sanghwa-na/llama2-13b.kor: A Korean-Optimized LLaMA-2 Model

This model, developed by Sanghwa Na, is a 13 billion parameter language model built upon the robust LLaMA-2 architecture. It has been specifically fine-tuned to excel in Korean language understanding and generation tasks, differentiating it from general-purpose LLMs.

Key Capabilities

• Korean Language Proficiency: Optimized for processing and generating text in Korean.
• Instruction Following: Enhanced through training on Orca-style and Platypus datasets, allowing it to follow complex instructions effectively.
• LLaMA-2 Backbone: Benefits from the strong foundational capabilities of the LLaMA-2 model family.

Training Details

The model was trained using the transformers library and utilized a combination of Orca-style and Platypus datasets. This training approach aims to improve its ability to understand and respond to diverse prompts.

Prompt Format

Users should interact with the model using the following instruction-answer template:

### Instruction:
{Instruction}

### Answer:
{Answer}

Good For

• Applications requiring strong Korean language generation.
• Instruction-tuned tasks in Korean.
• Developers looking for a LLaMA-2 based model with specialized Korean language capabilities.