Complete the Evaluate Gen AI model and agent performance skill badge to demonstrate your ability to use the Gen AI evaluation service. You will evaluate models to select the best model for a given task, compare models against each other and evaluate the performance of agents. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete the assessment challenge lab, to receive a skill badge that you can share with your network. When you complete this course, you can earn the badge displayed here and claim it on Credly! Boost your cloud career by showing the world the skills you have developed!
This course delves into the complexities of assessing the quality of large language model outputs. It examines the challenges enterprises face due to the subjective and sometimes incorrect nature of LLM responses, including hallucinations and inconsistent results. The course introduces various evaluation metrics for different tasks like classification, text generation, and question answering, such as Accuracy, Precision, Recall, F1 score, ROUGE, BLEU, and Exact Match. It also explores evaluation methods offered by Vertex AI LLM Evaluation Services, including computation-based, autorater, and human evaluation, providing insights into their application and benefits. Finally, the module covers how to unit test LLM applications within Vertex AI.
Evaluation is important at every step of your Gen AI development process. In this course you will learn how to evaluate gen AI agents built using agent frameworks.
This lab tests your ability to develop a real-world Generative AI Q&A solution using a RAG framework. You will use Firestore as a vector database and deploy a Flask app as a user interface to query a food safety knowledge base.
Learn how to build your own Retrieval-Augmented Generation (RAG) solutions for greater control and flexibility than out-of-the-box implementations. Create a custom RAG solution using Vertex AI APIs, vector stores, and the LangChain framework.
Model Garden is a model library that helps you discover, test, and deploy models from Google and Google partners. Learn how to explore the available models and select the right ones for your use case. And how to deploy and interact with Model Garden models through the Google Cloud console and APIs.
Model tuning is an effective way to customize large models to your tasks. It's a key step to improve the model's quality and efficiency. Model tuning provides benefits such as higher quality results for your specific tasks and increased model robustness. You learn some of the tuning options available in Vertex AI and when to use them.
Complete the Develop solutions using Model Garden APIs skill badge to demonstrate your ability to use Vertex AI Model Garden features when building gen AI solutions. You will use partner APIs such as Anthropic Claude ands Meta Llama, deploy and programatically access foundation models like Gemma and Stable Diffusion XL and access Vertex AI Endpoints. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete the assessment challenge lab, to receive a skill badge that you can share with your network. When you complete this course, you can earn the badge displayed here and claim it on Credly! Boost your cloud career by showing the world the skills you have developed!
Complete the Edit images with Imagen skill badge to demonstrate your skills with Imagen's mask modes and editing modes to edit images according to certain prompts. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete the assessment challenge lab, to receive a skill badge that you can share with your network. When you complete this course, you can earn the badge displayed here and claim it on Credly! Boost your cloud career by showing the world the skills you have developed!
Generate engaging media with Google's foundation models for media. Create new images with Imagen, or edit your existing photos by adding details or outpainting to create a wider view. Replace backgrounds to put your products in new scenes. And learn the basics of generating videos with Veo!
Learn how to create Hybrid Search applications using Vertex AI Vertex Search to combine semantic searching with keyword search to return results based on both semantic meaning and keyword matching.
Ce cours présente une solution de génération augmentée par récupération (RAG) dans BigQuery permettant de réduire les hallucinations de l'IA. Il décrit un workflow RAG qui couvre la création d'embeddings, la recherche dans un espace vectoriel et la génération de réponses améliorées. Il explique aussi les raisons conceptuelles derrière ces étapes et leur implémentation pratique avec BigQuery. À la fin du cours, les participants seront à même de créer un pipeline de RAG à l'aide de BigQuery et de modèles d'IA générative tels que Gemini, ainsi que des modèles d'embeddings pour traiter leurs propres cas d'hallucinations de l'IA.
Learn how to leverage Gemini multimodal capabilities to process and generate text, images, and audio and to integrate Gemini through APIs to perform tasks such as content creation and summarization.
Learn a variety of strategies and techniques to engineer effective prompts for generative models
An LLM-based application can process language in a way that resembles thought. But if you want to extend its capabilities to take actions by running other functions you have coded, you will need to use function calling. This can also be referred to as tool use. Additionally, you can give a model the ability to search Google or search a data store of documents to ground its responses. In other words, to base its answers on that information. In this course, you’ll explore these concepts.
