RL Unplugged is a suite of Offline RL benchmarks with open-sourced Acme agents and datasets. It has datasets from popular domains such as, dm_control (🦾), atari (👾), and RWRL (🌎). Paper: bit.ly/2BEvNDw Github: git.io/JJUhd

The Chess Transformer: Mastering Play using Generative Language Models pdf: arxiv.org/abs/2008.04057

Work by: @andre_s_barreto, Shaobo Hou, Diana Borsa, David Silver, and Doina Precup.

Work by M, @drao64, Roland Hafner, Thomas Lampe, Abbas Abdolmaleki, Tim Hertweck, Michael Neunert, Dhruva Tirumala, Noah Siegel, Nicolas Heess and Martin Riedmiller!

I wrote a new piece on Configuring Google Colab like a pro. Connect your drive, github, and even remote in with VSCode so you feel like you're working on your own machine, except it's all free. I'd love to hear your tips and tricks I might have missed. link.medium.com/0LKJNS9l88

It you use matplotlib, but haven't yet read this extremely useful overview of the stuff you really need to know to use it properly, then stop whatever you're doing and read it now! A lot of people only know the icky Matlab-style API, but you'll see there are better options

Impressive low-cost robot that can be built with stuff that costs no more than $50 and uses the smartphone as the main computing platform. openbot.org youtube.com/watch?v=qc8hFL…

One of the hardest parts of contributing to open source is making a pull request on GitHub. I've written the guide I wish I'd had when I started, showing every step in the process from start to finish. docs.fast.ai/dev-setup

The online causal inference course at causalcourse.com officially launches today! The first week of lecture material and the first 6 chapters of the course textbook are now on the website. Everything is completely free :)

⿻ Andrew Trask 😉 x.com/jponttuset/sta…

Have you ever finished a Kaggle Competition and wondered what you can do next to boost your technical knowledge? Kaggle Grandmaster Vladimir Iglovikov breaks it down for you in this excellent blog post: link.medium.com/uzEkVWedE9

Someone is exploring this direction❗👌 Physically Embedded Planning Problems: New Challenges for RL “When humans play such games (such as Go, chess, or shogi), they do not just reason about the moves but also interact with their physical environment…” arxiv.org/abs/2009.05524

Here's a sample release - it tells you what changed, bugs fixed, etc, and provides a download of all the files at that point. github.com/fastai/fastrel…

This is great - thanks all for your fascinating stories! If you haven't gotten yourself a little Linux server yet, check out this thread! (PS: if you're paying >$5/month, you're overpaying - try vultr, dreamcompute, ovh, lightsail, digitalocean, ethernetservers...)

starting a thread of interesting python features / modules / libraries that I found over time 🧵

First post: A look into Hudl Statsbomb's aerial duels metric, and what makes a player 'good in the air' thomasdkelly.github.io/blog/football/…

Excited to launch AquaPlan, a new tool from my group (me & Tom Kelly) + developmentseed (Ryan Avery & @kathrynaberger) to help farmers, businesses, and planners to manage water and climate risks in agriculture (tinyurl.com/aquaplan). More in🧵+ blog (tinyurl.com/3fevysba)