We just launched a new research simulation competition on @kaggle! Write a bot or train an AI agent to control a football player https://t.co/DSGC4OU8Rp pic.twitter.com/h7cc39x0aJ
— Ben Hamner (@benhamner) September 29, 2020
We just launched a new research simulation competition on @kaggle! Write a bot or train an AI agent to control a football player https://t.co/DSGC4OU8Rp pic.twitter.com/h7cc39x0aJ
— Ben Hamner (@benhamner) September 29, 2020
Dynabench: a framework to test ML system by asking adversarial human annotators to break it.
— Yann LeCun (@ylecun) September 24, 2020
A good way to evaluate the robustness (or brittleness) of ML systems beyond the traditional training set/test set paradigm. https://t.co/RhoY8C7hVW
Want to show off your skills with TensorFlow and deep learning using TPUs? We'd love to see what you can build in response to this task on the Chinese MNIST dataset: https://t.co/Fp66iBWt5h
— Kaggle (@kaggle) September 24, 2020
finetuning data efficient gans 100-shot-obama to cartoonset
— AK (@ak92501) September 23, 2020
github: https://t.co/CZt3YeJ5Gd
dataset: https://t.co/IzSvkrsLyk pic.twitter.com/4vfzvmYpjd
Microsoft researchers propose a new natural language processing paradigm—one in which models are pretrained from scratch entirely within a specialized domain. Learn about the new BLURB benchmark, leaderboard, and state-of-the-art model for biomedical NLP: https://t.co/xHX7Z26pSa
— Microsoft Research (@MSFTResearch) August 31, 2020
Just finished assembling #DadaGP v1.0 --- a tokenized symbolic music dataset of 26181 GuitarPro songs. Totaling 115M tokens, about as big as WikiText-103. Includes GuitarPro5 encoder/decoder. Who wants to train a generator? #nlp #mir #languagemodel #transformer @huggingface pic.twitter.com/ocyrZwYHOg
— dadabots (@dadabots) August 29, 2020
New package announcement! Introducing the {friends} package 🙌
— Emil Hvitfeldt (@Emil_Hvitfeldt) August 27, 2020
Includes the entire transcript of all the 10 seasons of the beloved American sitcom Friends
Perfect for getting your feet wet with network analysis 🌐 and text analysis 📚https://t.co/0D8DagYqnt#rstats #tidytext pic.twitter.com/HUHC02zzUq
New competition at @kaggle.
— Vladimir Iglovikov (@viglovikov) August 24, 2020
I am the host :)
A new type of task. Predicting the trajectories of different agents in the future.
I would guess that to win you will need to be creative. Stacking 100500 models may not help.
Feel free to join!#MachineLearning #SelfDriving https://t.co/cYgG14HU0u
We're releasing a corpus of over 1 million snapshots of English-language privacy policies from over 130,000 websites spanning two decades, with an accompanying paper: https://t.co/oc1b8w1ERk
— Arvind Narayanan (@random_walker) August 24, 2020
By @RyanBMAmos, Günes Acar, @elenalucherini4, Mihir Kshirsagar, @jonathanmayer, and me. pic.twitter.com/7Kk9aLOOuh
To better understand the impact of noisy labels on #MachineLearning model training, we are announcing MentorMix, a new method to mitigate the impact of noisy labels, as well as a benchmark and dataset on real-world label noise. Learn more about it at: https://t.co/jbsxPz7xtv pic.twitter.com/FNQv7XPCWk
— Google AI (@GoogleAI) August 19, 2020
Turns out a lot of open-domain QA datasets have test set leakage. If you control for it, model performance drops by a mean absolute of 63%. Yikes! If we missed this for such a long time, I wonder if there are problems with other NLP datasets too. https://t.co/uPT2uYqou7
— Tim Dettmers (@Tim_Dettmers) August 7, 2020
️We've got a new Getting Started competition up! Check out "Contradictory, My Dear Watson: Detecting contradiction and entailment in multilingual text using TPUs"---we can't wait to see what you create 🕵️♀️🕵️♂️ https://t.co/8IUySU4TfM
— Kaggle (@kaggle) August 3, 2020