Releases: langchain-ai/langchain
Releases Β· langchain-ai/langchain
v0.0.210
What's Changed
- MD header text splitter returns Documents by @rlancemartin in #6571
- add mongo (HOLD) by @hwchase17 in #6437
- Allow callback handlers to opt into being run inline by @nfcampos in #6424
- StreamlitCallbackHandler by @tconkling in #6315
- added redis method to delete entries by keys by @ChrKahl in #6222
- Loader for OpenCityData and minor cleanups to Pandas, Airtable loaders by @rlancemartin in #6301
- Add tags in agent initialization by @vowelparrot in #6559
- Session to project by @vowelparrot in #6249
- fix: remove callbacks arg from Tool and StructuredTool inferred schema by @aledelunap in #6483
- bump 210 by @dev2049 in #6656
New Contributors
- @tconkling made their first contribution in #6315
- @aledelunap made their first contribution in #6483
Full Changelog: v0.0.209...v0.0.210
v0.0.209
What's Changed
- Relax string input mapper check by @vowelparrot in #6544
- [Feature][VectorStore] Support StarRocks as vector db by @dirtysalt in #6119
- Minor Grammar Fixes in Docs and Comments by @aevedis in #6536
- Remove unintended double negation in docstring by @BioGeek in #6541
- update pr tmpl by @dev2049 in #6552
- feat: faiss filter from list by @HenriZuber in #6537
- Wait for all futures by @vowelparrot in #6554
- Fix whatsappchatloader - enable parsing new datetime format on WhatsApp chat by @suyuchenxm in #6555
- Remove duplicate databricks entries in ecosystem integrations by @gengliangwang in #6569
- Change Data Loader Namespace by @kenzie28 in #6568
- Detailed using the Twilio tool to send messages with 3rd party apps incl. WhatsApp by @movaid7 in #6562
- add motherduck docs by @hwchase17 in #6572
- Upgrade the version of AwaDB and add some new interfaces by @ljeagle in #6565
- feat: interfaces for async embeddings, implement async openai by @BrendanGraham14 in #6563
- Dev2049/openllm by @dev2049 in #6578
- Dev2049/azure endpoint by @dev2049 in #6580
- Add missing word in comment by @BioGeek in #6587
- Clarifai integration by @minhajul-clarifai in #5954
- bump 209 by @dev2049 in #6593
- Fix callback forwarding in async plan method for OpenAI function agent by @rogeriochaves in #6584
New Contributors
- @dirtysalt made their first contribution in #6119
- @aevedis made their first contribution in #6536
- @BioGeek made their first contribution in #6541
- @HenriZuber made their first contribution in #6537
- @suyuchenxm made their first contribution in #6555
- @kenzie28 made their first contribution in #6568
- @movaid7 made their first contribution in #6562
- @BrendanGraham14 made their first contribution in #6563
- @minhajul-clarifai made their first contribution in #5954
- @rogeriochaves made their first contribution in #6584
Full Changelog: v0.0.208...v0.0.209
v0.0.208
What's Changed
- improve documentation on base chain by @hwchase17 in #6468
- Vector store support for Cassandra by @hemidactylus in #6426
- Update notebook for MD header splitter and create new cookbook by @rlancemartin in #6399
- docs/fix links by @dev2049 in #6498
- Fix link by @dev2049 in #6501
- Documentation Fix: Correct the example code output in the prompt templates doc by @harjotsinghparmar in #6496
- Fixed: 'readible' -> readable by @djm93dev in #6492
- Make streamlit import optional by @dev2049 in #6510
- Fix typo in docstring of format_tool_to_openai_function by @northern-64bit in #6479
- typo(llamacpp.ipynb): 'condiser' -> 'consider' by @TheOnlyWayUp in #6474
- Export trajectory eval fn by @vowelparrot in #6509
- Update index.mdx by @dataprofessor in #6326
- Add KuzuQAChain by @mewim in #6454
- Be able to use Codey models on Vertex AI by @HassanOuda in #6354
- Add async support for HuggingFaceTextGenInference by @mirceapasoi in #6507
- Feat: Add a prompt template parameter to qa with structure chains by @ElReyZero in #6495
- Integrate Rockset as Vectorstore by @anubhav94N in #6216
- Fix issue with non-list
