https://github.com/ai-mindset/knowledgesummarised.jl

Summarising information for a more efficient acquisition of knowledge
https://github.com/ai-mindset/knowledgesummarised.jl

Last synced: about 1 year ago
JSON representation

Summarising information for a more efficient acquisition of knowledge

• Host: GitHub
• URL: https://github.com/ai-mindset/knowledgesummarised.jl
• Owner: ai-mindset
• License: mit
• Created: 2024-03-01T15:05:47.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2025-01-24T17:07:07.000Z (over 1 year ago)
• Last Synced: 2025-03-22T12:45:03.276Z (over 1 year ago)
• Language: Julia
• Homepage:
• Size: 33.2 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 6
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Knowledge, summarised

[![SciML Code Style](https://img.shields.io/static/v1?label=code%20style&message=SciML&color=9558b2&labelColor=389826)](https://github.com/SciML/SciMLStyle)

[![Compile Julia Apps](https://github.com/inferential/ImmuneSummarised/actions/workflows/ci.yml/badge.svg?branch=main)](https://github.com/inferential/ImmuneSummarised/actions/workflows/ci.yml)

This project started as an efficient way to summarise some of the excellent [MicrobeTV](https://www.microbe.tv) podcast episodes, for my personal education. I chose [Immune](https://www.microbe.tv/immune/) as a starting point. It runs on a monthly cadence, each episode spans approx. 1 hour, where a panel of immunology experts discuss new publications. At the time of writing, the [Immune playlist](https://www.youtube.com/watch?v=jnvBvbTcwIQ&list=PLGhmZX2NKiNkNlShZ2YuHH1GkwdsnH4pr&pp=iAQB) contains approximately 66 hours of high-quality scientific content over 49 episodes (some older episodes seem to be missing from the YouTube playlist).

This initial aim acted as motivation for me to write a tool in Julia, which will help me learn what I'm interested in, at a faster pace. The final goal is to utilise open-source technology able to run locally on an affordable personal computer, to learn about immunology in a time efficient way, affordably.  

Should this experiment yields satisfactory results, the same principle could be applied to different playlists depending on interest. Hence broadening the scope by naming this tool `KnowledgeSummarised.jl` i.e. knowledge, summarised, using Julia.

Technology stack:

[Julia](https://julialang.org/) for development

[yt-dlp](https://github.com/yt-dlp/yt-dlp) to download podcast audio from its YouTube playlist  

[Whisper-Faster](https://github.com/Purfview/whisper-standalone-win/releases/tag/faster-whisper) for audio transcription  

[Ollama](https://ollama.ai/) with Llama2, for summarisation and information retrieval