Releases: langchain-ai/langchain
Releases · langchain-ai/langchain
v0.0.187
What's Changed
- Fixed docstring in faiss.py for load_local by @luckyduck in #5440
- Removes duplicated call from langchain/client/langchain.py by @patrickkeane in #5449
encoding_kwargs
for InstructEmbeddings by @Xmaster6y in #5450- MRKL output parser no longer breaks well formed queries by @ninjapenguin in #5432
- docs: cleaning by @leo-gan in #5413
- Added async _acall to FakeListLLM by @camille-vanhoffelen in #5439
- Feat: Add batching to Qdrant by @kacperlukawski in #5443
- Update psychicapi version by @Ayan-Bandyopadhyay in #5471
- Add maximal relevance search to SKLearnVectorStore by @mrtj in #5430
- add simple test for imports by @hwchase17 in #5461
- Ability to specify credentials wihen using Google BigQuery as a data loader by @nsheils in #5466
- convert the parameter 'text' to uppercase in the function 'parse' of the class BooleanOutputParser by @ARSblithe212 in #5397
- added n_threads functionality for gpt4all by @Vokturz in #5427
- Allow for async use of SelfAskWithSearchChain by @pors in #5394
- Allow ElasticsearchEmbeddings to create a connection with ES Client object by @jeffvestal in #5321
- SQLite-backed Entity Memory by @JoseHervas in #5129
- py tracer fixes by @agola11 in #5377
- Harrison/html splitter by @hwchase17 in #5468
- Feature: Qdrant filters supports by @kacperlukawski in #5446
- Harrison/matching engine vectorstore by @hwchase17 in #3350
- code splitter docs by @hwchase17 in #5480
- Bedrock llm and embeddings by @3coins in #5464
- add more vars to text splitter by @hwchase17 in #5503
- bump 187 by @dev2049 in #5504
New Contributors
- @luckyduck made their first contribution in #5440
- @patrickkeane made their first contribution in #5449
- @Xmaster6y made their first contribution in #5450
- @camille-vanhoffelen made their first contribution in #5439
- @nsheils made their first contribution in #5466
- @ARSblithe212 made their first contribution in #5397
- @Vokturz made their first contribution in #5427
- @JoseHervas made their first contribution in #5129
Full Changelog: v0.0.186...v0.0.187
v0.0.186
v0.0.185
What's Changed
- Add ToolException that a tool can throw. by @xming521 in #5050
- Harrison/text splitter by @hwchase17 in #5417
- New Trello document loader by @GDrupal in #4767
- DocumentLoader for GitHub by @UmerHA in #5408
- Harrison/spark reader by @hwchase17 in #5405
- Set old LCTracer to default to port 8000 by @vowelparrot in #5381
- Rename and fix typo in lancedb by @eddyxu in #5425
- Harrison/condense q llm by @hwchase17 in #5438
- adding MongoDBAtlasVectorSearch by @P-E-B in #5338
- Add more code splitters (go, rst, js, java, cpp, scala, ruby, php, swift, rust) by @ByronHsu in #5171
- bump 185 by @dev2049 in #5442
New Contributors
- @xming521 made their first contribution in #5050
- @GDrupal made their first contribution in #4767
- @eddyxu made their first contribution in #5425
- @P-E-B made their first contribution in #5338
Full Changelog: v0.0.184...v0.0.185
v0.0.184
What's Changed
- Add path validation to DirectoryLoader by @os1ma in #5327
- Fix: Handle empty documents in ContextualCompressionRetriever (Issue #5304) by @hanguofeng in #5306
- handle json parsing errors by @hwchase17 in #5371
- Use Default Factory by @vowelparrot in #5380
- Update PR template with Twitter handle request by @jacoblee93 in #5382
- fix: Blob.from_data mimetype is lost by @Digma in #5395
- Add async support to routing chains by @amaudruz in #5373
- Fix update_document function, add test and documentation. by @martinholecekmax in #5359
- Update llamacpp demonstration notebook by @mmagnesium in #5344
- Removed deprecated llm attribute for load_chain by @imeckr in #5343
- Harrison/llamacpp by @hwchase17 in #5402
- Add pagination for Vertex AI embeddings by @Jflick58 in #5325
- Reformat openai proxy setting as code by @sevendark in #5330
- Harrison/deep infra by @hwchase17 in #5403
- Harrison/prediction guard update by @hwchase17 in #5404
- Implemented appending arbitrary messages by @eavanvalkenburg in #5293
- docs:
ecosystem/integrations
update 2 by @leo-gan in #5282 - docs:
ecosystem/integrations
update 1 by @leo-gan in #5219 - Harrison/datetime parser by @hwchase17 in #4693
- bump version 184 by @hwchase17 in #5407
New Contributors
- @os1ma made their first contribution in #5327
- @jacoblee93 made their first contribution in #5382
- @Digma made their first contribution in #5395
- @amaudruz made their first contribution in #5373
- @martinholecekmax made their first contribution in #5359
- @mmagnesium made their first contribution in #5344
- @sevendark made their first contribution in #5330
Full Changelog: v0.0.183...v0.0.184
v0.0.183
What's Changed
- Fixes iter error in FAISS add_embeddings call by @ninjapenguin in #5367
