nikvaessen

s3prl

Self-Supervised Speech Pre-training and Representation Learning Toolkit

whisperX

WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)

whisper.cpp

Port of OpenAI's Whisper model in C/C++

w2v2-batch-size

Code for paper "The effect of batch size on contrastive self-supervised speech representation learning"

nikvaessen

tinygrad

You like pytorch? You like micrograd? You love tinygrad! ❤️

tinygrad-wav2vec2

A wav2vec 2.0 implementation using TinyGrad

ParallelWaveGAN

Unofficial Parallel WaveGAN (+ MelGAN & Multi-band MelGAN & HiFi-GAN & StyleMelGAN) with Pytorch

dscore

Diarization scoring tools.

LibriMix

An open source dataset for source separation

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

w2v2-speaker-few-samples

Research code for the paper "Training speaker recognition systems with limited data" at https://arxiv.org/abs/2203.14688

pytorch-lightning

The lightweight PyTorch wrapper for ML researchers. Scale your models. Write less boilerplate

jiwer

Python library for calculating Word Error Rate (WER), a common measure of speech recognition performance

debug-hydra-lightning

disjoint-mtl

Research code for "Towards multi-task learning of speech and speaker recognition" at https://arxiv.org/pdf/2302.12773.pdf

MLonHPC_May2023

Contains the material for the Machine Learning on HPC systems course on 16-05-2023

awk-course

Material for the "Introduction to awk programming" course at Heidelberg University

nanoGPT

The simplest, fastest repository for training/finetuning medium-sized GPTs.

triple_accel

Rust edit distance routines accelerated using SIMD. Supports fast Hamming, Levenshtein, restricted Damerau-Levenshtein, etc. distance calculations and string search.

2022-repo-mt-w2v2

PYLLR

Python toolkit for likelihood-ratio calibration of binary classifiers

kagi-docs

Documentation for products made by Kagi Inc

VocalAdversary2022

data

A PyTorch repo for data loading and utilities to be shared by the PyTorch domain libraries.

d2l-en

Interactive deep learning book with multi-framework code, math, and discussions. Adopted at 300 universities from 55 countries including Stanford, MIT, Harvard, and Cambridge.

w2v2-speaker

Research code for the paper "Fine-tuning wav2vec2 for speaker recognition" found at https://arxiv.org/abs/2109.15053

transformers

🤗 Transformers: State-of-the-art Natural Language Processing for Pytorch, TensorFlow, and JAX.

dutch-digit-parser

SBCSAE-preprocess

Preprocessing and downloading scripts for the Santa Barbara Corpus of Spoken American English (SBCSAE).