ComfyUI Node: LoRA Save 🍌

LoRA Save 🍌| LoRA Save 🍌:

The LoraSave| LoRA Save 🍌 node is designed to save a LoRA (Low-Rank Adaptation) model to a specified file format, ensuring that the model's parameters are correctly scaled and stored. This node is particularly useful for AI artists who work with custom-trained LoRA models and need to save their work for future use or sharing. The node takes into account the strength of the model and clip adjustments, applying necessary scaling to the parameters before saving. This ensures that the saved model retains the desired characteristics and can be reliably loaded and used in subsequent projects. By providing a straightforward interface for saving LoRA models, this node simplifies the workflow for AI artists, allowing them to focus more on creative aspects rather than technical details.

LoRA Save 🍌| LoRA Save 🍌 Input Parameters:

lora

This parameter represents the LoRA model that you want to save. It includes the model's parameters and the strength settings for both the model and the clip. The strength settings determine how much influence the LoRA model has on the final output. If both strengths are set to 1, the model is saved as-is. If they are different, the parameters are scaled accordingly before saving. This ensures that the saved model reflects the desired adjustments.

file_name

This is the name of the file where the LoRA model will be saved. It is a string parameter and does not support multiline input. The default value is "merged". This name will be used as the base name for the saved file, and it will be combined with the specified extension to form the complete file name. Choosing a descriptive file name can help you easily identify the saved model later.

extension

This parameter specifies the file format for saving the LoRA model. Currently, the only supported option is "safetensors". This format is chosen for its efficiency and safety in storing model parameters. The extension ensures that the file is saved in a compatible format that can be easily loaded and used in future projects.

LoRA Save 🍌| LoRA Save 🍌 Output Parameters:

None

This node does not produce any output parameters. Its primary function is to save the LoRA model to a file, and it does not return any values upon completion. The success of the operation can be confirmed by checking the specified save location for the presence of the saved file.

LoRA Save 🍌| LoRA Save 🍌 Usage Tips:

• Ensure that the lora parameter includes the correct model and strength settings before saving to get the desired results.
• Use a descriptive file_name to easily identify the saved model later, especially if you are working with multiple models.
• Verify that the save location has sufficient permissions and space to store the file to avoid any saving issues.

LoRA Save 🍌| LoRA Save 🍌 Common Errors and Solutions:

"Saving LoRA to <path> failed"

• Explanation: This error occurs when the node is unable to save the LoRA model to the specified path.
• Solution: Check if the save path is correct and accessible. Ensure that you have write permissions for the directory and that there is enough disk space available.

"Invalid LoRA model parameters"

• Explanation: This error indicates that the provided LoRA model parameters are not valid or incomplete.
• Solution: Verify that the lora parameter includes all necessary model parameters and strength settings. Ensure that the model is correctly loaded and configured before attempting to save.

"Unsupported file extension"

• Explanation: This error occurs if an unsupported file extension is specified.
• Solution: Ensure that the extension parameter is set to "safetensors", as this is the only supported format for saving LoRA models.