- revert bad json by @hwchase17 in #5370
- bump to 183 by @hwchase17 in #5372
Full Changelog: v0.0.182...v0.0.183
v0.0.182
What's Changed
- Better docs for weaviate hybrid search by @hsm207 in #5290
- Add instructions to pyproject.toml by @eyurtsev in #5138
- docs: improve flow of llm caching notebook by @malandis in #5309
- Fix typos by @russellpwirtz in #5323
- docs: added link to LangChain Handbook by @leo-gan in #5311
- add enum output parser by @hwchase17 in #5165
- Add Chainlit to deployment options by @constantinidan in #5314
- Fixing blank thoughts in verbose for "_Exception" Action by @svdeepak99 in #5331
- fix: remove empty lines that cause InvalidRequestError by @mbchang in #5320
- Sample Notebook for DynamoDB Chat Message History by @KBB99 in #5351
- added cosmos kwargs option by @eavanvalkenburg in #5292
- feat: support for shopping search in SerpApi by @aymenfurter in #5259
- Add SKLearnVectorStore by @mrtj in #5305
- bump 182 by @dev2049 in #5364
New Contributors
- @russellpwirtz made their first contribution in #5323
- @constantinidan made their first contribution in #5314
- @KBB99 made their first contribution in #5351
- @aymenfurter made their first contribution in #5259
- @mrtj made their first contribution in #5305
Full Changelog: v0.0.181...v0.0.182
v0.0.181
What's Changed
- remove extra "\n" to ensure that the format of the description, examp… by @pengqu123 in #5232
- Resolve error in StructuredOutputParser docs by @mwinterde in #5240
- Added the option of specifying a proxy for the OpenAI API by @ymaurer in #5246
- OpenSearch top k parameter fix by @dev2049 in #5216
- Fixed regression in JoplinLoader's get note url by @alondmnt in #5265
- Docs link custom agent page in getting started by @JanilsWoerst in #5250
- Zep sdk version by @dev2049 in #5267
- Add C Transformers for GGML Models by @marella in #5218
- Add visible_only and strict_mode options to ClickTool by @cancan101 in #4088
- Add Multi-CSV/DF support in CSV and DataFrame Toolkits by @NickL77 in #5009
- OpenAI lint by @dev2049 in #5273
- Added pipline args to
HuggingFacePipeline.from_model_id
by @solomspd in #5268 - Support bigquery dialect - SQL by @HassanOuda in #5261
- feat: add Momento as a standard cache and chat message history provider by @malandis in #5221
- Fixed typo: 'ouput' to 'output' in all documentation by @deepblue in #5272
- Tedma4/twilio tool by @tedma4 in #5136
- LLM wrapper for Databricks by @mengxr in #5142
- Add an example to make the prompt more robust by @pengqu123 in #5291
- Update CONTRIBUTION guidelines and PR Template by @eyurtsev in #5140
- Fixed passing creds to VertexAI LLM by @lkuligin in #5297
- bump 181 by @dev2049 in #5302
New Contributors
- @ymaurer made their first contribution in #5246
- @JanilsWoerst made their first contribution in #5250
- @marella made their first contribution in #5218
- @cancan101 made their first contribution in #4088
- @solomspd made their first contribution in #5268
- @HassanOuda made their first contribution in #5261
- @deepblue made their first contribution in #5272
- @tedma4 made their first contribution in #5136
- @mengxr made their first contribution in #5142
Full Changelog: v0.0.180...v0.0.181
v0.0.180
What's Changed
- Harrison/modelscope by @hwchase17 in #5156
- Reuse
length_func
inMapReduceDocumentsChain
by @zachschillaci27 in #5181 - Update Cypher QA prompt by @tomasonjo in #5173
- Improve weaviate vectorstore docs by @hsm207 in #5201
- tfidf retriever by @dev2049 in #5114
- standardize json parsing by @hwchase17 in #5168
- fixing total cost finetuned model giving zero by @tommasodelorenzo in #5144
- Fixes scope of query Session in PGVector by @ninjapenguin in #5194
- Output parsing variation allowance by @dibrale in #5178
- Allow readthedoc loader to pass custom html tag by @ByronHsu in #5175
- add a Iugu loader by @rasiqueira in #5162
- Add Joplin document loader by @alondmnt in #5153
- nit by @dev2049 in #5208
- add option to pass openai key to langchain plus command by @agola11 in #5213
- Log warning by @vowelparrot in #5192
- Add Delete Session Method by @vowelparrot in #5193
- Add 'status' command to get server status by @vowelparrot in #5197
- Harrison/vertex by @hwchase17 in #5049
- fix a mistake in concepts.md by @leo-gan in #5222
- Create async copy of from_text() inside GraphIndexCreator. by @maspotts in #5214
- Remove API key from docs by @kbressem in #5223
- Change Default GoogleDriveLoader Behavior to not Load Trashed Files (issue #5104) by @NickL77 in #5220
- Allow to specify ID when adding to the FAISS vectorstore. by @atisharma in #5190
- Bibtex integration for document loader and retriever by @eyurtsev in #5137
- Add MiniMax embeddings by @archongum in #5174
- Weaviate: Add QnA with sources example by @hsm207 in #5247
