|
|
|
|
| LEADER |
05591nmm a2200385 u 4500 |
| 001 |
EB002207462 |
| 003 |
EBX01000000000000001344663 |
| 005 |
00000000000000.0 |
| 007 |
cr||||||||||||||||||||| |
| 008 |
240503 ||| eng |
| 050 |
|
4 |
|a QA76.76.A65
|
| 100 |
1 |
|
|a Kansal, Aarushi
|
| 245 |
0 |
0 |
|a Building generative AI-powered apps
|b a hands-on guide for developers
|c Aarushi Kansal
|
| 250 |
|
|
|a [First edition]
|
| 260 |
|
|
|a New York, NY
|b Apress
|c 2024
|
| 300 |
|
|
|a 175 pages
|b illustrations
|
| 505 |
0 |
|
|a Includes bibliographical references and index
|
| 505 |
0 |
|
|a LangSmith Config -- Run a Simple App -- The Pirate App -- Setting Up -- Feedback -- Why? -- How? -- Feedback Collection -- Analysis -- Datasets -- Evaluations -- Summary -- Chapter 8: Prompt Engineering Techniques -- What Is Prompt Engineering? -- Chain of Thought -- What Is It? -- Design -- Zero-Shot CoT -- Tree of Thought -- Design -- Structure of the ToT Framework -- Self-Evaluation and Critique -- Thought Decomposition and Expansion -- The Role of the Evaluator -- Deliberate Reasoning -- Backtracking in the ToT Process -- Tree Search Techniques -- Dual Roles of the AI Model -- Chain of Note
|
| 505 |
0 |
|
|a Loading Your Data -- Transforming Your Structured Data -- Embeddings and Storage -- Memory -- What's Next? -- Summary -- Chapter 3: Chains, Tools and Agents -- High-Level Concepts -- Chains -- Tools -- Building a Custom Tool -- Agents -- ReAct -- zero-shot-react-description -- conversational-react-description -- react-docstore -- self-ask-with-search -- The App -- Summary -- Chapter 4: Guardrails and AI: Building Safe + Controllable Apps -- Why Guardrails? -- NeMo Guardrails -- Keeping Your Bot on Topic -- Moderating Your Bot -- Preventing Hallucination -- Implementing Guardrails
|
| 505 |
0 |
|
|a Intro -- Table of Contents -- About the Author -- About the Technical Reviewer -- Chapter 1: Introduction to Generative AI -- What Is Generative AI? -- Model Types -- Transformers Explained -- RNNs -- LSTMs -- Transformers -- Diffusion Explained -- The Core Idea -- How Diffusion Models Work -- What's Next? -- Summary -- Chapter 2: LangChain: Your Swiss Army Knife -- The Whats and Whys -- Chatbot -- Memory -- Types of Memory -- Retrieval -- Diving into RAG -- Embeddings Explained -- Vector Embeddings -- Vector Stores Explained -- What Else Is RAG Good For? -- The App -- Prerequisites
|
| 505 |
0 |
|
|a Reward Modeling (RM) -- Reinforcement Learning Algorithms -- Human Preference Comparison -- Iterative Training -- AI Alignment and Safety -- Challenges and Considerations -- PEFT -- How Does PEFT Work -- Low-Rank Adaptation (LoRA) -- Decomposing LoRA's Mechanism -- Summary -- Chapter 6: Finetuning: Hands on -- Refresher -- 4-Bit NormalFloat (NF4) Data Type -- Double Quantization -- Paged Optimizers -- What Is Llama 2? -- Fine-Tuning -- Setup -- Llama 2 Model -- Formatting -- Summary -- Chapter 7: Monitoring -- What Is LangSmith? -- Examples? -- Why? -- Quickstart -- Getting a LangSmith Key
|
| 505 |
0 |
|
|a Keeping the Bot on Topic -- Blocking a User -- Actions -- Using This Config -- Under the Hood -- User Interaction -- Next Step -- BotIntent -- Embeddings -- Summary -- Chapter 5: Finetuning: The Theory -- Let's Talk Foundational Models -- The Whys of Fine-Tuning? -- The Whats of Fine-Tuning -- Starting Point: The Pre-trained Model -- Preparation for Fine-Tuning -- Fine-Tuning Process -- During Training -- Fine-Tuning Strategies -- After Fine-Tuning -- Network Level Changes -- The Hows of Fine-Tuning -- Reinforcement Learning with Human Feedback (RLHF) -- Supervised Fine-Tuning (SFT)
|
| 653 |
|
|
|a Logiciels d'application / Développement
|
| 653 |
|
|
|a Artificial intelligence / http://id.loc.gov/authorities/subjects/sh85008180
|
| 653 |
|
|
|a Intelligence artificielle
|
| 653 |
|
|
|a artificial intelligence / aat
|
| 653 |
|
|
|a Application software / Development / http://id.loc.gov/authorities/subjects/sh95009362
|
| 041 |
0 |
7 |
|a eng
|2 ISO 639-2
|
| 989 |
|
|
|b OREILLY
|a O'Reilly
|
| 024 |
8 |
|
|a 10.1007/979-8-8688-0205-8
|
| 776 |
|
|
|z 9798868802041
|
| 776 |
|
|
|z 9798868802058
|
| 856 |
4 |
0 |
|u https://learning.oreilly.com/library/view/~/9798868802058/?ar
|x Verlag
|3 Volltext
|
| 082 |
0 |
|
|a 005.3
|
| 520 |
|
|
|a Generative AI has gone beyond the responsibility of researchers and data scientists and is being used by production engineers. However, there is a lot of confusion where to get started when building an end-to-end app with generative AI. This book consolidates core models, frameworks, and tools into a single source of knowledge. By providing hands-on examples, the book takes you through the generative AI ecosystem to build applications for production. The book starts with a brief and accessible introduction to transformer models before delving into some of the most popular large language models and diffusions models (image generation). These models are the foundations of both AI and your potential new apps. You will then go through various tools available to work with these models, starting with Langchain, a framework to develop foundational models, which is the next building block you should grasp after understanding generative AI models. The next chapters cover databases, caching, monitoring, etc., which are the topics necessary to build larger-scale applications. Real-world examples using these models and tools are included. By the end of this book, you should be able to build end-to-end apps that are powered by generative AI. You also should be able to apply the tools and techniques taught in this book to your use cases and business
|