What’s the difference between CBOW and Skip-gram, in terms of performance? Turns out nothing if you fix bugs in popular implementations. I love this genre of papers. https://t.co/GegvyGXr5O pic.twitter.com/ikMgrluLMt
— Delip Rao (@deliprao) June 9, 2022
We are *so* excited to announce vetiver 🏺
— Julia Silge (@juliasilge) June 9, 2022
I ❤️ working on software for people's real, applied work, and this project offers such a great solution for tasks that have been so challenging! https://t.co/SSe90xIjBp
Techniques for training large neural networks, by @lilianweng and @gdb: https://t.co/S5xDaO9nFN
— OpenAI (@OpenAI) June 9, 2022
DETR++: Taming Your Multi-Scale Detection Transformer
— AK (@_akhaliq) June 8, 2022
abs: https://t.co/kOQ5V4vC3C
DETR++, a new architecture that improves detection results by 1.9% AP on MS COCO 2017, 11.5% AP on RICO icon detection, and 9.1% AP on RICO layout extraction over existing baselines pic.twitter.com/Kt3EQRXwuH
Thanks everyone for debunking those!
— Sebastian Raschka (@rasbt) June 7, 2022
1) VSCode is hard to set up on a new computer since you have to install plugins individually. Not anymore: https://t.co/74pdUoAdDW
2) PyCharm doesn't support remote development. Apparently, also a thing from the past: https://t.co/pmKozDxBQd
Yes, machine learning is everywhere. But this one application where it really delivers, my ugly handwriting and all. (Fun fact: my students implemented and live-demoed sth similar as their class project). pic.twitter.com/6VU8BsGcFX
— Sebastian Raschka (@rasbt) June 6, 2022
I have one note on iOS notes app where I add random ideas / thoughts / todos / questions one per line to the top as they happen. Once in a while I look at and pop interesting stuff upwards. Most sink down. I’d normally forget 75% of what’s on there and find the practice valuable.
— Andrej Karpathy (@karpathy) June 3, 2022
EfficientFormer: Vision Transformers at MobileNet Speed
— AK (@_akhaliq) June 3, 2022
abs: https://t.co/lfdbJdS46J
EfficientFormer-L1, achieves 79.2% top-1 accuracy on ImageNet-1K with only 1.6 ms inference latency on iPhone 12, which is even a bit faster than MobileNetV2 (1.7 ms, 71.8% top-1) pic.twitter.com/4zcvunXe57
Woke up this morning to the alerts about the following pepper that supposedly comes up with a new NN algorithm that far outperforms all others on tabular datasets:https://t.co/KGzMHUVbVT
— Bojan Tunguz (@tunguz) June 2, 2022
1/8
Teaching Models to Express Their Uncertainty in Words
— AK (@_akhaliq) May 31, 2022
abs: https://t.co/rKcZNhBLt5
GPT-3 model can learn to express uncertainty about its own answers in natural language – without use of model logits. The model generates both an answer and a level of confidence pic.twitter.com/X37qwDGami
X-ViT: High Performance Linear Vision Transformer without Softmax
— AK (@_akhaliq) May 30, 2022
abs: https://t.co/A6HZ2vXKDB pic.twitter.com/kArY0Tm4VE
GIT: A Generative Image-to-text Transformer for Vision and Language
— AK (@_akhaliq) May 30, 2022
abs: https://t.co/iFly0pcoXM
model surpasses the human performance for the first time on TextCaps (138.2 vs. 125.5 in CIDEr) pic.twitter.com/vn9LV98Dwr
