Fine-tuningOpen SourceCommunity A Qwen Fine-Tune Aims for Humanlike Chat, Sidesteps 'AI Assistant' Tone A Reddit user's LoRA fine-tune on a Qwen model shows how a targeted dataset can produce more 'humanlike' chat than a generic model.