Underdog Woof TTS โ 1-click Pinokio launcher ๐ถ
Underdog Woof is a tiny English text-to-speech model from Conway Research. This launcher runs it on your own machine
in one click, and no GPU is needed.
๐ชถ The whole model is one 40.6 MB file (38M parameters)
๐ป Runs on CPU using only numpy. No PyTorch, no CUDA
๐๏ธ Two voices: Juno and Mira
๐ข It reads numbers, dates, times, prices and codes the way people say them ("415-555-0199", "5:30 p.m.", "March 3rd")
๐ฒ Temperature and seed controls: the same text and seed always give the same audio
๐พ Every clip is saved as a 24 kHz WAV (Outputs button in the sidebar)
๐ Comes with an API, with Python, JavaScript and curl examples in the README
Install โ Start โ Open Web UI. That's it.
Model: https://huggingface.co/ConwayResearch/Underdog-Woof
Try it in your browser first: https://underdog.ai/woof
Launcher: https://github.com/PierrunoYT/Underdog-Woof-TTS
โ ๏ธ The model is non-commercial (Underdog Non-Commercial License 1.0). Credit it as "Voice by Underdog Woof".
Underdog Woof TTS โ 1-click Pinokio launcher ๐ถ
Underdog Woof is a tiny English text-to-speech model from Conway Research. This launcher runs it on your own machine
in one click, and no GPU is needed.
๐ชถ The whole model is one 40.6 MB file (38M parameters)
๐ป Runs on CPU using only numpy. No PyTorch, no CUDA
๐๏ธ Two voices: Juno and Mira
๐ข It reads numbers, dates, times, prices and codes the way people say them ("415-555-0199", "5:30 p.m.", "March 3rd")
๐ฒ Temperature and seed controls: the same text and seed always give the same audio
๐พ Every clip is saved as a 24 kHz WAV (Outputs button in the sidebar)
๐ Comes with an API, with Python, JavaScript and curl examples in the README
Install โ Start โ Open Web UI. That's it.
Model: https://huggingface.co/ConwayResearch/Underdog-Woof
Try it in your browser first: https://underdog.ai/woof
Launcher: https://github.com/PierrunoYT/Underdog-Woof-TTS
โ ๏ธ The model is non-commercial (Underdog Non-Commercial License 1.0). Credit it as "Voice by Underdog Woof".