For my AI experiments I personally use ComfyUI, which is easy to use and flexible with its node-based interface. It also has a lot of ready-made workflows, and when a model is not available the program knows where to download it. Simple and easy.
Support for ACE-Step (an open-source music generation model jointly developed by ACE Studio and StepFun) has recently been added and it generates various music genres, including General Songs, Instrumentals, and Experimental Inputs, all supported by multiple languages.
ACE-Step is really very interesting (read at least the presentation on his home page) as it allows rich extensibility for the OSS community: through fine-tuning techniques like LoRA and ControlNet, developers can customize the model according to their needs, whether audio editing, vocal synthesis, accompaniment production, voice cloning, or style transfer applications. The model is a meaningful milestone for the music/audio generation genre.
The possibility of inserting lyrics for the singing is remarkable, very effective.
Add the ability to edit existing music and imagine the endless possibilities by yourself.
What else to say, you have to try it!
This is a raw short video of a 3 minute song generated in 145 seconds with a simple prompt on my poor PC (6GB VRAM).
Crazy. From China, fist of fury . For free, on your PC.
Leave a Reply