releases like these are always fun cause you get a full overview of all the langchain components… vectordb, llms, embeddings, document loaders, tools, databases
@hwchase17
-
Advanced Techniques for LLM Data Connection Beyond Semantic Search
By
–
When connecting LLMs to your data, semantic search is a great first pass – but to get even better performance you can turn to more advanced techniques: Contextual Compression
Self Query
Time weighting Going live in 30m to talk about this and more! -
LangChain Integrates Nebula Graph for Natural Language Database Queries
By
–
Question/Answering over Nebula Graph Graph databases have a lot of complex information & relationships in them… @langchain makes it easy to interact with them in natural language! Big thanks to @wey_gu for adding @NebulaGraph integration Docs: https://
python.langchain.com/en/latest/modu
les/chains/examples/graph_nebula_qa.html
… -
Self Query Technique Improves Retrieval in JavaScript
By
–
Self query is awesome technique for improving retrieval – glad to see it making its way to JS
-
LangChain integrates Vectara for advanced retrieval capabilities
By
–
LangChain x Vectara The @vectara platform is a fully managed retrieval service, allowing for high quality retrieval that can we accessed via an API This can now help power your @langchain app! Thanks to @awadallah @ofermend for the integration
-
Deterministic Triggers vs Agent-Driven Event Handling
By
–
probably something different. these are triggered deterministically, you probably want something like you described to be triggered as needed by the agent
-
LangChain Prompt Composition Features for Reusability
By
–
Python Documentation: https://
python.langchain.com/en/latest/modu
les/prompts/prompt_templates/examples/prompt_composition.html
… JS/TS Documentation: https://
js.langchain.com/docs/modules/p
rompts/prompt_templates/prompt_composition
… We want to make prompts as composable and reusable as possible – would love feedback on this implementation! -
PipelinePrompts formatting order and application guide
By
–
When `.format` is called, the PipelinePrompts are first formatted in order, and are then used in future formatting steps with their respective `name` arguments Finally, the FinalPrompt.format is called using any previously formatted values as neccesary
-
PipelinePrompt: Sequencing Prompt Templates for AI Workflows
By
–
The way this works is you define a `PipelinePrompt` with two components: – FinalPrompt: the final prompt template to be formatted
– PipelinePrompts: a sequence of tuples of (name, PromptTemplate) The `name` argument is how the formatted prompt will be passed to future prompts