Anubis 70 B v1

Introduction

Anubis-70B-v1 is a highly creative and balanced language model developed by TheDrummer, part of the Hugging Face model repository. It is notable for its natural dialogue generation and ability to create engaging, human-like interactions in role-playing scenarios. The model is optimized for use with the Llama 3 Chat template.

Architecture

Anubis-70B-v1 is a fine-tuned version of a larger language model designed to provide a balance between creativity and performance. It is optimized for generating natural-sounding dialogue and prose, capable of delivering outputs comparable to models with a significantly larger parameter count, such as 120 billion parameters.

Training

The model has been fine-tuned on the L3.3 dataset, with a focus on delivering high-quality, creative outputs that enhance role-playing experiences. The training emphasizes natural flow, adherence to prompts, and creative writing, making it suitable for scenarios requiring human-like interaction and storytelling.

Guide: Running Locally

To run Anubis-70B-v1 locally:

1. Clone the repository: Access the model on Hugging Face and clone the repository.
2. Install dependencies: Ensure all required libraries and dependencies are installed.
3. Download the model files: Use the provided links to download the model weights and configuration files.
4. Load the model: Use a framework like PyTorch to load the model for inference.
5. Run inference: Prepare your inputs and run the model to generate outputs.

For optimal performance, it is recommended to use cloud GPUs such as NVIDIA A100 or V100 to handle the model's computational requirements.

License

The model is released under a custom license. For detailed information on permitted uses and restrictions, please refer to the license file included in the model's repository.