Toheader in GmailSendMessage Tool by @jmisilo in #6242 - Update model token mappings/cost to include 0613 models by @eoriont in #6122
- Infino integration for simplified logs, metrics & search across LLM data & token usage by @naman-modi in #6218
- Harrison/multi tool by @hwchase17 in #6518
- bump to ver 208 by @hwchase17 in #6540
New Contributors
- @hemidactylus made their first contribution in #6426
- @harjotsinghparmar made their first contribution in #6496
- @djm93dev made their first contribution in #6492
- @northern-64bit made their first contribution in #6479
- @TheOnlyWayUp made their first contribution in #6474
- @dataprofessor made their first contribution in #6326
- @mewim made their first contribution in #6454
- @mirceapasoi made their first contribution in #6507
- @ElReyZero made their first contribution in #6495
- @anubhav94N made their first contribution in #6216
- @jmisilo made their first contribution in #6242
- @eoriont made their first contribution in #6122
- @naman-modi made their first contribution in #6218
Full Changelog: v0.0.207...v0.0.208
v0.0.207
What's Changed
- add FunctionMessage support to
_convert_dict_to_message()in OpenAI chat model by @thehunmonkgroup in #6382 - fix openai qa chain by @dev2049 in #6487
- Add Alibaba Cloud OpenSearch as a new vector store by @zhaoshengbo in #6154
- release 207 by @dev2049 in #6488
- fix: change ddg to DDGS by @Undertone0809 in #6480
New Contributors
- @thehunmonkgroup made their first contribution in #6382
- @zhaoshengbo made their first contribution in #6154
Full Changelog: v0.0.206...v0.0.207
v0.0.206
What's Changed
- Include placeholder value for all secrets, not just kwargs by @dqbd in #6421
- expose docs chains by @hwchase17 in #6453
- Run eval in eval mode by @vowelparrot in #6447
- Update arize_callback.py by @hakantekgul in #6433
- Add Trajectory Eval RunEvaluator by @vowelparrot in #6449
- fix anthropic chat model mutating input list by @agola11 in #6457
- Remove extra word in the introduction documentation by @andreybavt in #6450
- Add example for question answering over documents with OpenAI Function Agent by @homanp in #6448
- Add
_similarity_search_with_relevance_scoresinPineconeby @neo in #6446 - Update web_base.ipynb by @jackfrost1411 in #6430
- Remove backticks without clear purpose from docs by @ykerus in #6442
- Fix Custom LLM Agent example by @nirga in #6429
- Update introduction.mdx by @StandardAI in #6425
- docs
retrieversfixes by @leo-gan in #6299 - Fixed a link typo /-/route -> /-/routes. and change endpoint format by @kylehh in #6186
- Incorrect argument count handling by @swappysh in #5543
- Harrison/functions in retrieval by @hwchase17 in #6463
- Fix for #6431 - chatprompt template with partial variables giing validation error by @Avinash-Raj in #6456
- Update SinglStoreDB vectorstore by @volodymyr-memsql in #6423
- Fix broken links in autonomous agents docs by @gadkins in #6398
- Fix the issue where ANTHROPIC_API_URL set in environment is not takin⦠by @zengbo in #6400
- Improve error message by @alonsosilvaallende in #6275
- Harrison/unstructured page number by @hwchase17 in #6464
- feat: use latest duckduckgo_search API to call by @Undertone0809 in #6409
- fix: llm caching for replicate by @brycedrennan in #6396
- Update serpapi.py Support baidu list type answer_box by @Gavin-YYC in #6386
- fix neo4j schema query by @HubertKl in #6381
- bump version to 206 by @hwchase17 in #6465
New Contributors
- @dqbd made their first contribution in #6421
- @andreybavt made their first contribution in #6450
- @ykerus made their first contribution in #6442
- @nirga made their first contribution in #6429
- @StandardAI made their first contribution in #6425
- @swappysh made their first contribution in #5543
- @Avinash-Raj made their first contribution in #6456
- @gadkins made their first contribution in #6398
- @alonsosilvaallende made their first contribution in #6275
