Hugging Face's logo

dkounadis
/
artificial-styletts2

Text-to-Audio
Audiocraft
English
audiogen
styletts2
shift-tts
sound
audio-generation
text-to-speech
mimic3
Model card Files Community
4
artificial-styletts2 / README.md
Dionyssos's picture
Dionyssos
Audionar
02bf1ff
preview code
|
raw
history blame
4.48 kB
metadata
license: cc-by-nc-sa-4.0
language:
  - en
pipeline_tag: text-to-audio
tags:
  - audiocraft
  - audiogen
  - styletts2
  - shift-tts
  - sound
  - audio-generation
  - text-to-speech
  - mimic3

Audionar - StyleTTS2 of speakers pregenerated by another TTS

Beta Text 2 Speech Tool

SHIFT TTS / AudioGen

Phonetic variation of SHIFT TTS blend to AudioGen soundscapes

Listen Voices

Native English / Non-native English: Accents / Foreign languages

CUDA_DEVICE_ORDER=PCI_BUS_ID HF_HOME=/data/.hf7/ CUDA_VISIBLE_DEVICES=0 python demo.py

Flask API

Build virtualenv & run api.py

Above TTS Demo is a standalone script that loads TTS & AudioGen models and synthesizes a txt. We also provide a Flask api.py that allows faster inference with loading only once the TTS & AudioGen.

Clone

git clone https://huggingface.co/dkounadis/artificial-styletts2

Install

cd artificial-styletts2
virtualenv --python=python3.10 .env0
source .env0/bin/activate
pip install -r requirements.txt

Flask API - open a 2nd terminal

CUDA_DEVICE_ORDER=PCI_BUS_ID HF_HOME=/data/.hf7/ CUDA_VISIBLE_DEVICES=0 python api.py

Following examples need api.py to be running. Set this IP to the IP shown when starting api.py.

Foreign Lang TTS

This will produce the following video.

# git lfs pull # to download assets/ocr.jpg
python tts.py --text assets/ocr.txt --image assets/ocr.jpg --soundscape "battle hero" --voice romanian

Landscape 2 Soundscapes

The following needs api.py to be already running on a tmux session. 

# TTS & soundscape - output .mp4 saved in ./out/
python landscape2soundscape.py

For SHIFT demo / Collaboration with SMB

  • YouTube Videos

01

02

03

05

06

07

08

10

11

12

SoundScape Live Demo - Paplay

Flask API for playing sounds live

CUDA_DEVICE_ORDER=PCI_BUS_ID HF_HOME=/data/dkounadis/.hf7/ CUDA_VISIBLE_DEVICES=4 python api.py

Describe any sound via text, the tts & soundscape is played back

python live_demo.py  # type text & plays AudioGen sound & TTS

Audiobook

Create audiobook from .docx. Listen to it - YouTube male voice / female voice

#  audiobook will be saved in ./tts_audiobooks
python audiobook.py