroozazarkhalili",
"id": 80390531,
"node_id": "MDQ6VXNlcjgwMzkwNTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/80390531?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/behroozazarkhalili",
"html_url": "https://github.com/behroozazarkhalili",
"followers_url": "ht... | [] | open | false | [] | null | 1 | 2026-09-17T20:04:06 | 2026-09-17T21:04:15 | null | null | CONTRIBUTOR | [] | null | null | null | null | ### Describe the bug
`datasets.utils._filelock.FileLock` silently discards every keyword argument except `lock_file`. Passing `timeout=` has no effect, and `is_singleton=True` raises instead of returning the cached lock.
The wrapper declares a variadic signature:
```python
# src/datasets/utils/_filelock.py
def __ini... | null | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8637/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8637/timeline | null | null | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | {
"blocked_by": 0,
"total_blocked_by": 0,
"blocking": 0,
"total_blocking": 0
} | null | false |
https://api.github.com/repos/huggingface/datasets/issues/8636 | https://api.github.com/repos/huggingface/datasets | https://api.github.com/repos/huggingface/datasets/issues/8636/labels{/name} | https://api.github.com/repos/huggingface/datasets/issues/8636/comments | https://api.github.com/repos/huggingface/datasets/issues/8636/events | https://github.com/huggingface/datasets/pull/8636 | 5,485,310,470 | PR_kwDODunzps8AAAABD5lZ2g | 8,636 | fix(ci): harden GitHub Actions workflows (#8635) | {
"login": "hf-security-analysis[bot]",
"id": 265538906,
"node_id": "BOT_kgDOD9PNWg",
"avatar_url": "https://avatars.githubusercontent.com/u/25720743?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hf-security-analysis%5Bbot%5D",
"html_url": "https://github.com/apps/hf-security-analysis",
... | [] | closed | false | [] | null | 1 | 2026-09-17T09:08:15 | 2026-09-21T09:50:14 | 2026-09-21T09:50:12 | null | NONE | [] | null | null | false | {
"url": "https://api.github.com/repos/huggingface/datasets/pulls/8636",
"html_url": "https://github.com/huggingface/datasets/pull/8636",
"diff_url": "https://github.com/huggingface/datasets/pull/8636.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/8636.patch",
"merged_at": "2026-09-21T09:50... | Automated hardening of the workflow files flagged on #8635.
> [!WARNING]
> **This narrows what the workflow can reach.** Job permissions were declared in `.github/workflows/ci.yml`, `.github/workflows/release-conda.yml`. Each job now gets only the scopes its steps were read to need — if one of them does something this... | {
"login": "paulinebm",
"id": 155966238,
"node_id": "U_kgDOCUvbHg",
"avatar_url": "https://avatars.githubusercontent.com/u/155966238?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/paulinebm",
"html_url": "https://github.com/paulinebm",
"followers_url": "https://api.github.com/users/paulin... | {
"url": "https://api.github.com/repos/huggingface/datasets/issues/8636/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/datasets/issues/8636/timeline | null | null | null | null | null | true |
End of preview. Expand in Data Studio
- Downloads last month
- 38