OpenThaiGPT 13b 0.1.0-beta Overview

OpenThaiGPT 13b 0.1.0-beta is a 13-billion parameter language model built upon the LLaMA architecture, specifically fine-tuned for the Thai language. Developed by OpenThaiGPT Volunteers, the Artificial Intelligence Entrepreneur Association of Thailand (AIEAT), and the Artificial Intelligence Association of Thailand (AIAT), this model is designed to understand and follow instructions translated into Thai.

Key Capabilities

• Thai Instruction Following: Optimized to process and respond to instructions provided in Thai.
• LLaMA-based Architecture: Leverages the robust LLaMA framework for its foundational language understanding.
• Community-Driven Development: Supported by a collaborative effort from Thai AI associations and volunteers.

Good For

• Research in Thai NLP: Ideal for academic and research purposes focusing on large language models in the Thai context.
• Thai Language Applications: Suitable for developing applications that require instruction-tuned capabilities for Thai users.
• Exploring LLaMA in Thai: Provides a specific implementation for those interested in LLaMA's performance with Thai language data.

Licensing

The source code is released under the Apache Software License 2.0. However, the model weights are currently restricted to research use only due to the Facebook LLaMA's weight license, with a commercial use license anticipated for future releases.