v0.1.13-pre.0

What's Changed

• googleai: add support for system instructions in vertex too by @eliben in #930
• googleai: refactor shared test a bit and make it more resilient to model randomness by @eliben in #936
• googleai: update dependency version by @eliben in #939
• anthropic: Add support for beta header by @tmc in #967
• anthropic: Add improved handling of other error event types by @tmc in #1012
• openai: reflect max_token field deprecation by @tmc in #1013
• openai: add completion_tokens_details to usage response by @tmc in #1014
• openai: add reasoning tokens to usage response by @tmc in #1015
• openai: add reasoning tokens to streaming response by @tmc in #1016
• examples: Add openai o1-preview example by @tmc in #1017
• llms: GenerateFromSinglePrompt struct literal uses unkeyed fields by @iocron in #926
• ci: upgrade golangci lint version by @wangjiancn in #928
• Add perplexity completion example by @sklinkert in #977
• agents: optimize the mrkl prompt by @Alan796 in #976
• fix(docs): Type method missing from OutputParser by @chxru in #972
• outputparser: improve BooleanOutputParser by @amitaifrey in #978
• outputparser: improve DefinedOutputParser by @amitaifrey in #980
• textsplitter: add option to join table rows by @corani in #981
• googleai: add support for JSONMode and ResponseMIMEType by @mheck136 in #984
• openai: add support for sending Structured Output request for chat completions api by @takeyamakenta in #986
• redisvector: fix score threshold option by @acrmp in #1003
• milvus: add skipFlushOnWrite option by @fatedier in #992
• outputparser: use strings.Cut in structured parser by @anuraaga in #997
• examples: replace deprecated agents.Initialize by @treywelsh in #1009
• bedrock: Fix unchecked error when stream channel is closed by @lucaronca in #998
• llms: Add fake package by @devalexandre in #935
• examples: replaced compose v1 with compose v2 in readme for pgvector example by @elnoro in #969
• fix(mistral): supports the default llms.WithTools implementation by @douglarek in #970
• vectorstores: add mongovector by @prestonvasquez in #1005
• lint: fixup lint settings, refactor package by @tmc in #1018

New Contributors

• @iocron made their first contribution in #926
• @Alan796 made their first contribution in #976
• @chxru made their first contribution in #972
• @amitaifrey made their first contribution in #978
• @mheck136 made their first contribution in #984
• @takeyamakenta made their first contribution in #986
• @acrmp made their first contribution in #1003
• @anuraaga made their first contribution in #997
• @treywelsh made their first contribution in #1009
• @lucaronca made their first contribution in #998
• @prestonvasquez made their first contribution in #1005

Full Changelog: v0.1.12...v0.1.13-pre.0

v0.1.12

What's Changed

• llms/googleai: fix vertex do not return usage info by @wangjiancn in #904
• bugfix: Anthropic Function Calling by @lwlee2608 in #899
• llms/anthropic: Fix Content set incorrectly for anthropic llm requests by @JckHoe in #908
• llms/googleai: fix convert tool error by @wangjiancn in #903
• llms/googleai: Fix usage population tests, max token test, update dpep by @tmc in #910
• docs/groq: by @devalexandre in #907
• llms/watsonx: update module by @h0rv in #902
• docs: fix text_splitters mdx code example never closed by @4lxprime in #900
• textsplitter: add WithHeadingHierarchy option to markdown splitter to retain heading hierarchy in chunks by @iwilltry42 in #898
• examples: point watsonx example at current main by @tmc in #911
• outputparser: add Defined parser to extract a struct from LLM output by @erictse in #856
• anthropic: Improve streaming message handling in anthropic client by @tmc in #912
• googleai: Settle on GOOGLE_API_KEY for google auth env var name by @tmc in #913
• llms: Increase default tokens by @tmc in #914
• examples: Add example readmes by @tmc in #916
• examples: Improve top level readme by @tmc in #919
• anthropic: Spruce up tool calling example by @tmc in #921
• examples: point to 0.1.12-pre.0 by @tmc in #922

New Contributors

• @JckHoe made their first contribution in #908
• @4lxprime made their first contribution in #900
• @iwilltry42 made their first contribution in #898
• @erictse made their first contribution in #856

