Biały Wilk
Results
21
comments of
Biały Wilk
> > If you use LoRA on the base models, you need to finetune the embeddings and the lm_head if you use the ChatML template. > > 使用ChatML方式。 > >...