llama2-ko-13b-instruct-v1 This model is a fine-tuned version of meta-llama/Llama-13b-hf on an instruction-following dataset(670k)