Spaces:

Launchpad

/

lofi-bytes

like 0

Runtime error

Hi,

We’re Launchpad, a technology club at UC Berkeley focused on building creative ML projects. This is a demo of our finetuned MusicTransformer model that generates lofi tracks from input MIDI samples. We have a website (https://callaunchpad.github.io/lofi-bytes-app/) that sends API requests to this Space, but current runtime is extremely slow since we’re running on cpu.

We would like to apply for a GPU grant for faster inference. Our model is quite lightweight, <= 16GB VRAM should be more than enough for our use case. Thanks!