AIGC-Audio/Make-An-Audio-3 · Apply for community grant: Academic project (gpu)

Generative AI For Audio org Jun 19

This project is a new type of text-to-audio model that utilizes Flow Matching compared to past LDM/DDPM/DDIM projects. Our program has advantages in terms of intelligibility and match with text. Now our project has been open-sourced on Huggingface, and we have designed a demo that can be used by everyone. This has some GPU requirements, so we would like to apply for a free GPU.

Unfortunately, we accidentally switched to the CPU version and need to reapply for GPU usage.

Thank you for your understanding and support.

hysts

Jun 19

Hi @guowenxiang , I just reassigned ZeroGPU.

ysharma

Jun 19

Hey @guowenxiang this is a great demo! Congratulations on the release.
You can link your project/paper on the Space to drive more visibility to your project and also it would be great to fill out the model card for the ease of use for the community : https://huggingface.co/AIGC-Audio/Make-An-Audio-3
🥳 Also, we have included this demo in our today's Comms from the official Gradio SM handles (LI and X).