Complete the Extend Gemini with controlled generation and Tool use skill badge to demonstrate your proficiency in connecting models to external tools and APIs. This allows models to augment their knowledge, extend their capabilities and interact with external systems to take actions such as sending an email. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete the assessment challenge lab, to receive a skill badge that you can share with your network. When you complete this course, you can earn the badge displayed here and claim it on Credly! Boost your cloud career by showing the world the skills you have developed!"
In this course, you'll learn to develop generative agents that answer questions using websites, documents, or structured data. You will explore Vertex AI Applications and understand the advantages of data store agents, including their scalability and security. You'll learn about different data store types and also discover how to connect data stores to agents and add personalization for enhanced responses. Finally, you'll gain insights into common search configurations and troubleshooting techniques.
Complete the Build basic Conversational Agents with Playbooks and Flows skill badge to demonstrate your proficiency in building virtual agents using traditional NLU and generative-based features. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete the assessment challenge lab, to receive a skill badge that you can share with your network. When you complete this course, you can earn the badge displayed here and claim it on Credly! Boost your cloud career by showing the world the skills you have developed!
Explore Playbooks and their implementation of the ReAct pattern for building Conversational Agents. You will learn how to construct a Playbook, set up goals and instructions to build a chatbot in natural language, and learn to test and deploy your solution.
Demonstrate the ability to create and deploy deterministic virtual agents using Dialgflow CX and augment responses by grounding results on your own data integrating with Vertex AI Agent Builder data stores and leveraging Gemini for summarizations. You will use the following technologies and Google Cloud services: Vertex AI Agent Builder Dialogflow CX Gemini
Demonstrate the ability to create and deploy generative virtual agents with natural language using Vertex AI Agent Builder and augment responses by integrating Gemini responses with third party APIs and your own data stores You will use the following technologies and Google Cloud services: Vertex AI Agent Builder Gemini Cloud Functions
In this course you will learn the key architectural considerations that need to be taken into account when designing for the implementation of Conversational AI solutions. Please note Dialogflow CX was recently renamed to Conversational Agents and CCAI Insights was renamed to Conversational Insights.
This is an introductory course to all solutions in the Conversational AI portfolio and the Gen AI features that are available to transform them. The course also explores the business case around Conversational AI, and the use cases and user personas addressed by the solution. Please note Dialogflow CX was recently renamed to Conversational Agents and this course is in the process of being updated to reflect the new product name for Dialogflow CX.
Demonstrate your ability to implement updated prompt engineering techniques and utilize several of Gemini's key capacilities including multimodal understanding and function calling. Then integrate generative AI into a RAG application deployed to Cloud Run. This course contains labs that are to be used as a test environment. They are deployed to test your understanding as a learner with a limited scope. These technologies can be used with fewer limitations in a real world environment.
Ce cours présente les outils et les bonnes pratiques MLOps pour déployer, évaluer, surveiller et exploiter des systèmes de ML en production sur Google Cloud. Le MLOps est une discipline axée sur le déploiement, le test, la surveillance et l'automatisation des systèmes de ML en production. Les participants s'entraîneront à utiliser l'ingestion en flux continu de Vertex AI Feature Store au niveau du SDK.
Ce cours présente les outils et les bonnes pratiques MLOps pour déployer, évaluer, surveiller et exploiter des systèmes de ML en production sur Google Cloud. Le MLOps est une discipline axée sur le déploiement, le test, la surveillance et l'automatisation des systèmes de ML en production. Les ingénieurs en machine learning utilisent des outils pour améliorer et évaluer en permanence les modèles déployés. Ils collaborent avec des data scientists (ou peuvent occuper ce poste) qui développent des modèles permettant de déployer de manière rapide et rigoureuse les solutions de machine learning les plus performantes.
This course explores the different products and capabilities of Customer Engagement Suite (CES) and Conversational agents. Additionally, it covers the foundational principles of conversation design to craft engaging and effective experiences that emulate human-like experiences specific to the Chat channel.
Ce cours présente les avantages liés à l'utilisation de Vertex AI Feature Store, ainsi que la manière d'améliorer la précision des modèles de ML et de déterminer les colonnes de données présentant les caractéristiques les plus utiles. Ce cours inclut également du contenu et des ateliers portant sur l'ingénierie des caractéristiques à l'aide de BigQuery ML, Keras et TensorFlow.
Ce cours porte sur la création de modèles de ML à l'aide de TensorFlow et Keras, l'amélioration de la précision des modèles de ML et l'écriture de modèles de ML pour une utilisation évolutive.
