Dear all, as the talk is based upon a lot of content, please find related links here:
-
Convert text into vectors: Word2Vec online demo: https://remykarem.github.io/word2vec-demo/
-
Word2vec (2013): https://en.wikipedia.org/wiki/Word2vec
-
Classical example: King-Queen vs Man-Woman vectorization: https://www.researchgate.net/figure/The-classical-king-woman-man-queen-example-of-neural-word-embeddings-in-2D-It_fig1_332679657
-
Try it yourself: "Semantic Calculator" based on vector values: http://vectors.nlpl.eu/explore/embeddings/en/calculator/#
-
Demo: Amazon Commprehend (e.g. Sentiment Analysis) example (AWS Account needed): https://us-west-2.console.aws.amazon.com/comprehend/home?region=us-west-2#home
-
Large Language Models - Youtube Video: https://www.youtube.com/watch?v=5sLYAQS9sWQ
-
Wikipedia: https://en.wikipedia.org/wiki/Large_language_model
-
One of the largest publicly available text datasets (825 GiB): https://pile.eleuther.ai/
-
Model size comparizon (graphical): https://lifearchitect.ai/models/#model-bubbles
-
"Transformers, explained" Youtube Video: https://www.youtube.com/watch?v=SZorAJ4I-sA
-
Illustrated Guide to Transformers: https://www.youtube.com/watch?v=4Bdc55j80l8
-
"Attention is all you need" paper on Transformers (2017): https://arxiv.org/abs/1706.03762
-
Semi-supervised learning: https://blog.roboflow.com/what-is-semi-supervised-learning/
-
Billboard chart for Large Language Models: https://s10251.pcdn.co/wp-content/uploads/2023/06/2023-Alan-D-Thompson-AI-Billboard-Rev-1.png
-
Falcon 40B LLM on Huggingface: https://huggingface.co/tiiuae/falcon-40b-instruct
-
Very simple Huggingface text completion: https://github.com/typex1/Huggingface-Text-completion-simple/
-
Falcon LLMs company: https://en.wikipedia.org/wiki/Technology_Innovation_Institute
-
Demo: SageMaker Notebook: Text Generation with Falcon: https://github.com/aws/amazon-sagemaker-examples/blob/main/introduction_to_amazon_algorithms/jumpstart-foundation-models/text-generation-falcon.ipynb
-
Demo: Amazon Bedrock Sentiment Analysis: https://github.com/typex1/Amazon-Review-Analysis-Bedrock
-
Should you Prompt, RAG or fine tune? https://medium.com/@pandey.vikesh/should-you-prompt-rag-tune-or-train-a-guide-to-choose-the-right-generative-ai-approach-5e264043bd7d
-
What is Retrieval Augmented Generation (RAG)? https://www.youtube.com/watch?v=T-D1OfcDW1M
-
Amazon Bedrock Knowledge Base = RAG with Bedrock: https://aws.amazon.com/blogs/aws/knowledge-bases-now-delivers-fully-managed-rag-experience-in-amazon-bedrock/
-
Amazon Bedrock RAG workshop: https://github.com/aws-samples/amazon-bedrock-rag-workshop
-
Jupyter Notebook used in the talk: https://github.com/aws-samples/amazon-bedrock-rag-workshop/tree/main/02_Semantic_Search
-
Twitch video on RAG with SageMaker and Amazon Kendra: https://www.twitch.tv/aws/video/1959038426
-
GitHub repo used in the above presentation: https://github.com/aws-samples/generative-ai-on-aws-immersion-day/tree/main/lab4
-
Amazon Partyrock example: "Text rewriter" https://partyrock.aws/u/cabcookie/RjOhsSpr0/Text-Rewriter-An-AI-Assistant-for-Professionally-Editing-Content/snapshot/7CHKa58sa
-
Files and links from Generative AI official webinar on Youtube: https://github.com/typex1/AWS-GenAI-Foundations-slides/
-
Free Bedrock related courses on AWS Skillbuilder: https://explore.skillbuilder.aws/pages/16/learner-dashboard?ctl99=l-_en~se-Bedrock
-
Free SageMaker Lab Linux platform (25GB permanent storage, 4 CPUs): https://studiolab.sagemaker.aws/
-
Free of charge AWS Workshops related to Bedrock (you only need to bring your own AWS account): https://workshops.aws/card/bedrock
