Spent my last weekend walking on random streets of Paris and that made me realize one of my core beliefs:

'There are no wrong turns, only left or right. Failing is part of the process to progress.'

And I stumbled across lots of amazing experiences like this one👇🏼

This should be impossible!

In just 3 lines of code, this open-source library can clean any ML dataset!

You can:
- flag outliers
- find label errors
- identify near duplicates
- perform active-learning
- find out of distribution samples
- And more ...

Introducing Cleanlab 🚀

A…

Large Language Models (in 2023)

An excellent summary of the research progress and developments in LLMs.

I appreciate that Hyung Won Chung made this content publicly available. It's a great way to catch up on some important themes like scaling and optimizing LLMs.

talk:…

Really love this slide in Andrej Karpathy' State of GPT talk.

It summarizes elegantly how solutions for finding information are evolving. We have two extremes in retrieval systems like Google/Bing (retrieval only) and now LLMs (memory only).

In between these two, we have what is…

Using LLMs to summarize large datasets can be hard! LangChain x Mendable partnered to analyze user questions on our documentation. We're open sourcing notebooks showing 2 approaches that use both Anthropic's new Claude-2 and OpenAI ..
blog.langchain.dev/llms-to-improv…

PyGWalker

Turn your pandas DataFrame into a Tableau dashboard.

• Launch from Jupyter Notebook
• Tableau-like visualizations
• Run in Kaggle and Colab

Tableau is $800 per year.

PyGWalker is $0.

Get the code on GitHub:

ChatGPT is a cheat code, but it doesn't have acces to the latest scientific papers.

With LangChain and arXiv.org, it does.

Let me show you how you can set this up in under 10 minutes.

#AI

Agents 🤖 built with the OpenAI function API can do advanced data analysis out of the box 🕵️

We’re excited to introduce a brand-new cookbook highlighting these tasks + limitations! 🧑‍🍳👇

- Vector db auto-retrieval
- Joint Text-to-SQL + Semantic Search

gpt-index.readthedocs.io/en/latest/exam…

What happens if we train LLMs for multiple epochs?

The question I asked multiple times in the past finally got answered in this new preprint,
'To Repeat or Not To Repeat: Insights from Scaling LLM under Token-Crisis'.

1/6