Full Changelog: v0.1.11...v0.1.12

v0.1.11

What's Changed

• llms: add watsonx by @h0rv in #577
• pgvector restore WithConnectionURL option by @tmc in #832
• examples: Update to v0.1.10 by @tmc in #833
• agents: GetTools added to Agent interface by @AugustDev in #831
• examples: add redis vectorstore example by @CrazyWr in #834
• llms/googleai: batch embedding calls by @johanbrandhorst in #825
• fix: better handling of errors in openai-like streaming messages by @douglarek in #839
• llms/googleai: upgrade genai SDKs and fix build issue vs. new version by @eliben in #854
• fix: handle nil pointer exception in chat_prompt_template by @ankitraj2330 in #879
• httputil: debug transport layer by @AugustDev in #845
• llms: Add improved yaml and json marshaling by @tmc in #880
• tools: Improve tool choice population and parsing by @tmc in #883
• memory: move mongo chat history into the separate module by @CrazyWr in #843
• llms: Tidy up marshaling by @tmc in #884
• anthropic: Add tool/function calling support by @lwlee2608 in #882
• milvus_vectorstore: store metadata with JSON data type & support search with filter by @CrazyWr in #844
• readme: add scheme to vectorstores/chroma URL by @codefromthecrypt in #846
• llms: Fix ErrIncompleteEmbeddings Message Typos by @wgeorgecook in #872
• googleai: add support for array type in convertToolSchemaType by @johnknott in #865
• marita: update Authorization by @devalexandre in #864
• Fix typos map key value of the modelToContextSize by @zhanluxianshen in #874
• memory: Add Zep Cloud integration by @paul-paliychuk in #746
• openai: add the field detail in ImageURLContent struct by @douglarek in #853
• weaviate: Fix up weaviate tests by @tmc in #885
• deps: clean up direct dep on zerolog by @tmc in #886
• googleai: Fix: Resolve Type Error in googleai convertTools by @wangjiancn in #888
• openai: enable optional return of usage data while streaming by @wangjiancn in #887
• llms/googleai: Support more option.ClientOption and update google ai sdk by @wangjiancn in #841
• googleai: Remove unused err, fix typo by @tmc in #891
• examples: Correct API key usage for Groq integration by @Ashad001 in #893
• ollama: Fix JSON format bug issue when not streaming by @doslindos in #892

New Contributors

• @h0rv made their first contribution in #577
• @johanbrandhorst made their first contribution in #825
• @ankitraj2330 made their first contribution in #879
• @lwlee2608 made their first contribution in #882
• @codefromthecrypt made their first contribution in #846
• @wgeorgecook made their first contribution in #872
• @johnknott made their first contribution in #865
• @zhanluxianshen made their first contribution in #874
• @paul-paliychuk made their first contribution in #746
• @wangjiancn made their first contribution in #888
• @Ashad001 made their first contribution in #893
• @doslindos made their first contribution in #892

Full Changelog: v0.1.10...v0.1.11-pre.0

v0.1.10

What's Changed

• examples: fix openai toolcall example after refactor by @eliben in #783
• googleai: implement tool support by @eliben in #794
• googleai: support tool calls in googleai/vertexai by @eliben in #795
• googleai: add example of tool calling by @eliben in #796
• examples: update langchaingo to v0.1.9 by @tmc in #800
• llms: add llama3 models to bedrock by @semioz in #805
• examples: Specify embedding model when instantiating LLM in Pinecone example code by @nikhilnarayanan623 in #816
• feat: support gemini pro system prompt by @douglarek in #814
• scraper: fix concurrent scraped links map access by @vdpoora in #813
• ernie: Remove ERNIE-Bot-8K Model by @sxk10812139 in #812
• llms: llms/googleai/option.go With methods added that was not implemented yet by @Mannetjie in #808
• openai: Add metadata fields in OpenAI calls by @lowjiansheng in #811
• llms: Add usability methods to parts types by @tmc in #818
• memory: Add mongodb memory implementation by @CrazyWr in #810
• examples: Add groq completion example by @sklinkert in #804
• embeddings: Add Jina by @devalexandre in #809
• bugfix: incorrect error return in applyClientOptions function by @0xd219b in #807
• feat: bedrock anthropic stream support by @douglarek in #821
• vectorstores/pgvector: support both pgx and pgxpool. by @joeychilson in #802
• memory: add sqlite3 backend by @mvrilo in #798
• embeddings: add voyageai embedder by @joeychilson in #792
• agent: Update openai agent tool calling (#765) by @txfsxq in #772
• llms: Maritaca AI by @devalexandre in #762
• vectorstores: add redisvector implementation by @CrazyWr in #777
• feat: add WithKeepSeparator option for RecursiveCharacter by @zhangi in #721
• docs: re-add llamafile example by @devalexandre in #693
• openai: add toolcalls support when streaming by @ChrisCPoirier in #763

