Matthew Baas

danbooru-pretrained. Pretrained pytorch models for the Danbooru2018 dataset

simple-speaker-embedding. A speaker embedding network in Pytorch that is very quick to set up and use for whatever purposes.

transfusion-asr. Transcribing Speech with Multinomial Diffusion, training code and models.

simple-asgan. Training code and trained checkpoints for ASGAN.

simple-autovc. A simple, performant re-implementation of AutoVC

lmao. Language Model Accessor Object - Neural autocomplete for Overleaf

simple-sashimi. A simple-to-use standalone Pytorch implementation of the SaShiMi speech synthesis model

simple-speech-commands. A pretrained Pytorch classifier for the Google Speech Commands dataset that is very quick to set up and use.

simple-pi-motion. Simple Raspberry Pi with camera module motion detection with Telegram notification

mos-finetune-ssl. MOS predictor network with torchhub integration

HierSpeechpp. Fork of HierSpeechpp with torchhub integration

Experiment-K. Random experiments, data, and testing 'out-there' ideas involving machine learning and stats

simple-diffwave. DiffWave, adapted to be simple to use and train.

DiffWave-unconditional. Pytorch Reimplementation of DiffWave unconditional generation: a high quality waveform synthesizer.