Cloud hardware AWS assignment notes

This commit is contained in:
Adog64 2025-04-24 11:58:38 -04:00
parent 29767107a1
commit cd559cd401
3 changed files with 42 additions and 0 deletions

View File

@ -0,0 +1,42 @@
---
title: ECE09488 AWS Assignment Notes
author: Aidan Sharpe
date: April 24th, 2025
geometry: margin=1in
colorlinks: true
---
Video: [Amazon AI Conclave 2024 Generative AI Keynote | AWS Events](https://www.youtube.com/watch?v=p_CwWdPTdm8&list=PL2yQDdvlhXf8xcKr0-BHEyg_8VB4tWdu1&index=12)
## Overall vision
- Enable easy deployment of generative AI in the cloud
- "Drag and drop" foundational models
- Make comparing models easy
## New services
- Amazon Bedrock
- Model evaluation
- Knowledge bases
- Agents
- Amazon Titan Foundation Models
- Titan Text Embeddings
- Titan Text Lite
- Titan Text Express
- Titan Multimodal Embeddings
- Titan Image Generator
- Amazon Q
## Use cases
- Choose between models from different providers
- AI21Labs Jurassic
- Amazon Titan
- Anthropic Claude
- Cohere Command + Embed
- Meta Llama 2
- Stability AI Stable Diffusion
- Stitch different types of models together
- Text and images
## Demos
- Demonstrated automatic model evaluation on a specific task type (text summarization) on several different metrics
- Demonstrated knowledge bases using text stored in an S3 bucket to give information to models