Nomic Blog
"Open source, open data, open training code, fully reproducible and auditable text embedding model"
alphacep/vosk-api: Offline speech recognition API for Android, iOS, Raspberry Pi and servers with Python, Java, C# and Node
"Vosk is an offline open source speech recognition toolkit. [...] Vosk models are small (50 Mb) but provide continuous large vocabulary transcription, zero-latency response with streaming API, reconfigurable vocabulary and speaker identification." Bindings for various languages, "scales from small devices like Raspberry Pi or Android smartphone to big clusters."
eyung/Singling: Java application for sonification of linguistic data.
"Application for the sonification of text which can be transformed according to various triggers and parameters to facilitate the learning and analysis of literacoustics, reading by listening."
Rainbow Zero by Spinfoam Games
"Rainbow Zero is a... toy? widget? thingy? that allows you to explore a part of the space defined by the GloVe word vectors."
Applied Language Technology - YouTube
spacy 3 tutorials
Robustness Gym
"Despite impressive performance on standard benchmarks, deep neural networks often fail when deployed to real-world systems, due to distribution shifts, training artifacts, and noisy data. To address these vulnerabilities, we introduce Robustness Gym: a simple and extensible toolkit for robustness testing that supports the entire spectrum of evaluation methodologies, from adversarial attacks to rule-based data augmentations."
nipunsadvilkar/pySBD: đđŻpySBD (Python Sentence Boundary Disambiguation) is a rule-based sentence boundary detection that works out-of-the-box.
"a rule-based sentence boundary detection that works out-of-the-box"
ToxMod â Modulate
"detects toxic, disruptive, or otherwise problematic speech in real-time, and gives you the option to have our software respond immediately, or to escalate to your moderation team"
Hateful Memes Challenge winners
"Hate speech can come in many forms, including memes that combine text and images. This kind of multimodal content can be particularly challenging for AI to detect because it requires a holistic understanding of the meme." that is not the reason that hate speech is difficult to detect, and it's actually harmful that you think it's the reason, sorry
Natalie Schluter on Twitter: "Hey #NLProc and #linguistics people! I'm looking for key examples of government/national cultural institutions (not universities) being successful in the preservation of national/indigenous languages, and maybe even truly ena
language preservation and nlproc
UniMorph: Schema and datasets for universal morphological annotation
"a collaborative effort to improve how NLP handles complex morphology in the worldâs languages. The goal of UniMorph is to annotate morphological data in a universal schema that allows an inflected word from any language to be defined by its lexical meaning, typically carried by the lemma, and by a rendering of its inflectional form in terms of a bundle of morphological features from our schema."
on vocal cloning â Are.na
everest's bibliography on text-to-speech and vocal cloning
In-browser topic modeling
from david mimno, lda in the browser (may be helpful for workshops?)
NLP Course | For You
looks good!
StereoSet
"StereoSet is a dataset that measures stereotype bias in language models. StereoSet consists of 17,000 sentences that measures model preferences across gender, race, religion, and profession."
CCMatrix: A billion-scale bitext data set for training translation models
"CCMatrix is the largest data set of high-quality, web-based bitexts for training translation models. With more than 4.5 billion parallel sentences in 576 language pairs pulled from snapshots of the CommonCrawl public data set, CCMatrix is more than 50 times larger than the WikiMatrix corpus that we shared last year."
đŚđ¤đŚ Encoder-decoders in Transformers: a hybrid pre-trained architecture for seq2seq
this looks promising
alexwarstadt/blimp: The Benchmark of Linguistic Minimal Pairs
"a challenge set for evaluating what language models (LMs) know about major grammatical phenomena in English" it warms my heart to see an ngram baseline in there, haha
Autocat
very cool online text classifier generator (just upload your data and then you can pip install your model!)
Semantic Specialization of Distributional Representation Models
another tutorial from emnlp-19
Data Collection and End-to-End Learning for Conversational AI
overview + materials for emnlp-19 workshop
Measuring gender imbalances in reporting on the creative industries
Google AI Blog: Announcing Two New Natural Language Dialog Datasets
"In the movie-oriented CCPE dataset, individuals posing as a user speak into a microphone and the audio is played directly to the person posing as a digital assistant. The âassistantâ types out their response, which is in turn played to the user via text-to-speech. [...] The Taskmaster-1 dataset makes use of both the methodology described above as well as a one-person, written technique to increase the corpus size and speaker diversityâabout 7.7k written âself-dialogâ entries and ~5.5k 2-person, spoken dialogs. For written dialogs, we engaged people to create the full conversation themselves based on scenarios outlined for each task, thereby playing roles of both the user and assistant."
Learn how to make BERT smaller and faster
"ways to make huge models like BERT smaller and faster": quantization, pruning, distillation
Universal Dependencies
"Universal Dependencies (UD) is a framework for consistent annotation of grammar (parts of speech, morphological features, and syntactic dependencies) across different human languages. UD is an open community effort with over 200 contributors producing more than 100 treebanks in over 70 languages."
Other Orders
"Recommendation engines like the ones powering the endless feeds on Twitter, Facebook and YouTube, are designed to maximize ad revenue, and therefore to keep you online for as long as possible. In doing so they promote the most reactionary content on their platforms. Yet, these recommendation systems are nothing more than sorting mechanisms. Other Orders provides an alternate set of sorts, optimized for other outcomes."
The Annotated Transformer
BPEmb
"a collection of pre-trained subword embeddings in 275 languages, based on Byte-Pair Encoding (BPE) and trained on Wikipedia. Its intended use is as input for neural models in natural language processing"
trees are harlequins, words are harlequins â the transformer ... âexplainedâ?
"The Transformer is nothing more than an architecture where the core functional unit is attention. You stack attention layers on top of attention layers, just like you would do with CNN or RNN layers."
Tsvetshop: Home
"Yulia Tsvetkov's research group at Language Technologies Institute of Carnegie Mellon University. Our work focuses on natural language processing, particularly cross-lingual approaches, low-resource settings, and social good."
gpt-2-poetry
kyle mcdonald's take