ComfyUI Node: MultipleLoraLoader3 🍌

MultipleLoraLoader3 🍌| MultipleLoraLoader3 🍌:

The MultipleLoraLoader3| MultipleLoraLoader3 🍌 node is designed to facilitate the simultaneous loading and application of multiple LoRA (Low-Rank Adaptation) models to a primary model and an optional CLIP model. This node is particularly useful for AI artists who want to blend the effects of several LoRA models to achieve more nuanced and complex modifications to their base models. By allowing the user to specify the strength and application of each LoRA model, the node provides a high degree of customization and control over the final output. The node also includes normalization options to ensure that the combined strength of the applied LoRA models is balanced, preventing any single model from disproportionately influencing the result.

MultipleLoraLoader3 🍌| MultipleLoraLoader3 🍌 Input Parameters:

model

This parameter represents the primary model to which the LoRA models will be applied. It is a required input and serves as the base model that will be modified by the LoRA models.

normalize

This is a boolean parameter that determines whether the strengths of the LoRA models should be normalized. If set to True, the combined strength of the applied LoRA models will be scaled to match the value specified in normalize_sum. The default value is False.

normalize_sum

This parameter is a float that specifies the target sum for the normalized strengths of the LoRA models. It is only relevant if normalize is set to True. The default value is 1.0, with a minimum of -50.0 and a maximum of 50.0, adjustable in steps of 0.01.

lora_name_0

This parameter allows you to specify the name of the first LoRA model to be applied. It is a required input and can be set to "None" if no model is to be applied in this slot.

strength_model_0

This float parameter sets the strength of the first LoRA model's effect on the primary model. The default value is 1.0, with a minimum of -20.0 and a maximum of 20.0, adjustable in steps of 0.01.

apply_0

This boolean parameter determines whether the first LoRA model should be applied. The default value is True.

lora_name_1

This parameter allows you to specify the name of the second LoRA model to be applied. It is a required input and can be set to "None" if no model is to be applied in this slot.

strength_model_1

This float parameter sets the strength of the second LoRA model's effect on the primary model. The default value is 1.0, with a minimum of -20.0 and a maximum of 20.0, adjustable in steps of 0.01.

apply_1

This boolean parameter determines whether the second LoRA model should be applied. The default value is True.

lora_name_2

This parameter allows you to specify the name of the third LoRA model to be applied. It is a required input and can be set to "None" if no model is to be applied in this slot.

strength_model_2

This float parameter sets the strength of the third LoRA model's effect on the primary model. The default value is 1.0, with a minimum of -20.0 and a maximum of 20.0, adjustable in steps of 0.01.

apply_2

This boolean parameter determines whether the third LoRA model should be applied. The default value is True.

clip_optional

This optional parameter represents the CLIP model to which the LoRA models will be applied. If not provided, only the primary model will be modified.

MultipleLoraLoader3 🍌| MultipleLoraLoader3 🍌 Output Parameters:

MODEL

This output parameter represents the primary model after the specified LoRA models have been applied. It reflects the cumulative modifications made by the LoRA models based on the specified strengths and application settings.

CLIP

This output parameter represents the optional CLIP model after the specified LoRA models have been applied. If no CLIP model was provided as input, this output will be None.

MultipleLoraLoader3 🍌| MultipleLoraLoader3 🍌 Usage Tips:

• To achieve a balanced effect when applying multiple LoRA models, consider enabling the normalize option and setting an appropriate normalize_sum value.
• Experiment with different strengths for each LoRA model to find the optimal blend that enhances your primary model without overpowering it.
• If you do not want to apply a particular LoRA model, set its apply parameter to False or its lora_name to "None".

MultipleLoraLoader3 🍌| MultipleLoraLoader3 🍌 Common Errors and Solutions:

"LoRA model not found"

• Explanation: This error occurs when the specified LoRA model name does not exist in the designated folder.
• Solution: Ensure that the LoRA model name is correctly specified and that the model file is located in the appropriate folder.

"Strength sum is zero"

• Explanation: This error occurs when the combined strength of the applied LoRA models is zero, which can happen if all strengths are set to zero or if all apply parameters are set to False.
• Solution: Adjust the strengths of the LoRA models or ensure that at least one apply parameter is set to True.

"Normalization scale is invalid"

• Explanation: This error occurs when the normalization scale cannot be computed, usually due to a zero or negative normalize_sum.
• Solution: Ensure that normalize_sum is set to a positive value and that the combined strength of the LoRA models is not zero.