New Contributors

• @nikhilnarayanan623 made their first contribution in #816
• @douglarek made their first contribution in #814
• @vdpoora made their first contribution in #813
• @Mannetjie made their first contribution in #808
• @lowjiansheng made their first contribution in #811
• @sklinkert made their first contribution in #804
• @0xd219b made their first contribution in #807
• @txfsxq made their first contribution in #772
• @zhangi made their first contribution in #721
• @ChrisCPoirier made their first contribution in #763

Full Changelog: v0.1.9...v0.1.10

v0.1.9

What's Changed

• googleai: vertex - upgrade dep version and increase default max tokens by @eliben in #742
• googleai: combine options for googleai and vertex by @eliben in #743
• googleai: add safety/harm threshold settings by @eliben in #744
• github: update lint workflow to newer version of Go by @eliben in #745
• vectorstores/milvus: Update testcontainer image by @devalexandre in #741
• tools/sqldataase: update postgres image by @devalexandre in #740
• chains: Update mysql testcontainer image by @devalexandre in #739
• vectorstores/qdrant: Update testcontainer image by @devalexandre in #737
• doc: fix typo by @XiaoConstantine in #758
• examples: clarify openai-function-call-example by @eliben in #751
• schema: move Chat* types to llms by @eliben in #757
• examples: remove 'replace' lines in go.mod files by @eliben in #782
• schema: remove leftover doc comment part by @eliben in #784
• vectorstores/pinecone: update to latest v0.4.1 by @andrewyang17 in #789
• llms/openai: Fixed no response when streaming func with tools calling by @shawti in #788
• llms/ollama: support 'keep_alive' parameter by @CrazyWr in #787
• fix: fix example path by @LooJee in #780
• tools: serpapi api key option by @AugustDev in #773
• Fix OpenAI createChat bug by @Pramodh-G in #759
• docs: resolve rendering issue with code block in getting started docs by @schneedotdev in #754

New Contributors

• @XiaoConstantine made their first contribution in #758
• @andrewyang17 made their first contribution in #789
• @CrazyWr made their first contribution in #787
• @LooJee made their first contribution in #780
• @AugustDev made their first contribution in #773
• @Pramodh-G made their first contribution in #759
• @schneedotdev made their first contribution in #754

Full Changelog: v0.1.8...v0.1.9

v0.1.8

What's Changed

• Improvements on agents package API by @haochunchang in #551
• llms: add caching functionality for Models by @corani in #564
• httputil: Add httputil package to provide some common helpers by @tmc in #702
• examples: Use new debugging helper in example by @tmc in #703
• examples: Fix up import in openai debugging example by @tmc in #704
• tests: Perform env-var checking skips sooner as to speed up tests by @tmc in #706
• llms/anthropic: adds full support for messages api by @joeychilson in #707
• llms/bedrock: Fixed error when using Claude3 model and giving MessageContent with images by @mashiike in #713
• readme: Include contributors by @tmc in #714
• llms/cloudflare: Implement Cloudflare Workers AI LLM by @rajaiswal in #679
• add new example for OCR using Claude3's Vision feature with Bedrock by @mashiike in #715
• llms: Add mistral hosted inference llm implementation by @aannirajpatel in #717
• vectorstores/weaviate: Update testcontainer image by @tmc in #719
• tooling: Update minimum go version to 1.22, update golangci-lint by @tmc in #722
• openai: Render single text content parts directly by @tmc in #734
• all: set explicit 1.22 version in go.mod by @eliben in #727
• googleai: increase default max tokens setting by @eliben in #726
• examples: Fix and tidy examples, add nvidia example by @tmc in #735
• llms: Implement tool calling, including parallel tool call request support by @tmc in #709
• openai: WithEmbeddingModel option is incorrectly designating the desired model to use for embedding by @devalexandre in #731
• llms: Add Seed option to all supporting backends by @devalexandre in #732

New Contributors

• @joeychilson made their first contribution in #707
• @mashiike made their first contribution in #713
• @rajaiswal made their first contribution in #679
• @aannirajpatel made their first contribution in #717

Full Changelog: v0.1.7...v0.1.8

v0.1.7

What's Changed

• llms/openai: Azure: add check for model parameter by @janezkenda in #691
• llms/openai: Optionally supply ResponseFormat by @shawti in #690

New Contributors

• @janezkenda made their first contribution in #691
• @shawti made their first contribution in #690

