Josh Meyer

Deep learning for Text to Speech

🐸STT - a deep learning toolkit for Speech-to-Text, battle-tested in research and production

Use your data to create a speech recognition system in Kaldi. Fast.

Markdown template for Dataseets for Datasets

An example directory for running Multi-Task Learning training on Kaldi neural networks. In Kaldi-speak, this is an egs dir for nnet3 training.

Code for email classification in TensorFlow.

Forced Alignments for Common Voice

A Python3 program for converting Japanese words and numbers into phonemes.

A recipe for creating a Speaker Identification system built on Kaldi.

A living document for all things Common Voice.

Scripts for exporting Kaldi labeled data into TensorFlow

A set of scripts for running the bare bones of Kaldi's nnet2.

A site for things

Custom GLSL shaders for Ghostty terminal

No description.

Joshua's CV in LaTeX

Tool for scraping text from websites

MCP server that displays images inline in Ghostty/Kitty terminals, built for Claude Code

arxiv-sanity lite: tag arxiv papers of interest get recommendations of similar papers in a nice UI using SVMs over tfidf feature vectors based on paper abstracts.

No description.

No description.

No description.

Turn a git repo into a flat dir of files while preserving structure. Useful to feed into LLM context window

No description.

No description.

A super simple LaTeX template for little everyday things

A guide for researching ways of funding open source projects.

No description.

Text from the LJSpeech Corpus

No description.

No description.

Some small scripts for playing with and vizualizing neural nets and Gaussians.

I submitted this paper to Interspeech 2018. The paper was not accepted. The reviewer comments are included in the repo.

No description.

list of conferences for ASR research

My submission to the 2018 Machine Learning in Speech and Language Processing Workshop

No description.

Swift package for on-device speech-to-text with NVIDIA Parakeet TDT 0.6B v3 compiled to Core ML. Runs on macOS 14+ / iOS 17+, selectable across ANE / GPU / CPU.

🎙️ Lightweight macOS menubar app for voice-to-text dictation using OpenAI Whisper API. Hold-to-record with Fn key, instant transcription, auto-paste. Built with Hammerspoon.

Transform your voice into text instantly with the power of OpenAI, Groq, and Deepgram APIs. Say goodbye to macOS dictation limitations and hello to lightning-fast, accurate transcription! ⚡️

Get up and running with Kimi-K2.5, GLM-5, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

My personal Claude assistant that runs in Apple containers. Lightweight, secure, and built to be understood and customized for your own needs.

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!

A lightweight, powerful framework for multi-agent workflows

Personal tmux configuration with Emacs-style keybindings

No description.

https://hf.co/hexgrad/Kokoro-82M

No description.

Artie Bias Corpus: an audio corpus + code for detecting demographic bias

No description.

Benchmark suite for LLMs from Fireworks.ai

Split {Japanese, English} text into sentences.

🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production

Golang bindings for Coqui's speech-to-text library

Linguistic processing for Common Voice

Open models for Coqui STT

Common Voice is part of Mozilla's initiative to help teach machines how real people speak.

This project can help user to build their own container for OVHcloud AI Training Platform

Command line tool to create corpora for Common Voice

Coqui STT Model Manager - install, manage and try out Coqui STT models from the Model Zoo

Voice control add-on for WebThings Gateway

A TensorFlow implementation of Baidu's DeepSpeech architecture

Template for model cards

🐍 Coqui's machine learning job scheduler

No description.

Maltese Automatic Speech Recognition

CPU inference engine that delivers unprecedented performance for sparse models

No description.

a tool that leverages rich metadata and lineage information in MLMD to build a model card

my ~/.bashrc file

No description.

An Open Source Machine Learning Framework for Everyone

:microphone: A simple web interface for building voice assistants with Rasa

single-digit numbers and yes / no in different languages

Implementation code of non-parallel sequence-to-sequence VC

matplotlib backend for graph output in unicode terminals using drawille

eSpeak fork providing significant changes and improvements to the upstream eSpeak project

No description.

No description.

take in English text and map each character onto a 2 / 3 / or 4-byte UTF8 encoding

Collection of Log files and decoded test clips from common voice

A set of alphabets for training DeepSpeech with data from Common Voice

Client for the snakepit machine learning job scheduler

A curated list of speech and natural language processing resources

This is now the official location of the Kaldi project.

Converts 'western' style numbers to a Japanese context.

Some scripts and commands for working with Automatic Speech Recognition

Snips Python library to extract meaning from text

A parser for Japanese number (Kanji, arabic) in the natural language.

pykakasi is re-imprelemtation of kakasi library by Python.

No description.

Turkic text to speech

files relevant for training and testing ASR on a section of librispeech

TensorFlow implementation for the CIFAR-100 dataset

text files needed to run kaldi train / test on Kyrgyz audiobook

No description.

This is now the official location of the Merlin project.

NVDA, the free and open source Screen Reader for Microsoft Windows

No description.

No description.

Master thesis of Ondrej Platek: Automatic speech recognition using Kaldi. Supervised by Filip Jurcicek.

No description.

A LaTeX template for University of Arizona Letterhead

No description.

An atomic implementation of emacs keybindings for the Atom text editor.

No description.

A collection of tools to analyze tongue surface contours in ultrasound images

No description.