Inspiration
The groundbreaking AI Music research and momentum in the space. It's exciting to think about what this next era in music will look like with easy localization of songs, crowdsourcing creation to fans, and being able to create unique and interesting vocals instantly as a producer.
What it does
Transforms any song on Youtube with a chosen artist's voice.
How we built it
I readjusted a google colab that one of my ML friends made into a python webhook and then built a next.js app around it so anyone can easily make songs!
Challenges we ran into
Cloud deployment for the API has been a nightmare and took up a majority of the hackathon time. Still trying to get the generation down to <3 minutes.
Accomplishments that we're proud of
Getting the frontend with a working backend.
What we learned
ML Compute for complicated custom ML models is a headache.
What's next for Beat Llama
Building out a B2B api so any artist can easily localize their tracks, building out monetization for artists so they can collect proceeds from tracks generated with their voice.
Built With
- modal
- next.js
- python
Log in or sign up for Devpost to join the conversation.