Releases: activeloopai/deeplake
Releases · activeloopai/deeplake
v2.8.3 🌈
🧭 What's Changed
- Fixes type mismatch for expiration(#1858) @levongh
- Flag to disable wandb integration (#1863) @farizrahman4u
- Fixes wandb+local datasets (#1861) @hakanardo
- [DL-668] Make pytorch() work with views (#1855) @farizrahman4u
- [AL-1949] Make experimental pytorch dataloader consistent with existing implementation (#1853) @AbhinavTuli
- [DL-650] Better error handling when not passing a tensor name to ds.append (#1817) @adolkhan
- Update docs URL in readme (#1857) @FayazRahman
- Speedup conversion of hub storage datasets->deeplake for experimental features (#1856) @levongh
- [DL-611] New API reference (#1830) @FayazRahman
- Wandb update: report datasets created with deepcopy (#1848) @farizrahman4u
- [Bugfix] 1828 raising UserNotLoggedInException when invalid path is provided (#1829) @adolkhan
- [DL-655] Added min and max length options (#1841) @adolkhan
⚙️ Who Contributes
@AbhinavTuli, @FayazRahman, @adolkhan, @davidbuniat, @farizrahman4u, @hakanardo and @levongh
v2.8.1 🌈
🧭 What's Changed
- Ensure that new format for chunk id isn't used for encoders with version <= 2.7.6 (#1850) @AbhinavTuli
⚙️ Who Contributes
v2.8.0 🌈
🧭 What's Changed
- Release Candidate 0 for new experimental dataloader and queries (#1819) @AbhinavTuli
- [AL-1946] Fix delete group + reset bug (#1843) @AbhinavTuli
- [DL-652] Add append_empty arg to ds.append (#1846) @farizrahman4u
- Avoid printing syncing labels message when no labels were added (#1845) @FayazRahman
- [DL-684] Fix ds.reset bug with local datasets (#1842) @FayazRahman
- Use staging visualizer in tests. Correct dev visualizer url. (#1838) @khustup
- Changes default chunk id size to 8 bits from 4 bits to reduce possibility of collisions (#1835) @AbhinavTuli
- wandb integration (#1739) @farizrahman4u
⚙️ Who Contributes
v2.7.5 🌈
🧭 What's Changed
- [AL-1775] Point Cloud htype (#1685) @adolkhan
- [AL-1912] Don't allow generic htypes with link (#1824) @AbhinavTuli
- [Bugfix] Fixes rechunking with hub link + cloud paths (#1825) @AbhinavTuli
- Enable progressbar for syncing labels (#1820) @FayazRahman
- [Bug fix] Ensure None/"ENV" isn't added to used_creds_keys for linked data (#1823) @AbhinavTuli
⚙️ Who Contributes
v2.7.4 🌈
🧭 What's Changed
- Fix get_incompatible_dtype bug (#1814) @farizrahman4u
- [AL-1888] Enable rechunking for text like htypes (#1815) @AbhinavTuli
- [AL-1858] Treat empty list as None (#1813) @AbhinavTuli
- Older reporting configurations were not properly handling username (#1806) @zomglings
⚙️ Who Contributes
v2.7.3 🌈
🧭 What's Changed
- [AL-1884] Fixes bug with ds.reset for newly added/deleted tensors (#1797) @AbhinavTuli
- [DL-618] Appending to class labels with text using multiple workers (#1794) @FayazRahman
- [AL-1848] New agreements handling (#1796) @AbhinavTuli
- [DL-590] S3: Always show retry warnings (#1807) @farizrahman4u
- [DL-620] Prevent saving of dataset views for public datasets when user is not logged in (#1803) @farizrahman4u
⚙️ Who Contributes
v2.7.2 🌈
🧭 What's Changed
- [DL-593] Bugout correctly identifying the user's username when tokens are used (#1792) @adolkhan
- Fix double indexing when saving strided views (#1793) @farizrahman4u
🚀 New
- Gcp support for connected datasets (#1736) @ProgerDav
⚙️ Who Contributes
v2.7.1 🌈
🧭 What's Changed
- [AL-1855] Adds support for JWT token while reading from http urls (#1780) @AbhinavTuli
- Fix visualizer links for local views (#1791) @FayazRahman
- Unhide auto and util submodules in docs (#1790) @FayazRahman
- [AL-1813] Clarify whether a user is logged-in or not in authentication/permission errors (#1750) @adolkhan
- Fix linked video playback test (#1789) @FayazRahman
⚙️ Who Contributes
v2.7.0 🌈
🧭 What's Changed
- Disable locking for optimized views + new view urls (#1788) @farizrahman4u
- Optimize data loading for fixed shape tensors (#1734) @AbhinavTuli
- [AL-1809] refactor .data (#1769) @adolkhan
- Fixes handling of empty tensors in pytorch (#1786) @AbhinavTuli
- Make pbar on shuffle buffer filling readable (#1787) @farizrahman4u
- Enable parallelization for view optimization (#1785) @farizrahman4u