Le cours commence par une discussion sur les données : vous découvrirez comment améliorer leur qualité et effectuer des analyses exploratoires. Ensuite, nous vous présenterons Vertex AI AutoML et vous expliquerons comment créer, entraîner et déployer un modèle de machine learning (ML) sans écrire une ligne de code. Vous découvrirez également les avantages de BigQuery ML. Enfin, nous verrons comment optimiser un modèle de ML, et en quoi la généralisation ainsi que l'échantillonnage peuvent vous aider à évaluer la qualité des modèles de ML destinés à un entraînement personnalisé.
Ce cours présente Vertex AI Studio, un outil permettant d'interagir avec des modèles d'IA générative, de prototyper des idées commerciales et de les envoyer en production. Au moyen d'un cas d'utilisation immersif, de leçons captivantes et d'un atelier pratique, vous allez découvrir le cycle de vie de la requête au produit. Vous apprendrez également à utiliser Vertex AI Studio pour les applications multimodales Gemini, la conception de requêtes, le prompt engineering (ingénierie des requêtes) et le réglage de modèles. L'objectif est de vous permettre d'exploiter tout le potentiel de l'IA générative dans vos projets avec Vertex AI Studio.
A Business Leader in Generative AI can articulate the capabilities of core cloud Generative AI products and services and understand how they benefit organizations. This course provides an overview of the types of opportunities and challenges that companies often encounter in their digital transformation journey and how they can leverage Google Cloud's generative AI products to overcome these challenges.
Ce cours présente les solutions d'IA et de machine learning (ML) de Google Cloud permettant de développer des projets d'IA prédictive et générative. Il décrit les technologies, produits et outils disponibles tout au long du cycle de vie des données à l'IA, en englobant les éléments de base, le développement et les solutions d'IA. Son but est d'aider les data scientists, les développeurs d'IA et les ingénieurs en ML à améliorer leurs compétences et connaissances par le biais d'expériences d'apprentissage captivantes et d'exercices pratiques.
In this course, you'll use text embeddings for tasks like classification, outlier detection, text clustering and semantic search. You'll combine semantic search with the text generation capabilities of an LLM to build Retrieval Augmented Generation (RAG) solutions, such as for question-answering systems, using Google Cloud's Vertex AI and Google Cloud databases.
This course explores Google Cloud technologies to create and generate embeddings. Embeddings are numerical representations of text, images, video and audio, and play a pivotal role in many tasks that involve the identification of similar items, like Google searches, online shopping recommendations, and personalized music suggestions. Specifically, you’ll use embeddings for tasks like classification, outlier detection, clustering and semantic search. You’ll combine semantic search with the text generation capabilities of an LLM to build Retrieval Augmented Generation (RAG) systems and question-answering solutions, on your own proprietary data using Google Cloud’s Vertex AI.
Avec ce cours, explorez les technologies de recherche, les outils et les applications optimisés par l'IA. Découvrez la recherche sémantique, qui utilise les embeddings vectoriels (ou "plongements vectoriels"), la recherche hybride, qui combine les approches sémantique et par mots-clés, et la génération augmentée par récupération (RAG), qui réduit les hallucinations générées par l'IA en agissant comme un agent ancré. Enfin, acquérez une expérience pratique de Vertex AI Vector Search afin de créer votre moteur de recherche intelligent.
In this course, you'll use text embeddings for tasks like classification, outlier detection, text clustering and semantic search. You'll combine semantic search with the text generation capabilities of an LLM to build Retrieval Augmented Generation (RAG) solutions, such as for question-answering systems, using Google Cloud's Vertex AI and Google Cloud databases.
Learn how to design, develop, and deploy customer conversational solutions using Contact Center Artificial Intelligence (CCAI). You'll also learn some best practices for integrating conversational solutions with your existing contact center software, establishing a framework for human agent assistance, and implementing solutions securely and at scale.
Text Prompt Engineering Techniques introduces you to consider different strategic approaches & techniques to deploy when writing prompts for text-based generative AI tasks.
This content is deprecated. Please see the latest version of the course, here.
Ce cours présente Vertex AI Studio, un outil permettant d'interagir avec des modèles d'IA générative, de prototyper des idées commerciales et de les envoyer en production. Au moyen d'un cas d'utilisation immersif, de leçons captivantes et d'un atelier pratique, vous allez découvrir le cycle de vie de la requête au produit. Vous apprendrez également à utiliser Vertex AI Studio pour les applications multimodales Gemini, la conception de requêtes, le prompt engineering (ingénierie des requêtes) et le réglage de modèles. L'objectif est de vous permettre d'exploiter tout le potentiel de l'IA générative dans vos projets avec Vertex AI Studio.
