ChatGPT

Motivation

I use a Mac at home but a Windows PC at work. After using a PC at the office on weekdays, I often mix up the keyboard shortcuts for copying and pasting (e.g., Ctrl-c vs. Cmd-c) when I switch to my Mac at home. Conversely, on Mondays at the office, I sometimes reflexively press Cmd+c. Incidentally, I use an HHK (US layout) keyboard in both environments.

I decided to remap the Mac shortcuts Cmd-c/v/x/z/a/s/f to Ctrl-c/v/x/z/a/s/f. To achieve this, I used an application called Karabiner-Elements.

During the setup process, I consulted ChatGPT to verify the steps and check how to handle any potential issues.

Introduction

I built a RAG using HyDE (Hypothetical Document Embeddings), a method to improve RAGs. This post summarizes my trial of HyDE. The LLM used was gpt-4o-mini to keep costs down.

Introduction

By yesterday, I had extracted astronomy-related entries from Wikipedia and created a vector database and keyword base for RAG. Here, I will use those databases to build the RAG system.

The LLMs used are ChatGPT (gpt-4o) and Llama-3-ELYZA-JP-8B.