CantoneseLLM v2: Reasoning in a Low-Resource Language
Evaluation across the training stages shows that chat-vector merging transfers instruction following but preserves the donor model's reasoning language, while SFT with limited Cantonese reasoning data substantially shortens or removes reasoning traces and reduces benchmark performance.
Tsz-Chung Cheng, Chung-Shing Cheng, Chaak-ming Lau et al.
· 0 citations