- Fix typo in docstring of RetryWithErrorOutputParser by @mwinterde in #5244
- bump 180 by @dev2049 in #5248
New Contributors
- @ninjapenguin made their first contribution in #5194
- @dibrale made their first contribution in #5178
- @ByronHsu made their first contribution in #5175
- @rasiqueira made their first contribution in #5162
- @alondmnt made their first contribution in #5153
- @maspotts made their first contribution in #5214
- @kbressem made their first contribution in #5223
- @NickL77 made their first contribution in #5220
- @atisharma made their first contribution in #5190
- @archongum made their first contribution in #5174
Full Changelog: v0.0.179...v0.0.180
v0.0.179
What's Changed
- docs: fix minor typo + add wikipedia package installation part in human_input_llm.ipynb by @amicus-veritatis in #5118
- solving #2887 by @tommasodelorenzo in #5127
- Improve PlanningOutputParser whitespace handling by @TMRolle in #5143
- Add ElasticsearchEmbeddings class for generating embeddings using Elasticsearch models by @jeffvestal in #3401
- Adding Weather Loader by @iamadhee in #5056
- Add MosaicML inference endpoints by @dakinggg in #4607
- Empty check before pop by @edwardzjl in #4929
- Add async versions of predict() and predict_messages() by @jlowin in #4867
- fix: fix current_time=Now bug for aadd_documents in TimeWeightedRetriever by @mbchang in #5155
- Docs: updated getting_started.md by @DanQuin in #5151
- Clarification of the reference to the "get_text_legth" function in ge… by @DanQuin in #5154
- docs: added missed
document_loaders
examples by @leo-gan in #5150 - Add Typesense vector store by @jasonbosco in #1674
- Vectara by @ofermend in #5069
- Beam by @NolanTrem in #4996
- Update rellm_experimental.ipynb by @eltociear in #5189
- example usage by @jeffvestal in #5182
- adjust docarray docstrings by @jupyterjazz in #5185
- bump 179 by @dev2049 in #5200
New Contributors
- @tommasodelorenzo made their first contribution in #5127
- @TMRolle made their first contribution in #5143
- @jeffvestal made their first contribution in #3401
- @dakinggg made their first contribution in #4607
- @edwardzjl made their first contribution in #4929
- @jlowin made their first contribution in #4867
- @DanQuin made their first contribution in #5151
- @jasonbosco made their first contribution in #1674
- @ofermend made their first contribution in #5069
- @NolanTrem made their first contribution in #4996
- @jupyterjazz made their first contribution in #5185
Full Changelog: v0.0.178...v0.0.179
v0.0.178
What's Changed
- fix: revert docarray explicit transitive dependencies and use extras instead by @malandis in #5015
- Improving Resilience of MRKL Agent by @svdeepak99 in #5014
- Improve pinecone hybrid search retriever adding metadata support by @lbsnrs in #5098
- Add the usage of SSL certificates for Elasticsearch and user password authentication by @CodeDevNinja in #5058
- add get_top_k_cosine_similarity method to get max top k score and index by @hwaking in #5059
- PowerBI major refinement in working of tool and tweaks in the rest by @eavanvalkenburg in #5090
- fix: add_texts method of Weaviate vector store creats wrong embeddings by @Shawn91 in #4933
- update langchainplus client and docker file to reflect port changes by @agola11 in #5005
- Fixed import error for AutoGPT e.g. from langchain.experimental.auton… by @ankitarya1019 in #5101
- Update serpapi.py by @venetisgr in #4947
- changed ValueError to ImportError by @leo-gan in #5103
- fix: assign current_time to datetime.now() if current_time is None by @mbchang in #5045
- Add Mastodon toots loader by @imrehg in #5036
- Add OpenLM LLM multi-provider by @r2d4 in #4993
- Pass Dataset Name by Name not Position by @vowelparrot in #5108
- Fixes issue #5072 - adds additional support to Weaviate by @jettro in #5085
- Improve effeciency of TextSplitter.split_documents, iterate once by @eyurtsev in #5111
- WhyLabs callback by @jamie256 in #4906
- Add AzureCognitiveServicesToolkit to call Azure Cognitive Services API by @whiskyboy in #5012
- Add link to Psychic from document loaders documentation page by @Ayan-Bandyopadhyay in #5115
- bump 178 by @dev2049 in #5130
New Contributors
- @malandis made their first contribution in #5015
- @lbsnrs made their first contribution in #5098
- @CodeDevNinja made their first contribution in #5058
- @hwaking made their first contribution in #5059
- @Shawn91 made their first contribution in #4933
- @ankitarya1019 made their first contribution in #5101
- @venetisgr made their first contribution in #4947
- @imrehg made their first contribution in #5036
- @r2d4 made their first contribution in #4993
- @jettro made their first contribution in #5085
- @jamie256 made their first contribution in #4906
- @whiskyboy made their first contribution in #5012
Full Changelog: v0.0.177...v0.0.178