- @brycedrennan made their first contribution in #6396
- @Gavin-YYC made their first contribution in #6386
- @HubertKl made their first contribution in #6381
Full Changelog: v0.0.205...v0.0.206
v0.0.205
What's Changed
- support memory for functions by @Fei-Wang in #6165
- fix prod docs build by @dev2049 in #6402
- nit by @dev2049 in #6350
- changes to llm chain by @hwchase17 in #6328
- Harrison/refactor functions by @hwchase17 in #6408
- bump version to 205 by @hwchase17 in #6410
Full Changelog: v0.0.204...v0.0.205
v0.0.204
What's Changed
- Guardrails output parser: Pass LLM api for reasking by @irgolic in #6089
- Harrison/gdrive enhancements by @hwchase17 in #6375
- Extend
ArgillaCallbackHandlersupport by @alvarobartt in #6153 - Add the ability to run the map_reduce chains process results step as async by @MalhotraVijay in #6181
- fix link of callbacks on modules page by @xleven in #6323
- Fixed PermissionError on windows by @MIDORIBIN in #6170
- Harrison/myscale self query by @hwchase17 in #6376
- Harrison/metaphor search fix by @hwchase17 in #6387
- Harrison/zep mem by @hwchase17 in #6388
- Fix class promotion by @leo-gan in #6187
- Harrison/functions docs improvements by @hwchase17 in #6389
- Add option to save/load graph cypher QA by @tomasonjo in #6219
- Fix output final text for HuggingFaceTextGenInference when streaming by @janpawellek in #6211
- OpenAI functions dont work with async streaming... #6225 by @ju-bezdek in #6226
- searx_search: updated tools and doc by @blob42 in #6276
- Fix integration tests for Faiss vector store by @hp0404 in #6281
- [hotfix] Deep Lake fails on newer version due to hardcode by @davidbuniat in #6383
- Fix latest clickhouse vector schema change by @haoch in #6385
- Fix typo in the CAI critique prompt by @dulacp in #6123
- Correct AzureSearch Vector Store not applying search_kwargs when searching by @CameronVetter in #6132
- Fixed an unhandled error that was raised when DynamoDB did not have any chat history. by @ngo275 in #6141
- Fix LLM types so that they can be loaded from config dicts by @janpawellek in #6235
- add max_context_size property in BaseOpenAI by @ikebo in #6239
- Add markdown to specify important arguments by @jackfrost1411 in #6246
- Update web_base.ipynb for guiding purposes by @jackfrost1411 in #6248
- Iterate through filtered file types instead of all listed files by @tedhli in #6258
- Fixes typo in Vectara.similarity_search by @georgiansarghi in #6277
-
- Fix pass system_message argument in new feature openai_functions_agent by @lyntcelec in #6297
- Added gpt-3.5-turbo 0613 16k and 16k-0613 pricing by @matias-biatoz in #6287
- rm pandas from arize by @hwchase17 in #6392
- Harrison/typesense fix by @hwchase17 in #6391
- Harrison/chroma fix by @hwchase17 in #6390
- Update web_base.py _fetch() method For SiteMapLoader by @jackfrost1411 in #6256
New Contributors
- @MalhotraVijay made their first contribution in #6181
- @xleven made their first contribution in #6323
- @MIDORIBIN made their first contribution in #6170
- @janpawellek made their first contribution in #6211
- @dulacp made their first contribution in #6123
- @CameronVetter made their first contribution in #6132
- @ngo275 made their first contribution in #6141
- @ikebo made their first contribution in #6239
- @tedhli made their first contribution in #6258
- @georgiansarghi made their first contribution in #6277
- @lyntcelec made their first contribution in #6297
- @matias-biatoz made their first contribution in #6287
Full Changelog: v0.0.203...v0.0.204
v0.0.203
What's Changed
- Dev2049/docs docusaurus by @dev2049 in #6300
- rm ignore_vercel by @dev2049 in #6302
- update api link by @dev2049 in #6303
- update readme by @dev2049 in #6304
- basic redirect by @dev2049 in #6309
- more redirect by @dev2049 in #6314
- Update dev container by @bamurtaugh in #6189
- Del linkcheck readme by @dev2049 in #6317
- Fixes #6282 by @paschembri in #6283
- nit by @dev2049 in #6305
