Chat-Haruhi-Suzumiya aims to revive anime characters as interactive chatbots using large language models, targeting anime fans and researchers interested in character role-playing. It allows users to engage in conversations with AI-powered characters that mimic their personalities, speech patterns, and plot contexts, offering an immersive experience.

How It Works

The project leverages fine-tuned large language models to achieve zero-shot role-playing. It processes character-specific data, including dialogue scripts and personality traits, to generate contextually relevant and in-character responses. The architecture supports various LLM backends and embedding models, enabling flexibility in deployment and customization for different characters and languages.

Quick Start & Requirements

• Install via pip: pip install transformers openai tiktoken langchain datasets chatharuhi
• Alternatively, clone the repository: git clone https://github.com/LC1332/Haruhi-2-Dev and adjust imports.
• Colab notebooks are available for quick testing with various LLMs and characters.
• Requirements include Python, Hugging Face libraries, and potentially API keys for services like OpenAI.

Highlighted Details

• Supports over 140 characters, with ongoing expansion and community contributions.
• Offers integration with multiple LLM providers (OpenAI, GLMPro, Qwen, Ernie) and embedding models.
• Provides tools for data extraction, preprocessing, and fine-tuning for custom characters.
• Achieved second place in two AI hackathons.

Maintenance & Community

The project is actively developed by a large team recruited from open-source communities like DataWhale. Recent updates include new datasets and fine-tuned models. Links to community discussions or support channels are not explicitly provided in the README.

Licensing & Compatibility

The project is licensed under Apache 2.0, permitting commercial use. However, users must adhere to the licensing terms of any underlying LLMs, APIs, or character copyrights used.

Limitations & Caveats

The project is described as "under construction" with ongoing refactoring for ChatHaruhi 2.0. Some features like image and sound integration have been removed in the current version and are planned for future research. API usage can be costly due to the extensive use of few-shot examples.