ComfyUI Node: CD Tuner 🍌

CD Tuner 🍌| CD Tuner 🍌:

The CD_Tuner| CD Tuner 🍌 node is designed to fine-tune the details and contrast of your AI-generated models, providing you with greater control over the visual output. By adjusting specific parameters, this node allows you to enhance or reduce details and contrast in your models, making it a powerful tool for AI artists looking to refine their creations. The node works by modifying the weights and biases of certain layers within the model, applying these changes during a specified range of timesteps. This targeted approach ensures that the adjustments are applied precisely when needed, resulting in more nuanced and visually appealing outputs.

CD Tuner 🍌| CD Tuner 🍌 Input Parameters:

model

This parameter represents the AI model you wish to fine-tune. It is a required input and serves as the base model that will undergo adjustments.

detail_1

This parameter controls the first level of detail adjustment by reducing the weight and increasing the bias of the initial convolutional layer. The value ranges from -10 to 10, with a default of 0. Adjusting this parameter can either enhance or diminish the fine details in your model's output.

detail_2

This parameter manages the second level of detail adjustment, specifically targeting the GroupNorm layer before the final convolutional layer. The value ranges from -10 to 10, with a default of 0. Modifying this parameter allows for further refinement of the model's details.

contrast_1

This parameter adjusts the contrast by increasing the bias of the first channel in the final convolutional layer. The value ranges from -20 to 20, with a default of 0. By tweaking this parameter, you can enhance or reduce the contrast in your model's output, making it more visually striking or subdued.

start

This integer parameter defines the starting timestep for applying the adjustments. The value ranges from 0 to 1000, with a default of 0. It determines when the fine-tuning process begins during the model's execution.

end

This integer parameter sets the ending timestep for the adjustments. The value ranges from 0 to 1000, with a default of 1000. It specifies when the fine-tuning process concludes, ensuring that the adjustments are applied within a controlled timeframe.

CD Tuner 🍌| CD Tuner 🍌 Output Parameters:

MODEL

The output is the fine-tuned model, which incorporates the specified adjustments to details and contrast. This modified model can then be used for further processing or generation tasks, providing enhanced visual quality based on the input parameters.

CD Tuner 🍌| CD Tuner 🍌 Usage Tips:

• To achieve a more detailed output, increase the detail_1 and detail_2 parameters gradually and observe the changes.
• For a more dramatic visual effect, adjust the contrast_1 parameter to higher positive or negative values.
• Use the start and end parameters to apply adjustments only during specific phases of the model's execution, allowing for more precise control over the final output.

CD Tuner 🍌| CD Tuner 🍌 Common Errors and Solutions:

"Invalid model input"

• Explanation: This error occurs when the provided model is not compatible with the CD_Tuner node.
• Solution: Ensure that the input model is correctly formatted and compatible with the node's requirements.

"Parameter out of range"

• Explanation: This error happens when one of the input parameters is set outside its allowed range.
• Solution: Verify that all input parameters are within their specified ranges and adjust them accordingly.

"Adjustment application failed"

• Explanation: This error indicates that the node failed to apply the specified adjustments to the model.
• Solution: Check the model's structure and ensure that the layers targeted by the adjustments exist and are accessible.