lwd Mirau R P 14 B

Introduction

MIRAU is a sophisticated character roleplaying model designed to enhance the consistency and expressiveness of roleplaying interactions. It leverages a story flow thought chain approach, allowing for dynamic narrative construction where users can simultaneously direct and act within the plot, requiring significant control over the system.

Architecture

The model operates by dividing its output into two sections: the model's inner monologue and the visible final response. This separation allows for enhanced character consistency and long-context memory capabilities, ensuring that dialogues remain engaging and coherent.

Training

The MIRAU model is built on the Qwen/Qwen2.5-14B-Instruct base model and utilizes the transformers library for its architecture, specifically the AutoModelForCausalLM. The training involves implementing system prompts with first-person narratives and managing dual roles within the interaction.

Guide: Running Locally

1. Clone the Repository:
   Access the model files from the Hugging Face repository.

2. Install Requirements:
   Ensure that Python and the transformers library are installed.

3. Download the Model:
   Load the model via the Hugging Face transformers library.

4. Run the Model:
   Execute the model locally using a Python script, ensuring the environment has sufficient resources.

5. Cloud GPUs:
   For optimal performance, especially with large models like this, consider using cloud GPUs such as those offered by AWS, Google Cloud, or Azure.

License

The license for this model is currently unknown, and users should verify the licensing terms before utilizing the model in different applications.