Skip to content

'Talk to your slide deck' (Multimodal RAG) using foundation models (FMs) hosted on Amazon Bedrock and Amazon SageMaker

License

Notifications You must be signed in to change notification settings

llauraa23/multimodal-rag-on-slide-decks

 
 

Repository files navigation

About this repository

This is a 3-part series on the topic 'talk to your slide deck' using multimodal foundation models hosted on Amazon Bedrock and Amazon SageMaker.

  • Part 1 of the series implements a solution design that first converts slides into embeddings and stores these embeddings in a vector store (Amazon OpenSearch Serverless). When a user queries this data, LLaVA 1.5 is invoked and inference returned to user.
  • Part 2 of the series follows a different solution design. This approach will generate and store LLaVA 1.5 inferences in a vector store (Amazon OpenSearch Serverless) and use those stored inferences to respond to user queries.
  • Part 3 of the series will compare the two approaches.

About

'Talk to your slide deck' (Multimodal RAG) using foundation models (FMs) hosted on Amazon Bedrock and Amazon SageMaker

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 94.8%
  • Jupyter Notebook 5.1%
  • Python 0.1%