Ofir Press is the creator of AliBi attention and Self-Ask prompting. Here are the papers that are discussed in the podcast: AliBi Attention: https://arxiv.org/abs/2108.12409 Self-Ask Prompting: https://arxiv.org/abs/2210.03350

This research was introduced in "Gorilla: Large Language Model Connected with Massive APIs" by Patil et al. 2023.

Here is a paper summary review: https://www.youtube.com/watch?v=LkV5DTRNxAg&t=1213s

The 1.21 release is here! Here is the TL;DR of the new features: • ContainsAny and ContainsAll operators added – Convenient, new operators to simplify complex queries. • Multi-tenancy improvements – Experimental tenant deactivation for efficiency, performance improvements. • New vectorizer modules: 1. text2vec-gpt4all provides fast transformer inference on CPUs; and 2. multi2vec-bind vectorizes multi-modal data from up to 7 modalities. • Performance improvements – A suite of improvements to search, indexing and backup performance. • Hybrid search algorithm refinement - Improved scoring stability for small limits in hybrid search.

Check out the blog post for an in-depth overview of each feature and links to the documentation: https://weaviate.io/blog/weaviate-1-21-release

Here is how to build a retrieval augmented generation (RAG) chatbot using the new Llama 2 model:

• Replicate for the endpoint to llama13b-v2-chat • LlamaIndex for the LLM framework and query engine • Weaviate for the vector store

Love how this shows the value of search and retrieval augmented generation! Perfectly shows the value of Weaviate's AutoCut, Re-Rankers, and Hybrid Rank Fusion feature. (: