learning center
At Aurelio AI we believe in the power of online learning. That's why we dedicate a lot of our efforts to developing educational material. Here is some of the latest content and talks that members of the Aurelio team have published.
AI
Getting Started with NVIDIA's AI Workbench
Everything you need to know about the foundation of NVIDIA's AI offerings, the AI Workbench.
Semantic Chunkers
Getting Started with Semantic Chunking
Everything you need to know about different types of semantic chunking for RAG.
Setup a Remote NVIDIA AI Workbench Node using EC2
How to run GPU-accelerated ML workloads with ease using NVIDIA's AI Workbench and a CUDA-enabled EC2 instance.
Semantic Chunkers
Semantic Chunking of Videos for AI
How to intelligent process video for downstream AI applications using semantic chunking.
LLMs
Mixtral 8X7B Intro
An introduction to the first open-weights model to outperform GPT-3.5 and how to use it as an adbanced AI agent.
Events
Pinecone x LangChain SF Hacking Hours
Join James and Harrison at LangChain HQ in San Francisco for a night of hacking with Pinecone and LangChain.
RAG
Mitigate Hallucinations Through RAG (AWS Blog)
We explore how to mitigate hallucinations in Large Language Models (LLMs) using Retrieval Augmented Generation (RAG) and the Pinecone vector database in Amazon SageMaker JumpStart.
Events
AI Day Tokyo 2023
Join James and a packed lineup of experts in AI at AI Day Tokyo in Shibuya on December 2023.
RAG
Alternative Embedding Models to OpenAI
A look at a few of the best retrieval models for Retrieval Augmented Generation (RAG) and how we use them.
RAG
RAG and Rerankers with Cohere AI
James explains how we can use Cohere AI's reranking models to massively enhance the performance of Retrieval Augmented Generation (RAG) pipelines.
Events
Oslo: Large Language Models and Search (Tekna)
Large Language Models (LLMs) are revolutionalizing how machines understand and generate human-like text. In this Tekna meetup, James (Aurelio AI) and Jo Kristian Bergum (Vespa) explain how to get them working with retrieval.
MLOps
Deploying HuggingFace Models on NVIDIA-Enabled EKS Nodes
Bogdan shares how he used battle-tested container orchestration for our GPU-enabled inference workloads, with baked-in telemetry from Prometheus.
LLMs
Fine-tuning OpenAI's GPT 3.5
Fine-tuning for GPT 3.5 turbo is finally here! The latest update gives OpenAI users the ability to create their own custom GPT-3.5 model that has been tuned towards a particular dataset. James explains how it works here.
Conversational AI
Nvidia's NeMo Guardrails: Full Walkthrough
James explores Nvidia's NeMo Guardrails library for building chatbots. Allowing us to create semi or fully deterministic shields that use against specific behaviors, conversation topics, or even to trigger particular actions.
AI Agents
Superagent.sh August Meetup
Simonas shares his thoughts on the Superagent.sh library and how it helps us develop agent frameworks for our clients.
LLMs
Supercharging Llama-2 with RAG
James and the LangChain team explore how we add external knowledge to Large Language Models using Retrieval Augmented Generation (RAG) using the Pinecone vector database.
Conversational AI
Podcast Chatbot using LangChain Agents and GPT-3.5
Conversational agents are set to become the future of natural language interfaces. Learn how they work in this LangChain walkthrough with James.
LLMs
Retreival Augmented Chat using GPT-4 and Pinecone
GPT-4 is the completion model from OpenAI. This walkthrough from James demonstrates how to ground it with real world information using a vector database.
LLMs
Building the Future with LLMs, LangChain, & Pinecone
Join Harrison Chase (founder of LangChain), and James Briggs, (Aurelio and Pinecone), as we explore the new age of AI and it's biggest enabler — LangChain.
Vector Search
Medical Search Engine with SPLADE + Sentence Transformers
A walkthrough from James on hybrid search applied to the Medical domain. Hybrid search is a powerful solution in language-only and multi-modal retrieval systems.