Releases: iterative/datachain
Releases · iterative/datachain
0.7.2
What's Changed
- feat(dc): make expand work if some fields are optional by @shcheklein in #607
- fix(tests): minor fix in comment by @shcheklein in #618
- test_query_e2e: do not interrupt running script multiple times by @skshetry in #614
- sqlitedb: retain
db_file
in_connect
method to ensure proper connection ondb.clone()
by @skshetry in #620 - [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #623
- Bump astral-sh/setup-uv from 3 to 4 by @dependabot in #625
- dataset_select_paginated: do not clone on each paginated query by @skshetry in #629
- docs: document prefetch by @skshetry in #630
- fix type issues produced by new version of pydantic by @mattseddon in #638
- Finish SQL functions refactoring by @dreadatour in #543
Full Changelog: 0.7.1...0.7.2
0.7.1
What's Changed
- Increase resolution of
train_test_split
by @0x2b3bfa0 in #611 - Update ultralytics models (take two) by @dreadatour in #613
Full Changelog: 0.7.0...0.7.1
0.7.0
What's Changed
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #608
- Bump codecov/codecov-action from 4 to 5 by @dependabot in #609
- Implement pre-fetching in map() and gen() by @rlamy in #521
Full Changelog: 0.6.11...0.7.0
0.6.11
What's Changed
- Fixing np array dumping by @ilongin in #599
- Revert "Update ultralytics models" by @shcheklein in #610
Full Changelog: 0.6.10...0.6.11
0.6.10
What's Changed
- Fix for instantiation of pulled dataset by @ilongin in #573
- update pytest-servers to 0.5.8 by @mattseddon in #596
- Fix 'limit' + 'offset' methods for SQLite warehouse by @dreadatour in #595
- docs: move overrides inside docs/ dir by @skshetry in #601
- feat(dc): add explode function by @shcheklein in #581
- Split Dataset into train/test/val by @dreadatour in #604
- Update ultralytics models by @dreadatour in #592
Full Changelog: 0.6.9...0.6.10
0.6.9
What's Changed
- Update README.rst, added missing preposition. by @gromag in #583
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #582
- readme next by @dmpetrov in #584
- doc(agg): add an example with custom UDF spec by @shcheklein in #586
- fix(dc): collect function return types by @shcheklein in #589
- Adding
uuid
to dataset version by @ilongin in #587
New Contributors
Full Changelog: 0.6.8...0.6.9
0.6.8
What's Changed
- pin version of clickhouse to 24.6 (match clickhouse-cloud) by @mattseddon in #570
- feat(cli): enhance dataset listing with additional options for studio by @amritghimire in #561
- Add 'BoundingBox', 'Pose2D' and 'Pose3D' models by @dreadatour in #557
Full Changelog: 0.6.7...0.6.8
0.6.7
What's Changed
- Make main default branch for CI action by @amritghimire in #569
- Dataset pull fixes by @ilongin in #560
Full Changelog: 0.6.6...0.6.7
0.6.6
What's Changed
- clarify semantics in order_by docstring by @mattseddon in #518
- chore(hf): remove unused arguments by @shcheklein in #556
- fix(tests): hf test can be flakey if runs in parallel by @shcheklein in #559
- [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #562
- feat: setup plausible by @yathomasi in #565
- guarantee order if order_by directly precedes collect in a chain by @mattseddon in #558
- Support for
UInt32
by @ilongin in #568
Full Changelog: 0.6.5...0.6.6
0.6.5
What's Changed
- fix(hf): use proper source when we create a file entry by @shcheklein in #555
Full Changelog: 0.6.4...0.6.5