Thanks for sharing the code!
I'm trying to reproduce the results, but I couldn't find the dataset or data preparation instructions in the README. Could you please clarify how to obtain the data used in the paper?
Also, the paper only provides the fine-tuning layer number for Qwen2.5-VL-7B; could you provide experimental settings for other models, such as Qwen2.5-VL-2B/3B?
Thanks for sharing the code!
I'm trying to reproduce the results, but I couldn't find the dataset or data preparation instructions in the README. Could you please clarify how to obtain the data used in the paper?
Also, the paper only provides the fine-tuning layer number for Qwen2.5-VL-7B; could you provide experimental settings for other models, such as Qwen2.5-VL-2B/3B?