Nodes/ComfyUI_FL-SongGen/FL Song Gen Style Transfer
ComfyUI Node

FL Song Gen Style Transfer

A ComfyUI node in FL Song Gen with 10 inputs and 3 outputs.

By filliptm·Created 7 months ago·Updated 4 months ago· 62
FL Song Gen Style Transfer
  • model
  • reference_audio
  • mixed_audio
  • vocal_audio
  • bgm_audio
lyrics[intro-short] ; [verse] Hello world.This is a test ; [chorus] Singing along.Making music ; [outro-short]
description
duration60
temperature0.90
cfg_coef1.5
top_k50
gen_typemixed
seed-1
CategoryFL Song Gen

Inputs (10)

NameTypeDefaultDescription
modelSONGGEN_MODELLoaded SongGeneration model
lyricsSTRING[intro-short] ; [verse] Hello world.This is a test ; [chorus] Singing along.Making music ; [outro-short]Formatted lyrics with section tags
reference_audioAUDIOReference audio for style (max 10 seconds used)
descriptionoptSTRINGOptional style description to combine with reference audio (e.g., 'female, pop, emotional')
durationoptFLOAT6030–270Target duration in seconds
temperatureoptFLOAT0.900.1–2Sampling temperature (higher = more random)
cfg_coefoptFLOAT1.50.5–5Classifier-free guidance strength
top_koptINT501–500Top-k sampling (lower = more focused)
gen_typeoptCOMBOmixedOutput type: mixed, separate (all tracks), vocal only, or bgm only
seedoptINT-1-1–2147483647Random seed (-1 for random)

Outputs (3)

NameTypeDescription
mixed_audioAUDIO
vocal_audioAUDIO
bgm_audioAUDIO