F5 CYBER AUDIO - LOCAL SETUP

1. Install Python 3.12 from https://www.python.org/downloads/ if needed.
2. Put app.py and requirements.txt in the same folder.
3. On macOS, also download Install and Start.command and Start Audio.command.
   In Terminal, type bash followed by a space, drag Install and Start.command into
   the window, and press Return. For later launches do the same with Start Audio.command.
4. On Windows: open Terminal in that folder and run:
   py -3.12 -m venv .venv
   .venv\Scripts\python -m pip install -r requirements.txt
   .venv\Scripts\python app.py
5. On Linux: use python3.12 -m venv .venv, then .venv/bin/python -m pip install
   -r requirements.txt, then .venv/bin/python app.py.
6. The app opens at http://127.0.0.1:7860. Leave Terminal running.

FIRST USE
Download the audio briefing PDF from https://f5cyber.com.
Upload it in the local app. Review its extracted narration.
Optionally upload 10-20 seconds of clear speech in your own voice (WAV preferred).
Confirm permission and select Generate audio. Listen and download the WAV.
No fixed audio duration: it follows the narration length.
The first generation downloads several GB of model files and may take time.
Apple silicon, NVIDIA GPU, or CPU are selected automatically; CPU can be slow.
Close Terminal to stop the app. Reopen with Start Audio.command on macOS.

PRIVACY AND OUTPUT
No cloud audio API or subscription is required. Your documents and voice samples
are processed locally. Models are downloaded from Hugging Face. No public share
link is created. Audio is watermarked by Chatterbox. Review facts, acronyms and
pronunciation before publishing. Do not use a voice without permission.
Generated WAVs are stored in output/; downloaded models in models/.
Scanned/encrypted PDFs are not supported. Prefer the F5 Cyber audio PDF, which
includes narration.txt; generic text PDFs may need editing after extraction.
If port 7860 is already occupied, close the other local app before launching.

Official Chatterbox source and MIT license:
https://github.com/resemble-ai/chatterbox