Build a question box that answers from the law itself, puts the section under every answer, and refuses rather than guesses.
Most retrieval-augmented generation tutorials stop at "embed some documents and ask a model." This book goes further. You build, by hand and one stage at a time, the pipeline behind a real, deployed ask box that answers learner drivers' road-law questions from each state's own manual and vehicle code. Every chapter compares your version with the production code it is drawn from.
What you'll build
Who this book is for
Developers who can read TypeScript and want to put a language model into a product where a wrong answer costs something. There is no agent loop here: your code decides every next step, and each model call is one request and one reply.
Every listing is complete and runnable, and every figure about the production pipeline traces to its source, a design document or a pull request. You finish with a pipeline you understand line by line and a method for deciding what a model may be trusted to do.
Le informazioni nella sezione "Riassunto" possono far riferimento a edizioni diverse di questo titolo.
I tuoi diritti di garanzia legale: Ogni consumatore nell'UE beneficia di almeno 2 anni di garanzia legale per qualsiasi bene di consumo venduto nell'UE. Scopri di più
Da: PBShop.store US, Wood Dale, IL, U.S.A.
PAP. Condizione: New. New Book. Shipped from UK. THIS BOOK IS PRINTED ON DEMAND. Established seller since 2000. Codice articolo L0-9798164069698
Quantità: Più di 20 disponibili
Da: Grand Eagle Retail, Bensenville, IL, U.S.A.
Paperback. Condizione: new. Paperback. Build a question box that answers from the law itself, puts the section under every answer, and refuses rather than guesses.Most retrieval-augmented generation tutorials stop at "embed some documents and ask a model." This book goes further. You build, by hand and one stage at a time, the pipeline behind a real, deployed ask box that answers learner drivers' road-law questions from each state's own manual and vehicle code. Every chapter compares your version with the production code it is drawn from.What you'll buildA single, well-shaped request to a model, then an answer written only from the laws you give it, naming the ones it usedA check written in plain code, with no model in it, that refuses any answer citing a law it wasn't given or stating a number no cited law containsRetrieval by words (BM25) and by meaning (embeddings), fused with reciprocal rank fusionA small model that rephrases the learner's question in the law's own vocabulary, and a reranker that chooses the few laws worth readingLookups that pull in the whole section, the court's fine schedule and the agency's own pageA corpus fetched verbatim and dated, a pipeline that degrades step by step when a stage fails, an evaluation that measures the whole thing, and tests that never call the modeWho this book is forDevelopers who can read TypeScript and want to put a language model into a product where a wrong answer costs something. There is no agent loop here: your code decides every next step, and each model call is one request and one reply.Every listing is complete and runnable, and every figure about the production pipeline traces to its source, a design document or a pull request. You finish with a pipeline you understand line by line and a method for deciding what a model may be trusted to do. This item is printed on demand. Shipping may be from multiple locations in the US or from the UK, depending on stock availability. Codice articolo 9798164069698
Quantità: 1 disponibili
Da: California Books, Miami, FL, U.S.A.
Condizione: New. Print on Demand. Codice articolo I-9798164069698
Quantità: Più di 20 disponibili
Da: PBShop.store UK, Fairford, GLOS, Regno Unito
PAP. Condizione: New. New Book. Delivered from our UK warehouse in 4 to 14 business days. THIS BOOK IS PRINTED ON DEMAND. Established seller since 2000. Codice articolo L0-9798164069698
Quantità: Più di 20 disponibili
Da: CitiRetail, Stevenage, Regno Unito
Paperback. Condizione: new. Paperback. Build a question box that answers from the law itself, puts the section under every answer, and refuses rather than guesses.Most retrieval-augmented generation tutorials stop at "embed some documents and ask a model." This book goes further. You build, by hand and one stage at a time, the pipeline behind a real, deployed ask box that answers learner drivers' road-law questions from each state's own manual and vehicle code. Every chapter compares your version with the production code it is drawn from.What you'll buildA single, well-shaped request to a model, then an answer written only from the laws you give it, naming the ones it usedA check written in plain code, with no model in it, that refuses any answer citing a law it wasn't given or stating a number no cited law containsRetrieval by words (BM25) and by meaning (embeddings), fused with reciprocal rank fusionA small model that rephrases the learner's question in the law's own vocabulary, and a reranker that chooses the few laws worth readingLookups that pull in the whole section, the court's fine schedule and the agency's own pageA corpus fetched verbatim and dated, a pipeline that degrades step by step when a stage fails, an evaluation that measures the whole thing, and tests that never call the modeWho this book is forDevelopers who can read TypeScript and want to put a language model into a product where a wrong answer costs something. There is no agent loop here: your code decides every next step, and each model call is one request and one reply.Every listing is complete and runnable, and every figure about the production pipeline traces to its source, a design document or a pull request. You finish with a pipeline you understand line by line and a method for deciding what a model may be trusted to do. This item is printed on demand. Shipping may be from our UK warehouse or from our Australian or US warehouses, depending on stock availability. Codice articolo 9798164069698
Quantità: 1 disponibili