Pondhouse Data Blog
Advanced RAG: Improving Retrieval-Augmented Generation with Hypothetical Document Embeddings (HyDE)
HyDE is a technique used to improve the performance of RAG models by generating hypothetical document embeddings based on the query and using them to retrieve relevant documents from the knowledge base.
GenAI: Technological Masterpiece or Ecological Disaster?
Since 2022 GenAI changed how we are working and already assists us in many ways. But is GenAI the revolutionary pathway to our future of worklife or is their carbon footprint the shurtcut to our ecological apocalypse?
11 Proven Strategies to Reduce Large Language Model (LLM) Costs
LLMs have become everyday tools for many businesses and individuals. However, the costs of running these models can quickly add up. In this guide, we will explore some strategies to help you save on LLM costs.
How to boost your database performance with OpenAIs new v3 embeddings
OpenAI has released its new embedding models, which have been trained using the new MRL technique. Discover how this not only enhances the accuracy of semantic searches but also can improve the performance of vector databases
Advanced RAG: Increase RAG Quality with ColBERT Reranker and llamaindex
Finding the right documents during retrieval is probably the most important aspect of your RAG pipeline. This guide demonstrates how to use 'reranking' for
Advanced RAG: Recursive Retrieval with llamaindex
With recursive retrieval, RAG can generate more coherent and contextually relevant responses. This guide introduces you to the concept of recursive retrieval and how to implement it with llamaindex.
How to Set Up a Secure, Self-Hosted Large Language Model with vLLM & Caddy
Running your own LLM provides a lot of flexibility and control over your data. This guide introduces you to the seamless integration of vLLM and Caddy web server, enabling HTTPS encryption for a robust, private AI environment.
Improving Retrieval Augmented Generation: A Step-by-Step Evaluation of RAG Pipelines
RAG pipelines are one of the corner-stones of modern AI applications. Evaluating there performance is detrimental for making them robust and production ready.
Showing 41 to 50 of 50 results