ComfyUI Node: LLM Messenger

LLM Messenger:

The LF_LLMMessenger node is designed to facilitate seamless communication with a local Language Model (LLM). This node is particularly useful for AI artists who want to integrate conversational AI capabilities into their projects. By leveraging this node, you can manage and process chat data, ensuring that interactions with the LLM are smooth and efficient. The primary goal of the LF_LLMMessenger is to handle chat payloads, extract relevant information, and provide structured outputs that can be easily utilized in various creative applications. This node simplifies the process of interacting with LLMs, making it accessible even to those without a deep technical background.

LLM Messenger Input Parameters:

messenger

The messenger parameter is a JSON string that contains the chat payload. This payload includes the dataset and configuration necessary for the node to process the chat data. The dataset typically consists of nodes representing different characters and their respective chat histories, while the configuration specifies the current character and other relevant settings. The messenger parameter is crucial for the node's execution as it provides the necessary context and data for processing. Ensure that the JSON string is correctly formatted to avoid errors during execution.

LLM Messenger Output Parameters:

chat_data

The chat_data output is a JSON object representing the entire chat history. This includes all messages exchanged between the user and the LLM, providing a comprehensive record of the conversation.

chat_history_string

The chat_history_string output is a formatted string that presents the chat history in a readable format. Each message is prefixed with either "User" or the character's name, making it easy to follow the conversation flow.

last_message

The last_message output is a string containing the most recent message in the chat. This can be useful for quickly referencing the latest interaction without parsing the entire chat history.

last_user_message

The last_user_message output is a string containing the most recent message sent by the user. This helps in identifying the user's last input, which can be crucial for context in ongoing conversations.

last_llm_message

The last_llm_message output is a string containing the most recent message generated by the LLM. This allows you to quickly access the LLM's latest response, which can be useful for various applications.

styled_prompt

The styled_prompt output is a string that combines the chat history with additional styling or formatting. This can be used to create visually appealing representations of the conversation, enhancing the user experience.

character_name

The character_name output is a string that identifies the current character involved in the chat. This is important for context, especially when multiple characters are available in the dataset.

outfit_name

The outfit_name output is a string that specifies the current outfit of the character. This can be useful for applications where visual representation or character customization is involved.

location_name

The location_name output is a string that indicates the current location of the character. This adds another layer of context to the conversation, making it more immersive.

style_name

The style_name output is a string that describes the current style or theme of the conversation. This can be used to adjust the tone or presentation of the chat.

timeframe_name

The timeframe_name output is a string that specifies the current timeframe or period in which the conversation is set. This can be useful for historical or futuristic chat scenarios.

LLM Messenger Usage Tips:

• Ensure that the messenger parameter is a correctly formatted JSON string to avoid execution errors.
• Utilize the chat_history_string output to create readable and user-friendly representations of the chat.
• Leverage the styled_prompt output to enhance the visual appeal of the conversation in your applications.
• Use the character_name, outfit_name, location_name, style_name, and timeframe_name outputs to add rich context and detail to your chat scenarios.

LLM Messenger Common Errors and Solutions:

It looks like the chat is empty!

• Explanation: This error occurs when the messenger parameter is either empty or incorrectly formatted.
• Solution: Ensure that the messenger parameter is a valid JSON string containing the necessary dataset and configuration.

Invalid config format

• Explanation: This error indicates that the configuration part of the messenger parameter is not a valid JSON string.
• Solution: Verify that the configuration is correctly formatted as a JSON string within the messenger parameter.

You must choose a character

• Explanation: This error occurs when the configuration does not specify a current character.
• Solution: Ensure that the configuration includes a valid currentCharacter field.

Character with id <id> not found in dataset

• Explanation: This error indicates that the specified character ID does not exist in the dataset.
• Solution: Verify that the character ID in the configuration matches an existing character in the dataset.

It looks like the chat with <character_name> is empty

• Explanation: This error occurs when the chat history for the specified character is empty or missing.
• Solution: Ensure that the chat history for the specified character is present and correctly formatted in the dataset.