Mastering Multimodal RAG & Embeddings with Amazon Nova & Bedrock

Level: Beginner Duration: 5 Hours Students Enrolled: 1500+ Average Rating: 4.8

About this Course

This course provides a strong foundation in natural language processing by delving into:

• Word embeddings, tokenization, Byte Pair Encoding (BPE), and data sampling techniques.
• Utilization of Amazon's Titan Text Embeddings for effective text representation, enhancing AI application performance.
• Integration of various data modalities using Amazon Nova and Bedrock for developing advanced, AI-powered solutions.

Learning Outcomes

Upon completion, you will be able to:

• Understand Embeddings: Learn how embeddings enhance NLP and LLM capabilities.
• Explore Multimodal RAG: Master retrieval-augmented generation with multimodal data.
• Utilize Amazon Nova & Bedrock: Leverage Amazon Nova and Bedrock for AI-powered solutions.

Course Curriculum

The comprehensive curriculum covers:

1. Embedding in NLP and LLMs

• Introduction to the course
• Understanding word Embeddings and Tokenization
• Implementing Byte-Pair Encoding (BPE)
• Data sampling with a sliding window

2. Amazon Bedrock & Amazon Titan Text Embeddings model

• Exploring Embedding model on Amazon Bedrock

3. Multimodal LLMs

• Multimodals and Transformers for vision
• Understanding CLIP
• Text Generation Multimodals

4. Multimodal RAG

• What is RAG
• What is Multimodal RAG
• Building Multimodal RAG with Amazon Bedrock, Amazon Nova and LangChain

Instructor

• Suman Debnath: Principal Developer Advocate for Machine Learning at AWS.

Pricing

• Enrollment: Free