Hello everyone 👋 and welcome to
#TrendingTuesday. Every Tuesday we'll post a list of trending GitHub repos in a cool and interesting topic.
There's lots of chatter around
#ChatGPT, so this week we're bringing you a list of the fastest-growing ML repos on GitHub.
In place number 10 we have repo that indexes some of the most important ML and NLP papers with pointers to open articles and blogposts that deconstruct them in a way that most developers can understand. Have a look if you want to gain a deeper understanding of the main technology behind today's most famous chatbot.
Next, in place number 9, we have another great educational repo that traces the learning journey to follow in order to become a research scientist. This guide is targeted to developers with basic programming or CS knowledge, but very motivated in deep learning or NLP research.
In number 8 we have a GPT-related repo that is targeted to practitioners and ML engineers. GPT-Index consists of a set of data structures designed to make it easier to use large external knowledge bases with LLMs.
Mage-ai falls in the ML/data/infra category and claims to be the modern replacement for Airflow, a very popular open source workflow platform for data engineering pipelines. With 3.3k stars and 21 contributors, it's worth checking it out!
This is a relatively new repo by Andrej Karpathy who was formerly Director of AI at Tesla. The repo contains a series of YouTube videos where viewers code and train neural network. It's really cool to see two "zero-to-hero" type of repos in a single week. Jupyter notebooks are included in the box.
Place number five is an amazing project that envisions to replace ChatGPT by creating a personal assistant able to write emails, cover letters, use APIs, and much more. It's a really ambitious project that is also quickly onboarding new contributors. If you have some ML experience and you want to get your hands dirty by contributing to a new project, this might be a good place to start.
In place number four we have a very sleek app to interface with ChatGPT from the comfort of your desktop. The project was built in Rust 👌
The third most starred repo was created by researchers at Google Brain and Harvard University, who created a playbook for tuning neural networks. This repo is helping ML practitioners get on top of the tuning process by sharing good tricks and best-practices from one of the best AI labs in the world.
Spot number two is also by Andrej Karpathy 🤯. This is a repo designed to to train and finetune GPT networks. It is a less educational version of his previous repo karpathy/minGPT. It is definitely worth studying to really understand GPT networks.
The top place is for f/awesome-chatgpt-promts which is an
#awesome list that collects prompt examples to use ChatGPT like a PRO. Courtesy of the developer who also snapped the “@f” username on GitHub
#TrendingTuesday was brought to you by the team at Quine.
Which topics would you like to see trending repos for? Let us know in the comments :)
Top comments (0)