YuE2 Particle Sliders v2

YuE2 Particle Sliders v2

Adds 16 sliders that steer the voice and genre of YuE2-3B audio generation, for people making English text to audio clips who want to shape the sound from one prompt.

Made by ntc-ai on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: YuE2 Particle Sliders v2 is a set of 16 adapters for the YuE2-3B text to audio model. Each slider shifts the voice or the genre of the generated sound while the prompt stays the same. A second set of standard LoRAs approximates the sliders for tools that only load those.

Who it suits: It suits people who generate audio with YuE2-3B and want a quick way to change how a voice or style comes out, without rewriting the prompt each time.

Industry | Who uses it | How they use it | Example
Music | Hobbyist producers | Try different voices and genres on the same lyrics | A producer renders one lyric sheet with a female voice, then a male voice
Research | Audio model researchers | Compare the sliders with the standard LoRA approximation and the base model | A researcher plays the three versions of the same clip side by side
Software | Tool builders | Add voice and genre controls to a local audio generator | A developer loads the converted sliders in a C++ app

If you need audio for a commercial product, you can skip it, since the license is non-commercial.

Where to use it:

• yue2.cpp, the C++ and GGML port of YuE2, using the converted slider files
• Ruach Studio, which runs those converted sliders
• Windows, macOS or Linux with Python, for downloading the files

Use case | Who | What they get back
Change the singer's voice | Producer | Clips with a female or a male voice from one prompt
Change the genre | Producer | A clip in a style such as pop from the same lyrics
Check slider against base | Researcher | Matching clips with the slider on, a LoRA version, and off

How to set it up:

• Python on your computer, to download the files
• The YuE2-3B base model, since these sliders are built for it
• yue2.cpp or Ruach Studio if you want to use the converted files

1. Install the Hugging Face hub package.
2. Download the repository files.
3. For yue2.cpp or Ruach Studio, take the converted files that goldhub hosts in a repository named Ruach_Studio_Models, which carries the same license.

pip install -U huggingface_hub
hf download ntc-ai/yue2-particle-sliders

How to use it: You load a slider alongside YuE2-3B, then generate with a caption, lyrics and a seed. The page compares three runs of the same input, about 20 seconds each: sliders on, the Distill LoRA, and Off for the plain base model. For the loading steps, follow the docs of yue2.cpp or Ruach Studio.

hf download ntc-ai/yue2-particle-sliders

Tool | What it does
hf download | Fetches the repository files to your computer
yue2.cpp | Runs YuE2 in C++ and GGML, and loads the converted sliders
Ruach Studio | An app that runs the converted sliders and credits this repository

Example: You keep one caption, one set of lyrics and one seed, then render three clips of about 20 seconds. One uses the slider for a female voice, one uses the Distill LoRA, and one has the slider off. You play them in turn and hear how much the slider changes the voice.

For newcomers and experts: A slider is a dial for one trait. Here the traits are the singing voice and the genre. The simplest first step is to play the sample clips on the model page, where the same input is rendered with the slider on, with the Distill version and with it off.

For experts, the sliders are built natively for YuE2-3B, and the standard LoRAs are distilled approximations of them. The trade is that a LoRA loads in ordinary tools but follows the slider only roughly. The Off clip is the base model, so it acts as your control.

YuE2-3B is the base model
Sixteen sliders adjust voice and genre
Distilled LoRAs approximate the sliders
You render with the same caption, lyrics and seed
You compare On, Distill and Off

Good to know:

• Language: English
• Samples: the page shows female, male and pop examples, among the 16 sliders
• Running them: loading the sliders is done in yue2.cpp or Ruach Studio, so check their own docs for the steps
• License: CC BY-NC 4.0, as Hugging Face lists it
• Cost to run: none named, apart from your own hardware
• Data it reaches: downloading uses the web and writes files to your disk, and the source names no other access
• Not for: commercial use, since the license is the non-commercial CC BY-NC 4.0

On Hugging Face, 13 November 2026

• Task: text to audio
• Downloads in the last 30 days: 0
• Likes: 22
• License: cc-by-nc-4.0
• Languages: en
• Based on: m-a-p/YuE2-3B
• Tags: yue2, lora, music-generation, comfyui, text-to-audio

Links

• Model: https://huggingface.co/ntc-ai/yue2-particle-sliders
• Files: https://huggingface.co/ntc-ai/yue2-particle-sliders/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download ntc-ai/yue2-particle-sliders

Then follow the usage in its model card. Nothing in this kit runs by itself.