Full Changelog: v0.1.6...v0.1.7

v0.1.6

What's Changed

• feat: add new textspliter option: lenfunc by @whyiug in #609
• googleai: fix options need add default value by @Abirdcfly in #625
• chains: add test with GoogleAI by @eliben in #628
• chains: fix issue with overriding defaults in chainCallOption by @eliben in #632
• chains: Update comments, bump example dependencies and clarify chain example by @eliben in #633
• examples: Fix llm_math_chain example by @eliben in #634
• chains: fix add ignore StreamingFunc by @Abirdcfly in #639
• chore: Pinning chroma-go ahead of major new release by @tazarov in #640
• docs: Add huggingface documentation by @devalexandre in #641
• agents: Fix agent stream callback by @devinyf in #614
• feat: add JSON format option to ollama by @corani in #647
• feat: run integration tests for vector databases using testcontainers-go modules by @mdelapenya in #648
• embeddings: Add Amazon Bedrock embeddings by @sansmoraxz in #643
• vectorstores: Add support for OpenAI Organization ID header in Chroma by @AshDevFr in #646
• change mrkl-prompt to python version by @devinyf in #653
• examples: Point examples to v0.1.5 by @tmc in #656
• vectorstores: fix pgvector issues and add more test by @Abirdcfly in #617
• googleai: return err not log.Fatal when stream get error by @Abirdcfly in #663
• docs: fixup a bunch of links by @codyoss in #659
• docs: Add example about retriever by @devalexandre in #651
• Create FUNDING.yml by @tmc in #669
• docs: Add documentation for splitters by @devalexandre in #649
• tool: sql_database generate sql-query filter redundant text by @devinyf in #612
• llms/anthropic: Implement WithBaseURL and WithHTTPClient for Anthropic LLM by @GRVYDEV in #671
• This example hangs if the JSON data is either marked as human or AI, it only works for me when marked as system by @timmattison in #670
• vectorstores/milvus: add WithMetricType by @chenquan in #674
• chore: Chroma dependency update by @tazarov in #676
• vectorstores/milvus: fixed the bug that the collection was created in… by @chenquan in #675
• Implementation proposal for using llamafile by @devalexandre in #677
• documentloaders: add AssemblyAI document loader by @marcusolsson in #668
• llms/bedrock: Add AWS Bedrock LLMs by @sansmoraxz in #666

New Contributors

• @tazarov made their first contribution in #640
• @devalexandre made their first contribution in #641
• @sansmoraxz made their first contribution in #643
• @AshDevFr made their first contribution in #646
• @codyoss made their first contribution in #659
• @GRVYDEV made their first contribution in #671
• @timmattison made their first contribution in #670
• @chenquan made their first contribution in #674
• @marcusolsson made their first contribution in #668

Full Changelog: v0.1.5...v0.1.6

v0.1.5

What's Changed

• docs: Fix vertex example refs by @tmc in #595
• agents: [OneShotZeroAgent] Change regex to allow multi-line inputs by @terwey in #593
• vectorstores: Added the option to specify a HNSW index for improving performance by @pieterclaerhout in #589
• callbacks: Implement logging for text types, fixes #596 by @terwey in #597
• vectorstores: fix pgvector return nothing when embedding table has ve… by @Abirdcfly in #548
• agents: improve TestConversationalWithMemory testing by @tmc in #602
• README: add link to new blog post about langchaingo by @eliben in #606
• fix: stackoverflow caused by textsplitter by @whyiug in #608
• chroma: update chroma-go functions to pass context by @dddbliss in #610
• tools: run integration tests with testcontainers-go by @mdelapenya in #519
• llms: Remove omitempty from temperature params by @tmc in #621

New Contributors

• @terwey made their first contribution in #593
• @dddbliss made their first contribution in #610

