rinna/qwen2.5-bakeneko-32b-instruct-v2

rinna/qwen2.5-bakeneko-32b-instruct-v2 is a 32.8 billion parameter instruction-tuned causal language model developed by rinna, based on the Qwen2.5 architecture. This model is fine-tuned using Chat Vector and Odds Ratio Preference Optimization (ORPO) to enhance instruction-following capabilities. It demonstrates performance comparable to a reasoning model on Japanese MT-Bench without requiring additional reasoning processes, making it suitable for complex Japanese language tasks.