Ce cours présente les modèles de diffusion, une famille de modèles de machine learning qui s'est récemment révélée prometteuse dans le domaine de la génération d'images. Les modèles de diffusion trouvent leur origine dans la physique, et plus précisément dans la thermodynamique. Au cours des dernières années, ils ont gagné en popularité dans la recherche et l'industrie. Ils sont à la base de nombreux modèles et outils Google Cloud avancés de génération d'images. Ce cours vous présente les bases théoriques des modèles de diffusion, et vous explique comment les entraîner et les déployer sur Vertex AI.
Dans ce cours, vous allez apprendre à créer un modèle de sous-titrage d'images à l'aide du deep learning. Vous découvrirez les différents composants de ce type de modèle, comme l'encodeur et le décodeur, et comment l'entraîner et l'évaluer. À la fin du cours, vous serez en mesure de créer vos propres modèles de sous-titrage d'images et de les utiliser pour générer des sous-titres pour des images.
Ce cours offre un aperçu de l'architecture encodeur/décodeur, une architecture de machine learning performante souvent utilisée pour les tâches "seq2seq", telles que la traduction automatique, la synthèse de texte et les questions-réponses. Vous découvrirez quels sont les principaux composants de l'architecture encodeur/décodeur, et comment entraîner et exécuter ces modèles. Dans le tutoriel d'atelier correspondant, vous utiliserez TensorFlow pour coder une implémentation simple de cette architecture afin de générer un poème en partant de zéro.
Ce cours présente l'architecture Transformer et le modèle BERT (Bidirectional Encoder Representations from Transformers). Vous découvrirez quels sont les principaux composants de l'architecture Transformer, tels que le mécanisme d'auto-attention, et comment ils sont utilisés pour créer un modèle BERT. Vous verrez également les différentes tâches pour lesquelles le modèle BERT peut être utilisé, comme la classification de texte, les questions-réponses et l'inférence en langage naturel. Ce cours dure environ 45 minutes.
Ce cours présente le mécanisme d'attention, une technique efficace permettant aux réseaux de neurones de se concentrer sur des parties spécifiques d'une séquence d'entrée. Vous découvrirez comment fonctionne l'attention et comment l'utiliser pour améliorer les performances de diverses tâches de machine learning, dont la traduction automatique, la synthèse de texte et les réponses aux questions.
This course on Integrate Vertex AI Search and Conversation into Voice and Chat Apps is composed of a set of labs to give you a hands on experience to interacting with new Generative AI technologies. You will learn how to create end-to-end search and conversational experiences by following examples. These technologies complement predefined intent-based chat experiences created in Dialogflow with LLM-based, generative answers that can be based on your own data. Also, they allow you to porvide enterprise-grade search experiences for internal and external websites to search documents, structure data and public websites.
In this course you will learn how to use the new generative AI features in Dialogflow CX to create virtual agents that can have more natural and engaging conversations with customers. Discover how to deploy generative fallback responses to gracefully handle errors and omissions in customer conversations, deploy generators to increase intent coverage, and structure, ingest, and manage data in a data store. And explore how to deploy and maintain generative AI agents using your data, and deploy and maintain hybrid agents in combination with existing intent-based design paradigms.
(This course was previously named Multimodal Prompt Engineering with Gemini and PaLM) This course teaches how to use Vertex AI Studio, a Google Cloud console tool for rapidly prototyping and testing generative AI models. You learn to test sample prompts, design your own prompts, and customize foundation models to handle tasks that meet your application's needs. Whether you are looking for text, chat, code, image or speech generative experiences Vertex AI Studio offers you an interface to work with and APIs to integrate your production application.
This course will help ML Engineers, Developers, and Data Scientists implement Large Language Models for Generative AI use cases with Vertex AI. The first two modules of this course contain links to videos and prerequisite course materials that will build your knowledge foundation in Generative AI. Please do not skip these modules. The advanced modules in this course assume you have completed these earlier modules.
