ComfyUI Node: Pre CFG exponentiation

Pre CFG exponentiation:

The Pre CFG exponentiation node is designed to modify the conditioning data used in AI models by applying an exponentiation operation. This node is particularly useful for fine-tuning the influence of conditioning and unconditioning data on the model's output, allowing for more nuanced control over the generated results. By adjusting the exponentiation factor, you can amplify or diminish the impact of certain conditions, leading to more precise and desired outcomes in your AI-generated art. This node is essential for artists looking to experiment with different levels of conditioning influence to achieve specific artistic effects.

Pre CFG exponentiation Input Parameters:

model

This parameter represents the AI model that will be modified by the node. It is a required input and ensures that the node has a model to apply the exponentiation operation on.

do_on

This parameter determines which parts of the conditioning data the exponentiation operation will be applied to. It has three options: both, cond, and uncond. The default value is both. Selecting both applies the operation to both conditioning and unconditioning data, cond applies it only to the conditioning data, and uncond applies it only to the unconditioning data. This allows for targeted adjustments based on your specific needs.

exponent

This parameter specifies the exponent value used in the exponentiation operation. It is a floating-point number with a default value of 0.8. The minimum value is 0.0, and the maximum value is 10.0, with a step size of 0.05 and rounding to two decimal places. Adjusting this value changes the degree to which the conditioning data is modified, providing fine control over the model's behavior.

Pre CFG exponentiation Output Parameters:

model

The output is the modified AI model with the applied exponentiation operation. This model will now use the adjusted conditioning data during its processing, leading to potentially different and more refined outputs based on the specified exponentiation settings.

Pre CFG exponentiation Usage Tips:

• Experiment with different exponent values to see how they affect the output. Lower values will have a subtler effect, while higher values can significantly amplify or diminish the conditioning influence.
• Use the do_on parameter to target specific parts of the conditioning data. For example, if you only want to adjust the unconditioning data, set do_on to uncond.

Pre CFG exponentiation Common Errors and Solutions:

TypeError: 'NoneType' object is not subscriptable

• Explanation: This error occurs when the input model is not properly provided or is None.
• Solution: Ensure that a valid model is passed to the node as the model parameter.

ValueError: Exponent value out of range

• Explanation: This error occurs when the exponent value is set outside the allowed range (0.0 to 10.0).
• Solution: Adjust the exponent value to be within the specified range.

RuntimeError: Invalid 'do_on' option

• Explanation: This error occurs when an invalid option is provided for the do_on parameter.
• Solution: Ensure that the do_on parameter is set to one of the valid options: both, cond, or uncond.