Releasing blog + sample to run text to speech with real-time-factor (RTF) of 6 using Tacotron2 + WaveGlow. #TensorRT 7. Learn more: https://t.co/3sCdrIKx1V
— NVIDIA AI Developer (@NVIDIAAIDev) January 6, 2020
Releasing blog + sample to run text to speech with real-time-factor (RTF) of 6 using Tacotron2 + WaveGlow. #TensorRT 7. Learn more: https://t.co/3sCdrIKx1V
— NVIDIA AI Developer (@NVIDIAAIDev) January 6, 2020
TensorFlow has a suite of tools for optimizing your models for faster inference: https://t.co/h3qichyArj
— François Chollet (@fchollet) January 6, 2020
This includes post-training weight quantization, and gradual weight pruning during training for your Keras models. pic.twitter.com/rN5o2516wL
This is neat 😊
— Radek Osmulski (@radekosmulski) January 6, 2020
✅Train a model
✅Realize you want some functionality the framework doesn't provide
✅Define said functionality in your notebook
✅🥳🎉💃
(this is using @fastdotai v2) pic.twitter.com/l56UIJomuI
Pynndescent, an approximate nearest neighbor search library, got a major update recently. Index construction is now multicore by default. Querying is now much faster -- competitive with some of the fastest ANN libraries around.
— Leland McInnes (@leland_mcinnes) January 5, 2020
(1/4) pic.twitter.com/FECOTp0dh8
Looks useful: Keras-OCR (3rd-party package) provides out-of-the-box OCR models and an end-to-end training pipeline to build new OCR modelshttps://t.co/q7rN49VjvD
— François Chollet (@fchollet) January 4, 2020
Labeling, transforming, and structuring training data sets for #MachineLearning: https://t.co/r074wWx4G8 by @bigdata @OReillyMedia
— Kirk Borne (@KirkDBorne) January 4, 2020
——————#BigData #DataScience #DataLabeling #DataStrategy #AI #DeepLearning #Annotation #Tagging pic.twitter.com/xKbJyFVGhg
I just published the first release of Altair Viewer – this lets you easily view Altair charts without a web connection & in any development environment – no notebook required!
— Jake VanderPlas (@jakevdp) December 29, 2019
Give it a spin: https://t.co/jl4qUd4sgg pic.twitter.com/WUMJzzDva0
If you're using Transformers from source, we've rolled out 2 nice beta features (TBR in January)
— Thomas Wolf (@Thom_Wolf) December 27, 2019
💥Ultra-fast Bert/GPT2 tokenizers (up to 80x faster)
🦄Easy/versatile sequence generation for generative models: top-k/nucleus/temperature sampling, penalized/greedy, beam search... pic.twitter.com/KNAmDbQPk3
Here's my capsule review of Cloud Run / Cloud SQL / Cloud Build, from a Heroku stan:
— jacobian (@jacobian) December 26, 2019
👍🏻 native Docker, no weirdness
👍🏻 don't pay for idle time!
👍🏻 cool access control, traffic options
👎🏻 the build/release/deploy pipeline is complex af
👎🏻 docs are terrible
👎🏻 cli ux is worse
Exactly one year ago was the first public release of FastAPI.
— Sebastián Ramírez (@tiangolo) December 24, 2019
Today I have a new Christmas present for my fellow developers!
Typer: the FastAPI of CLIs 🚀https://t.co/KozKhOWrBr
BackPACK: Packing more into backprop
— Thomas Lahore (@evolvingstuff) December 24, 2019
"we introduce BackPACK, an efficient framework built on top of PyTorch, that extends the backpropagation algorithm to extract additional information from first- and second-order derivatives"https://t.co/KzO1zq9eNahttps://t.co/ubtvZu1G5P pic.twitter.com/sX9Rp6We1w
The code for hosting papers on the excellent @aclanthology is open source and free to use: https://t.co/HiWqGxbqIN
— Graham Neubig (@gneubig) December 22, 2019
Does anyone want to do a global search and replace of ACL to ACM to show @TheOfficialACM how they could easily host all their papers open access for virtually free?