Releases: qdrant/qdrant-client
v1.12.1
Change Log
Bug Fixes 🐌
- #810 - fix distance matrix api in local mode with the default vector by @joein
- #812 - fix plain prefetch conversion to grpc by @joein
- #812 - fix query batch points offset in local mode by @joein
- #813 - raise an error on attempt to insert a non-existing named vector in local mode by @hh-space-invader
- #817 - fix DBSF rescoring on empty response in local mode by @joein
- #816 - fix resolving prefetch in local mode by @joein
Thanks to everyone, who've contributed to this release!
@joein @hh-space-invader
v1.12.0
Change Log
Features 🪶
- #759 - Facets API by @coszio
- #767 - Distance matrix API by @agourlay
- #804 - add mmap geo index by @joein
Bug Fixes 🦟
- #797 - fix query embeddings for sparse models by @hh-space-invader
- #800 - fix async grpc info call by @joein
- #802 - add local_files_only to sparse embeddings by @joein
Thanks to everyone, who've contributed to this release!
1.11.3
Change Log
Bug Fixes 🦟
- #787 - fix update vectors in local mode by @joein
- #785 - fix creating shard key in async grpc client by @joein @VanaMartin
- #790 - fix conversion of sharding method in collection params by @joein
- #782 - replace np.NINF with np.inf in local mode for numpy v2.x by @icfly2
- #791 - fix cluster collection info by making resharding operations field optional by @joein
Thanks to everyone, who've contributed to this release!
@VanaMartin @icfly2 @joein
v1.11.2
Change Log
Bug Fixes 🪰
- #763 - handle tuple conversion in grpc by @hh-space-invader
- #772 - fix nested prefetch in local mode by @paul2302 @hh-space-invader
- #773 - fix rest get_collection call for a collection with grpc-built payload indices, make parametrized integer index params optional by @joein
- #775 - fix
with_vectors
in local mode when retrieving by default vector name by @joein
Thanks to everyone, who've contributed to this release!
@paul2302 @hh-space-invader @joein
v1.11.1
Change Log
Features 📢
- #753 - update
fastembed
to 0.3.6
Bug Fixes 🪲
- #748 - fix
modifier.None
conversion by @joein - #750 - fix parametrized payload indices conversion by @joein
Deprecations ⏳
- #742 - deprecate
rest
property inQdrantRemote
in favour ofhttp
by @hh-space-invader - #740 - deprecate async grpc fixtures in
QdrantClient
in favour ofAsyncQdrantClient
by @hh-space-invader
Thanks to everyone, who've contributed to this release!
@hh-space-invader @joein
v1.11.0
Change Log
Features 📢
- #541 - added init_options property to recreate qdrant client instances from an existing one by @Anush008
- #686 - add bm42 and bm25 models support for
add
andquery
methods by @joein - #706 - add retry mechanism to
migrate
by @hh-space-invader - #678 - add info endpoint by @joein
- #690 - add query points group api by @agourlay @coszio
- #703 - add distribution based score fusion by @coszio
- #705 - add random sampling mechanism by @coszio
- #707 - extend struct interfaces to support raw string via types.Document by @joein
Bug Fixes 🪲
- #697 - fix conversion of datetime objects to grpc by @joein
- #715 - fix in-place modification of queries in
query_points
by @joein - #717 - unlock pyright, fix type hints by @hh-space-invader
- #721 - add missing timeouts to count, scroll and retrieve methods by @hh-space-invader @joein
- #730 - make upload_points method in AsyncQdrantClient synchronous by @hh-space-invader
Thanks to everyone, who've contributed to this release!
@agourlay @coszio @Anush008 @generall @hh-space-invader @joein
v1.10.1
v1.10.0
Change Log
Features 🐯
-
Universal query API - [Docs]
- Support search, recommendations, discovery, ordering by payload in a single request
- Perform hybrid search by merging results from different queries with fusion
- Build multi-stage queries with re-scoring to achieve better performance
-
Support for multivectors - [Docs]
- Use late interaction (such as ColBERT) models natively in Qdrant
- Store dynamic number of vectors per point with shared payload
-
Support for
float16
datatype for dense vectors - [Docs] -
Support for
float16
anduint8
datatype for sparse vectors index - [Docs] -
Support for Inverse Document Frequency (IDF) in sparse vectros - [Docs]
- Provides an option to automatically account for IDF in sparse vectors search
- Unlocks the ability to stream updates for BM25 and BM42 embeddings
Bug Fixes 🦟
- Fix exclusion of ids in discovery api
- Fix StartFrom conversion
Thanks to everyone, who've contributed to this release!
@agourlay @coszio @tellet-q @generall @joein
v1.9.2
Changelog
Features 🪄
- #612 Add gpu support for fastembed and possibility to set onnx providers by @joein
- #623 Replace black with ruff by @generall
Fixes 🪛
- #627 Save and return copies of data in local mode by @joein
- #637 Allow string
RecommendStrategy
with grpc by @coszio - #640 Fix integer index grpc conversion by @joein
- #646 Fix boolean filters when using pydantic<2.6 by @joein
Thanks to everyone who contributed to the current release: @generall @coszio @joein