Full Changelog: v0.1.4...v0.1.5

v0.1.4

What's Changed

• Add GenerationInfo to GenerateContent by @eliben in #488
• Clean up GeneratePrompt -- unused code, and simplify caller by @eliben in #489
• fix unmarshalling of chatmessage by @ctrlaltdel121 in #491
• feature: add new gemini llm supporting new experimental interface by @mrothroc in #497
• Add Ollama test and remove custom prompt for chat by @eliben in #495
• ollama: Clean up Ollama options by @eliben in #496
• googleai: set up options and embeddings by @eliben in #498
• Extend Model / GenerateContent interface to support multi-content chats by @eliben in #500
• Add multi-turn tests for new GenerateContent capabilities by @eliben in #501
• ollama: implement Model interface with GenerateContent by @eliben in #502
• Test streaming support in GenerateContent in openai and ollama by @eliben in #503
• googleai: implement streaming and some generation options by @eliben in #504
• Add function invocation into GenerateContent responses by @eliben in #512
• googleai: fix streaming termination by @eliben in #515
• Add callback handler method for GenerateContent start and end by @eliben in #514
• Finalize the majority of the model/LLM interface refactoring by @eliben in #521
• all: clean up now unused types Generation and LLMResult by @eliben in #523
• all: clean up and update comments by @eliben in #524
• docs: add example to embeddings by @haochunchang in #518
• Deprecate the LLM interface by @eliben in #525
• googleai: fix test comparisons to use case-insensitive regexps by @eliben in #528
• googleai: add more model parameters and update comments by @eliben in #529
• googleai: move downloadImageData to separate file by @eliben in #530
• Link to LangChain Python (deserves credit) by @carsonkahn-external in #509
• vectorstores: Weaviate make additional fields configurable by @jvgrootveld in #522
• ollama: add WithHttpClient option by @elnoro in #505
• googleai: propagate more options and add test by @eliben in #532
• llms: configure multiple response candidates for models that support them by @eliben in #533
• vectorstores: added pgvector example by @elnoro in #534
• weaviate: add option to deduplicate documents while adding by @corani in #527
• weaviate: honor vectorstores.WithEmbedder by @corani in #526
• vectorstores: pgvector add option WithConn by @Abirdcfly in #499
• memory: fix cutMessages in ConversationWindowBuffer by @Abirdcfly in #506
• textsplitter: render code blocks and reference links by @corani in #493
• templates: Add f-string based templating implementation by @chyroc in #479
• agents: add openai functions agent by @lujingxuansc in #459
• bug fix: openai_functions_agent not compatible with the new llms.Model interface by @devinyf in #536
• llms: rename convenience function to GenerateFromSinglePrompt by @eliben in #537
• googleai: test GenerateFromSinglePrompt convenience function by @eliben in #538
• docs: add pgvector page by @elnoro in #535
• googleai: add initial Vertex (GCP) implementation of Model by @eliben in #540
• googleai: move the PaLM provider into googleai by @eliben in #541
• embeddings: add cybertron local embeddings by @corani in #531
• googleai: add embeddings to vertex by @eliben in #546
• googleai: refactor to better separate generated code by @eliben in #547
• huggingface: allows to set custom inference endpoint by @mheers in #545
• cybertron: add example for cybertron embeddings by @corani in #543
• Move utility function by @eliben in #552
• googleai: add codegen tool to generate vertex.go from googleai.go by @eliben in #553
• googleai: clean up comments and order files with consistent naming by @eliben in #554
• llms: officially deprecate Call method on the interface by @eliben in #555
• llms: Add convenience helper for text parts by @tmc in #549
• feat: Added Qdrant vectorstore support by @Anush008 in #544
• googleai: rename constructors to New for consistency by @eliben in #570
• llms: convenience methods to create Parts by @eliben in #571
• embeddings: fix variable typo in huggingface_test by @ion232 in #569
• vectorstores: Wrapped the PGVector store init function in a single transaction by @pieterclaerhout in #563
• vectorstores: Added a unique constraint for the collection name when using PGVector by @pieterclaerhout in #559
• vectorstores: Added an option to PGVector to set the collection metadata by @pieterclaerhout in #557
• docs: add example and update conceptual guide URL for agents by @haochunchang in #539
• examples: add new vertex and googleai examples by @eliben in #572
• docs: qdrant-examples by @Anush008 in #568
• vectorstores: Implemented support for deduplicating documents when adding for PGVector by @pieterclaerhout in #566
• Update openai examples to new APIs by @eliben in #573
• examples: use llms.TextPart helper when it's useful by @eliben in #574
• bump examples dependency to v0.1.4-alpha.0 by @eliben in #575
• examples: update all to new llms interfaces by @eliben in #576
• vertex: add tests by @eliben in #582
• googleai: share tests between the two main providers by @eliben in #584
• googleai: add tests by @eliben in #586
• vectorstores: Added the missing comment on the exported WithCollectionMetadata function by @pieterclaerhout in #578
• vectorstores: Added an option to specify the vector dimensions for the embeddings by @pieterclaerhout in #579
• googleai: add test with BinaryPart image by @eliben in #587
• readme: update example with non-deprecated call by @eliben in #588
• vectorstores: Removed the custom ID option for embeddings when using PGVector by @pieterclaerhout in #590

New Contributors

• @ctrlaltdel121 made their first contribution in #491
• @mrothroc made their first contribution in #497
• @haochunchang made their first contribution in #518
• @carsonkahn-external made their first contribution in #509
• @jvgrootveld made their first contribution in #522
• @elnoro made their first contribution in #505
• @lujingxuansc made their first contribution in #459
• @devinyf made their first contribution in #536
• @mheers made their first contribution in #545
• @Anush008 made their first contribution in #544
• @ion232 made their first contribution in #569
• @pieterclaerhout made their first contribution in #563

Full Changelog: v0.1.3...v0.1.4