Learn best practices, patterns and processes for developers and DevOps teams who design and implement LLM-based applications using Google Gemini.
Overview
Syllabus
Introduction
- Build LLM-based applications with Google Gemini
- What you should know
- Using cloud services
- Understanding Google Gemini
- Use Google AI Studio
- Use Vertex AI Studio
- Use Colab Notebooks
- Use Gemini Code Assist in cloud workstations
- Use Google AI Studio to test prompts
- Use system instructions with prompts
- Design and test language model prompts
- Design and test multimodal prompts
- Design prompts in Cloud Code for APIs
- Using the Gemini API: Set up
- Using the Gemini API: Testing prompts
- Using function calling with Gemini
- Programming multimodal use cases
- Use the Gemini File API
- Use embeddings with Gemini
- Set up a RAG pattern with Gemini
- Implement a RAG pattern with Gemini
- Understand model grounding
- Ground a model with Google Search
- Ground with a semantic retriever
- Understand model evaluation
- Perform model evaluation
- Fine-tune a Gemini model
- Use Vertex AI Model Garden
- Deploy a GenAI cloud architecture: Document summaries
- Deploy a GenAI cloud architecture: Knowledge base
- Preview of Vertex AI Agent Builder
- Next steps
Taught by
Lynn Langit
