Text Generation for Applications Using Vertex AI Studio

(1 day)

 

Generative AI is being used to develop new products and services across multiple industries, such as personalized marketing communications, chatbots for interacting with customers, and virtual assistants. For example, It can also be used to create chatbots that can answer customer questions and provide customer support.

In this course, you will explore the use of text generation models using Vertex AI Studio on Vertex AI and learn how to incorporate those models into your application using the PaLM API and client libraries. You will learn how to design and tune prompts to ensure the best outputs for your applications and discuss how to fine-tune foundational models to improve model output quality.

Course Objectives

  • Understand Vertex AI generative AI options for your applications
  • Explore Vertex AI Studio to interact with foundation models
  • Design and tune prompts for your generative AI use cases
  • Implement the PaLM API into your applications using the Python SDK
  • Fine-tune foundation model weights to improve model output quality

Audience

Application developers leveraging generative AI in their applications and machine learning practitioners supporting the development of GenAI-powered applications.

Prerequisites

Basic understanding of one or more of the following:

  • Programming in Python
  • Leveraging APIs in applications
  • Basic familiarity with Google Cloud and Vertex AI as covered in the Google Cloud Big Data and Machine Learning Fundamentals course

Course Outline

Module 1: Generative AI on Vertex AI

  • Vertex AI on Google Cloud
  • Generative AI options on Google Cloud
  • Introduction to the Course Use Case (Text Generation)

Module 2: Vertex AI Studio

  • Introduction to Vertex AI Studio
  • Available models and use cases
  • Designing and testing prompts in the Google Cloud console
  • Data governance in Vertex AI Studio
  • Lab: Getting Started with Vertex AI Studio’s User Interface

Module 3: Prompt Design

  • Why is prompt design so important?
  • Zero-shot vs. few-shot prompting
  • Providing additional context and instruction-tuning
  • Best practices
  • Lab: Question Answering with Generative Models on Vertex AI

Module 4: Implementing the PaLM API

  • Lab: Getting Started with the Vertex AI PaLM API and Python SDK
  • Introduction to the PaLM API
  • Utilizing generative models using the Python SDK
  • Understanding model parameters for text generation
  • Lab: Use the PaLM API to Integrate GenAI into Applications

Module 5: Fine-Tuning Models

  • Scenarios to use model tuning
  • Workflow for model tuning
  • Preparing your model tuning dataset
  • Create a model tuning job
  • Loading a tuned model
  • Demo: Fine-Tuning Models for Your Specific Use Case