Manifold: A Model-Agnostic Visual Debugging Tool for ML
— ML Review (@ml_review) January 19, 2020
By Uber ( @lezhili )
Github: https://t.co/bkYedhVzEq
Blog: https://t.co/xH6Ay757R5 pic.twitter.com/7nvakmHCu7
Manifold: A Model-Agnostic Visual Debugging Tool for ML
— ML Review (@ml_review) January 19, 2020
By Uber ( @lezhili )
Github: https://t.co/bkYedhVzEq
Blog: https://t.co/xH6Ay757R5 pic.twitter.com/7nvakmHCu7
"Advbox: a toolbox to generate adversarial examples that fool neural networks -- https://t.co/Honf6FaSzv " Looks like a nice, new, and comprehensive toolbox for experimenting with DL model security (github link here: https://t.co/PNd9ujpHDv) pic.twitter.com/UsMadk7NUu
— Sebastian Raschka (@rasbt) January 18, 2020
Check out the new #GPU-accelerated #dataviz tools from @rapidsai, cuxfilter: browser-based dashboard for @ProjectJupyter notebooks, cuDataTile, @datashader integration, and more. https://t.co/Ji1j4dyPFJ
— RAPIDS AI (@rapidsai) January 16, 2020
torchtext v0.5: improved data API, unsupervised text tokenization
— PyTorch (@PyTorch) January 16, 2020
- bindings for SentencePiece
- New: enwiki9, revisions to PennTreebank, WikiText103, WikiText2, IMDb (looking for feedback)
Release Notes: https://t.co/vSlmggYh8G
torchaudio v0.4: more transforms, datasets, backend support
— PyTorch (@PyTorch) January 16, 2020
- LibriSpeech and Common Voice loaders
- Filters (biquad), batched / jittable transforms (MFCC, gain, dither), more augmentation
- interactive speech recognition demo with voice detectionhttps://t.co/Tm5Fge2LyF
torchvision v0.5: quantization, production
— PyTorch (@PyTorch) January 16, 2020
- ResNets, MobileNet, ShuffleNet, GoogleNet and InceptionV3 now have quantized counterparts with pre-trained models, scripts for quantization-aware training.
- All models are TorchScript-ready and ONNX-readyhttps://t.co/qbItIUpZFT
v1.4: customizable mobile builds, Distributed Model Parallelism via experimental RPC API, Java Bindings, Chaining LRSchedulers
— PyTorch (@PyTorch) January 15, 2020
Summary: https://t.co/jgVsN6EWQ0
Release Notes: https://t.co/EwhjMbVel5
Last release for Python 2 (bye bye!)
PyTorch Metric Learning now available on Anaconda!
— Kevin Musgrave (@LightningSource) January 15, 2020
Installation:
"conda install pytorch-metric-learning -c metric-learning"
View more details here:https://t.co/GMmfQvbkwr#DeepLearning #MachineLearning #PyTorch #Python #ComputerVision #conda pic.twitter.com/AG4MPziSDG
Hey weather data geeks: @NOAAClimate has a Dataset Gallery https://t.co/PdmuS61BAX
— Sharon Machlis (@sharon000) January 14, 2020
( #rstats weather geeks probably already know about @rOpenSci 's rnoaa package. But if you don't: https://t.co/IuXRfF7XtC ) pic.twitter.com/wvgFHnSPUT
Rough.js is a JavaScript library to draw lines and shapes roughly https://t.co/3SLiy4BiiK pic.twitter.com/xcd6S6JRLw
— Nathan Yau (@flowingdata) January 13, 2020
We have open-sourced wav2letter@anywhere, an inference framework for online speech recognition that delivers state-of-the-art performance. https://t.co/1W7PsVu8tO pic.twitter.com/9peZPbUNu4
— Facebook AI (@facebookai) January 13, 2020
🔥 Introducing Tokenizers: ultra-fast, extensible tokenization for state-of-the-art NLP 🔥
— Julien Chaumond (@julien_c) January 13, 2020
➡️https://t.co/BiOfFHfUdL pic.twitter.com/M8eT59A3gg