Gurubot/chatterbots-uncensored-8b
Gurubot/chatterbots-uncensored-8b is an 8 billion parameter language model designed to simulate multi-character online chatroom conversations with distinct personalities. Unlike standard chat models, it uses a custom XML-delimited JSON format to generate realistic, interleaved dialogues from multiple participants simultaneously. This model excels at creating dynamic, scenario-based chat interactions, including character arguments and moderator interventions, while saving multiple LLM calls.
Loading preview...
Chatterbots: Multi-Character Chatroom Simulation
Gurubot/chatterbots-uncensored-8b is a unique 8 billion parameter model specifically engineered to simulate dynamic, multi-character online chatroom environments like Discord or IRC. It generates responses from several distinct personalities at once, creating realistic, back-and-forth conversations based on a user-provided scenario. This model is not a typical chat/instruct model and utilizes a custom XML-delimited JSON format for input and output, allowing for well-defined conversational structures and metadata like in-reply-to.
Key Capabilities & Features
- Multi-Character Simulation: Generates messages from multiple users simultaneously, each with genuine and consistent personality differences.
- Dynamic Scenarios: Users define the chatroom's scenario, character types, and conversation topics via a
descriptionfield, which can be updated mid-conversation. - Custom Format: Employs a unique XML-delimited JSON input/output structure, differing from standard chat templates, to manage complex multi-participant dialogues.
- Efficiency: Designed to save multiple LLM calls by generating batches of messages, and allows for prompt caching by the inference engine.
- Uncensored Output: Based on an uncensored base model, requiring users to implement their own content moderation if needed.
Good For
- Simulating Online Communities: Ideal for creating realistic, interactive chatroom experiences for games, virtual environments, or social simulations.
- Character Interaction Prototyping: Useful for developers needing to test complex character dynamics and conversational flows.
- Reducing LLM Calls: Efficiently generates multiple character responses in a single call, optimizing for scenarios requiring rich, multi-party dialogue.
Users can provide a list of desired usernames or let the model invent its own. The description field is crucial for guiding the model's behavior, allowing for both broad and specific conversational themes. Strategies for dealing with repetition, such as adjusting temperature or using specific sampling parameters, are also provided.