Problem on custom_fintune

[ZhihuaGao]

首先我基于训练代码训练了语言模型为qwen2.5 0.5B，视觉编码器为 siglip-so400m-patch14-384的多模态1模型，测试一切正常，然后我基于该训练好的siglip-so400m-patch14-384-qwen2-0_5b_instruct-finetune模型

进行custom_fintune，运行的时候报出下面的错误

ValueError: The checkpoint you are trying to load has model type tinyllava but Transformers does not recognize this architecture. This could be because of an issue with the checkpoint, or because your version of Transformers is out of date.

请问该如何解决啊？

[ZhihuaGao]

@bhneo @huangleiBuaa @jiajunlong @eltociear 而且hf上的模型使用transformers库直接推理的方式也都会报出ValueError: The checkpoint you are trying to load has model type tinyllava but Transformers does not recognize this architecture. This could be because of an issue with the checkpoint, or because your version of Transformers is out of date.这个错误，升级transformers也不能解决；

[Mactor2018]

Same problem, have you ever resolved it?

[ZhangXJ199]

For this issue, please refer to:
#162