- Improve the performance of add_texts interface and upgrade the AwaDB from 0.3.2 to 0.3.3 by @ljeagle in #6316
- fix eval guide links by @dev2049 in #6319
- Harrison/deeplake new features by @hwchase17 in #6263
- Better Entity Memory code documentation by @lukestanley in #6318
- Handle Managed Motorhead Data Key by @softboyjimbo in #6169
- changed height in the nb example by @fpingham in #6327
- fix links to prompt templates and example selectors by @masafumimori in #6332
- DocArray as a Retriever by @jupyterjazz in #6031
- make modelname_to_contextsize as a staticmethod by @Ja-sonYun in #6040
- Improve AnalyticDB Vector Store implementation without affecting user by @wangxuqi in #6086
- Add oobabooga/text-generation-webui support as a llm by @lonestriker in #5997
- Allow GoogleDrive to authenticate via application default credentials on Cloud Run/GCE etc without service key by @MarkEdmondson1234 in #6035
- qdrant: search by vector by @otwieracz in #6043
- Harrison/error zero tools by @hwchase17 in #6340
- Harrison/faiss score by @hwchase17 in #6341
- Custom Anthropic API URL by @zengbo in #6221
- Make lckwargs private by @nfcampos in #6344
- update web_base.py to have verify option by @jackfrost1411 in #6107
- Add new openai 0613 model costs by @kdcokenny in #6110
- Add ignore vercel preview script by @dev2049 in #6320
- Harrison/mmr support for opensearch by @hwchase17 in #6349
- Update MD header text splitter notebook by @rlancemartin in #6339
- Add self query retriever example with MD header splitting by @rlancemartin in #6359
- bump 203 by @dev2049 in #6372
New Contributors
- @bamurtaugh made their first contribution in #6189
- @paschembri made their first contribution in #6283
- @masafumimori made their first contribution in #6332
- @Ja-sonYun made their first contribution in #6040
- @lonestriker made their first contribution in #5997
- @MarkEdmondson1234 made their first contribution in #6035
- @otwieracz made their first contribution in #6043
- @zengbo made their first contribution in #6221
- @jackfrost1411 made their first contribution in #6107
- @kdcokenny made their first contribution in #6110
Full Changelog: v0.0.202...v0.0.203
v0.0.202
What's Changed
- Add Tags for LLMs by @vowelparrot in #6229
- Feature/add acreom loader by @inferense in #5780
- Support chat history persistence in AutoGPT by @alonroth in #5716
- Improve Error Message for failed callback by @vowelparrot in #6247
- Harrison/openai functions by @hwchase17 in #6223
- Harrison/openai functions by @hwchase17 in #6261
- Fix: change the chatgpt plugin retriever metadata format by @JaysonAlbert in #5920
- Update output format of MosaicML endpoint to be more flexible by @dakinggg in #6060
- ArxivAPIWrapper - doc_content_chars_max by @hp0404 in #6063
- bump to 202 by @hwchase17 in #6262
New Contributors
- @inferense made their first contribution in #5780
- @alonroth made their first contribution in #5716
- @JaysonAlbert made their first contribution in #5920
Full Changelog: v0.0.201...v0.0.202
v0.0.201
What's Changed
- Add Run Collector Callback by @vowelparrot in #6133
- Update Name by @vowelparrot in #6136
- Update readthedocs_documentation.ipynb by @maver1ck in #6148
- typo: 'following following' to 'following' by @thiswillbeyourgithub in #6163
- Add docs for tags by @nfcampos in #6155
- feat: Add support for the Solidity language by @0xJord4n in #6054
- feat: add content_format param to ConfluenceLoader.load() by @haoqixu in #5922
- count tokens for new OpenAI model versions by @kylrth in #6195
- Fix typo
pandocstopandocby @gyroflaw in #6203 - Harrison/use functions agent by @hwchase17 in #6185
- bump version to 201 by @hwchase17 in #6233
New Contributors
- @thiswillbeyourgithub made their first contribution in #6163
- @0xJord4n made their first contribution in #6054
- @kylrth made their first contribution in #6195
- @gyroflaw made their first contribution in #6203
Full Changelog: v0.0.200...v0.0.201