I.N.F.E.R.N.O.S.

Interactive Neural Framework for Efficient Realtime Network Operations on Streams

🔥 Welcome to Infernos, where data comes to life in real-time! 🔥

Overview

Harness the power of I.N.F.E.R.N.O.S. to transform audio, video, and text streams with state-of-the-art inference in an instant. Embrace a blazing-fast future, free from lag.

News

Work is underway to scale TTS performance 50x. Stay tuned.

I am going to present at FOSDEM 2024 (RTC track). See you all there!

Features

Interactive: Infernos isn't just another tool; it's an experience. Speak in one voice and marvel as it's automatically translated into a completely different tone or even language, and then seamlessly transmitted in real-time during phone or web meetings.
Neural Power: With deep learning at its core, Infernos is optimized for top-notch performance.
Multimodal Support: Whether it's audio, video, or text, Infernos handles them with elegance.
Efficiency: Designed for low-latency, high-throughput operations.
Realtime: Don't wait. Experience the magic as it unfolds.

Quick Start

Clone the repository:

git clone https://github.com/sippy/Infernos.git

Navigate to the project directory and install dependencies:

cd Infernos && pip install -r requirements.txt

Light the fire:

python Infernos.py -f

Use SIP device or software such as Linphone to place a SIP call to sip:anything@localhost:5060. Replace localhost with a local IP of machine running Infernos if testing over LAN.

Ready to experience real-time inferencing?

Contribute

Feeling the warmth? 🔥 Eager to stoke the flames of Infernos? Delve into our contribution guidelines and join the firestorm!

License & Acknowledgements

Powered by the 2-clause BSD license. A heartfelt shoutout to the community for their priceless insights and tireless contributions.

Stay on the lookout for more sizzling updates, and always remember: Infernos makes the future sizzle!