The current CPU-based version on HuggingFace has slow inference, you can access the GPU-based mirror on
ModelScope
Video demo
0:00 / 0:00
Dataset
Generate by emotion condition
Valence (reflects negative-positive levels of emotion)
Low
High
Arousal (reflects the calmness-intensity of the emotion)
Low
High
Generate chords (coming soon)
Generate
Generate by feature control
Pitch SD
Low
High
Mode
Minor
Major
Tempo (BPM)
↺
40
228
Octave (±12)
↺
-24
24
Volume (dB)
↺
-5
10
Generate chords (coming soon)
Generate
The emotion to which the current template belongs
Q1
Q2
Q3
Q4
Save template
Audio
Download MIDI
Download PDF score
Download MusicXML
Download MXL
ABC notation
Staff
Feedback: the emotion you believe the generated result should belong to
Q1
Q2
Q3
Q4
Clear
Submit
output