Hugging Face Overview
Overview
scb10x/llama3.2-typhoon2-1b-instruct is a 1 billion parameter instruction-tuned large language model, built upon the Llama3.2-1B foundation. Developed by scb10x, this model is primarily focused on Thai language processing while also supporting English. It is designed to handle instruction-following and function-calling tasks effectively.
Key Capabilities
- Strong Thai Language Performance: Demonstrates competitive results in Thai-specific benchmarks such as IFEval-TH, MT-Bench TH, and Thai Code-Switching.
- Function Calling: Shows notable performance in both Thai and English function-calling tasks, outperforming Qwen2.5 1.5B and the base Llama3.2 1B instruct model in Thai function calls.
- Bilingual Support: Proficient in both Thai and English, making it suitable for mixed-language applications.
- Instruction Following: Achieves solid scores in instruction-following evaluations for both languages.
Good For
- Applications requiring a compact yet powerful model for Thai language understanding and generation.
- Use cases involving function calling where the model needs to interact with external tools or APIs, particularly in a Thai context.
- Scenarios demanding code-switching capabilities between Thai and English.
- Developers looking for an instruction-tuned model with a focus on Southeast Asian language proficiency.