Nodes/ID-LoRA-LTX2.3-ComfyUI/ID-LoRA One-Stage Sampler
ComfyUI Node

ID-LoRA One-Stage Sampler

Generate audio+video with speaker identity transfer using the ID-LoRA one-stage pipeline.

By ID-LoRA·Created 5 months ago·Updated about a month ago· 148
ID-LoRA One-Stage Sampler
  • pipeline
  • conditioning
  • first_frame
  • reference_audio
  • Video
seed42
height512
width512
num_frames121
num_inference_steps30
frame_rate25.0
video_guidance_scale3.0
audio_guidance_scale7.0
auto_resolutiontrue
max_resolution512
CategoryID-LoRA

Inputs (14)

NameTypeDefaultDescription
pipelineID_LORA_PIPELINELoaded ID-LoRA pipeline.
conditioningID_LORA_CONDITIONINGEncoded prompt conditioning.
seedINT420–2147483647
heightINT51264–2048
widthINT51264–2048
num_framesINT1211–1000
num_inference_stepsINT301–200
frame_rateFLOAT25.01–120
video_guidance_scaleFLOAT3.00–30
audio_guidance_scaleFLOAT7.00–30
auto_resolutionBOOLEANtrueAuto-detect resolution from first-frame aspect ratio.
max_resolutionINT51264–2048Maximum long-side resolution for auto-resolution (ignored when auto_resolution is off).
first_frameoptIMAGEOptional first-frame image for face conditioning.
reference_audiooptAUDIOOptional reference audio for speaker identity transfer.

Outputs (1)

NameTypeDescription
VideoVIDEOGenerated video with audio.