YuE<T>

Models & Types · AiDotNet.Audio.Generation

YuE full-song music generation model with vocals and accompaniment.

Inheritance: NeuralNetworkBase<T> → AudioNeuralNetworkBase<T> → YuE<T>

Implements: IAudioGenerator<T>

For Beginners

YuE is like having a virtual band that can write and perform an entire song. You give it lyrics and a style (“pop, female vocalist, upbeat”) and it generates a complete song with singing, instruments, and production. Unlike most AI music tools that only make short clips, YuE can create full-length songs.

Usage:

How It Works

YuE (Yuan et al., 2025) generates complete songs with vocals and accompaniment from lyrics and genre/style tags. It uses a dual-AR architecture: a lyrics-conditioned language model generates semantic tokens, then a second stage produces acoustic tokens, enabling generation of songs lasting several minutes.

Constructors

Constructor Summary
YuE(NeuralNetworkArchitecture<T> architecture, YuEOptions options, IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>> optimizer) Creates a YuE model in native training mode.
YuE(NeuralNetworkArchitecture<T> architecture, string modelPath, YuEOptions options) Creates a YuE model in native training mode.

Properties

Property Summary
bool IsOnnxMode { get; }
bool SupportsAudioContinuation { get; }
bool SupportsAudioInpainting { get; }
bool SupportsTextToAudio { get; }
bool SupportsTextToMusic { get; }
double MaxDurationSeconds { get; }

Methods

Method Summary
AudioGenerationOptions<T> GetDefaultOptions()
ModelMetadata<T> GetModelMetadata()
Task<Tensor<T>> GenerateAudioAsync(string prompt, string negativePrompt, double durationSeconds, int numInferenceSteps, double guidanceScale, int? seed, CancellationToken cancellationToken)
Task<Tensor<T>> GenerateSongAsync(string lyrics, string styleTags, double durationSeconds, int? seed, CancellationToken cancellationToken) Generates a complete song asynchronously.
Tensor<T> ContinueAudio(Tensor<T> inputAudio, string prompt, double extensionSeconds, int numInferenceSteps, int? seed)
Tensor<T> GenerateAudio(string prompt, string negativePrompt, double durationSeconds, int numInferenceSteps, double guidanceScale, int? seed)
Tensor<T> GenerateMusic(string prompt, string negativePrompt, double durationSeconds, int numInferenceSteps, double guidanceScale, int? seed)
Tensor<T> GenerateSong(string lyrics, string styleTags, double durationSeconds, int? seed) Generates a complete song from lyrics and style tags.
Tensor<T> InpaintAudio(Tensor<T> audio, Tensor<T> mask, string prompt, int numInferenceSteps, int? seed)