brunopistone / flan-t5-multi-language

Multi Language HuggingFace Flan-T5 XXL Sharded using Amazon SageMaker

The scope of this repo is to provide an end-to-end example for using HuggingFace Flan T5 XXL model on AWS by using Amazon SageMaker Real-Time Endpoint, Amazon Comprehend, and Amazon Translate for managing multi-language inputs.

Reference Blogs

1. Deploy FLAN-T5 XXL on Amazon SageMaker
2. Create Your Own Large Language Model Playground in SageMaker Studio
3. Architect personalized generative AI SaaS applications on Amazon SageMaker

Repository Content

1. notebook: Use Amazon SageMaker Studio Notebooks for testing the end-to-end solution in the notebook Deploy-LLM-Model
2. project: Automate the creation of an MLOps Pipeline for Model deployment by using SageMaker Project

Getting Started

1. Run Deploy-LLM-Model
2. Run the streamlit app streamlit run flan-t5-playground.py --server.port 6006

SageMaker Project

Please refer to the instructions

Architecture