book

Learning LangChain

by Mayo Oshin, Nuno Campos

February 2025

Beginner to intermediate

296 pages

6h 39m

English

O'Reilly Media, Inc.

Book available

Read now

Unlock full access

Includes

Includes Quizzes

Brief Primer on LLMsInstruction-Tuned LLMsDialogue-Tuned LLMsFine-Tuned LLMsBrief Primer on PromptingZero-Shot PromptingChain-of-ThoughtRetrieval-Augmented GenerationTool CallingFew-Shot PromptingLangChain and Why It’s ImportantWhat to Expect from This BookConventions Used in This BookUsing Code ExamplesO’Reilly Online LearningHow to Contact UsAcknowledgments
Getting Set Up with LangChainUsing LLMs in LangChainMaking LLM Prompts ReusableGetting Specific Formats out of LLMsJSON OutputOther Machine-Readable Formats with Output ParsersAssembling the Many Pieces of an LLM ApplicationUsing the Runnable InterfaceImperative CompositionDeclarative CompositionSummary
The Goal: Picking Relevant Context for LLMsEmbeddings: Converting Text to NumbersEmbeddings Before LLMsLLM-Based EmbeddingsSemantic Embeddings ExplainedConverting Your Documents into TextSplitting Your Text into ChunksGenerating Text EmbeddingsStoring Embeddings in a Vector StoreGetting Set Up with PGVectorWorking with Vector StoresTracking Changes to Your DocumentsIndexing OptimizationMultiVectorRetrieverRAPTOR: Recursive Abstractive Processing for Tree-Organized RetrievalColBERT: Optimizing EmbeddingsSummary
Introducing Retrieval-Augmented GenerationRetrieving Relevant DocumentsGenerating LLM Predictions Using Relevant DocumentsQuery TransformationRewrite-Retrieve-ReadMulti-Query RetrievalRAG-FusionHypothetical Document EmbeddingsQuery RoutingLogical RoutingSemantic RoutingQuery ConstructionText-to-Metadata FilterText-to-SQLSummary
Building a Chatbot Memory SystemIntroducing LangGraphCreating a StateGraphAdding Memory to StateGraphModifying Chat HistoryTrimming MessagesFiltering MessagesMerging Consecutive MessagesSummary
Architecture #1: LLM CallArchitecture #2: ChainArchitecture #3: RouterSummary
The Plan-Do LoopBuilding a LangGraph AgentAlways Calling a Tool FirstDealing with Many ToolsSummary
ReflectionSubgraphs in LangGraphCalling a Subgraph DirectlyCalling a Subgraph with a FunctionMulti-Agent ArchitecturesSupervisor ArchitectureSummary
Structured OutputIntermediate OutputStreaming LLM Output Token-by-TokenHuman-in-the-Loop ModalitiesMultitasking LLMsSummary
PrerequisitesInstall DependenciesLarge Language ModelVector StoreBackend APICreate a LangSmith AccountUnderstanding the LangGraph Platform APIData ModelsFeaturesDeploying Your AI Application on LangGraph PlatformCreate a LangGraph API ConfigTest Your LangGraph App LocallyDeploy from the LangSmith UILaunch LangGraph StudioSecuritySummary

Testing Techniques Across the LLM App Development CycleThe Design Stage: Self-Corrective RAGThe Preproduction StageCreating DatasetsDefining Your Evaluation CriteriaRegression TestingEvaluating an Agent’s End-to-End PerformanceProductionTracingCollect Feedback in ProductionClassification and TaggingMonitoring and Fixing ErrorsSummary
Interactive ChatbotsCollaborative Editing with LLMsAmbient ComputingSummary

Content preview from Learning LangChain

Chapter 6. Agent Architecture

Building on the architectures described in Chapter 5, this chapter will cover what is perhaps the most important of all current LLM architectures, the agent architecture. First, we introduce what makes LLM agents unique, then we show how to build them and how to extend them for common use cases.

In the artificial intelligence field, there is a long history of creating (intelligent) agents, which can be most simply defined as “something that acts,” in the words of Stuart Russell and Peter Norvig in their Artificial Intelligence (Pearson, 2020) textbook. The word acts actually carries a little more meaning than meets the eye:

Acting requires some capacity for deciding what to do.
Deciding what to do implies having access to more than one possible course of action. After all, a decision without options is no decision at all.
In order to decide, the agent also needs access to information about the external environment (anything outside of the agent itself).

So an agentic LLM application must be one that uses an LLM to pick from one or more possible courses of action, given some context about the current state of the world or some desired next state. These attributes are usually implemented by mixing two prompting techniques we first met in the Preface:

Tool calling: Include a list of external functions that the LLM can make use of in your prompt (that is, the actions it can decide to take) and provide instructions on how to format its choice ...