lyc728 opened this issue 2 months ago · 0 comments
请问,在SFT qwen-vl时,如果想微调一个chat类的模型,在三阶段时用的微调模型是二阶段的,还是qwen-vl-chat模型呢?