v4.2.7 🌈

Deeplake new release brings the following major changes:

###New features:

• BM25 and Inverted index now support full unicode
• New text index type - Exact, for comparison and IN operations
• Added ability to run batch queries
• Struct types now support other rich types, including images
• Ability to access raw bytes from the dataset/column
• Improved tagging and branching apis to support message and timestamp

Bugfixes and improvements:

• Improved compaction consistency during parallel data modification
• Fixed chunking strategy to avoid too large chunks

v4.2.3 🌈

• Added compaction of datafiles
• Improved images ingestion and streaming
• Added new index type for text comparison
• Improved handling of async operations
• Support async iteration on ds.batches

v4.2.1 🌈

Deeplake 4.2 release brings major changes to the deeplake format, while keeping backward compatibility.

• Commit compaction - Deeplake automatically compacts the small commits. Making dataset opens faster.
• Improved and stabilized deeplake async api - open_read_only_async, open_async, query_async, commit_async
• deeplake.like now copies dataset and columns metadata
• ds[row] is now convertible to Python dict

v4.1.17 🌈

• Fixed dataset size when concurrently deleting the same row.

v4.1.16 🌈

Improvements in version-control, new data types and observability

• Improved experience with version control and tagging
• Allow merging branches
• Integrated with opentelemetry for better observability
• Added float16 and bfloat16 data types

v3.9.44 🌈

🧭 What's Changed

• Updated versions for 3.9.44 release (#3040) @activesoull
• add labelbox iam integration support (#3043) @tyesayan

⚙️ Who Contributes

@activesoull and @tyesayan

v4.1.12 🌈

Big release bringing major features to deeplake users:

• Allow creating branches in deeplake datasets
• Allow running queries with directly passing credentials to the storage
• Support all JOIN types and cross product in deeplake query engine
• Improve labelbox integration. Allow authenticating by IAM roles

Additionally, small fixes and improvements are done:

• Optimize the way deeplake stores dataset metadata
• Allow MAXSIM comparison on dynamic columns

v4.1.11 🌈

Fixed bugs:

• Permanent redirect error message from S3 client on different location resource loading
• Incorrect dtype in virtual columns
• Stuck or crash when doing multiple iterations with break
• Descriptive error message on network request failures

v3.9.43 🌈

🧭 What's Changed

• Fixed ds.visualize for local datasets. (#3039) @khustup2

⚙️ Who Contributes

@activesoull and @khustup2

v3.9.42 🌈

🧭 What's Changed

• added exception if class_names are empty (#3038) @activesoull
• Fix palette (#3037) @activesoull
• Updated versions for 3.9.42 release (#3034) @activesoull

⚙️ Who Contributes

@activesoull and @davidbuniat