This isn't my project, and I have no affiliation with it; I’m simply offering a straightforward way to get started and providing instructions that save you from having to grapple with complex code compilation or read through lengthy GitHub articles - because, in reality, it’s much simpler than that. My only contribution is a quick-launch bat file (so you don't have to struggle with PowerShell) and an example of what can be achieved with this model.installation1) download the archive and unpack it (preferably in the root folder C)2) click run.batthis file containsstart "" "audiocpp_server.exe" --ui --ui-management --backend cpu --port 8081start <a target="_blank" rel="ugc" href="http://127.0.0.1:8081">http://127.0.0.1:8081</a>if this port is busy for you, then change its value for example to 80823) after launch, go to the models tab and download the model weights and vae<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/49d6852e-cba5-4294-b3c7-179ea1c30290/original=true/49d6852e-cba5-4294-b3c7-179ea1c30290.jpeg" />4) after that you can use the model by selecting it in the studio tab<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/77c68fd2-2238-473d-963c-086939587650/original=true/77c68fd2-2238-473d-963c-086939587650.jpeg" />link to repository <a target="_blank" rel="ugc" href="https://github.com/0xShug0/audio.cpp">GitHub - 0xShug0/audio.cpp: An all-in-one, pure C++ inference engine for audio models, powered by ggml. Supports TTS, STT, VAD, voice conversion, music generation, and more, with highly optimized performance. No Python dependency. · GitHub</a>

audio.cpp Yue2 3B (song gen) fast C++

YuE2 3B audio music model.This C++ version runs about twice as fast as the Comfy workflow and requires no installation.Here is my take on this model: it is roughly 8–10 times better than ACESTEP 1.5 and about 4–5 times better than ACESTEP 1.5 XL. In terms of sound quality, it outperforms Suno V6 Pro twofold, even though it often falls short regarding musical ideas. Nevertheless, the actual sound of the instruments, the timbres, and the overall audio quality are significantly better in Yue2 3B. Essentially, while this model may be less inventive compositionally, it surpasses its competitors specifically in terms of sound quality.

Files with SHA256: 75fc014c23977a1ad59cce69bdf0095b03e9fda7126e4557dea95af8fd263a17