namelessai/audio-sr · Apply for community grant: Personal project (gpu)

Audiophiles and developers have been working together to create the AudioSR model, but there has never been a working demo for everyone because of the amount of VRAM it requires. The only demo is a paid one on Replicate. In the interest of bringing the power of AI to everyone, I humbly request a grant for a Nvidia A100 large processor, or at least 2 hours of free compute time a day. This should provide enough VRAM to run the space well enough for one person at a time to use it. (ZeroGPU does not have enough VRAM for the model.)