streamspeech

streamspeech: generative AI that talks!

Features List

• Text to Speech

• Speech to Text

• Custom user voices

• API Free Development (local LLM or text generation model)

• Image to text to speech

Installation

The simplest way to install the latest release is as follows:

pip install streamspeech

To install the development version:

Open the Terminal/CMD/Git bash/shell and enter

pip install git+https://github.com/Nelson-Gon/streamspeech.git

# or for the less stable dev version
pip install git+https://github.com/Nelson-Gon/streamspeech.git@dev

Otherwise:

# clone the repo
git clone git@github.com:Nelson-Gon/streamspeech.git
cd streamspeech
pip install -e . 

Usage

python -m streamspeech

The above will open a web browser with the app. NOTE: At the moment, one needs to sign up for an API Key to use this app.

Acknowledgements

streamspeech is made possible by the PaLM API, speechbrain by NVIDIA, and streamlit

To report any issues, suggestions or improvement, please do so at issues.

If you would like to cite this work, please use:

Nelson Gonzabato (2023) streamspeech: Generative AI that talks! https://github.com/Nelson-Gon/streamspeech

Thank you very much.

“Ideas come and go, stories stay” – NNT, The Black Swan