Retrieval Augmented Generation + Numpy array processing with C

Join us in an evening with Large Language Models (LLMs), numpy arrays, C, Python, demos and practical tips! The event is at Vespa.ai, Prinsens gate 49, 3rd floor.

There will be snacks and refreshments as usual. And two presentations!

Accelerating tensor processing with C
This presentation will be a quick dive into the world of C and Python, where we'll see how to accelerate tensor (numpy array) processing with C. In just 15 minutes, we'll touch on the power of SIMD/AVX, and the C Foreign Function Interface (CFFI), and see how Github Actions and cibuildwheel can streamline builds. There will also be a light-hearted rant on the complexities of optimizing code for various operating systems, Python versions, CPU architectures, and more. Expect a blend of code, insights and a brief intro to various concepts.

Speaker: Iver Jordal (Nomono). He develops audio processing pipelines that combine AI and Digital Signal Processing (DSP).

Improving the Usefulness of Large Language Models with Retrieval Augmented Generation
Large Language Models (LLMs) like GPT can give useful answers to many questions, but there are also well-known issues with their output: The responses may be outdated, inaccurate, or outright hallucinations, and it’s hard to know when you can trust them. And they don’t know anything about you or your organization's private data (we hope).

RAG - “Retrieval Augmented Generation” - can help reduce the problems with “hallucinated” answers, and make the responses more up-to-date, accurate, and personalized - by injecting related knowledge, including non-public data.

We’ll go through what RAG means, demo some ways you can implement it - and warn of some traps you still have to watch out for.

Speaker: Andreas Eriksen. He is a Vespa Developer and Solutions Engineer working on large-scale search and recommendation applications