NLP & vision transformers are very expensive to train, so we now focus more on fine-tuning. But should we select by model size, data size, upstream accuracy? Turns out that (maybe intuitively) upstream accuracy is the best predictor for downstream acc: https://t.co/mR8Y8HNmAe pic.twitter.com/o3JD20b750
— Sebastian Raschka (@rasbt) February 10, 2022
GPT-NeoX-20B, 20 billion parameter large language model made freely available to public, with candid report on strengths, limits, ecological costs, etc.
— Gary Marcus (@GaryMarcus) February 10, 2022
Genuinely Open AI https://t.co/kwcCtNQiGD
GiraffeDet: A Heavy-Neck Paradigm for Object Detection
— AK (@ak92501) February 10, 2022
abs: https://t.co/Mrtg47qSpG pic.twitter.com/OHpF2tcGmq
A treasure trove of various variational autoencoders implemented in PyTorch: https://t.co/JoBocYFqpE pic.twitter.com/7fZZ8k4js2
— Sebastian Raschka (@rasbt) February 8, 2022
Some people requested the blog post version so here it is. Hopefully this is easier to read 😀https://t.co/pFe4HxNiFf
— Chip Huyen (@chipro) February 8, 2022
Context Autoencoder for Self-Supervised Representation Learning
— AK (@ak92501) February 8, 2022
abs: https://t.co/ziVP3wwUAo pic.twitter.com/02PN3y6hp9
Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework
— AK (@ak92501) February 8, 2022
abs: https://t.co/UlC8dDNdMm pic.twitter.com/pLIrOyQOTp
Whoa! Pandas has a nifty function read_html for pulling a webpage, returning a list of dataframes representing the tables on it
— Ben Hamner (@benhamner) February 5, 2022
I wanted sunrise/sunset for SF, and thought I was going to have to get my hands dirty parsing https://t.co/LH3T1O8zRa
Nope! It's a pandas one-liner pic.twitter.com/UTQ4iwgRFx
I wrote a lot of hacky scripts in my lifetime for logging, hyperparameter sweeps, and submitting my jobs to our SLURM cluster. I may be biased, but using @gridai_ now is very refreshing. Just select your Gh repo, the hardware you want to use & the hparam ranges, and off you go https://t.co/qdfYEvWGqk pic.twitter.com/u4TWrAfl7T
— Sebastian Raschka (@rasbt) February 4, 2022
This is sweet 🥧 !https://t.co/fUarOS3CNn
— Martin Görner (@martin_gorner) February 4, 2022
Finally a solid way of of teaching a neural network to know what it does not know.
(OOD = Out Of Domain, i.e. not one of the classes in the training data.) Congrats @SharonYixuanLin @xuefeng_du @MuCai7 pic.twitter.com/x3DEX7Y6Hl
😍 Visualize and compare repositories' stars over time with this xkcd-esque chart generator courtesy of @tim_qian and @BytebaseHQ: https://t.co/z2O4shSTdn 📈🚀 pic.twitter.com/F1VlQy9ayF
— GitHub (@github) February 4, 2022
CoST: Contrastive Learning of Disentangled Seasonal-Trend Representations for Time Series Forecasting
— AK (@ak92501) February 4, 2022
abs: https://t.co/4jk8lps3pJ pic.twitter.com/WJmZRWrYMk