Avec l'essor de l'utilisation de l'intelligence artificielle et du machine learning en entreprise, il est de plus en plus important de développer ces technologies de manière responsable. Pour beaucoup, le véritable défi réside dans la mise en pratique de l'IA responsable, qui s'avère bien plus complexe que dans la théorie. Si vous souhaitez découvrir comment opérationnaliser l'IA responsable dans votre organisation, ce cours est fait pour vous. Dans ce cours, vous allez apprendre comment Google Cloud procède actuellement, en s'appuyant sur des bonnes pratiques et les enseignements tirés, afin de vous fournir un framework pour élaborer votre propre approche d'IA responsable.
Suivez les cours Introduction to Generative AI, Introduction to Large Language Models et Introduction to Responsible AI, et obtenez un badge de compétence. Votre réussite au quiz final démontrera que vous comprenez les concepts de base relatifs à l'IA générative. Un badge de compétence est un badge numérique délivré par Google Cloud. Il atteste de votre expertise sur les produits et services Google Cloud. Partagez votre badge de compétence en rendant votre profil public et en l'ajoutant à votre profil sur les réseaux sociaux.
Ce cours de micro-apprentissage, qui s'adresse aux débutants, explique ce qu'est l'IA responsable, souligne son importance et décrit comment Google l'implémente dans ses produits. Il présente également les sept principes de l'IA de Google.
Ce cours de micro-apprentissage, qui s'adresse aux débutants, explique ce que sont les grands modèles de langage (LLM). Il inclut des cas d'utilisation et décrit comment améliorer les performances des LLM grâce au réglage des requêtes. Il présente aussi les outils Google qui vous aideront à développer votre propre application d'IA générative.
Welcome to "CCAI Operations and Implementation", the fourth course in the "Customer Experiences with Contact Center AI" series. In this course, learn some best practices for integrating conversational solutions with your existing contact center software, establishing a framework for human agent assistance, and implementing solutions securely and at scale. In this course, you'll be introduced to Agent Assist and the technology it uses so you can delight your customers with the efficiencies and accuracy of services provided when customers require human agents, connectivity protocols, APIs, and platforms which you can use to create an integration between your virtual agent and the services already established for your business, Dialogflow's Environment Management tool for deployment of different versions of your virtual agent for various purposes, compliance measures and regulations you should be aware of when bringing your virtual agent to production, and you'll be given tips from virtua…
Welcome to "Virtual Agent Development in Dialogflow CX for Software Devs", the third course in the "Customer Experiences with Contact Center AI" series. In this course, learn how to develop more customized customer conversational solutions using Contact Center Artificial Intelligence (CCAI). In this course, you'll be introduced to more advanced and customized handling for virtual agent conversations that need to look up and convey dynamic data, and methods available to you for testing your virtual agent and logs which can be useful for understanding issues that arise. This is an intermediate course, intended for learners with the following type of role: Software developers: Codes computer software in a programming language (e.g., C++, Python, Javascript) and often using an SDK/API.
Welcome to "Virtual Agent Development in Dialogflow CX for Citizen Devs", the second course in the "Customer Experiences with Contact Center AI" series. In this course, learn how to develop customer conversational solutions using Contact Center Artificial Intelligence (CCAI). In this course, you'll be introduced to adding voice (telephony) as a communication channel to your virtual agent conversations using Dialogflow CX.
Ce cours de micro-apprentissage, qui s'adresse aux débutants, explique ce qu'est l'IA générative, décrit à quoi elle sert et souligne ce qui la distingue des méthodes de machine learning traditionnel. Il présente aussi les outils Google qui vous aideront à développer votre propre application d'IA générative.
Welcome to "CCAI Conversational Design Fundamentals", the first course in the "Customer Experiences with Contact Center AI" series. In this course, learn how to design customer conversational solutions using Contact Center Artificial Intelligence (CCAI). You will be introduced to CCAI and its three pillars (Dialogflow, Agent Assist, and Insights), and the concepts behind conversational experiences and how the study of them influences the design of your virtual agent. After taking this course you will be prepared to take your virtual agent design to the next level of intelligent conversation.
Earn a skill badge by completing the Create Conversational AI Agents with Dialogflow CX quest, where you will learn how to create a conversational virtual agent, including how to: define intents and entities, use versions and environments, create conversational branching, and use IVR features. A skill badge is an exclusive digital badge issued by Google Cloud in recognition of your proficiency with Google Cloud products and services and tests your ability to apply your knowledge in an interactive hands-on environment. Complete this skill badge quest, and the final assessment challenge lab, to receive a skill badge that you can share with your network.