0.15.4

What's Changed

• Improve Database tool by @andreibondarev in #743
• Allow explicitly setting tool_choice on the Assistant instance by @andreibondarev in #744
• Add support for bulk embedding in Ollama by @bricolage in #735
• fix: LLM::Azure not applying full default_options by @namiwang in #746
• Langchain::Assistant supports MistralAI LLM by @andreibondarev in #748
• 0.15.4 by @andreibondarev in #749

New Contributors

• @namiwang made their first contribution in #746

Full Changelog: 0.15.3...0.15.4

0.15.3

What's Changed

• Fix OpenAI#embed when text-embedding-ada-002 is used by @andreibondarev in #741
• 0.15.3 by @andreibondarev in #742

Full Changelog: 0.15.2...0.15.3

0.15.2

What's Changed

• Add Assistant#add_messages() method and fix Assistant#messages= method by @andreibondarev in #731
• Update Gemini Key instructions by @palladius in #734
• Bump rexml from 3.3.3 to 3.3.6 by @dependabot in #736
• 0.15.2 by @andreibondarev in #737

Full Changelog: 0.15.1...0.15.2

0.15.1

What's Changed

• drop Colorize due to GPL license by @Azdaroth in #727
• Improve Langchain::Tool::Weather tool by @andreibondarev in #728
• remove to_bool gem by @Azdaroth in #729
• 0.15.1 by @andreibondarev in #730

New Contributors

• @Azdaroth made their first contribution in #727

Full Changelog: 0.15.0...0.15.1

0.15.0

What's Changed

• Rewrite Assistant tools definition for easier configuration by @dghirardo in #708
• Get rid of requiring activesupport and add deep_merge gem by @andreibondarev in #709
• Improvements to Langchain::Assistant support for Ollama by @andreibondarev in #711
• Update default chat completions model to gpt-4o-mini by @edudepetris in #712
• Update CHANGELOG.md by @andreibondarev in #713
• Introducing a factory/adapter patterns to clean up the Langchain::Assistant supporting multiple different LLMs. by @andreibondarev in #714
• Fix Langchain::Assistant::LLM::Adapters::Ollama class by @andreibondarev in #715
• Revert Style/ArgumentsForwarding Rubocop rule by @andreibondarev in #703
• Add Support for Ollama WebUI Security by @rience in #721
• Bump rexml from 3.3.1 to 3.3.3 by @dependabot in #719
• Weaviate initializer does not require api_key by @andreibondarev in #722
• Fix GoogleGemini#chat method by @andreibondarev in #723
• Fix Langchain::Assistant when llm is Anthropic by @andreibondarev in #724
• 0.15.0 by @andreibondarev in #725

New Contributors

• @edudepetris made their first contribution in #712
• @rience made their first contribution in #721

Full Changelog: 0.14.0...0.15.0

0.14.0

What's Changed

• Update milvus.rb by @andreibondarev in #691
• Bump anthropic gem by @andreibondarev in #692
• Track tokens count at the Assistant level by @dferrazm in #693
• Instructions are ignored for Anthropic if tools are not used by @victorivanov in #695
• Issue 697 - Show API error when calling .complete on Anthropic client by @Jbrito6492 in #698
• Delete token_length validators by @andreibondarev in #704
• Assistant that does Function Calling with the Ollama Mistral model by @andreibondarev in #702
• Fix for Ollama Assistant that does not use tools by @andreibondarev in #707
• 0.14.0 by @andreibondarev in #706

New Contributors

• @victorivanov made their first contribution in #695
• @Jbrito6492 made their first contribution in #698

Full Changelog: 0.13.5...0.14.0

0.13.5

What's Changed

• ids need to be passed as strings to Chroma#remove_texts() by @andreibondarev in #673
• Update loader.rb by @michelson in #676
• fix: GoogleGemini generation_config param by @mazenkhalil in #665
• State machine for Assistant's execution phases by @dghirardo in #677
• chore: update replicate.rb by @eltociear in #680
• Swap test pdf file by @andreibondarev in #687
• Add Milvus#remove_texts() method by @andreibondarev in #689
• 0.13.5 by @andreibondarev in #690

New Contributors

• @mazenkhalil made their first contribution in #665
• @dghirardo made their first contribution in #677
• @eltociear made their first contribution in #680

Full Changelog: 0.13.4...0.13.5

0.13.4

What's Changed

• accept non ascii url by @michelson in #660
• Fix standardrb gem by @andreibondarev in #663
• Be explicit about unsupported file types by @andreibondarev in #664
• Update URL to CHANGELOG by @ahx in #666
• Update comments for hnswlib similarity_search and similarity_search_by_vector methods by @yash-learner in #669
• Fix Chroma#remove_texts method by @andreibondarev in #670
• Fix News Retriever non-UTF-8 characters by @andreibondarev in #671
• 0.13.4 by @andreibondarev in #672

New Contributors

• @ahx made their first contribution in #666
• @yash-learner made their first contribution in #669

Full Changelog: 0.13.3...0.13.4

0.13.3

What's Changed

• aws bedrock streaming by @agomezcampero in #630
• Remove activesupport since it's not needed by @andreibondarev in #639
• Langchain::Assistant should not require thread parameter by @andreibondarev in #641
• OpenAI response with tools by @agomezcampero in #645
• Update README.md by @palladius in #647
• Exclude tiktoken_ruby from automatic max_tokens validation by @andreibondarev in #649
• New Langchain::Tool::Tavily tool by @andreibondarev in #650
• Support Ollama streamed responses by @dferrazm in #644
• fix undefined text completions in anthropic response by @agomezcampero in #653
• fix tool use inputs when streaming aws bedrock by @agomezcampero in #654
• fix bug when empty input json by @agomezcampero in #655
• 0.13.3 by @andreibondarev in #656

New Contributors

• @agomezcampero made their first contribution in #630

Full Changelog: 0.13.2...0.13.3

0.13.2

What's Changed

• Fix NewsRetriever tool by @andreibondarev in #625
• Adding GCP scope to auth in google_vertex_ai.rb by @palladius in #623
• Bump rexml from 3.2.6 to 3.2.8 by @dependabot in #626
• feat(hugging_face): Add support for multiple models and dimensions by @bricesanchez in #624
• Langchain::Assistant support for Anthropic LLM by @andreibondarev in #627
• Add support for .xls (Excel 1997-2004) by @bricolage in #599
• Add embedding method for Langchain::LLM::GoogleGemini by @swerner in #631
• 0.13.2 by @andreibondarev in #632

New Contributors

• @bricesanchez made their first contribution in #624
• @swerner made their first contribution in #631

Full Changelog: 0.13.1...0.13.2