LumiOpen/Llama-Poro-2-8B-Instruct

LumiOpen's Llama-Poro-2-8B-Instruct is an 8 billion parameter instruction-following chatbot model, fine-tuned from Llama 3.1 8B. Developed in collaboration with AMD Silo AI, TurkuNLP, and HPLT, it excels in conversational AI and instruction following in both Finnish and English. This model demonstrates efficient multilingual capability addition through continued pretraining, supervised fine-tuning, and Direct Preference Optimization, achieving significant improvements in Finnish instruction-following benchmarks while maintaining strong English performance.