quick question: LM model is different from text clf model, correct?

Hi,

I guess LM to generate words for augmentation is one, and text clf model is another, and thus LM model's parameters cannot applied to text clf model.
Am I correct?

Thanks,