LoraModelLoader - ComfyUI Built-in Node Documentation

The LoraModelLoader node applies trained LoRA (Low-Rank Adaptation) weights to a diffusion model. It modifies the base model by loading LoRA weights from a trained LoRA model and adjusting their influence strength. This allows you to customize the behavior of diffusion models without retraining them from scratch, including a bypass mode that leaves the base model weights unchanged.

Inputs

Parameter	Description	Data Type	Required	Range
`model`	The diffusion model the LoRA will be applied to.	MODEL	Yes	-
`lora`	The LoRA model to apply to the diffusion model.	LORA_MODEL	Yes	-
`strength_model`	How strongly to modify the diffusion model. This value can be negative (default: 1.0).	FLOAT	Yes	-100.0 to 100.0
`bypass`	When enabled, applies LoRA in bypass mode without modifying base model weights. Useful for training and when model weights are offloaded (default: False).	BOOLEAN	Yes	True or False

Note: When strength_model is set to 0, the node returns the original model without applying any LoRA modifications.

Outputs

Output Name	Description	Data Type
`model`	The modified diffusion model with LoRA weights applied.	MODEL

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 81eb2a9b0376fe7453f6e7e422414472e80a3d1b92bb6874b91df6de8aed0d9a

​Inputs

​Outputs

Inputs

Outputs