-
Monitor and analyze brands using language models
Description I have built a brand monitoring application that leverages both large and specialized smaller language models. The entire application runs on CPU, relying on external APIs for specific tasks, making it a cost-free solution for users.
Read more… -
Enhancing decision making with RAG chatbot
Description I recently built an AI-powered virtual assistant specializing in NATO’s Alternative Analysis (AltA) methodology. This chatbot leverages Large Language Models (LLMs) combined with Retrieval-Augmented Generation (RAG) to provide accurate responses to questions that general-purpose models might struggle with.
Read more… -
Prototype-based learning. Part II: LVQ family of models in PyTorch
Description In my journey to bridge psychology and data science, I discovered Learning Vector Quantization (LVQ) and immediately saw its potential connection to human category learning. This realization led me to dive deeper, implementing LVQ from scratch, initially by coding all the operations using PyTorch, and then by abstracting the optimization and learning processes to fully leverage PyTorch’s features.
Read more…