The current CPU-based version on HuggingFace has slow inference, you can access the GPU-based mirror on ModelScope
0:00 / 0:00
Dataset

Generate by emotion condition
Valence (reflects negative-positive levels of emotion)
Arousal (reflects the calmness-intensity of the emotion)

Generate by feature control
Pitch SD
Mode
40 228
-24 24
-5 10
The emotion to which the current template belongs

Cite

@inproceedings{Zhou2025EMelodyGen,
  title     = {EMelodyGen: Emotion-Conditioned Melody Generation in ABC Notation with the Musical Feature Template},
  author    = {Monan Zhou and Xiaobing Li and Feng Yu and Wei Li},
  month     = {Mar},
  year      = {2025},
  publisher = {GitHub},
  version   = {0.1},
  url       = {https://github.com/monetjoe/EMelodyGen}
}
Feedback: the emotion you believe the generated result should belong to