VLM Style Rewriter

Transforms text to specific styles using advanced models, enhancing context and quality.

VLM Style Rewriter:

The VLMStyleRewriter is a specialized node designed to enhance and transform text by rewriting it according to specific stylistic guidelines. This node is particularly useful in scenarios where you need to adapt existing text for different contexts or audiences, such as making it more cinematic, technical, narrative, artistic, or documentary in style. It leverages advanced text models like Qwen2.5 to perform these transformations, ensuring that the rewritten text aligns with the desired style while maintaining the original message's integrity. The node is ideal for the second round of a two-round processing system, where the first round involves analyzing an image using a vision model, and the second round focuses on refining the text output for specific use cases. By using this node, you can achieve more polished and contextually appropriate text outputs, enhancing the overall quality and effectiveness of your AI-generated content.

VLM Style Rewriter Input Parameters:

context

The context parameter provides the necessary environment or session information required for the node to operate. It is essential for maintaining state and ensuring that the text rewriting process is consistent with the ongoing session's requirements. This parameter does not have specific minimum or maximum values as it is typically managed by the system.

input_text

The input_text parameter is the original text that you want to rewrite. It serves as the foundation for the transformation process, and its content will be modified according to the specified style and instructions. There are no restrictions on the length of the input text, but longer texts may require more processing time.

rewrite_instruction

The rewrite_instruction parameter allows you to provide specific guidance on how the text should be rewritten. This can include particular changes or emphasis you want to be applied to the text. The instructions should be clear and concise to ensure the desired outcome.

style

The style parameter determines the stylistic approach used for rewriting the text. Options include "cinematic," "technical," "narrative," "artistic," "documentary," and "custom." Each style option applies a different set of guidelines to the text, influencing its tone and structure. The "custom" option allows you to use your own instructions without predefined guidance.

max_tokens

The max_tokens parameter specifies the maximum number of tokens (words or word parts) that the rewritten text can contain. This helps control the length of the output and ensures it remains within a manageable size. The exact range of values depends on the capabilities of the underlying text model.

temperature

The temperature parameter controls the randomness of the text generation process. A lower temperature results in more deterministic outputs, while a higher temperature introduces more variability and creativity. The value typically ranges from 0.0 to 1.0, with a default value around 0.7 for balanced creativity and coherence.

system_prompt

The system_prompt parameter provides a predefined prompt that sets the context for the text rewriting process. If not specified, a default prompt is used, which positions the system as an expert in rewriting text for specific purposes. This parameter helps guide the model's behavior and focus during the rewriting task.

trigger_word

The trigger_word parameter allows you to specify a word or phrase that should be emphasized or used as the main subject identifier in the rewritten text. This can help ensure that the output remains focused on a particular theme or concept.

avoid_terms

The avoid_terms parameter lets you list words or phrases that should be avoided in the rewritten text. This is useful for steering clear of specific language or topics that may not be appropriate for the intended audience or context.

response_cleanup

The response_cleanup parameter determines the level of post-processing applied to the rewritten text. Options include "none," "basic," "standard," and "strict," each applying different levels of text cleaning, such as trimming whitespace or removing unwanted characters. This ensures the final output is polished and free of unnecessary artifacts.