A large-scale multilingual speech corpus for representation learning, semi-supervised learning and interpretation

The Gridspace-Stanford Harper Valley speech dataset. Created in support of CS224S.

A list of papers that studies out-of-distribution (OOD) detection and misclassification detection (MisD)

Confidence-Aware Learning for Deep Neural Networks (ICML2020)

PyTorch implementation of our CVPR2023 paper "OpenMix: Exploring Out-of-Distribution samples for Misclassification Detection"

Test implementation of "Aligned Cross Entropy for Non-Autoregressive Machine Translation" https://arxiv.org/abs/2004.01655

NeMo: a toolkit for conversational AI

A simple cross attention that updates both the source and target in one step

[ACL'19] [PyTorch] Multimodal Transformer

Accompanying code for paper "Attention-Based Contextual Language Model Adaptation for Speech Recognition", submitted to ACL 2021.

Windows alt-tab on macOS

Code for "TAPIR: Learning Adaptive Revision for Incremental Natural Language Understanding with a Two-Pass Model", Findings of ACL 2023

Repo for the FB AI Speech team.

open-source Mandarian biased word dataset

End-to-end MOdeling of ASR (Automatic Speech Recognition)

The code for our ACL2022 findings paper: CRACSpell: A Contextual Typo Robust Approach with Copy Mechanism to Improve Chinese Spelling Correction

Source code for the paper "PLOME: Pre-training with Misspelled Knowledge for Chinese Spelling Correction" in ACL2021

NeuSpell: A Neural Spelling Correction Toolkit

Many Natural Language Processing tasks rely on sentence boundary detection (SBD). Although amazing libraries like spacy provide state of the art SBD, they often depend on text extractors (e.g pdf t…

🧑‍🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including transformers (original, xl, switch, feedback, vit, ...), optimizers (adam, adabelief, sophia, ...), ga…

A python package for finding words that sound like other words. Useful for entity resolution and poetry, among other things.

Various speech datasets made available to the public

A curated list of awesome papers on contextualizing E2E ASR outputs

Graph2Seq is a simple code for building a graph-encoder and sequence-decoder for NLP and other AI/ML/DL tasks.