Hot take. #Python is a *fast* language. Given a finite amount of time to develop something, you can spend a greater portion of that time working on a more efficient algorithm with Python. This can beat a less efficient algorithm implemented in a compiled language.

More samples of this brave lady hiking Mt. Fuji wearing a Kimono by #Dalle

Yesterday I asked some Kagglers about their favorite creative ideas with real-world applicability that they found through Competitions ... They're very cool so I'm sharing them in a thread (add your own!) 👇

Deploying GPT-like language models on a chatbot is tricky. You might wonder • How to access the model? • Where to host the bot? In this 🧵I walk you through how easily I deployed a GPT-J-6B model by #EleutherAI on a #Telegram bot with Hugging Face and Gradio. For FREE 🚀

Mark Tenenholtz No company ever has a “dataset” for you to come and work with. 99.99% of the companies have a problem to solve and absolutely no idea what data they need to solve it. Not only there’s no “dataset”, but it’s your job to decide what data should be part of it in the first place.

Check out #Imagen, Google Brain’s new SOTA Text-to-Image model I’m especially fond of these prompts: “Android Mascot made from bamboo.” + “Intricate origami of a fox and a unicorn in a snowy forest.” 😉 Impressed that AK always finds the links before it is officially released❗️

AI can unlock joint human/computer creativity! Imagen is one direction we are pursuing: gweb-research-imagen.appspot.com "A high contrast portrait of a very happy fuzzy panda dressed as a chef in a high end kitchen making dough. There is a painting of flowers on the wall behind him."

A shocking revelation: Companies don't have "datasets" waiting for you. It's your job to create and maintain them. This is time-consuming and hard, but primarily responsible for the success of your machine learning system. Here are some thoughts: 1 of 7

🎓 Getting Started with Transformers Transformer is one of the most important concepts in ML and NLP today. I created this guide for students that want to get started with transformers. github.com/dair-ai/Transf…

Over the weekend I released fastxtend to pypi and wrote a post highlighting some of its features. Fastxtend is a collection of tools, extensions, and addons for fast.ai. benjaminwarner.dev/2022/06/06/int…

Big release of fastai today - Hugging Face Accelerate is now supported for distributed training thanks to Zach Mueller and Sylvain Gugger. That means you can now do distributed training in a notebook! 1/🧵

lexica.art 😍✨

"AudioLM: a language modeling approach to audio generation", w/ Zalán Borsos and colleagues. AudioLM achieves high-quality and long-term consistency, from audio only. Meaningful speech w/o text, consistent piano w/o MIDI. 📃: arxiv.org/abs/2209.03143 🔊: google-research.github.io/seanet/audiolm…

Epic!

Gather round, Twitter folks, it's time for our beloved **Alice's adventures in a differentiable wonderland**, our magical tour of autodiff and backpropagation. 🔥 Slides below 1/n 👇

At #pyberlin! PyBerlin

Jacek about using Haystack framework for NLP in Python

Meet the Princess of the Underworld in HADES II, the next game from our studio! Details about Early Access coming in 2023. 🌘Wishlist on Steam: supergiant.games/hades2steam 🌓Wishlist on Epic: supergiant.games/hades2egs 🌔More info: supergiant.games/hades2 #TheGameAwards #HadesGame

Our 43rd event already started

Evgenia Kochkina is giving a talk about "Meetings That Yield Results"