ComfyUI Node: Qwen Processor (Wrapper)

Qwen Processor (Wrapper):

The QwenProcessorWrapper is a versatile node designed to process both text and images, leveraging the capabilities of the QwenVLTextEncoderLoaderWrapper. This node is integral for tasks that require the transformation of textual and visual data into a format suitable for further processing or analysis. It handles a variety of functions such as text tokenization, image preprocessing including resizing and normalization, and the insertion of vision tokens. Additionally, it can label images with "Picture X:" and format templates, making it a powerful tool for creating structured data inputs for AI models. The QwenProcessorWrapper is particularly beneficial for users looking to seamlessly integrate text and image data processing in their workflows, ensuring that the data is prepared in a consistent and efficient manner.

Qwen Processor (Wrapper) Input Parameters:

processor

The processor parameter is a required input that specifies the processor to be used for handling the text and image data. It must be of type QWEN_PROCESSOR, which ensures that the node can perform its intended functions effectively. This parameter is crucial as it dictates the processing capabilities available to the node.

text

The text parameter is a required input that allows you to provide the textual data to be processed. It is of type STRING and supports multiline input, with a default value of "A beautiful landscape". This parameter is essential for tasks that involve text-to-image generation or any operation where text needs to be transformed or analyzed.

images

The images parameter is an optional input that accepts image data for processing. It is of type IMAGE and is used when the task involves image editing or when images need to be preprocessed alongside text. This parameter is important for workflows that require the integration of visual data.

mode

The mode parameter is an optional input that determines the operation mode of the node. It can be set to either "text_to_image" or "image_edit", with a default value of "text_to_image". This parameter influences how the node processes the input data, allowing you to tailor the processing to specific tasks.

add_picture_labels

The add_picture_labels parameter is an optional boolean input that, when enabled, adds "Picture X:" labels to multiple images. It has a default value of True and provides a tooltip explaining its function. This parameter is useful for organizing and labeling images in a structured manner, especially when dealing with multiple images.

system_prompt

The system_prompt parameter is an optional input that allows you to provide a system prompt for template formatting. It is of type STRING, supports multiline input, and has a default value of an empty string. This parameter is beneficial for customizing the processing template according to specific requirements.

return_tensors

The return_tensors parameter is an optional input that specifies the format of the returned tensors. It can be set to either "pt" (PyTorch) or "np" (NumPy), with a default value of "pt". This parameter is important for ensuring compatibility with the desired framework or further processing steps.

Qwen Processor (Wrapper) Output Parameters:

processed

The processed output is of type QWEN_PROCESSED and represents the processed data resulting from the node's operations. This output is crucial as it provides the transformed text and image data, ready for subsequent use in AI models or other processing nodes.

formatted_text

The formatted_text output is of type STRING and contains the text that has been formatted according to the node's processing logic. This output is important for understanding how the input text has been transformed and for verifying the correctness of the processing.

Qwen Processor (Wrapper) Usage Tips:

• Ensure that the processor parameter is correctly set to a valid QWEN_PROCESSOR to avoid processing errors and to utilize the full capabilities of the node.
• Use the mode parameter to switch between text-to-image and image-editing tasks, optimizing the node's performance for your specific needs.
• When working with multiple images, enable the add_picture_labels parameter to maintain clarity and organization in your output.

Qwen Processor (Wrapper) Common Errors and Solutions:

No processor provided!

• Explanation: This error occurs when the processor parameter is not set or is invalid, preventing the node from executing its processing functions.
• Solution: Ensure that a valid QWEN_PROCESSOR is connected to the node. Verify that the processor is correctly loaded and available for use.

Failed to load processor: <error_message>

• Explanation: This error indicates that the node was unable to load the specified processor, possibly due to missing dependencies or incorrect configuration.
• Solution: Check the installation of required libraries and ensure that the processor is correctly configured. If using Qwen2_5_VLProcessor, verify its availability and consider using AutoProcessor as a fallback.