- Allow passing creds explicitly for external views (#1783) @farizrahman4u
- [AL-1842] Update API reference for dataset views (#1781) @FayazRahman
- [Bug] Fixes bug with links and compute (#1778) @AbhinavTuli
- Adds ability to pop sample from dataset instead of popping from individual tensors (#1776) @AbhinavTuli
- [AL-1811] - empty tensor error handling (#1741) @adolkhan
- [AL-1816] Make fetching multiple images smarter (#1744) @adolkhan
- [AL-1851][AL-1852][AL-1853][AL-1854] Allow plat to provide creds explicitly while saving views to user dir (#1775) @farizrahman4u
- [Improvement] Retries in s3 get_object_from_full_url (#1777) @AbhinavTuli
- [AL-1817][AL-1842] API reference for ViewEntry class (#1773) @FayazRahman
- View fixes (#1774) @FayazRahman
- [BUGFIX] Fix double indexing when fetching shapes of samples in views (#1772) @farizrahman4u
- Fixes to linked tensors, partial reads (#1770) @AbhinavTuli
- Allow platform to specify username for saving queries against public datasets (#1771) @farizrahman4u
- [AL-1817][AL-1482] API reference updates (#1768) @FayazRahman
- [AL-1807][AL-1844][AL-1845][AL-1849] Dataset view updates (#1764) @farizrahman4u
- [API] Added example for changing credential management (#1766) @AbhinavTuli
- [BUGFIX] Make linked video test work without google creds (#1765) @FayazRahman
- Changed dev location to https://app-dev.activeloop.dev/ (#1767) @timfox456
- [Bug fix] Fixes tensor meta is_link fast forwarding (#1763) @AbhinavTuli
- [AL-1738] Adds ability to delete sample (#1673) @AbhinavTuli
⚙️ Who Contributes
@AbhinavTuli, @FayazRahman, @adolkhan, @davidbuniat, @farizrahman4u and @timfox456
v2.6.0 🌈
🧭 What's Changed
- Change default max chunk size (#1759) @farizrahman4u
- Removes sequence test (#1762) @AbhinavTuli
- [AL-1818] Adds ability to pad tensors in pytorch and compute (#1751) @AbhinavTuli
- Enable random assignment by default (#1742) @farizrahman4u
- deepcopy + s3 fixes (#1749) @farizrahman4u
- [AL-1748] Show a progress bar while pytorch shuffle buffer is getting filled (#1745) @AbhinavTuli
- Update class_labels with text (#1756) @FayazRahman
- Update README.md (#1755) @istranic
- Fixes a bug with Version control + Info modification (#1753) @AbhinavTuli
- [AL-1780] [Small] Add support for pathlib.Path arguments (#1683) @adolkhan
- [BUGFIX] Fix appending to nested class_label tensors with text (#1743) @FayazRahman
- Skip google drive root test if no --gdrive flag (#1747) @FayazRahman
- [AL-1781] Reduce duration of CI tests (#1713) @FayazRahman
- Fix linked video tests (#1740) @FayazRahman
- [AL-1785] Changes to linked tensor credential management + other fixes (#1726) @AbhinavTuli
- Prevent changing logging configuration (#1738) @FayazRahman
- [AL-1822] Register hub datasets created by deepcopy (#1732) @farizrahman4u
- [AL-1821] ds.get_view (#1729) @farizrahman4u
- [AL-1829] Support views from datasets with uneven tensors (#1737) @farizrahman4u
- Google drive fixes (#1730) @FayazRahman
- Support timestamps for linked videos (#1731) @FayazRahman
- [AL-1820] view.sample_indices (#1728) @farizrahman4u
- [Optimization] Make sample size calculation faster in pytorch (#1725) @AbhinavTuli
- spell check - Found pytorch misspelt while reading docs (#1724) @neel2299
- [AL-1806] Do partial reads to retrieve shape for samples (#1721) @AbhinavTuli
- Fix linked video playback test (#1723) @FayazRahman
- Rechunking fixes (#1701) @farizrahman4u
- [AL-1810][AL-1814] Pytorch fixes (#1719) @AbhinavTuli
- [AL-1808] Support playing linked videos (#1717) @FayazRahman
- [BUG] Fix setitem issue with list of length 1 (#1718) @AbhinavTuli
- [BUG] Fix tobytes for linked tensors (#1714) @FayazRahman
- [AL-1795] Changes .data to return both numeric and text value for class labels (#1716) @AbhinavTuli
- [AL-1787] Enable appending to class_label tensors with text (#1710) @FayazRahman
- Update README.md (#1712) @istranic
🐛 Bug Fixes
- Correctly passing credentials to gcp provider (#1715) @ProgerDav
⚙️ Who Contributes
@AbhinavTuli, @FayazRahman, @ProgerDav, @adolkhan, @davidbuniat, @farizrahman4u, @istranic, @mikayelh and @neel2299